Targeted Cancer Therapy

Genetic and epigenetic mechanisms initiate the progression of cancer. Therapeutic approaches to cancer include local control with surgery and/or radiation, with combination chemotherapy for systemic control. Historically, specific targeting of cancer cells with antihormonal therapies to estrogen receptor (breast cancer) and testosterone receptor (prostate cancer) has been successful. However, the “Achilles heel” of cancer treatment has been failure of current treatments because of the emergence of genetic mechanisms of drug resistance.

The image below depicts positive staining of tumor cell nuclei with monoclonal antibody to estrogen receptor.

The 'omics' technology coupled with ultrafast DNA sequencing has led to the identification of dysregulated, overlapping core oncogenic signaling pathways that drive cancer. Moreover, combinations of molecular techniques, including high-throughput shRNA or siRNA knockdown, have provided tools to validate discovered targets within the context of vulnerability and synthetic lethality. Effective targeting of aberrant-oncogene addicted and nonaddicted pathways with rational combinations of targeted therapies directed to the hallmarks of cancer will be the key to cure.

This article reviews current trends and future directions of targeted therapies that are most likely to make significant contributions to the fight against cancer.

Human cancers arise via a multistep mutagenic process reflective of genetic and epigenetic changes that drive progressive transformation of normal cells into highly malignant counterparts. Many types of human cancers have an age-dependent incidence implicating potentially 4-7 rate-limiting stochastic events. ^[1] This observation mirrors pathologic analyses of tumors that reveal the existence of premalignant lesions that progressively evolve from normal to invasive to metastatic cancer. ^[2]

Many human tumor types with distinct genotypes have 6 essential alterations in cell physiology that appear to collectively dictate the malignant phenotype. These cellular processes are self-sufficiency in growth signals (oncogene addiction), insensitivity to growth-inhibitory signals (loss of tumor suppressors), evading programmed cell death (anti-apoptosis), limitless replication potential (aberrant cell cycle), sustained angiogenesis, and invasion/metastasis. ^[3] Each of these 6 acquired physiologic capabilities occurs during tumor development because of reactivation and changing of existing cellular programs utilized during embryogenesis and development.

Two additional hallmarks have been proposed based on evading immune surveillance ^[4] and the cancer cell stress response phenotypes. ^[5] The final common pathway is the successful breaching of a hardwired anticancer defense program by cancer cells that survive and proliferate in an altered microenvironment.

Many oncogenes (RAS, PI3K, KIT) and tumor suppressors (P53, RB, PTEN) are frequently mutated in different human malignancies. High throughput genome sequencing of tumor DNA has elucidated low-frequency mutations in several genes that are highly likely to drive oncogenesis. For example, mutations in the human kinome approach approximately 20%, some of which are drivers and many of which are passengers, ^[6] but the frequency of mutations in other gene families contributing to cancer remains to be established. Somatic mutations vary in distinct cancer types (eg, brain, pancreas, breast, colon), as well as in a given tumor type, but they do appear to utilize common overlapping oncogenic pathways detected in the malignant phenotype. ^{[7, 8, 9]} The Cancer Genome Atlas Network analyzed the somatic mutation spectrum of four mRNA-expression subtypes of breast tumors, and found that basal-like tumors differed substantially from luminal A, luminal B and HER2- enriched tumors. ^[10] Interestingly, breast basal-like tumors shared a number of molecular characteristics common to ovarian cancer such as the types and frequencies of genomic mutations, suggesting a related etiology and potentially similar responsiveness to some of the same therapies.

Specific targeting of nodal points of these key oncogenic pathways is now a reality with robust techniques of target validation, such as siRNA/shRNA knockdown, dominant-negative mutants, synthetic lethality screening, and genetically engineered mouse models (GEMM) in combination with target-specific drugs (small molecule inhibitors [SMI], monoclonal antibodies [Mab], antisense oligonucleotides [ASO]) to confirm and advance targeted therapeutics.

Despite the complexity of genetic alterations in cancer, some cures have been achieved in certain malignancies with combination chemotherapy, such as childhood leukemia, Hodgkin lymphoma, testicular cancer, and diffuse large B-cell non-Hodgkin lymphoma. However, for many other types of advanced cancer (eg, lung, pancreas, colon, breast, prostate), no current standard of curative therapy exists. Hence, some key questions are as follows:

• How do we effectively treat these malignancies that arise from such varied nodal perturbations in overlapping oncogenic pathways?

• What are the genetic drug resistance mechanisms, and how do we target these?

• What is the role of cancer stem cells/progenitors in self-renewal and drug resistance?

The key to successful therapies is identification of critical, dysfunctional nodes in oncogenic networks whose effective inhibition will result in abrogation and/or reversal of the malignant state by apoptosis and/or differentiation. The specific targeted therapy or combination of therapies should be less toxic to normal tissue, coupled to a large therapeutic window that targets the ”context of vulnerability” of the tumor. ^[11]

Imatinib mesylate (IM) is an excellent example of an SMI that targets a major context of vulnerability of a defined genetic abnormality, such as the t(9;22) translocation (BCR-ABL) in chronic myeloid leukemia (CML) and c -Kit gain-of-function mutations in gastrointestinal stromal tumors (GIST), with a good safety profile. However, discovering and effectively targeting a complex of low-frequency mutant cancer phenotypes and aberrantly overexpressed oncogenes that drive most epithelial and nonepithelial cancer is far more complicated. To address this complexity, the targeting of the hallmarks of cancer in each tumor type is a rational way forward for cure and, in the process, provides answers to undiscovered biology.

What are the hallmarks of cancer? Can they be targeted with a rational combination of therapeutic agents with low toxicity and an adequate therapeutic window to reverse the growth of tumors? The original 6 hallmarks of cancer were proposed to include the following ^[3] :

• Cellular processes of self-sufficiency in growth signals

• Insensitivity to growth-inhibitory signals

• Evading programmed cell death

• Limitless replication potential

• Sustained angiogenesis

• Invasion/metastasis

For decades, cancer was studied by isolating malignant cells and ignoring the subverted conscripted stromal elements. However, a global understanding of cancer includes all elements of the tumor depicted by the 6 hallmarks, which generates a self-sustaining pseudoorgan. Within this pseudoorgan, there is likely to be molecular and phenotypic heterogeneity, cancer stem cells, and a variable sensitivity to therapy, implicating preexisting mechanisms of drug resistance.

Two additional hallmarks have been proposed, as follows:

• Immune evasion ^[4]

• A stress response phenotype, which is composed of metabolic (lactic acidosis), proteotoxic (heat-shock response), mitotic (chromosome instability), oxidative (reactive oxygen species), and DNA damage stresses (double-strand breaks) ^[5]

As tumors grow, they outstrip the nutrient supply, leading to a state of hypoxia, which promotes a switch to a less efficient glycolytic metabolism. This, in turn, leads to an acidic microenvironment that helps evasion of immune attack, as well as an increased level of reactive oxygen species (ROS) leading to increased DNA damage. Tumors grow despite DNA damage due to cell cycle checkpoint defects, which in turn leads to increased mitotic stress that promotes a state of aneuploidy. Chromosomal instability, in turn, elicits a state of proteotoxic stress that is compensated for by the heat-shock response pathway.

In the author's opinion, there are 2 additional hallmarks of cancer, as follows:

• Stromal subversion

• An inflammatory serum cytokine response that helps tumors grow and proliferate once the initiation is complete

A more prominent role for stromal fibroblasts in carcinogenesis than previously thought is appreciated. Evidence indicates that mutations arising in stromal fibroblasts and the consequent manifestation of paracrine factors promote growth and proliferation of cancer cells. ^[12] Hence, the paradigm that stromal fibroblasts are conscripted bystanders is changing, and their prominent role in cancer initiation and progression is now accepted. Similarly, an inflammatory serum cytokine profile manifested as the body’s response to the tumor and elaborated from the tumor stroma are also likely to impact on the cellular hallmarks of cancer in an adverse manner. ^[13]

The formation of a tumor is driven by a few select changes at the genetic and epigenetic level. The term "oncogene addiction" ^[14] describes tumor maintenance dependent on a continued activity of a defined mutated constitutively activated oncogene. Specific oncogene-dependent tumor development in mouse models (MYC, RAS, BCR-ABL) ^{[15, 16, 17]} and their reversal by removal or inhibition have been successfully shown. Targeting of these oncogenes with specific small molecule inhibitors in human malignancies (RAS and MYC) has been challenging. However, successful therapeutic targeting of oncogene addiction has been achieved, but only in relatively nonfrequent tumor types that have a defined molecular pathology. A few examples are highlighted in the discussion below.

BCR-ABL

The t(9;22) balanced reciprocal chromosome translocation generates the BCR-ABL oncogene. The fusion product (p210) is an unregulated tyrosine kinase that transforms cells; is the hallmark of CML (about 5000 cases per year in the United States) and, rarely, acute lymphoblastic leukemia (ALL) (p190); and promotes proliferation independent of the cytokine milieu. Imatinib mesylate (IM) is a tyrosine kinase (TK), ATP-site competitive SMI that binds the inactive form of the enzyme and has revolutionized the management of chronic CML ^[18] by changing the natural history of the disease, with a significantly superior survival compared to prior therapies.

Despite targeting the context of vulnerability in CML (BCR-ABL), IM does not appear to cure the disease because of underlying genetic mechanisms of resistance, including identification of more than 40 mutant forms of BCR-ABL, overexpression of BCR-ABL by amplification, other mechanisms (eg, upregulation of drug efflux pumps, downregulation of drug influx pumps, alternative overexpression of nonreceptor TK Lyn), and enrichment of CML stem cell progenitors that are resistant to BCR-ABL -directed therapies. ^[19]

Several kinase-domain, active-site mutants are IM resistant (Y235F/H, E255K/V), while the T315I (gatekeeper) is completely resistant to BCR-ABL -directed therapies (ie, IM, nilotinib, dasatinib, bosutinib). However, aurora kinase SMIs (MK-0457, XL228, AT9713) has been shown to effectively inhibit the T315I mutant enzyme. ^[20]

IM fails in approximately 25% of patients, who are given salvage therapy with second-line drugs such as dasatanib (dual Src-family/Abl kinase SMIs) and the IM derivative nilotinib. ^[21] Patients with accelerated and blast-phase CML do not respond well to BCR-ABL -directed therapies and allotransplantation, which is the only mode of cure recommended for eligible patients. Several dual Src-family/Abl kinase SMIs (INNO-406, PHA-739358, AP24534, SGX393, DC-2036) have been evaluated with activity against the T315I mutant. ^[20] Novel BCR-ABL allosteric or switch-site TKSMIs (Deciphera) are also in clinical development. ^[20] Further, targeting of the CML IM-resistant stem cell progenitor population is under active investigation. ^[22]

c-KIT/PDGFR

Gain-of-function mutations in c-Kit (exon 9, 11, 13, 17) and a platelet-derived growth factor receptor (PDGFR) (exon 14 and 18) are the drivers in the pathogenesis of gastrointestinal stromal tumor (GIST) (about 5000 cases per year in the United States). ^[23] IM is an effective drug in the management of advanced disease ^[24] and in the adjuvant setting, ^[25] with significantly prolonged survival, and has changed the natural history of the disease. Despite this success, complete responses are rare; most responses are good partial responses and stable disease.

IM eventually fails in approximately 50% of patients, who are then given salvage therapy with sunitinib malate (SM), a c-Kit/PDGFR/VEGFR TKI, which has a good response to exon 9 and wild-type patients. However, patients with exon 11 mutations in whom IM fails do not respond well to SM because of secondary resistant mutations within the activation loop of the kinase domain. ^[26] As with IM-resistant CML patients, IM-resistant GIST patients develop novel genetic mechanisms of drug resistance. In the "RTK switch," c-Kit is downregulated in association with upregulation of c-Met, and AXL oncogenic RTKs is a novel mechanism of resistance to c-Kit-directed therapies. ^[27] IM is also approved by the FDA for several other human malignancies that are driven by c-Kit/PDGFR (eg, hypereosinophilic syndrome, systemic mastocytosis, dermatofibrosarcoma protuberance, CMML). ^[28]

HER Family

The human epidermal growth factor receptor family (HER1, 2, 3, and 4) consist of RTKs that are overexpressed (HER1, 2, and 3; lung, head and neck, breast, and prostate cancers) or mutated (HER1 and 2; lung cancer and glioblastoma multiforme [GBM]), leading to constitutive activation in human epithelial malignancies. Of the various dimer pairs, the HER1/HER2 heterodimer is the most potent and activates RAS-MAPK and STAT pathways, as well as HER3 (inactive kinase domain), which activates the PI3K/AKT survival pathway. ^[29] HER family members were selected as the first molecules for targeted therapy, primarily because of the detection of the HER2 amplicon on chromosome 17q12-21 in approximately 20% of aggressive breast cancer patients with a poor prognosis. ^[30]

Trastuzumab (4D5) binds extracellular domain (ECD) IV of HER2 close to the membrane and was developed because of its high affinity, specificity, and efficacy in breast cancer cell lines and mouse models. It is approved by the FDA for the treatment of HER2+ metastatic breast cancer and in the adjuvant setting in combination with chemotherapy, because it has modest activity when used alone.

Neoadjuvant trastuzumab trials have disputed the fact that the Mab downregulates ERB2, since no decrease in ERB2 was seen in primary breast tumors. However, trastuzumab does downregulate HER2 signaling by masking a protease cleavage site on ECD IV and blocks shedding of the receptor. ECD shedding constitutively activates the TK domain, and blocking this correlates with positive responses to therapy. ^[31] Moreover, clinical data from neoadjuvant and adjuvant trials with trastuzumab have implicated additional activity through antibody-dependent cell cytotoxicity (ADCC). Trastuzumab is also indicated for treatment of HER2-overexpressing metastatic gastric or gastroesophageal junction adenocarcinoma.

In December 2017, FDA approved a biosimilar to Herceptin, trastuzumab-dkst (Ogivri) indicated for use in the treatment of human epidermal growth factor receptor (HER)–positive (HER+) breast cancer and also HER2+ metastatic stomach cancer (gastric or gastroesophageal junction adenocarcinoma). ^[32] In December 2018, the FDA also approved trastuzumab-pkrb (Herzuma) as a biosimilar to trastuzumab for patients with HER2-overexpressing breast cancer. ^[33]

A second HER2-targeting Mab, pertuzumab (2C4), binds to ECD II and interferes with heregulin (HRG)-induced HER2/HER3 heterodimerization and thus acts as an ERB2 dimerization inhibitor that may be useful in inhibiting HER3 overexpressing tumors (ovary, breast, prostate, colon, lung) by abrogating the activation of the PI3K/AKT pathway. ^[34] Since the approval of trastuzumab, the FDA has approved several HER1-directed Mabs (cetuximab, panitumumab), which disrupt autocrine loops in human malignancies (colon, head and neck) and have higher responses when combined with chemotherapy. ^[35] Several SMIs to the HER TK domain have also received FDA approval: erlotinib and gefitinib are HER1-specific (lungandpancreascancer), and lapatinib is a dual HER1/HER2 inhibitor (breast cancer resistant to trastuzumab). ^[36] The pan-HER inhibitor carnetinib and HER2-selective inhibitor CP-724,714 are currently undergoing clinical evaluation. ^[37] Similarly, the TKIs are effective alone only in patients with HER1 mutations (peripheral adenocarcinoma of the lung, GBM [de2-7 HER1vIII]) or addicted to HER signaling (breast and prostate cancer). A majority of TKIs appear to be additive when combined with chemotherapy. ^[38]

In May 2013, the US Food and Drug Administration (FDA) approved of the cobas EGFR Mutation Test, a companion diagnostic for erlotinib. This is the first FDA-approved companion diagnostic that can detect epidermal growth factor receptor (EGFR) gene mutations. The mutation test allows physicians to identify patients with non–small cell lung cancer (NSCLC) who are candidates for receiving erlotinib as first-line therapy.

The safety and effectiveness of the cobas EGFR Mutation Test was established with clinical data from the EURTAC study and showed progression-free survival in patients with NSCLC who had specific types of EGFR mutations (exon 19 deletions or exon 21 [L858R] substitution mutations) for 10.4 months when they received erlotinib treatment, compared with 5.4 months for those who received standard therapy. ^[39]

However, HER1 and/or HER2 targeted therapy (Mabs or Nibs) leads to drug resistance, including secondary mutations in the HER1 TK domain (T790M, D761Y), ^[40] KRAS mutations, ^{[35, 41]} and amplification and overexpression of c-Met ^[42] and HER3. ^[43]

Resistance mechanisms to HER-directed therapy maintain the activation of the same downstream signaling pathways (PI3K/AKT, RAS-MAPK), indicating escape at the RTK level by switching to another RTK and/or mutations within downstream signaling pathways. Biologic complexity challenge the need for novel therapies to overcome genetic-resistance mechanisms, including the design of irreversible inhibitors that target cysteines within the TK active site (Avila Therapeutics), pan-HER TKI (carnetinib), Mab targeting HER3 (Merrimack), multitargeted TKI (HER/VEGFR, MET/VEGFR), and multitargeting of downstream signaling pathways (PI3K/AKT inhibitor + RAS-RAF-MAPK pathway inhibitor).

c-MET

The hepatocyte growth factor/scatter factor (HGF/SF) activates c-Met, an RTK important for many normal cellular functions. Aberrant signaling through the HGF/MET-axis is implicated in a variety of human malignancies by enhancing anoikis resistance, invasion, antiapoptosis, and neoangiogenesis. Chemical carcinogen-induced chromosomal rearrangement (TPR-MET) in an osteosarcoma cell line identified MET as a transforming oncogene and subsequently also found it in some gastric cancers. ^[44] Single-allele germline missense mutations in c-Met are oncogenic in hereditary papillary renal cell carcinoma (RCC) type I and sporadic papillary RCC (13%). ^[45] The majority of the latter possess trisomy 7, with nonrandom duplication of mutant c-Met that drives the transformed phenotype.

Somatic c-Met mutations exist in a variety of other malignancies (gastric, liver, SCLC, NSCLC, H/N SCC), ^[46] but these are rare. Constitutive activation of MET due to gene amplification and/or overexpression is detected in many human malignancies. Aberrant autocrine and paracrine circuits also contribute to the malignant phenotype in several cancers (eg, GBM, breast cancer, sarcoma). Moreover, aberrant activation of MET is observed in metastatic lesions but not in the primary tumor (eg, colon cancer), and this may potentiate acquisition of secondary somatic mutations (Y1230C; Y1235D) during metastasis (eg, head/neck cancer). ^[47] Hence, oncogenic signaling in certain cancers due to "addiction" (mutations) or survival advantage (overexpression) provides a compelling rationale for targeting the HGF/MET axis. Strategies to effectively exploit this target have been explored.

Biologic antagonists of HGF (NK2, NK4, uncleavable HGF) and MET (decoy MET, Serma domain) have shown preclinical proof-of-concept inhibition of this axis, leading to decreased proliferation and tumor growth inhibition. However, these protein domain antagonists have not reached the clinic. ^[48]

Several monoclonal antibodies (Mabs) directed to HGF and MET have been generated. However, a Mab from Amgen to HGF (AMG102) showed promising activity in a GBM xenograft model. ^[49] AMG102 is a fully human Mab (IgG2) that has undergone phase I and II clinical trials (GBM, RCC). Therapeutic Mabs with unique mechanistic properties include L2G7 (anti-HGF), which crosses the blood-brain barrier, causing tumor regression in an intracranial glioma xenograft model ^[50] ; OA-5D5, a "one-armed" monovalent anti-MET Fab that has a profound intracerebral tumor-suppressive effect when given locally and is now entering clinical development ^[51] ; and DN-30 (anti-MET), which induces proteolytic cleavage of MET extracellular region, decreases the number of cell surface receptors, and generates a decoy effect by inhibiting HGF binding to MET. ^[52]

Small-molecule, intracellular kinase domain inhibitors have also been developed: 6 target the ATP-binding site (PF2341066, XL880, MK2461, MP470, SGX523, JNJ38877605), and 1 is an allosteric inhibitor (ARQ197). The first ATP-site, c-Met TKI to enter phase I clinical trials was the orally bioavailable XL880 (Exelixis). The agent also inhibits VEGFR2, PDGFR, RON, KIT, and Tie2 TK domains. ^[53] Phase II studies have been conducted in c-Met-dependent tumors (gastric cancer, head and neck SCC, papillary RCC). In this cohort of 18 selected patients, tumor shrinkage occurred in 14 patients within 2 months. Subsequently, 12 patients achieved stable disease and 1 a prolonged partial response. ^[47]

ARQ197 (ArQule) is a non-ATP-site competitive, selective SMI of the c-Met intracellular region. In a phase I study of 37 refractory solid tumor patients, 3 patients achieved a partial recovery and 20 achieved stable disease. ^[54] Patients treated for 6 weeks or less developed new metastases within 6 months, while those treated for 12 weeks or longer did not, implicating potential anti-invasive activity with prolonged therapy. A phase II study of ARQ197 versus gemcitabine in unresectable/metastatic pancreatic ductal adenocarcinoma is ongoing. ^[48]

PF2341066 (Pfizer) is an oral c-Met/ALK (anaplastic lymphoma kinase) ATP-site SMI that has undergone phase I studies, with the dosage ranging from 50 mg qd to 300 mg bid. The MTD was 250 mg bid, based on increased ALT and fatigue. The EML4-ALK fusion product is seen in approximately 4-5% of cases of adenocarcinoma of the lung and appears to be sensitive to PF2341066 (10 of 19 patients had a partial recover). Also, partial recovery was seen in a patient with inflammatory myofibroblastic tumor (ALK+ by IHC). ^[55]

Phase I studies are ongoing with MK2461 (Merck), MP470 (SuperGen), SGX523 (SGX), and JNJ38877605 (Johnson & Johnson).

PI3K/AKT

The phosphoinositide 3' kinase (PI3K) family consists of 3 classes of lipid kinases that have a regulatory subunit (p85) and a catalytic subunit (p110) that phosphorylate the 3'OH group of phosphoinositols. Class IA PI3Ks are implicated in cancer because of the presence of activating somatic mutations in the catalytic subunit p110a (PIK3CA) that are identified in approximately 30% of epithelial cancers (breast, colon, prostate, endometrial). ^[56]

Kinase domain (KD) and helical domain (HD) mutants predominate (approximately 80%); however, each type of mutant promotes constitutive PI3K signaling via distinct mechanisms. KD mutants (eg, H1047R activation loop) have increased lipid kinase activity and transform cells. ^[57] HD mutants (eg, E545K, E542K) disrupt the inhibitory intermolecular interaction between p85 and p110, thus activating the KD. However, KD mutants remain oncogenic when their Ras-binding domain is also mutated. ^[58] Moreover, rare mutations in p85a (PIK3R1) located within the inter-SH2 domain that binds the C2 domain of p110a abrogates the inhibitory effects on the catalytic domain, leading to constitutive PI3K activation. For example, in GBM, p85a and p110a mutations are mutually exclusive, implicating a redundancy in PI3K activation. ^[59]

At least 10 PI3K p110 KD-targeted agents (BEZ235, BGT226, BKM120, XL765, XL147, GDC0941, SF1126, PX-866, CAL-101, GSK1059615) have entered early-phase clinical trials, some of which possess dual PI3K-mTOR inhibitor properties (BEZ235, BGT226, XL765, SF1126), because the p110 KD shares similar structural features with the S/T kinase domain of mTOR. ^{[60, 61]}

BEZ235, BGT226, and BKM120 (Novartis) are being evaluated in refractory solid tumors (phase I) and aggressive breast cancer (phase II), since PI3K mutations are common (27%) in the latter. XL147 (Exelexis), a class 1 PI3K-selective ATP-site-competitive SMI, has been evaluated in patients with refractory solid tumors (NSCLC, 13; CRC, 7; BC, 7; sarcoma, 5; prostate cancer, 4) in 2 different dosing schedules (3 weeks on,1 week off; or continuous daily dosing [CDD]). Starting dose was 30 mg and escalated to 900 mg, with 600 mg declared as MTD.

A reversible hypersensitivity rash was the DLT for the first dose schedule. For CDD, dose escalation was from 100 mg to 400 mg. The plasma t_½ was 5 days. There was a slight increase in plasma insulin (fasting) levels, but there was only a modest increase in plasma glucose levels. Hair follicle PD markers p4EBP1, pAkt (T308), pAkt (S473), pPRAS40 (T246), and pS6 (S240/S544) decreased in a dose-dependent manner. In a patient with PDA skin, pAkt and p4EBP1 also decreased with therapy.

Six paired tumor biopsies (600 and 900 mg) indicated an 82% decrease and a 74% decrease in pAkt and p4EBP1, respectively. However, there was a modest decrease in Ki-67 and no change in apoptotic readout (Tunel assay). Interestingly, pErk also decreased in paired biopsies, indicating cross-talk with the Raf-Ras-MAPK pathway. Antitumor activity (RECIST) was mainly SD, with 1 patient experiencing a PR (NSCLC >16 wk with no PI3K mutations; and WT EGFR, 33% reduction). ^[62]

XL765 (Exelexis), a dual pan-PI3K/mTOR SMI, was evaluated (CRC 14, BC 4, others), with a starting dose of 15-120 mg bid or 70-100 mg qd, with MTD 50 mg for bid and 90 mg qd. Grade 3 adverse events were nausea/vomiting, fatigue, rash, and asthenia. The agent is highly protein bound (90%), with a plasma t_½ of 6.3 hours. The changes in plasma insulin and glucose levels were modest, with reversible inhibition of pAkt (S473) and pRas40 (T246) at 30 mg bid or 70 mg qd.

As with XL147, XL765 had a modest effect on proliferation rate (Ki-67) and apoptosis, with a decrease in pErk (60-74%), indicating cross-talk inhibition with the MAPK pathway. Further, antitumor activity was SD. An RCC patient has been studied for 182 weeks, and a patient with appendiceal carcinoma and one with rectal carcinoma were studied for 33 and 24 weeks, respectively. Combination studies are planned with temozolomide plus erlotinib for GBM. ^[63]

GDC0941 (Genentech) is a class I PI3K SMI that is active in many tumor types irrespective of Ras mutational status. Two dosing schedules were evaluated (3 wk on, 1 wk off, qd or bid CDD) in patients with refractory solid tumors (sarcoma 11, ovarian 6, CRC 4, endocrine 3, BC 2, prostate cancer 2). The MTD was not identified, but DLT was grade 3 headache (80 mg qd) and grade 3 was pleural effusion (30 mg bid). The plasma t_½ is in the range of 12-22 hours for both dosing schedules, with a dose-dependent decrease in pAkt. One patient with ovarian cancer with PI3KCA amplification has been studied for approximately 1 year with CA125 and a lymph node response (cycle 11). ^[64]

SF1126 (Semafore), a pan-PI3K and mTOR SMI that is also a vascular targeting agent because of the presence of the integrin-binding RGD motif, is being evaluated in refractory solid tumors (ie, CRC, NSCLC, GIST, CUP, RCC). It is given as an intravenous infusion twice per week for 4 weeks and is repeated once before response evaluation (RECIST, PET). It is well tolerated and shows SD as the main tumor response. ^[65]

PX-866 (Oncothyreon) is a pan-PI3k inhibitor that is a wortmannin analogue being evaluated in a phase I trial. ^[66]

Finally, CAL-101 (Calistoga) is a p110d-isoform selective SMI that has shown clinical activity (6/12 PR) in refractory CLL and NHL. However, the durability of these responses and target inhibition have not been determined. ^[67]

AKT, an S/T kinase, is directly activated in response to PI3K and consists of 3 isoforms, with Akt1 linked to survival and growth and Akt2 linked to invasion as major drivers in cancer. Rare mutations have been identified in all 3 Akt isoforms. ^[68] Somatic mutations in AKT1 are observed in 8% BC, 6% CRC, 2% ovarian cancer, and 5.6% squamous cell carcinoma of the lung. ^{[69, 70]} Mutation in the pleckstrin homology (PH) domain (E17K) allows continuous plasma membrane localization of Akt1, with constitutive pS473 in the absence of serum. However, T308 remains unphosphorylated and likely requires PI3K for full activation.

Currently, there is interest in targeting the PH domain, the kinase domain, and interdomain allosteric inhibitors. Perifosine (Keryx) is a PH domain SMI that has undergone extensive testing in early clinical studies, with no single-agent activity. ^[71]

GSK690693 (GSK), a KD SMI, entered a phase I solid-tumor study in 2008 but was discontinued because of a serious adverse event.

MK2206 (Merck) is an interdomain allosteric inhibitor with an IC50 of inhibition of 5, 12, and 65 nM for Akt1, 2, and 3 respectively. In a phase I dose-escalation study (30, 60, 90, 200, 300 mg), 18 patients were enrolled and treated qod, with tumor biopsies obtained before and after MTD (60 mg qod; N = 16). A small erythematous rash was the DLT, and no hematologic toxicity was observed. Other adverse events were pruritus, nausea, vomiting, diarrhea, and fatigue. The effect on glucose metabolism was mild. The t_½ on this dose schedule was 56 hours. Day 1 and 35 pAkt levels were measured in plucked hair and PBMCs where target inhibition was observed. PI3KCA mutations (exon 9 and 20) in circulating cells were conducted in patients with BC, CRC, and melanoma. The main antitumor response was SD with a patient with neuroendocrine tumor studied for 7 months (30 mg qod). ^[72]

c-RET

Gain-of-function mutations in the RET proto-oncogene leads to hereditary (98%) and sporadic (30%) medullary thyroid carcinoma (MTC), a rare calcitonin-producing tumor arising from parafollicular C cells. ^[73] Hereditary MTC is composed of 3 subtypes: MEN2A (MTC, pheochromocytoma, primary hyperparathyroidism), MEN2B (MTC, pheochromocytoma, ganglioneuromas, Marfanoid body habitus), and familial MTC (MTC). ^[74]

The RET (REarranged during transfection) proto-oncogene is located on chromosome 10q11.2 and consists of 21 exons and is activated after genetic rearrangement. ^[73] RET has an extracellular domain (4 cadherin-like repeats and a cysteine-rich region), a single membrane spanning region followed by an intracellular TK domain.

Ligand binding (GDNF, persepherin, artemin, neuturin) to the ECD leads to RTK dimerization mediated by activation of the cysteine-rich region and activation of downstream signaling pathways (MAPK, PI3K, JAK/STAT, PKC). ^[75]

Patients with MEN2A have missense mutations in exons 10 and 11, which affect one of the 6 cysteines in the ECD, leading to disulfide-mediated receptor dimerization and activation independent of ligand binding. In MEN2B, almost all patients have exon 16 mutations within the TK domain, with RTK activation in the monomeric form, while FMTC patients have exon 10, 11, 13, and 14 mutations. ^[74] Constitutively activated RET due to chromosomal translocations is detected in papillary thyroid carcinoma (PTC), which is more common than MTC. ^[76]

MTC, like GIST, represents a promising model for targeted RTK therapy, as chemotherapy and radiation therapy are ineffective. Several RET TK-directed therapies (ie, CEP-701, CEP-751, ZD6474 [vandetanib], XL184, AMG706 [motesanib]) have been evaluated in early-phase clinical trials. Vandetanib (ZD6474) targets RET, VEGFR2, and EGFR and is orally bioavailable with an IC50 100 nM for the RET TK. Phase I studies have shown that vandetanib has a half-life of more than 120 hours and is well tolerated at doses of up to 300 mg qd. However, the main adverse effects observed are GI (nausea, diarrhea), rash, hypertension, and asymptomatic QTc prolongation. ^[77]

Several phase II studies in MTC (locally advanced and metastatic) are under way. Of the first 20 patients studied, approximately 30% showed an objective RECIST response, while stable disease was observed in approximately 50% of patients. A greater than 50% decrease in plasma calcitonin level was observed in more than 80% of patients, which appears to correlate with an antitumor response. A randomized, placebo-controlled phase II registration trial is currently ongoing. ^[76]

Motesanib (AMG706) is a multitargeted TK inhibitor (VEGFR, c-Kit, PDGFR, RET) and was evaluated in a phase II study in locally advanced and metastatic MTC, with partial responses. ^{[78, 79]}

XL184 is an oral multitargeted TKI (RET, VEGFR2, c-Met) for which a phase I study has been completed in advanced solid cancers and MTC. ^[80] Patients (N = 55, 13 with MTC) received XL184 daily for 5 days of each 14D cycle or continuous daily dosing (CDD) (dose range, 0.08-11.52 mg/kg). Grade 3 DLTs were palmar-plantar erythema, transaminitis, and mucositis. The T_½ of XL184 was 59-136 hour. The best response in solid-tumor patients was SD of more than 6 months in 12 patients, while 3 MTC patients had a confirmed PR and 1 patient with neuroendocrine tumor had an unconfirmed PR. In all 13 MTC patients, the serum calcitonin and carcinoembryonic (CEA) levels were decreased. The MTD cohort was expanded to include more than 20 MTC patients. ^[80] A phase III registration trial of XL184 in MTC has been initiated. ^{[81, 82]}

The advantage of TKIs currently in trials that target RET is that they also target other RTKs that are implicated in proliferation and angiogenesis, and alternative RTK activation as a mechanism of drug resistance can be targeted simultaneously.

B-RAF

BRAF is a cytoplasmic S/T kinase that lies immediately downstream of RAS and is a key player of the RAS/RAF/MEK/ERK pathway, with a major activating mutation (V600E) detected in melanoma (50-70%), PTC (36-69%), CRC (5-12%), and NSCLC (1-4%). ^{[74, 83]} The V600E BRAF mutation constitutively activates ERK signaling that induces proliferation and promotes transformation. ^[84]

Sorafenib is a multitargeted KI that has activity against B-RAF, VEGFR-2/3, c-Kit, PDGFR, Flt-3, and FGFR1. In fact, sorafenib was initially tested as a B-RAF inhibitor in melanoma with no clinical activity. ^{[85, 86]} It is FDA approved for advanced RCC and HCC. ^{[61, 86]} However, phase II trials of sorafenib plus chemotherapy have shown modest activity in melanoma, with no correlation to BRAF mutational status. ^[86] Given the frequent occurrence of BRAF mutations in human cancer and the continued requirement for B-RAF activity in tumors in which it is mutated, efforts are under way to develop targeted inhibitors of B-RAF and its downstream effectors. These agents offer the possibility of greater therapeutic efficacy than the currently available systemic therapies for tumors driven by activating mutations in the MAPK pathway.

PLX4032 and PLX4720 (Plexxikon) are selective RAF inhibitors with a more than 10-fold higher selectivity for BRAF V600E than for BRAF WT; good oral bioavailability; and minimal toxicity in preclinical models. In 16 BRAF mutation-positive metastatic melanoma patients treated with PLX4032 doses at or above 240 mg bid, data showed PRs in 9 patients with more than 30% tumor regression by RECIST, with 7 confirmed responses. Disease control lasting up to 14 months with CDD has been achieved and is well tolerated, with MTD likely to be at 960 mg bid. The interim median PFS is 6 months, with many responding patients still receiving treatment. By contrast, no response was observed in patients without the BRAF mutation, and PFS was less than 2 months, which is consistent with historical data.

DLTs were rash, fatigue, and joint pains observed at 1,120 mg bid. Serious adverse events were observed in some patients after long-term treatment, including possibly drug-related cutaneous squamous cell carcinoma. ^[87] RAF265 (Novartis) is a potent RAF inhibitor, and preclinical studies have shown that it inhibits all 3 RAF isoforms (A, B, C), as well as mutant B-RAF, with high potency. Also, RAF265 possesses antiangiogenic activity through inhibition of VEGFR-2 and antitumor activity in xenograft models. RAF265 is now being investigated in Phase I clinical trials in malignant melanoma. ^[88]

XL281 (Exelexis) is a potent and highly selective inhibitor of RAF kinases (B, C not A isoform), is orally bioavailable, and showed activity in tumor xenograft models. In a phase 1 dose escalation of XL281 in patients with advanced solid malignancies, daily oral dosing (10 mg qd to 225 mg qd) on a 28-day cycle was evaluated. Of 48 evaluable patients (with melanoma, CRC, NSCLC, PTC), one patient with ocular melanoma carrying the c-Kit M541L mutation had a confirmed PR (40 weeks on study), and 19 patients continued to be studied for 12 weeks or more, for a clinical benefit rate of 36%. This included patients with and without mutations in components of the RAS/RAF signaling pathway (including the mutations BRAF V600E, NRAS Q61R, and KRAS G12S).

In the XL281 study, grade 3 adverse events were reported in 8 patients (3 fatigue, 3 vomiting, 2 diarrhea, and 1 each of nausea and arthralgia). The MTD was 150 mg qd [N=37], and DLTs were reported at 225 mg qd (grade 4 nausea, grade 3 fatigue, grade 3 diarrhea, grade 3 vomiting). Expansion of this cohort is currently ongoing at 150 mg bid.

PK studies have showed that XL281 exposure increases with dosage. Time to maximal plasma concentration is 2 hours; half-life is approximately 8 hours (5-22 hours); and steady-state levels are reached by day 8. There is an approximately 1.5-fold accumulation of XL281 after repeated dosing. PD studies (tumor, skin, hair follicles, pMEK, pERK) showed robust pathway inhibition across diverse tumor types, independent of RAS/RAF genotype status, with decreases in biomarkers for MAPK pathway activity ranging from 40-85%. There was increasing pathway inhibition with time in patients for whom serial hair samples were available. ^[89]

Other RAF S/T kinase inhibitors currently being evaluated in preclinical studies are SB-590885 (GSK) and AZ628 (AstraZeneca). ^{[86, 90]}

ALK

Anaplastic lymphoma kinase (ALK) is an RTK involved in neuronal cell differentiation and regeneration, as well as synapse formation and muscle cell migration. It belongs to the IGF-like family and is activated by its ligand plieotrophin. ^[91] ALK was identified as a tumor-associated antigen due to a chromosomal translocation t(2;5) involving the nucleophosmin (NPM1) gene on chromosome 5 in anaplastic large cell lymphoma (ALCL), a subset of T-cell non-Hodgkin lymphoma (NHL).

Since then, many other translocations of ALK have been identified, including 8 ALCL variants, 7 variants of inflammatory myofibroblastic tumor (IMT), and 1 NSCLC (6%; inv(2) with echinoderm microtubule-associated protein-like 4: EML4; EML4-ALK fusion). ^[92] In all cancers, translocations involving ALK fusion proteins lead to dimerization and activation of constitutive TK activity with subsequent transformation.

PF-02341066 (Pfizer), an oral TK SMI that specifically inhibits ALK and c-Met RTKs, was evaluated in a phase I targeted study (N = 37) of NSCLC (N = 27 with EML4-ALK fusion) and IMTs. Dose escalation ranged from 50 mg qd to 300 mg bid. MTD was chosen as 250 mg bid, because of DLTs of transaminitis and fatigue. Partial responses were seen in IMT patients (ALK + IHC) and in 10 of 19 patients with NSCLC with the EML4-ALK fusion. ^[55]

Other agents in early-phase trials are NVP-TAE684 (Novartis) and CEP-14083 (Cephalon). ^[92]

TRK

Neurotrophic receptor tyrosine kinase genes NTRK1, NTRK2, and NTRK3 encode the tropomyosin receptor kinase (TRK) proteins TRKA, TRKB, and TRKC, respectively. Chromosomal fusion of TRK have been identified among a wide variety of solid tumors that occur in adults and children. TRK fusions have been shown to activate chimeric proteins, which act as an oncogenic drivers by promoting cell proliferation and survival in tumor cell lines. ^[93]

Larotrectinib is a TRK inhibitor that was approved by the FDA in 2018. It is indicated for adults and children with solid tumors that have an NTRK gene fusion without a known acquired resistance mutation, are metastatic or where surgical resection is likely to result in severe morbidity, and have no alternative treatments or have progressed following treatment. Approval was based on results of 55 patients aged 4 months to 76 years. Seventeen unique TRK fusion-positive tumor types were represents across the study. The overall response rate was 75% according to independent review and 80% according to investigator assessment. At 1 year, 71% of the responses were ongoing and 55% of the patients remained progression-free. ^[93]

Another NTRK inhibitor, entrectinib, was approved by the FDA in August 2019. It is indicated for adults and children aged 12 years or older with solid tumors that have a neurotrophic tyrosine receptor kinase (NTRK) gene fusion without a known acquired resistance mutation. It is also indicated for adults with proto-oncogene tyrosine-protein kinase ROS1 (ROS1) NSCLC. Efficacy was evaluated in a pooled subgroup of 54 patients with unresectable or metastatic solid tumors with a NTRK gene fusion, who were enrolled in one of three multicenter, single-arm, open-label clinical trials (ALKA, STARTRK-1, STARTRK-2). Ninety-six percent of patients had metastatic disease, including 22% with CNS metastases, and 4% had locally advanced, unresectable disease. The ORR was 57%, and 45% had a DOR of at least 12 months. Entrectinib penetrates the CNS. Among the 54 adult patients, 4 had measurable CNS metastases at baseline as assessed by BICR and had not received radiation therapy to the brain within 2 months of study entry. Responses in intracranial lesions were observed in 3 of these 4 patients. ^[94]

Tumor cells have been established to have a program of gene silencing by epigenetic modification of the DNA and/or histones, especially of tumor suppressors. Several enzymes that epigenetically modify the nucleosome have now been established has anticancer targets, of which DNA methyltransferase (DNMT) and histone deacetylase (HDAC) have drugs approved for hematologic malignancies. Many other targets that help activate tumor suppressors are currently under development (eg, P53-MDM2 inhibitors).

DNA methyl transferase inhibitors

DNA methylation refers to the covalent modification of DNA by methylation at the 5-position of cytosine bases in a post-DNA synthesis reaction (S-phase) catalyzed by DNA methyltransferases (DNMT). Aberrant DNA methylation is associated with gene silencing in tumors. ^[95] Hypomethylating agents work by desilencing epigenetically modified DNA, thus promoting the expression of silenced tumor suppressors. ^[96] Demethylating agents are a group of nucleoside-analogues that induce transient DNA hypomethylation in the promoter regions of genes rich in CpG islands, by a mechanism in which the DNA/nucleoside-analogue complex binds to and inhibits DNMTs. ^[97]

The FDA has approved 2 DNA hypomethylating agents, 5-azacytidine (azacitidine) ^{[98, 99]} and 5-aza-2'-deoxycytidine (decitabine) ^{[100, 101]} for the treatment of higher-risk myelodysplastic syndrome (MDS). Azacitidine is a nucleoside analog with a ribose structure that gets incorporated into RNA and DNA via ribonucleotide reductase (RNR).

The CALGB conducted a randomized phase III crossover study of 5-AZA at 75 mg/m² SC for 7 days every 28 days, versus supportive care in patients with MDS. The overall response rate was 60% versus 5% for supportive care, with a median OS of 20 months compared to 14 months, respectively. ^{[98, 99]}

A second RCT study compared 5-AZA to conventional care regimens (CCR) with the same dosing for 5-AZA and showed a survival advantage (24 mo versus 15 mo), with a hazard ratio of 0.58 (95% CI: 0.43-0.77) ^[102] , favoring the 5-AZA arm. Decitabine is the deoxyribose analog of 5-AZA that gets incorporated selectively into DNA. It is also approved for higher-risk MDS and can be given as an IV infusion every 8 hours for 3 days every 6 weeks ^[100] or IV daily for 5 days every 28 days. ^[101]

Decitabine shows similar response rates and survival data to 5-AZA. The 5-day outpatient regimen is preferred for efficacy. ^[103]

HDAC inhibitors

The reversible acetylation of histones catalyzed by histone acetyltransferases (HATs) and histone deacetylases (HDACs) within the nucleosome structure modulates gene expression. In tumors, HDACs drive the equilibrium of this reaction in favor of deacetylation and tightening of histones, leading to epigenetic silencing. ^[104] DNA methylation and histone deacetylation work in concert in gene silencing because of direct binding interactions between DNMTs and HDACs.

The discovery and development of HDAC inhibitors (HDACIs; targeting class I and II HDAC and not class III) show effective binding of several distinct chemotypes to the zinc-ion bound catalytic pocket. ^[105] HDACIs are able to induce cell cycle arrest, promote differentiation, and hyperacetylate HSP90 and some of its client proteins (Bcr-Abl, HER2, Raf, Akt). The last effect appears to achieve a result similar to disruption of HSP90 function by HSP90 inhibitors. ^[61]

Vorinostat (suberoylanilide hydroxamine acid) is the first HADCI to be approved by the FDA for the treatment of advanced cutaneous T-cell lymphoma (CTCL) as a third-line therapy. In a phase II study of 74 patients (stage 1B and higher) with at least 2 previous treatments, patients who received 400 mg qd CDD had a response rate of 30%, with a mean duration of response of 168 days. The most common adverse events were nausea, diarrhea, fatigue, and anorexia. ^[106] The agent is currently in clinical trials for mesothelioma (single agent) and NSCLC (combined with chemotherapy).

A second HDACI (MS-275 or SNDX-275) is an oral benzamide that has a long half-life (33-80 hours) and was evaluated in patients with refractory solid malignancies and lymphomas on 3 dosage schedules: once every other week, biweekly for 3 weeks every 28 days, and once weekly for 3 weeks every 28 days. Twenty-seven patients received 149 or more courses of treatment. DLTs were hypophosphatemia and asthenia on the weekly and biweekly schedule, and there was no DLT on the every-other-week schedule. PKs revealed dose-dependent and dose-proportional increases. Two of 27 patients showed PRs, including 1 patient with metastatic melanoma who has been in a PR more than 5 years. Six patients showed prolonged SD. Histone H3 and H4 acetylation levels in peripheral blood mononuclear cells increased qualitatively with a high interpatient variability. The recommended dose for further disease-focused studies is 4 mg/m² given weekly for 3 weeks every 28 days, or 2-6 mg/m² once everyotherweek. ^[107] The agent is being developed for melanoma and NSCLC.

Belinostat (PXD-101, Curagen) is a novel HDC-I of class I and II histone deacetylases. This class of compounds has demonstrated anticancer activity in malignant mesothelioma. A phase II study of belinostat was conducted in patients with relapsed malignant pleural mesothelioma (N = 13) after progression with one prior chemotherapy regimen. Belinostat was administered at 1,000 mg/m² IV over 30 minutes on days 1-5, every-3-week cycle. A median of 2 cycles of therapy were administered. SD was seen in 2 patients, with no objective responses. Median survival was 5 months, with a median progression-free survival of 1 month. Major toxicities included nausea, emesis, fatigue, and constipation. One grade 5 toxicity of cardiac arrhythmia was possibly related to therapy. Evaluation of combination strategies or alternate dosing schedules may be necessary for further development of this novel agent in mesothelioma. ^[108]

A phase II study in relapsed/refractory DLBCL of PXD-101 is undergoing testing through the SWOG.

Based on the close biologic interactions between DNMTs and HDACs in modulating nucleosome structure and, in turn, gene expression, combination strategies have been evaluated. Combinations of 5-AZA or decitabine with valproic acid (HDACI) have been studied in higher-risk MDS and AML, with reported response rates (CR) being approximately 20% for patients with prior therapies and being approximately 50% in treatment-naive patients. Valproic acid has many side effects and has been replaced by other potent HDCAIs. MGCD0103, a class I HADCI, was combined with 5-AZA in patients with MDS/AML (relapsed/refractory or de novo). MGCD0103 at 90 mg PO 3 times per week plus 5-AZA was well tolerated, with a response rate of 50% in the de novo patient population, with a median time to response of 1 cycle (range, 1-3). ^[103] A second study of MS-275 plus 5-AZA is ongoing. ^[103]

P53-HDM2 inhibitors

Drugs targeting protein-protein interaction sites have huge therapeutic potential, but discovering specific SMIs is not trivial. Several recent success stories indicate discovery of SMIs with druglike potencies to "hotspots" of contact surfaces of proteins. These SMIs compete efficiently with the natural binding partners. ^[109] The human double minute 2 (HDM2) is a cancer target, as it inhibits p53 tumor suppressor activity. HDM2 binds to a 15-residue α-helical segment of p53 (Kd = 600 nM), the interface being mainly hydrophobic.

High-throughput screening and medicinal chemistry identified tetra-substituted imadazoles called Nutlins (Hoffman-La Roche). Nutlin-3 disrupts HDM2-p53 complexes with an IC₅₀ of 90 nM, has good preclinical activity, and is about to or has entered phase I trials. ^[110]

Johnson & Johnson pharmaceuticals screened 338,000 compounds for HDM2 binders and identified a series of benzodiazepinediones, which, when bound to HDM2, lead to dissociation from p53: Kd = 67 nM; IC₅₀ = 420nM- cancer-cell-based IC₅₀ = 10mM. Also, the SMI showed synergistic activity with doxorubicin against tumors in mice. ^[111]

JNJ-26854165, an oral HDM2 SMI, was evaluated in a phase I dose-escalation study (4 mg to 400 mg) of CDD every 21 days (N = 47; CRC 13, melanoma 7, BC 5, sarcoma 5, RCC 3, NET 3, others 11). Grade 3 adverse events were nausea, vomiting, diarrhea, anorexia, rash, asthenia, and QTc prolongation (at different doses) without hematologic toxicities. The MTD was not determined with this dosing schedule. There were responses based on RECIST, but patients with Hurthel cell cancer, ependymoma, and HER2+ BC had stable disease. PK studies showed a half-life of 50 hours, with Cmax of 2-6 hours. PD studies showed induction of p53 (skin biopsies), with increased nuclear p53 at 300 mg, while posttreatment tumor biopsies showed increased p53 staining by IHC. Serum MIC-1 level (transcription factor) increased with increasing doses, with skin and serum markers correlating with treatment. ^[112]

Clinical development of JNJ-26854165 will require an altered dosing schedule to avoid QTc prolongation and incorporation with rational combination therapies.

The balanced process of cell division and programmed cell death maintains cellular homeostasis. The pathways of cell survival and death have a pivotal role in cancer progression and response to therapy. Evidence is mounting that cells with defective apoptosis survive metabolic stress by utilizing autophagy. Therapeutic targeting of dysregulated apoptosis and autophagy provides a rationale to develop agents to promote cancer cell apoptosis.

Bcl-2 inhibitors

The Bcl-2 family consists of 25 apoptotic proteins that regulate the homeostatic balance between life and death. Malignant cells highjack the Bcl-2 family of pro- and anti-apoptotic proteins, primarily to inhibit apoptosis that secures immortality by overexpression of anti-apoptotic members, sequestration, and gene deletion of pro-apoptotic members. ^[113] Bcl-2 was discovered as a translocated overexpressed protein [t(14;18)] at the immunoglobulin heavy-chain locus in follicular NHL. ^[114] Subsequently, homologous, structurally defined members were identified by the presence of up to 4 conserved Bcl-2 homology (BH) domains, all of which possess α-helical segments.

Anti-apoptotic members (eg, Bcl-2, Bcl-Xl, Mcl-1, Bcl-B, Bcl-w, Bfl-1/A1) have conserved sequences in all 4 BH domains, while the pro-apoptotic members are divided into multi-BH domain members (eg, BAX, BAK, BOK) and BH3-only members (eg, BAD, NOXA or BID, BIM) that display sequence similarity only to the BH3 α-helical domain. The BH-3-only group is divided into 2 groups: inactivating BH3s (eg, BAD, NOXA) inhibit anti-apoptotic members (eg, Bcl-2, Mcl-1); and activating BH-3s (eg, BID, BIM) activate pro-apoptotic members (eg, BAX, BAK). Depending on the nature of the apoptotic stimulus, the BH-3-only subgroup’s death signal will either be inhibited by anti-apoptotic members or be delivered directly or indirectly to the mitochondrial members BAX/BAK. Activated BAX/BAK induces outer mitochondrial permeability, releasing mitochondrial factors that activate caspases, which execute the death program. ^[115]

The first agent to enter clinical trials targeting Bcl-2 was an antisense (oblimersen sodium, G3139, Genasense) phosphothiolate oligodeoxynucleotide that targets Bcl-2 mRNA. The agent was evaluated alone in CLL, with minimal response. ^[116] Subsequently, combinations with other anticancer modalities (chemotherapy, immunotherapy) have been evaluated in CLL, AML, MM, SCLC, NHL, and melanoma. A randomized phase III study with dacarbazine, with or without oblimersen, in advanced melanoma did not receive FDA approval. ^[117] A randomized phase III study with fludarabine plus cyclophosphamide, with or without oblimersen, in relapsed or refractory CLL with long-term follow-up appears to show a survival advantage in a subset of patients and is under FDAreviewforapproval. ^[118]

Several classes of small molecule inhibitors targeting the Bcl-2 anti-apoptosis members have entered early-phase clinical trials. The identification of BH3 domains with the critical α-helix that engages and regulates Bcl-2 family members has provided the critical interaction site currently being targeted as an anticancer therapy to treat myriad cancers. ^[115] The BAD BH-3 small molecule mimetic, designed by an NMR-guided fragment-based approach, discovered ABT-737 and derivative ABT-263 (Abbott), which are selective potent inhibitors of Bcl-2, Bcl-Xl, and Bcl-w, sequestering BH-3 domain pro-apoptotic proteins and thereby promoting Bax and Bak oligomerization leading to apoptosis. Unfortunately, high levels of Mcl-1 in cancer cells are associated with resistance to ABT-737. However, combination therapy with other agents may overcome this effect.

The orally bioavailable ABT-263 is currently in early-phase drug development as a single agent, as well as with immunotherapy and chemotherapy. ^[119] The agent is non-myelosuppressive, has a dose-dependent transient thrombocytopenia (targeting Bcl-Xl in megakaryocytes), and has single-agent activity in relapsed/refractory lymphoid malignancies, with minimal systemic toxicities. Clinical development of this agent is ongoing, in combination with other approved agents, for CML, SCLC, and CD20+ B-cell lymphoproliferative diseases. ^[120] GX15-070 (obatoclax, Gemin X), a pan-Bcl2 SMI with a lower affinity to Bcl-2 members, compared to ABT-263, is being evaluated in phase I studies in CLL. ^[121] A phase I trial of combination therapy with docetaxel for NSCLC established an MTD for GX15-070. ^[122] Thus far, no significant responses have been reported.

Levo-gossypol (AT-101, Ascenta), a natural product that inhibits Bcl-2, Bcl-Xl, and Mcl-1, is in phase II clinical trials for CLL (with rituximab) ^[123] and HRPC (with docetaxel). ^[124] GI toxicity was the dose-limiting toxicity in phase I studies of prostate cancer. A derivative of gossypol, apogossypol, is in preclinical development.

Finally, autophagy (self-eating) is another apoptotic pathway that Bcl-2 interferes with through inhibition of Beclin-1, an autophagy inducible gene. Autophagy maintains homeostasis by promoting cell death or survival, depending on the nutritional status of cells. The BH3 domain of Beclin-1 interacts with Bcl-2 or Bcl-Xl, and that drug disruption of the interaction between Bcl-2 and Beclin-1 can promote autophagy. Targeting autophagy in tumors is an active area of research that will likely lead to the discovery of novel SMIs that target Beclin-1. ^[125]

Antisense therapy to IAPs

The members of the inhibitor of apoptosis (IAP) gene family are potent inhibitors of caspases and, thus, protect cells from undergoing apoptosis. X-linked inhibitor of apoptosis (XIAP) is the most potent member of the IAP gene family, directly binding and inhibiting caspases ^{[3, 7, 9]} and suppressing apoptosis triggered intrinsically (mitochondria-related) and extrinsically (death receptor-related). Overexpression of XIAP suppresses apoptosis induced by a variety of apoptotic stimuli, including TNFα, FAS, serum or growth factor withdrawal, ischemia, and chemotherapy and radiotherapy.

XIAP can be targeted for proteasome-dependent degradation on pro-apoptotic stimuli because of its ubiquitin protein ligase activity associated with its RING finger domain. ^[126] This activity also promotes ubiquitin-proteasomal degradation of IAP binding partners, such as caspase-3, and enhances an XIAP anti-apoptotic effect in FAS-induced cell death. ^[127] Thus, XIAP regulation through the proteasome is complex. XIAP is highly expressed in many types of human tumors (GBM, AML, CRC, BC, pancreatic cancer, prostate cancer, gastric cancer) and appears to be a poor prognostic feature. Therefore, targeting XIAP in human malignancies is thought to be an important anticancer therapy option. ^[128]

AEG35156 is a second-generation antisense oligonucleotide [ASO] (19-mer that incorporates 2'-O-methyl with a phosphorothioate backbone), which targets XIAP mRNA to lower XIAP protein levels and, therefore, the apoptotic threshold of cancer cells, enhancing their sensitivity to intrinsic death and chemotherapy. In a phase I trial of advanced refractory cancer patients, AEG53156 was given as a 7-day or 3-day continuous infusion every 21 days. The MTD was 125 mg/m² (7-day) or 213 mg/m² (3-day), with DLTs including transaminitis, hypophosphatemia, and thrombocytopenia. The suppression of XIAP mRNA was maximal at 72 hours, with a mean suppression of 21%. In a patient with NHL, circulating lymphoma cells showed a large decrease, while a patient with melanoma and breast cancer had unconfirmed PRs.

A second phase I/II trial evaluated AEG35156 (12-350 mg/m²) IVI over 2 hours in combination with idarubicin plus Ara-C in relapsed or primary resistant AML. The suppression of XIAP mRNA increased with dose escalation, with more than 30% suppression at 350 mg/m², which appears to correlate with response (47% CR/CRp [incomplete platelet recovery]). Patients (10 of 11) refractory to single induction chemotherapy achieved a CR/CRp after reinduction with AEG35156 plus chemotherapy. With multiple doses of AEG35156, peripheral neuropathy was observed in 2 patients. ^[129] Currently, phase I/II trials are completed or ongoing with AEG35156 in combination with chemotherapy for pancreas cancer, BC, NSCLC, AML, NHL, and other solid tumors and in combination with Sorafenib in patients with hepatocellular carcinoma (HCC).

Survivin, another IAP gene family member involved in cell division and apoptosis inhibition, ^[130] is also being targeted by a second-generation ASO LY2181308 (Lilly/Isis) and is currently in phase I clinical trials. ^[131]

Death receptors

Tumor necrosis factor (TNF) and TNF-related apoptosis-inducing ligand or Apo2 ligand (TRAIL/Apo2L) members of the TNF superfamily induce apoptosis upon binding to cell surface death receptors TNFR1, TNFR2, CD95/FAS, and TRAIL receptors 1 and 2 (DR 4 and DR5), directly activating the caspase cascade via an initiator caspase (caspase-8) within the death-inducing signaling complex. The preferential expression of DRs on malignant cells provides for a potential target for cancer therapy.

Translational research has focused on the mechanism of apoptosis induced by TRAIL and drug resistance. In many cancers, only 1 of the 2 death-inducing TRAIL receptors is functional. Further, there is a need to avoid decoy receptor (DcR 1 and 2)-mediated neutralization of TRAIL leading to the development of receptor-specific TRAIL variants and agonist monoclonal antibodies. These agents are predicted to be more potent than native TRAIL and preferred for targeted treatment of specific tumor types. ^[132]

Due to liver toxicity in the first clinical trials with TNF-derived molecules, a newer version of TRAIL in phase I/II clinical trials (Genentech) uses purified extracellular region of TRAIL (residues 114–281) with zinc to stabilize a biologically active trimer at physiologic pH. TRAIL (0.5-15 mg/kg) administered as a 1 hour IVI on days 1-5 in a 21-day cycle (up to 8 cycles) demonstrated linear pharmacokinetics and no DLTs or liver toxicity, with a half-life of approximately 30 minutes. ^[133] Thirty-two patients were evaluable for response, with 1 patient with chondrosarcoma confirmed with a partial response (8 mg/kg), 17 patients with stable disease (53%), and 13 patients (41%) with progressive disease. A phase Ib study of rhTRAIL (4 and 8 mg/kg) plus rituximab (375 mg/m² weekly, up to 8 doses) in relapsed/refractory low-grade NHL showed no DLTs or AEs. ^[133] Five patients were eligible for responses: 2 CR, 2 PR, and 1 SD. ^[133]

Several therapeutic agonist Mabs targeting DR4 and DR5 are in phase I/II clinical trials: LBY135 (Novartis), AMG 665 (Amgen), lexatumumab, HGS-TR2J (HGS), and Apomab (Genentech) are DR5 agonist Mabs, while CS-1008 (Sankyo) and mapatumumab (HGS) are DR4 agonist Mabs. Clinical data indicate modest antitumor activity, good safety profile (no liver toxicity), and linear PKs and half-life lives of approximately 10–20 days, with no detection of human antihuman antibodies (HAHA). ^[134]

In a phase I study of lexatumumab, up to 10 mg/kg was safe in refractory solid malignancies, with SD in 12 patients (32%) that lasted a median of 4.5 months. ^[135] The safety of lexatumumab in combination with gemcitabine plus pemetrexed, plus doxorubicin or leucovorin, plus 5-FU, plus irinotecan, is being evaluated in a phase Ib trial. ^[136] Although PRs have been reported in both arms, randomized studies are needed to evaluate the efficacy of lexatumumab.

Mapatumumab, a fully humanized IgG1 DR4 Mab agonist, has been evaluated in patients with refractory solid malignancies. It was administered safely in a dosage of up to 10 mg/kg every 14 days with no adverse reactions. Nineteen of 49 patients enrolled in the study had SD, with 2 patients studied for 9 months. ^[137] Although DR4 is expressed on approximately 68% of the tumors, its expression alone does not predict response. ^[137]

Combination studies of mapatumumab with paclitaxel plus carboplatin in patients with advanced cancers found 4 of 28 patients experienced PRs. ^[138] In a third phase Ib study, mapatumumab was administered safely in a dosage of up to 30 mg/kg every 3 weeks in combination with gemcitabine plus cisplatin. In this study, 9 of 45 patients experienced a PR, and 13 of 45 patients achieved SD for more than 18 weeks. ^[139] In a phase II study of single-agent mapatumumab in NSCLC patients, the treatment was well tolerated, but none of the 32 patients had a meaningful response based on RECIST criteria, with approximately 29% achieving SD. ^[140]

For tumors to grow and metastasize, the development of a neovascular blood supply (angiogenesis) is paramount. Angiogenesis is a complex system composed of multiple proangiogenic and antiangiogenic factors that regulate new blood vessel formation. The vascular endothelial growth factor (VEGF) and its receptors have been implicated as key players in tumor-driven angiogenesis. However, there are multiple other key players and mechanisms that require scientific exploration for development of effective therapies targeting the complex process of neoangiogenesis.

VEGF

The VEGF family is composed of 6 secreted glycoproteins (VEGF-A, B, C, D, and E and placental growth factor) that activate 3 RTKs (VEGFR-1, 2, 3). VEGF-A (VEGF) are claimed to be a central mediator of tumor-mediated angiogenesis, and thus, targeting VEGF would be an effective anticancer therapy. ^[141] VEGF discovered as a vascular permeability factor exerts changes in endothelial cell (EC) morphology and cytoskeleton changes and stimulates EC migration and growth. ^[142]

Bevacizumab is a therapeutic monoclonal antibody that targets VEGF and has undergone clinical development as a single agent ^[143] and in combination with chemotherapy and targeted therapies. It is the only approved antiangiogeneic therapy for 4 human malignancies (CRC, NSCLC, BC, GBM). ^[144] Clinical trials have been completed with bevacizumab plus IFN-α versus IFN-α for RCC (phase III), which appears promising for the combination arm ^[145] , and bevacizumab plus R-CHOP for DLBCL (phase II; results pending [SWOG study]). Unfortunately, there are no validated predictive markers of response to bevacizumab. Its usefulness in the adjuvant setting and as a therapy after progression is under investigation. ^[146]

In September 2017, FDA has approved Mvasi (bevacizumab-awwb) as a biosimilar to Avastin (bevacizumab) for the treatment of multiple types of cancer including colorectal, lung, brain, kidney, and cervical cancers. This approval is the first biosimilar approved for the treatment of cancer in the United States. FDA noted that Mvasi has been approved as a biosimilar to Avastin but not as an interchangeable product. 

Several other methods of targeting VEGF are also under investigation: VEGF trap (Aflibercept), a fully human decoy receptor protein composed of the second Ig-ECD of VEGFR1, the third Ig-ECD of VEGFR2, and the Fc region of IgG1 ^[147] ; HuMV833, which is a humanized IgG Mab that binds the VEGF-A isoforms VEGF121 and VEGF165 with pM binding affinity and has completed a phase I study in patients with refractory solid tumors ^[148] ; and IMC-1C11, a Mab-targeting VEGFR2. ^[149] All 3 agents are in early-stage clinical development. Several ATP-site SMIs targeting the VEGFR (1, 2, and/or 3) TK domains are approved (sorafenib, sunitinib) or in early-phase clinical trials (vatalanib, motesanib, CP-547,632, pazopanib, ABT-869, cediranib). ^[144]

Tie

Angiopoietins are growth factors that promote angiogenesis, the formation of blood vessels from preexisting blood vessels. There are now 4 identified angiopoietins: Ang1, Ang2, Ang3, and Ang4. In addition, there are a number of proteins that are closely related to angiopoietins (ANGPTL2, 3, 4, 5, 6, 7). Ang1 and Ang2 are required for the formation of mature blood vessels, as demonstrated by mouse knockout studies. Tie are RTKs, named by their capacity to mediate cell signals by inducing phosphorylation of tyrosines, initiating the binding and activation of downstream, intracellular enzymes.

Which of the Tie receptors mediate functional signals downstream of Ang stimulation is somewhat controversial, but at least Tie2 is capable of physiologic activation as a result of binding the angiopoietins. ^[150] Tie2 was originally found to be overexpressed in tumor-associated blood vessels. Since then, Tie2 has also been shown to be overexpressed in leukemia and solid tumors (gastric cancer, BC, GBM). ^[151] Since cancer is a complex pseudo-organ with several coexisting cellular lineages, the aberrant expression of Tie2 by different tumor compartments makes this RTK an attractive target for cancer therapy. ^[152]

AMG 386 is a selective angiopoietin-1 and -2 neutralizing peptibody that inhibits angiogenesis by preventing interaction of angiopoietins with Tie2 receptors. A first-in-human study evaluated the safety, pharmacokinetics (PK), and tumor response of AMG 386 in adults with advanced solid tumors. Patients (N = 32) were treated with Q1W IV AMG 386 at 0.3, 1, 3, 10, and 30 mg/kg. A single DLT at 30 mg/kg of respiratory arrest, likely due to tumor burden, was thought to be related to AMG 386. Other treatment-related toxicities were grade 2 fatigue, nausea, and peripheral edema. PK was dose-linear, with a mean T_½ of 3.1-6.3 days.

Per RECIST criteria, 16 patients had SD (4-52 wk, median 8 wk). A patient with ovarian cancer was studied for more than 3 years, with 27% tumor shrinkage and reduction in serum CA-125. DCE-MRI showed significant vascular effect changes, with more than 20% reduction in 7 of 12 (58%) evaluable patients. In 5 of 6 evaluable patients, a decrease in SUV by FDG-PET was associated with a reduction vascular change by DCE-MRI.

A second open-label study evaluated AMG 386 plus FOLFOX-4 (F) or carboplatin plus paclitaxel (CP) or docetaxel (D) in patients (N = 16) with refractory solid tumors. Three cohorts of 6–9 patients received 1 full cycle of chemotherapy (2 weeks off or 3 weeks of D 75 mg/m² or CP). AMG 386 at a dosage of 10 mg/kg IV weekly was started on day 1 of cycle 2 for patients who did not experience a DLT to chemotherapy during cycle 1, and it was continued until disease progression or intolerance. AEs related to chemotherapy plus AMG 386 were neutropenia, thrombocytopenia, diarrhea, and vomiting. No neutralizing antibodies were observed, and chemotherapy did not appear to affect the PK profile of AMG 386. One patient receiving CP plus AMG 386 for refractory bladder cancer had a CR at week 8 and confirmed at week 16. ^[153]

Currently, a phase II study is evaluating the safety, tolerability, and PKs of AMG 386 in combination with AMG 706, bevacizumab, sorafenib, or sunitinib. Moreover, a safety study of AMG 386 (3 or 10 mg/kg) to treat HER2+ locally recurrent or metastatic BC with paclitaxel plus trastuzumab or capecitabine plus lapatinib is ongoing.

FGFR inhibitors

Fibroblast growth factor (FGF) and receptor (FGFR) RTK pathways are implicated in neo-angiogenesis, proliferation, and stem cell survival. The family is composed of 22 known ligands and 4 RTKs, including splice variants (1b, 1c, 2c, 3b, 3c and 4), with selective binding of overlapping sets of ligands to receptors. ^[154] The FGF pathway has been shown to be important in tumor initiation and progression due to amplifications, overexpression, and mutations (Apert syndrome). ^[155] It is therefore a potential unexplored avenue for anticancer therapy.

Germline mutations (S252W, P253R) in the ECD of FGFR2 are causative in Apert syndrome found in endometrial cancer (12-16% of patients with endometrial cancer). ^[155] The syndrome is a congenital human skeletal disorder characterized by craniosynostosis and syndactyly. These mutations cause inappropriate RTK activation due to high-affinity binding of FGF to its receptor. These mutations are also found in cervical cancer (4%) and SCLC (5%). 

Drugs that have been approved by the FDA that inhibit FGFR include erdafitinib, infigratinib, and pemigatinib. Erdafitinib is indicated for urothelial carcinoma that has fibroblast growth factor receptor-2 (FGFR2) or FGFR3 genetic alterations. Infigratinib and pemigatinib are indicated for cholangiocarcinoma with FGFR2 fusion or other rearrangement. Additionally, pemigatinib is indicated for relapsed or refractory myeloid/lymphoid neoplasms with FGFR1 rearrangement. 

The investigational FGFR inhibitors dovitinib, futibatinib, and bemarituzumab are also being evaluated for renal cell carcinoma, cholangiocarcinoma, and gastric cancers, respectively.  

Notch inhibitors

Notch-signaling pathway is evolutionarily conserved and plays major roles in embryonic vascular development and angiogenesis. Multiple components of the Notch pathway are expressed in the vasculature. ^[156] Mice deficient for a variety of these components display embryonic lethality with vascular remodeling defects. ^[156] Selective interruption of Notch signaling within tumors may provide an anti-angiogenic strategy. Notch-pathway components within the vasculature consist of 3 cell surface receptors (Notch1, 3, 4), 4 ligands (DLL4, Jagged1, 2, 3), and 3 downstream targets (Hey1, 2, L).

Ligand binding leads to the Notch intracellular domain (NICD) released from the endothelial cell plasma membrane by γ-secretase-dependent proteolytic cleavage of the Notch receptor. The NICD translocates to the nucleus, where it binds to the CSL (named mammalian C BF1/RBP-Jκ, Drosophila S u(H), and C elegans Lag-1) transcription repressor to activate transcription of target genes. The basic helix-loop-helix proteins (bHLH) Hairy/Enhancer of Split (Hes1, 5, 7) and Hes-related proteins (Hey1, 2, L) are the best-characterized downstream targets. Notch signaling from the tumor cells activates endothelial cells and, thus, initiates tumor angiogenesis. Accordingly, the neovasculature is able to stimulate the tumor growth and progression. ^[157]

The γ-secretase inhibitor IX (γ-SI IX) and soluble Jagged1 (sJag1) can protect the proteolytic cleavage by inhibiting the γ-secretase activity and inhibiting the interaction of Jagged1 and Notch, respectively, thereby preventing endothelial activation. Notch cross-signaling between tumor cells, stroma cells, and endothelial cells regulates the interaction of Notch ligands on tumor cells with receptors on endothelial cells, and vice versa. ^[156]

MK-0752 (Merck) is a potent oral γ-secretase inhibitor that was evaluated in a phase I/II study. Patients with refractory solid tumors or breast cancer were enrolled using an accelerated dose escalation, with 1 patient per dose level until the occurrence of grade 2 toxicity, then 3–6 patients per dose level. The agent was administered by daily oral dosing every 28 days. Once MTD was established, an additional 22 patients with advanced BC were enrolled in the phase II part of the study, based on the observation that activation of Notch signaling occurs in 40%, with an associated poor outcome.

In the phase I study, 2 patients were enrolled at 450 mg qd, and 5 were enrolled at 600 mg qd. DLTs were grade 3 diarrhea, constipation, nausea, and abdominal cramping (600 mg). In the phase II study, 14 patients with BC were enrolled at 450 mg qd. AEs included grade 2/3 fatigue requiring dose reduction in 6 patients, grade 3 diarrhea, nausea, and transaminitis. Mean PKs at 450 mg and 600 mg on day 1 were AUC_0–24hr = 1036 and 1065 µM·hr; C_max = 72 and 61 µM; C_24hr = 25 and 32 µM; and t_max = 3 and 7 hours. PD of γ-secretase inhibition showed a 12–78% (mean, 46%) decrease in plasma Abeta₄₀ at 4 hours on day 1, as compared to predose.

Continuous daily dosing of MK-0752 at 450 mg in BC patients was associated with significant toxicity, predominantly fatigue, and cannot be considered an MTD. ^[158] An intermittent dosing schedule is being explored. MK-0752 at all doses inhibits GS, as demonstrated by decreases in plasma Abeta₄₀. Currently, MK-0752 is being studied in combination with tamoxifen or letrozole to treat early stage BC, and a phase I/II study of MK-0752 followed by docetaxel in advanced or metastatic BC is ongoing.

RO4929097 (Roche) is a potent and selective inhibitor of γ-secretase, producing inhibitory activity of Notch signaling in tumor cells that does not block tumor cell proliferation or induce apoptosis but, rather, produces a less transformed, flattened, slower-growing phenotype. ^[159] The agent has now entered first-in-human studies in advanced cancer.

Integrin targeted agents

Integrins are cell surface adhesion receptors comprised of a- and b- subunits that form hetero-dimers. They are involved in outside-in and inside-out signaling that regulates normal cellular processes such vascular proliferation, adhesion, immune response, and wound repair. The molecular interactions between integrins and their respective ligands is the recognition of the Arg-Gly-Asp (RGD) motif present in the extracellular matrix components fibronectin, vitronectin, collagen, fibrinogen, and von Willebrand factor. Integrins cluster with RTKs and their adaptor proteins to form focal adhesion complexes. Many tumors overexpress integrins (a5b1, avb3, avb5), which promote angiogenesis and metastasis. Therefore, inhibition of ECM-ligand-integrin interactions with Mabs and peptides or peptidomimetics are considered useful anticancer agents to be developed in early-phase clinical trials. ^[160]

Unliganded avb3 can mediate anchorage-independent tumor growth and metastasis independent of FAK (focal adhesion kinase) and interacts with c-Src via the b3 subunit. An Src kinase inhibitor (SKI) or Src knockdown decreased soft agar colonies, indicating that avb3-induced anchorage-independent survival is dependent on c-Src. Therefore, Src inhibition would be of therapeutic importance in tumors overexpressing avb3 integrin. ^[161]

Volociximab (M200, BiogenIdec) is a chimeric IgG4 Mab that targets a5b1A (expressed on vascular endothelial cells) and inhibits endothelial cell-cell interactions, endothelial cell-ECM interactions, and angiogenesis. ^[162] A phase II study evaluated patients (N = 40) with advanced RCC (clear cell), with SD as the best response in 80% patients. The median TTP was 4 months, with approximately 80% alive at 6 months. ^[163] Studies of combinations with gemcitabine in metastatic PDA and with liposomal doxorubicin in platinum-resistant ovarian cancer are ongoing.

Vitaxin (etaracizumab, Abegrin) is a humanized Mab, which specifically binds aVb3, the vitronectin receptor, and interferes with blood vessel formation by inducing apoptosis of tumor endothelial cells in preclinical animal models. A phase I trial evaluated the safety and efficacy of vitaxin in patients with refractory solid tumors at doses of 10, 50, or 200 mg IV in cohorts of 3 patients. Three patients demonstrated SD. with one case lasting 22 months. ^[96]

A second identical study also evaluated a similar patient population in whom standard therapy with the above doses failed. Unlabeled vitaxin was infused on days 0 and 21 of a treatment cycle, and 11 patients received a pretherapy imaging dose of 1 mg of technetium-99m-vitaxin, with the use of gamma-camera imaging studies to evaluate the effect on tumor angiogenesis. There was no significant toxicity in both studies at the 3 dose levels. However, there were no objective antitumor responses, and imaging of the tumor vasculature was unsuccessful. ^[164]

A second phase I study was conducted in refractory solid tumors at doses ranging from 1 to 6 mg/kg. The half-life of etaracizumab ranged from 49-180 hours. There were no objective responses, but 5 patients experienced SD for more than 6 months. ^[165]

Despite these negative studies, vitaxin appears to be safe and potentially active, and combination with chemotherapy is warranted. CNTO-95 (Centocor) is a fully human Mab against av and has completed phase I studies. The phase I trial of CNTO-95 evaluated doses of 0.1, 0.3, 1.0, 3.0, and 10.0 mg/kg infused on days 0, 28, 35, and 42, with clinical assessments of DCE-MRI and FDG-PET in 24 patients with refractory solid cancer.

In the study, CNTO-95 caused 1 episode of grade III and 4 episodes of grade II infusion-related fever (responded to acetaminophen). Six patients received extended dosing, and one patient, at the 10.0 mg/kg dose with cutaneous angiosarcoma, achieved a PR for 9 months. Pretreatment and posttreatment tumor biopsies confirmed tumor cell av expression, with CNTO-95 penetration of the tumor and localization to tumor cells, in association with decreased Bcl-2 expression. The dose-dependent mean half-life ranged from 0.26-6.7 days. ^[166] Currently, CNTO-95 is in phase II studies in melanoma and hormone refractory prostate cancer.

Cilengitide (EMD-121974) is a cyclic pentapeptide inhibitor of avb3 and avb5, which was evaluated in a phase I study in patients with refractory solid tumors. The agent was administered 2 times per week every 28 days with no DLTs. ^[167] Phase II studies are ongoing in several solid tumor types (GBM, H/N, AML, melanoma, prostate).

Vascular targeting agents

Agents that act to deprive tumors of a blood supply fall into 2 categories: angiogenesis inhibitors and vascular-targeting agents (VTAs). VTAs have the potential advantage that they are able to damage the existing tumor vasculature, instead of prevention of further formation by neoangiogenesis. Vascular targeting agents (VTAs) are multifunctional agents that target the capillaries and vessels of solid tumors. The mechanism of action of VTAs is the disruption of the cytoskeleton and cell-cell interactions in ECs. This manifests as plasma leak, viscous blood flow, and initiation of thrombosis. Once localized to the tumor blood vessels, VTAs block the flow of oxygen and nutrients to the tumor by activating a thrombotic pathway. ^[168] There are currently 7 VTAs in clinical trials, with many more in preclinical development. Six small-molecule VTAs have entered phase II clinical trials as follows:

• Combretastatin A4 phosphate (CA4P, OXiGene)

• ZD6126 (Angiogene/AstraZeneca)

• Ave-8602 (Aventis/Ajinomoto)

• Soblidotin (TZT-1027)

• NGR-hTNF (Mol Med)

• Endo-Tag (MediGene)

Combretastatins (eg, CA4P) are structural analogs of colchicine that bind to tubulin, causing tubulin depolymerization destabilizing the EC cytoskeleton. Additionally, it also interferes with the VE-cadherin-b-catenin complex, resulting in loss of cell-cell interactions. A phase I study evaluated 3 dose levels given by IV, with DLTs of dyspnea, neurologic disturbance, cardiac ischemia, and intestinal ischemia. Partial responses were observed in thyroid cancer, sarcoma, and adrenocortical cancers. ^[169] The agent was evaluated in a phase II study of anaplastic thyroid cancer alone and in combination with chemotherapy. ^[170] Another study is evaluating CA4P plus chemotherapy, or CA4P plus chemotherapy and bevacizumab, by measuring the apparent diffusion coefficient by MRI. ^[171]

ZD6126 is a water-soluble phosphate prodrug of N-acetylcolchinol that disrupts the endothelial tubulin cytoskeleton, causing selective occlusion of tumor vasculature and extensive tumor necrosis. ZD6126 is converted in vivo into N-acetylcolchinol, which binds to and destabilizes the tubulin cytoskeleton of endothelial cells in tumor blood vessels. This may result in tumor endothelial cell apoptosis, the selective occlusion of tumor blood vessels, cessation of tumor blood flow, and tumor necrosis. A phase I study was conducted to evaluate the dose and administration schedule of ZD6126 in patients with refractory solid tumors. Patients received a 10-min, single-dose IVI every 14 or 21 days. Subsequent doses were escalated based on the incidence of AEs within the first cycle of administration.

Forty-four patients were enrolled and received ZD6126 (5−112 mg/m² in the 21-day schedule, N = 35; 40−80 mg/m² in the 14-day schedule, N = 9). Common AEs were dose-related abdominal pain, nausea, and vomiting. DLT was abdominal pain and cardiac toxicity (decreased left ventricular ejection fraction [LVEF], increased troponin, myocardial ischemia, ECG signs of myocardial ischemia) at 112 mg/m² in the 21-day schedule. PK studies confirmed that ZD6126 is rapidly hydrolyzed to ZD6126 phenol.

In the study, there was no difference in PKs of ZD6126 phenol on repeat administration or between the 2 dosing regimens. DCE-MRI demonstrated antivascular effects of ZD6126. The MTD was given every 2 or 3 weeks at 80 mg/m². In approximately 11% (5 out of 44) of patients, ZD6126 was associated with cardiac events categorized as dose-limiting toxicities (one patient asymptomatic). No meaningful responses were described based on RECIST criteria. ^[172]

Paclitaxel encapsulated in a cationic liposomal formulation known as EndoTAG-1 apparently binds preferentially to proliferating endothelial cells and interferes with the tumor vasculature. In a phase II study, 200 patients with advanced pancreas received 3 different doses of EndoTAG-1 (11, 22, 44 mg/m²) in combination with once-weekly gemcitabine (1,000 mg/m²) for 7 weeks. The 1-year survival with EndoTAG-1 plus gemcitabine nearly doubled, compared to gemcitabine alone (33-36% versus 17.5%). The median survival for the combination was 9.4 months (44 mg/m²) and 8.7 months (22 mg/m²), compared to 6.8 months for gemcitabine. ^[173] This combination is being studied in a phase III randomized clinical trial of advanced PDA.

The tumor-homing peptide NGR (Asn-Gly-Arg) selectively binds CD13, a cell surface aminopeptidase overexpressed on the tumor vasculature. NGR linked to human TNF generates a unique peptide-protein fusion product known as NGR-hTNF, which functions as a VTA. In preclinical studies, murine NGR-TNF showed a biphasic dose response with significant antitumor activity (0.005 mg/kg). The equivalent human dose of 0.2 mg/m² was the starting dose used in a phase I dose-escalation study. Seventeen dose levels were tested (0.2-60 mg/m²) in 70 patients with MTD, defined as 45 mg/m² when given as a single dose every 21 days. DLT was dyspnea and acute infusion reaction.

A phase Ib trial subsequently evaluated NGR-TNF plus doxorubicin based preclinical synergy studies, where vascular barrier alterations increased uptake of chemotherapy. In this study, 15 patients with refractory solid tumors were administered NGR-hTNF (0.2, 0.4, 0.8, 1.6 mg/m²) and doxorubicin (60-75 mg/m²), both given IV every 3 weeks. Nine of 15 patients had prior anthracyclin therapy, and no DLTs were observed. Adverse events were neutropenic fever (N = 2) and decreased ejection fraction (N = 2). There were no PK interactions, and shedding of soluble TNFRs did not increase to 0.8 mg/m². There was 1 PR and 10 SDs that lasted a median of 5.6 months. The dose of 0.8 mg/m² of NGR-hTNF plus doxorubicin 75 mg/m² was selected for phase II clinical development. ^[174]

Other VTAs in clinical development targeting are TZT-1027 (NSCLC, sarcoma) and the flavanoid derivative 5,6-dimethylxanthenone-4 acetic acid (DMXAA, AS-1404) (NSCLC, ovary cancer, HRPC). ^[61]

The ability of tumor cells to possess unlimited replication potential is linked to maintenance of telomeric DNA (repetitive TTAGGG sequences of DNA located on the ends of human chromosomes). Normal cells shorten their telomeres during each round of replication due to the "end-replication problem," in comparison to tumor cells, where telomere length is stabilized by reactivation of the reverse transcriptase telomerase (hTERT) and a multiprotein complex called shelterin. Telomerase also may play roles such as regulation of the chromatin state and DNA damaged responses. ^[175] Cells lacking hTERT have a diminished capacity for DNA double-break repair. ^[176]

The first antitelomerase is an oligonucleotide-based molecule GRN163L (imetelstat sodium, Geron) targeting the telomerase RNA template that has entered clinical trials. Another approach of targeting telomeres is the G-quadruplex with ligands such as RHPS4. ^[177] However, in the opinion of the author, targeting G-quadruplex structures is unlikely to lead to effective agents because of the nonspecific nature of such drugs with considerable toxicity. Limitless replication potential is also due to aberrant cell cycle control in tumor cells, which are driven by overexpression of cell cycle (CDKs), checkpoint (CHKs) mitotic kinases, and abnormal DNA damage repair responses. Small molecule inhibitors targeting these processes have entered clinical trials and appear to be more promising anticancer agents than current antitelomerase targeting agents.

CDK inhibitors

Advances in cell cycle regulatory mechanisms over the last 10 years has illustrated the importance of aberrations in key players that contribute to cancer pathogenesis. The mammalian cell cycle consists of 4 distinct phases occurring in a well-defined order, each of which has to be completed successfully before the next phase begins. Progression through major cell cycles is mediated by sequential assembly and activation of a family of S/T kinases, the cyclin-dependent kinases (CDKs). The timing of CDK activation is determined by phosphorylations-dephosphorylations and by association of specific cyclins, which bind and activate specific CDKs.

The cyclin family is divided into 2 main classes: G1 cyclins (C, D1-3, E) accumulation is rate limiting for progression from G1 to S phase; mitotic or G2 cyclins (A and B) are involved in the control of G2/M transition and mitosis. Activated CDKs phosphorylate and inhibit the tumor suppressor protein, Rb, committing the cell to G1 and to S phase progression by increasing the activity of E2F transcription factors known to promote cell proliferation. The D-type cyclins and their partner kinases CDK4/6 have proto-oncogenic properties, and their activity is tightly regulated by negative control by 2 families of CDK inhibitors. Members of the INK4 family (p16INK4A, p15INK4B, p18INK4C, p19INK4D) interact specifically with CDK4/6, while the CIP/KIP inhibitors (p21CIP1/WAF1, p27KIP1, p57KIP2) inhibit a broader spectrum of CDKs.

The interplay between p16INK4A, cyclin D/CDK4/6, and pRb/E2F/p53 together is a functional unit collectively known as the pRb/p53 pathway. Each of the major components of this mechanism may become deregulated in cancer, and accumulating evidence points to the pRb/p53 pathway as a candidate obligatory target in the multistep oncogenesis of possibly most human tumors. ^[178] Major advances in the understanding of cell cycle regulatory mechanisms have provided a better understanding of the molecular interactions involved in human cancer pathogenesis. ^[179]

Discovery and development of ATP-site SMIs to CDKs as antiproliferative agents is based on the hypothesis that selective cell growth arrest and/or apoptosis could be induced due to impaired control of cell cycle progression in malignant cells. Several CDK ATP-site SMIs, the first generation (flavopiridol, UCN-01, bryostatin, CYC202, BMS387032, E7070) and the second generation (SNS-032, AT7519, P276-00, ZK304709, R-547, PD-0332991, AG-24322, JNJ-7706621, GPC-286199, Bay 80-3000), are currently in early-phase clinical trials. ^[180]

Flavopiridol (Sanofi-Aventis), a natural product, is a potent pan-CDK SMI that blocks cell cycle progression at the G1/S and G2/M boundaries. Because of nonideal PK properties, dosing has changed by starting with a loading dose followed by continuous IV infusion. Initial testing in early clinical human trials with infusional flavopiridol showed activity in patients with NHL, RCC, prostate cancer, CRC, and gastric carcinomas. Main side effects were secretory diarrhea and a proinflammatory syndrome associated with hypotension. ^[181] Phase 2 trials with flavopiridol in several tumor types, in other schedules, and in combination with standard chemotherapies are also being conducted, particularly in relapsed/refractory CLL. ^[182]

UCN-01 (7-OH staurosporine), the second CDK SMI, has entered clinical trials. It inhibits protein kinase C (PKC) activity, promotes cell cycle arrest by accumulation in p21/p27, induces apoptosis in several preclinical models, and abrogates the G2 checkpoint by inhibition of CHK1. The last of these represents a novel strategy to combine UCN-01 with DNA-damaging agents. In the initial UCN-01 clinical trial (continuous infusion for 72 hours), a prolonged half-life of approximately 600 hours (100 times longer than in preclinical models) was observed. The MTD was 42.5 mg/m² per day for 3 days. DLTs were nausea/vomiting, hypoxemia, and symptomatic hyperglycemia. One patient with melanoma achieved a PR (8 months). Another patient with refractory ALCL had no evidence of disease at more than 4 years. Bone marrow and tumor samples in some patients showed loss in adducin phosphorylation, a substrate of PKC. ^[183] Phase I trials with shorter infusions have been completed.

Phase I trials with BMS 387032 and R-Roscovitine (CYC202) have shown good tolerability; however, absence of pharmacodynamic end points to confirm target inhibition has been a problem. ^[180]

Second-generation CDK SMIs are more specific to the CDKs they inhibit and appear to have improved PK and PD profiles. Two phase I studies of AT7519 (Astex), a multitargeted CDK SMI (CDK 1, 2, 4, 5, 9), in refractory solid tumors have been conducted with 2 different dosing schedules. The MTD was 28.8 mg/m² for schedule 1, while the MTD has not been reached for schedule 2. Adverse events were cyclical neutropenia, mucositis, and fatigue. DLT was QTc prolongation at 34 mg/m².

The second dosing schedule has not observed QTc prolongation with dose escalation. At the first dose level in schedule 1, one heavily pretreated patient with NSCLC achieved a PR of approximately 80% and lived for more than 12 months. Three patients with pancreas cancer post gemcitabine therapy survived for more than 6 months.

The PK profile across all dose levels for AT7519 showed multiphasic elimination with a long terminal half-life (8-12 h) and only modest interpatient variation. Biomarker modulation by AT7519 was evaluated by IHC of PCNA (proliferating cell nuclear antigen), a substrate of CDK2 in the proliferating layer of the skin. At the 28.8 mg/m²/day dose (N = 4), the trend was inhibition. In all 4 patients, apoptosis marker M30:M65 cytokeratin fragment ratio in the serum showed a consistent increase at 28.8 mg/m². Biological activity was observed at 28.8 mg/m²: PCNA levels were reduced in all 4 patients. Ki67 levels were reduced in 2 of 4 patients. ^[184] This agent is being developed in relapsed/refractory CLL, where the target is CDK9, which modulates RNA polymerase II phosphorylation.

CHK inhibitors

The cell cycle is a critical regulator of the process of cell proliferation and growth and cell division after DNA damage. Progression through the cell cycle is monitored by surveillance mechanisms known as cell cycle checkpoints that govern the transition from quiescence (G0) to proliferation, which ensures fidelity of the genetic transcript. Dysfunction in cell cycle checkpoints leads to genomic instability and contributes to tumor progression. Cancer therapy (chemotherapy and radiation) activates cell cycle checkpoints. Checkpoint kinases (CHK1 and CHK2) are a family of S/T kinases that are part of the DNA damage recognition and response pathways and represent attractive targets to be combined with established cancer therapies.

SMIs that target CHKs have demonstrated impressive preclinical activity by sensitizing tumors to a variety of DNA-damaging agents and increasing activity in preclinical mouse models. The most advanced agents are now in phase I clinical trials (XL-844, AZD7762, PF-477736). ^[185]

MTK inhibitors

The mitotic (M) phase of the cell cycle is tightly regulated by CDK1, pololike kinases (Plk-1, 2, 3), NimA-related protein kinase 2 (Nek2), and Auroras (A, B, C), with a complex biological process whereby a complete copy of the duplicated genome is precisely segregated by the microtubule spindle apparatus into 2 daughter cells. Auroras are S/T mitotic kinases (MTKs) that associate with chromosomes, chromosome-associated proteins. and cytoskeletal components that drive cell division and are thus critical regulators of genomic stability.

They appear at specific locations during the M phase, as follows: ^[186]

• Aurora A, known as the polar kinase, associates with the duplicating centrosomes.

• Aurora B, known as the equatorial kinase, is a chromosomal passenger protein (CPP).

• Aurora C, also a CPP, first localizes to centromeres and then to the midzone of mitotic cells.

Mitotic errors lead to genomic instability, a hallmark of malignancy. In experimental model systems, overexpression of Aurora induces spindle defects, chromosomal missegregation, and malignant transformation. Conversely, downregulation of Aurora expression causes mitotic arrest and apoptosis in a variety of tumor cell lines (breast, colon, pancreatic, ovarian, gastric, leukemia).

Dual Aurora-A/Aurora-B inhibitors demonstrate a cellular phenotype related to Aurora-B inhibition, characterized by rapid inhibition of serine-10 phosphorylation on histone H3 and aberrant mitosis leading to failed cytokinesis and endoreduplication. Aurora inhibitor-treated cells adopt apolyploid morphology and ultimately undergo apoptosis. Several potent inhibitors of Aurora kinases, VX-680/MK-0457, PHA739358, AZD1152 (Aurora-B selective), MLN8054 (Aurora-A selective), and AT9283 (Aurora-A/Aurora-B), have progressed into clinical trials in both refractory solid and hematologic malignancies. ^[187]

Pololike kinases (PLKs) are S/T kinases that play a key role in cell division and checkpoint regulation of mitosis. Human tumors (approximately 80%) overexpress PLKs, which is associated with a poor prognosis and a lower survival rate. The overexpression of PLKs in human tumors, but not in healthy nondividing cells, makes them potential selective targets for cancer therapeutics. PLK ATP-site SMIs inhibit different subphases of the M-phase (centrosome maturation, spindle formation, chromosome separation, cytokinesis) and induce mitotic catastrophe that severely perturbs cell cycle progression, leading to cancer cell apoptosis.

Several PLK SMIs are in early clinical development (BI 2536, BI 6727, GSK461364, ON 019190.Na, HMN-214) ^{[188, 189]} or preclinical development (ZK-thiazolidinone, NMS-1, CYC-800, DAP-81, LC-445). ^[190] PLK SMIs may offer a new targeted antitumor therapy for cancer patients.

Kinesin inhibitors

Kinesin spindle proteins (KSP) are microtubule-based motors that belong to a superfamily, which mediate centrosome separation and bipolar spindle assembly and maintenance. Inhibition of KSP function leads to cell cycle arrest at mitosis, with the formation of monoastral microtubule arrays and, ultimately, to apoptosis. ^[191] Several KSP inhibitors are currently being evaluated in phase I/ II clinical trials that may provide opportunities for the development of novel anticancer agents as an alternative to microtubule-targeting drugs and/or combinations.

Most of the SMIs are ATP uncompetitive and bind in an allosteric loop L5 binding pocket. Ispinesib is an allosteric small-molecule KSP inhibitor that has completed phase I trials and is now in phase II trials. ^[192] Mutations that attenuate ispinesib binding to KSP have been identified, which highlights the need for inhibitors that target different binding sites. New classes of selective KSP SMIs active against ispinesib-resistant forms of KSP are being developed. These ATP-competitive SMIs do not bind in the nucleotide binding pocket but compete with ATP binding via a novel allosteric mechanism ascertained from generation of resistant cells, site-directed mutagenesis, and photo-affinity labeling studies. ^[193]

Two phase II studies with ispinesib (SB-715992), one in chemonaive advanced HCC ^[194] and the other in refractory melanoma, ^[195] were conducted by NCI Canada. Fifteen patients with HCC were treated with SB-715992 at 18 mg/m² IV over 1 hour every 3 weeks. Stable disease in 7 patients was the best response, with SB-715992 plasma concentrations being comparable to those observed in the phase I studies. No correlation was observed between intensity of KSP staining and clinical outcome. ^[194] In the melanoma study, 17 patients were treated with ispinesib as above. The best response was SD for a median duration of 2.8 months. ^[195]

PARP inhibitors

Normal eukaryotic cells have evolved DNA repair pathways that help preserve genomic integrity by virtue of intrinsic (metabolism) and extrinsic (toxins, radiation) DNA damaging agents. Five recognized DNA repair pathways include the following:

• Nucleotide excision repair (NER)

• Double-strand break repair (DSBR)

• Mismatch repair (MMR)

• Base excision repair (BER)

• Direct repair (DR)

These pathways help repair damaged DNA. In malignant cells, DNA repair pathways are disrupted and abrogate apoptosis by being resistant to DNA damaging agents used as anticancer agents (eg, chemotherapy and radiotherapy). Poly (ADP-ribose) polymerases (PARP) are a family of highly conserved enzymes that play a key role in signaling DNA single-strand breaks (SSB) through the BER pathway and double-strand breaks (DSB) through the DSBR pathway.

PARP-1 has 3 functional domains:

• DNA binding domain (2 Zn-finger motifs that bind damaged DNA)

• Auto-modification domain (possesses a BRCA-1C-terminal domain and acceptor of ADP-ribose polymers)

• Catalytic domain (catalyses synthesis of long and branched poly [ADP-ribose] polymers from NAD+)

The net negative charge on the poly ADP-ribose polymer opens up the damaged DNA to allow access to the other components of the repair process. Subsequently, the polymer is rapidly removed by poly (ADP-ribose) glycohydrolase, which leads to release and inactivation of PARP-1. In cancer cells, inhibition of PARP-1 would lead to unrepaired SSBs and in proliferating cells converted to DSBs at the replication fork. Accumulation of multiple DSBs is a potent signal to initiate apoptosis. ^[196] However, the DNA damage response pathway will activate ATM or ATR (S/T kinases) to the strand breaks, with subsequent activation and recruitment of CHK1/CHK2 (S/T kinases), histone H2AX, FANCD2 (Fanconi anemia protein), BRCA1, and BRCA2, leading to cell cycle arrest and DNA repair. ^[197]

Despite the ongoing DNA repair, there is a sufficient therapeutic window to enhance apoptosis by PARP-1 inhibition alone in cancer cells that have an inherited repair deficiency (eg, BRCA1, BRCA2) or in combination with DNA damaging agents when no such deficiency exists. ^[198]

The PARP-1 NAD+ binding-site-based SMIs have entered phase I/II clinical trials. The first PARP-1 SMI to enter clinical trials was AG-014699 (Pfizer) given with temozolomide (100 and 200 mg/m²) in a phase I dose-escalation study (N = 33). The rationale for using temozolomide was that it confers BER resistance and that inhibiting PARP-1 would abrogate this effect. Patients with refractory solid tumors received escalating doses of AG-014699, with 100 mg/m²/d of temozolomide given 5 times every 28 days to establish the PARP inhibitory dose (PID). No DLTs were observed up to a dose of 12 mg/m² of AG-014699, and PID was established at 12 mg/m² based on 74-97% inhibition of peripheral blood lymphocyte PARP activity.

This regimen of AG-014699 was then evaluated in the phase II component in patients with metastatic melanoma at PID, which showed increases in DNA single-strand breaks and evidence of activity. However, due to myelosuppression, the temozolomide dose was reduced by 25% to 150 mg/m² in 12 of 40 patients. No CRs were observed, but there was an 18% PR with a doubling of TTP. ^[199] There is high interest in PARP-1 inhibition in breast and ovarian cancer driven by BRCA mutations and in GBM with temozolomide and radiation therapy.

Several PARP-1 inhibitors are now in early-phase clinical trials (BSI-201, INO-1001, KU-59436, ABT-888, GPI-21016). ^[96] BSI-201 (BiPar Sciences), a PARP-1 SMI, was evaluated in a randomized phase II study of triple negative BC patients (2 or fewer prior therapies) in combination with gemcitabine/carboplatin (G/C). Gemcitabine (1,000 mg/m²) and carboplatin (AUC = 2) were given on days 1 and 8, and BSI-201 (5.6 mg/kg IV biweekly) was given on days 1, 4, 8, and 11 every 21 days. Analyses of a planned 120 patients showed that BSI-201 plus G/C had improved ORR (48% versus 16%), CBR (62% versus 21%), median PFS (6.9 mo versus 3.3 mo), and median OS (9.2 mo versus 5.7 mo), compared with G/C alone. Non-heme and heme AEs did not differ between arms. ^[200]

Malignant cells have acquired genetic programs of invasion and metastasis that lethally compromise survival.

TGF b inhibitors

Transforming growth factor b (TGFb) is a secreted protein that governs cell growth, differentiation, migration, apoptosis, and ECM production. The family is composed of TGFb1, TGFb2, and TGFb3 and bone morphogenetic proteins (BMPs). TGFb binding to TGFbRII (receptor S/T kinase) promotes the formation of heterotetramers with TGFbRI (receptor S/T kinase), the latter of which is phosphorylated and activated by the former. Activated TGFbRI phosphorylates receptor-activated Smads (R-Smads) that complex with Smad4, which translocate to the nucleus, leading to target gene expression. A negative feedback pathway of Smad 6 and 7 inhibit the activation of R-Smads.

In normal cellular processes, TGFb inhibits cell proliferation. However, mutations within the TGFb signaling pathway promote cell proliferation by elaborating TGFb secretion in attempt to gain inhibitory control. Malignant cells secrete increased levels of TGFb that act on normal cells in the tumor mass and on immune cells, creating a fertile environment of immune tolerance, neo-angiogenesis, epithelial-to-mesenchymal transition, ECM deposition, and invasion metastasis. ^[201] Elevated levels of TGFb are associated with advanced stage and poor prognosis in many tumor types (PDA, RCC, BC, HCC, H/N, Kaposi sarcoma, MM, bladder cancer, PC, melanoma, SCLC, NSCLC). ^[202]

Therefore, inhibition of the TGFb signaling pathway with selective agents is likely to abrogate tumor-stroma interactions and prevent invasion and metastasis. Three distinct agents are in clinical development: anti-TGFb1 monoclonal antibody (GC-1008 for RCC, melanoma), TGFbRI/II S/T kinase SMI (LY2109761), and ASO (AP-12009: TGFb2 specific phosphorothioate, Antisense Pharma) directed to TGFb2. AP-12009 has been evaluated extensively in clinical trials of refractory high-grade gliomas (HGG: WHO grades III and IV, N = 145) as they overexpress TGFb2.

One study compared 2 doses of AP-12009 (10 µM = AP-10 or 80 µM = AP-80) and TMZ or PCV (in TMZ failures) in relation to safety, response rate, and survival. Patients were randomized into 3 treatment groups. AP-12009 was administered intratumorally by convection-enhanced delivery with up to 11 treatment cycles (7d-on, 7d-off/cycle). More survivors were observed in the 2 AP-12009 groups after 1.5 and 2 years than in the control group. After 12 months, CR+PR and best tumor control rate (CR+PR+SD) were observed in the AP-10 group. The study is the first specific molecular targeted agent in HGG to show efficacy equal to that of TMZ. A phase III study of AP-12009 in combination with standard antitumor therapeutics as first-line therapy is planned. ^[203]

IGF-IR inhibitors

The insulinlike growth factor (IGF) signaling axis is composed of ligands (IGF-I, IGF-II, insulin), RTKs (IGF-IR, IGF-IIR, IR) and binding proteins (IGFBP 1-4), which signal through the PI3K/Akt/mTOR and ERK/MAPK pathways. The IGF-IR is an RTK related to the insulin receptor (IR); the former regulates fetal growth and linear growth of the skeleton and other organs, while the latter regulates glucose metabolism. ^[204] Both IGF-IR and IR are mitogenic to cancer cells, implicating both receptors as targets for cancer therapy in many cancer types (BC, PC, CRC, HCC, melanoma, MM, mesothelioma, GBM, PDA). The IGF-IR is well established as a mitogen (IRS-1), but its recruitment of multiple adapter proteins, particularly IRS-2, is associated with invasion and metastasis. Increasing the IGFBP-1 levels, which bind IGF-I with high affinity, has also been shown to decrease metastasis in mice xenogratfs bearing MCF-7 breast cancers. ^[205]

Anti-IGF/IGF-IR targeted therapy in human malignancies includes monoclonal antibodies (CP-751,871, AMG-479, IMC-A12, R1507, BIIB022) and RTK SMIs (XL-228, OSI-906, NDGA). ^[206] CP-751,871 (figitumumab, Pfizer), a fully humanized IgG2 Mab to the IGF-IR that inhibits IGF-I binding to IGF-IR, decreases receptor autophosphorylation and downregulates cell surface expression by internalization. In a phase I dose-escalation study, CP-751,871 was given IV every 21 days in patients with refractory solid tumors. The MTD was not reached, but maximal dose evaluated was 20 mg/kg. The most common AEs were hyperglycemia, hyperuricemia, transaminitis, anorexia, nausea, and fatigue. With therapy, serum insulin and HGH levels increased, while IGF-IR expressing circulating tumor cells (CTCs) decreased but rebounded at the end of day 21. The best response was at 20 mg/kg, where 10 of 15 patients experienced SD. ^[207]

A phase II study evaluated carboplatin plus paclitaxel with or without CP-751,871 in patients with refractory solid tumors (N = 42), which showed 15 objective responses (RECIST), including 2 CRs in an NSCLC and ovarian carcinoma patient. ^[208] In a phase II randomized study of NSCLC patients, patients were randomized (2:1) to paclitaxel (T, 200 mg/m²), carboplatin (C, AUC=6), and CPI-751,871 (I, 10 or 20 mg/kg), or to TC alone every 3 weeks for up to 6 cycles. Forty-three of 85 patients receiving TCI (51%, p < 0.001) and 21 of 58 patients (36%) on TC alone had objective responses. Thirteen of 18 TCI patients (72%, p < 0.001) with squamous cell carcinoma responded to treatment, compared to 42% response rate with TC alone, including 6 responses in patients with bulky disease, 2 patients with no further evidence of disease, and a reversal of a superior vena caval obstruction. The hazard ratio for PFS (1.18) favors the TCI arm. ^[209]

AMG-479 (Amgen), a humanized anti-IGF-IR Mab, was also evaluated in a phase I study of 53 patients with refractory solid tumors or NHL given IV every 2 weeks. Increased levels of neutrophil IGF-IR binding and increases of serum IGF-I levels from baseline were observed in patients treated at 12 and 20 mg/kg. The maximal dose tested was 20 mg/m², with tumor responses of 1 durable CR and 1 unconfirmed PR in 2 patients with Ewing/primitive neuroectodermal tumors and 1 PR and 1 minor response in 2 patients with neuroendocrine tumors. ^[210] The agent is in phase II studies with Ewing sarcoma, NHL, and desmoplastic small round-cell tumors. Combination studies are also ongoing with gemcitabine or panitumumab.

In a phase I dose-escalation study, IMC-A12 (ImClone), an anti-IGF-IR human IgG1 Mab, was administered every other week to 16 patients with refractory solid tumors at dosage levels of 6 mg/kg (N = 5), 10 mg/kg (N = 9), and 15 mg/kg (N = 2). IMC-A12 was administered IV every other week, followed by a 2-week observation period during a 6-week cycle. AEs were grade 1 or 2, with the exception of one patient who experienced a grade 3 QTc prolongation. The best responses were SD in a patient with ovarian cancer (> 6 mo), one with thymoma (5 mo), one with lung cancer (3.5 mo), and one with prostate cancer (3.5 mo).

The half-life of IMC-A12 was 139 to 211 hours, with no detectable anti-IMC-A12 antibodies. The recommended phase-2 dose was 10 mg/kg every other week. ^[211] A phase 1 combination trial of IMC-A12 plus temsirolimus in patients with refractory solid malignancies and lymphoma is ongoing. Also, there are several phase II studies for sarcoma, prostate cancer, thymoma, and HER2+ BC. The objective is to evaluate the antitumor activity of capecitabine plus lapatinib with or without IMC-A12 in stages IIIB, IIIC, or IV that have progressed on trastuzumab-containing treatment.

The safety and pharmacokinetics of every-3-week administrations of R1507 (Genentech/Roche), a human Mab selective for IGF-1R, was explored in a phase I dose-escalation study in patients with refractory solid tumors or lymphomas. Multiple ascending doses of R1507 were administered as a 1-hour infusion every 3 weeks until the development of DLT or PD. Twenty-one patients were enrolled (dose range, 1-16 mg/kg). AE included infection (6 patients), fatigue (4 patients), rash, fever, arthralgia, cough, diarrhea, abdominal pain, and back pain. No DLTs were reported. Best activity was SD (N = 10) (median, 33 days). The half-life of approximately 8 days supports weekly dosing for future trials, at the dosage of 16 mg/kg every 3 weeks. ^[212] Early-phase clinical trials with BIIB022 (BiogenIdec) (Anti-IGF-IR Mab) in refractory solid tumors or in combination with paclitaxel plus carboplatin in patients with NSCLC are ongoing.

Small-molecule ATP-site tyrosine kinase inhibitors (XL-288 and OSI-906) have entered early-phase trials. OSI-906 (OSI) is a potent SMI of IGF-IR, an RTK activated by IGF-

I, which is overexpressed in numerous cancers and implicated in chemoresistance. Thirty-two patients have been treated at 10, 20, 40, 75, 150, and 300 mg qd and at 20, 40, and 75 mg bid. No DLTs have been observed.

At doses of OSI-96 of 10-150 mg, linear PK and median t_½ of 2.18-4.30 hours was measured. There appears to be no relationship between glucose and/or insulin level and OSI-906 plasma concentration. SD longer than 12 weeks was observed in 7 of 20 patients, including 1 patient each with thymic, adrenocortical, and colorectal cancer. Plasma concentrations of OSI-906 achieved levels required for antitumor efficacy shown in preclinical models. Since minimal toxicity was observed, further dose escalation is ongoing. ^[213]

SRC SMI

The nonreceptor tyrosine kinase (NRTK) c-SRC was the first proto-oncogene identified that has 4 functional/structural domains: SH1 protein kinase domain; SH2 phosphotyrosine binding domain; SH3 proline-rich binding domain; and SH4 membrane-associated domain. The C-terminal Tyr 530, when phopsphorylated, autoinhibits the SH1 domain, but the nonphosphorylated form interacts with the SH2/SH3 domains and activates SH1. Overexpressed c-Src in many human cancers (BC, CRC, NSCLC, PDA, ovarian, hematologic) has been shown to play an important role in mitosis, adhesion, motility, and invasive progression by activation of markers of invasion-metastasis: focal adhesion kinase (FAK), proline-rich tyrosine kinase 2 (Pyk2), Crk-associated substrate (p130Cas), and paxillin (Pax). ^[214]

Several Src TK SMIs are now FDA approved (dasatinib, bosutinib), and others are in early-phase clinical development (AZD-0530, XL999, XL228). Dasatinib is an oral Src-Abl TKI that is approved for imatinib-resistant CML and Bcr-Abl -positive ALL. It is also being evaluated as a multitargeted TKI in BC, NSCLC, CRC, and PDA at dosages of 70 mg bid and 100 mg qd. ^{[61, 215]}

Bosutinib is also an oral Src-Abl TKI that has completed a phase I study in patients with advanced solid tumors. Patients (N = 51) in cohorts of 3-6 received 50, 100, 200, 300, 400, 500 or 600 mg bosutinib orally on study day 1 and then once daily beginning on day 3. DLT of grade 3 diarrhea and grade 3 rash, with GI toxicity reported in the 500 mg dose, led to 400 mg being selected as the MTD. The half-life was approximately 17-21 hours, supporting a qd dosing regimen. SD was the best response, with 6 patients on study for more than 15 weeks (2 each with breast, colorectal cancer, NSCLC) and 1 with PDA more than 52weeks. ^[216]

Bosutinib was approved by the FDA in September 2012 for chronic-, accelerated-, or blast-phase Ph+ CML in patients resistant to or intolerant of other therapies, including imatinib. ^[217]

Approval was based on a single-arm, open-label, multicohort, phase I/II study of more than 500 patients with imatinib-resistant or -intolerant Ph+ CML. Separate cohorts were established for chronic-, accelerated-, and blast-phase CML previously treated with 1 or more prior tyrosine kinase inhibitors (ie, imatinib, imatinib followed by dasatinib and/or nilotinib). In 118 patients with chronic-phase CML, a major cytogenetic response was attained in 32% of patients, a complete cytogenetic response was attained in 24%, and a complete hematologic response was attained in 73%. At 2 years, the progression-free survival rate was 73% and the estimated overall survival rate was 83%. Responses were seen across Bcr-Abl mutations, including those associated with dasatinib and nilotinib resistance, except T3151. ^[218]

AZD-0530 (AstraZeneca) is also an oral Src-Abl that showed preclinical activity in inhibiting the phosphorylation of downstream mediators of motility and invasion [and focal adhesion kinase (FAK)], with inhibition of metastasis in vivo. Patients (N = 81) with refractory cancer received AZD0530 in doses of 50 to 250 mg qd in a 2-part phase I study.

Paired tumor biopsies were acquired for Src inhibition. Part I defined MTD, toxicity profile, and PK. Part 2 expanded the 50 mg, 125 mg, and 175 mg cohorts to evaluate modulation of phosphorylation of the Src substrates Pax and FAK by IHC. Also, markers of bone turnover were collected. Part 1: at 250 mg, DLTs occurred in 3 patients (leukopenia; septic shock with renal failure; asthenia), and at 200 mg, DLT was observed in 2 patients (febrile neutropenia; dyspnea). Part 2 confirmed that 50, 125, and 175 mg doses were tolerable. Eleven patients were on study for more than 3 months. In biopsies, Src inhibition modulated phosphorylation and/or cellular localization of tumor Pax (P = 0.067) and FAK (P = 0.002). ^[219]

AZD0530 has the potential to prevent metastases and invasion. Studies are ongoing that are evaluating safety and efficacy of AZD0530 in the following: breast cancer with metastatic bone disease; advanced ovarian cancer; recurrent osteosarcoma localized to the lung; recurrent stage IIIB or IV NSCLC previously treated with cisplatin or carboplatin; and metastatic or locally advanced gastric cancer.

XL999 (Exelixis) is a multitargeted SM TKI small-molecule inhibitor of VEGFR2, PDGFR, FGFR1, FLT-3, and SRC. Patients with refractory solid malignancies were treated with XL999 IV as a single 4-hour infusion on day 1, with PK sampling every 2 weeks in the absence of unacceptable toxicity. Twenty-three patients have been treated at 6 dose levels: 0.2, 0.4, 0.8, 1.6, 3.2, and 6.4 mg/kg (MTD).

In the XL999 study, 2 patients at 6.4 mg/kg experienced hypertension and grade 3 elevations in hepatic transaminases (1 with fatal cardiogenic pulmonary edema). At the 3.2 mg/kg, AEs were peri-infusion hypertension, perioral dyesthesias and dizziness, and grade 2 transaminase. PK analyses showed dose proportionality across all levels, with an elimination t_½ of about 24 hours (12–42 hours). In 22 evaluable patients at 2 months, there were 2 PRs, 1 minor response (28% reduction), and 4 SDs for 3–7 months. Based on safety and PK data, a weekly dosing schedule is being explored. ^[220]

XL228 (Exelixis) is an SMKI with a potent inhibitory profile against IGF-IR, SRC, Bcr-Abl (T315I), FGFR1-3, and Aurora kinases. A phase I dose-escalation trial evaluated 8 dose levels of XL228 (0.45-8.0 mg/kg) administered once or twice weekly. Of 40 evaluable patients, 1 patient with NSCLC had a confirmed PR and was on study for 48 weeks. Twelve additional patients had SD and were on study for 12 weeks or longer (SCLC, CRC, PDA, leiomyosarcoma). Three SAEs were observed (grade 3 vomiting, grade 2 hypotension and bradycardia, grade 3 diarrhea). In the once-weekly schedule, DLTs were observed in 2 of 5 patients (8.0 mg/kg) of grade 3/4 neutropenia, which established 6.5 mg/kg as the MTD.

In the twice-weekly schedule, 2 of 6 patients at the 2.7 mg/kg twice-weekly dose experienced DLTs (1 with grade 4 neutropenia; 1 with grade 3 neutropenia/grade 3 anemia/grade 2 thrombocytopenia). PD demonstrated inhibition of IGF-IR, SRC, Aurora B, and FGFR1 signaling in tumor samples from patients with SCLC and NSCLC. Analyses of peripheral blood cells, hair, and skin showed pathway inhibition post XL228. Transient modulation of glucose and insulin due to inhibition of IGF-IR and IR signaling was grade 1/2 asymptomatic hyperglycemia, which resolved within a few hours. ^[221] The study is enrolling patients to the Q1W MTD cohort, which includes patients with CRC, MM, and NSCLC.

FAK inhibitors

Focal adhesion kinase (FAK) is a multidomain NRTK that regulates cell adhesion, migration, and invasion by acting as a scaffold and signaling molecule. Growth factor and integrin receptor activation during cell adhesion recruit FAK to focal adhesions through its C-terminal focal adhesion targeting (FAT) domain. FAK is then activated by breaking the intra-molecular auto-inhibitory interaction between the N-terminal FERM (4.1, ezrin, radixin, moesin) homology and TK domains, with rapid autophosphorylation at Tyr397 in the linker between the FERM and TK domains and with consequent recruitment of Src to pTyr397 and subsequent phosphorylation of the activation loop of Src. Activated Src phosphorylates tyrosines within the C-terminus of FAK, which have docking sites for Grb2 and Cas. ^[222]

FAK is overexpressed in many tumor types (breast, prostate, colorectal, thyroid, liver, and brain), which correlates with an invasive phenotype and a poor prognostic marker. TAE226(Novartis), a bis-anilopyrimidine, is a potent ATP-site SMI of FAK, IR and IGF-IR that stabilizes an unusual α-helix conformation of the DGF motif. ^[223] The agent has good antitumor activity in mouse xenograft models of cancer (CRC, PDA, prostate) treated with bid dosing.

A phase I study evaluated 66 patients with refractory solid tumors, of which 32% had CRC. Dose escalation was performed from 25 mg bid to 225 mg qd fasting, or 100 mg bid to 150 mg bid with food. There were few grade 3 toxicities, with headache and nausea/vomiting being the DLTs. FDG-PET scans were performed at baseline and 2 weeks post TAE226. The MTD was declared at 125 mg bid, with the biologically active dose of 75 mg bid.

The best response was SD with 22% on treatment for 4 cycles or more and 17% on treatment for more than 6 cycles. ^[224] PF-562,271 (Pfizer) is a FAK/Pyk2 ATP-site SMI in phase I (N = 32: CRC, BC, NET, NSCLC, gastric cancer, SCLC, ovarian cancer). Dosage greater than 15 mg bid produced steady-state plasma levels exceeding levels of the agent required to inhibit FAK. A dose level of 105 mg bid has been reached, but MTD or recommended phase II dosing has not been achieved. The drug is well tolerated, with only 1 grade 3 AE (vomiting). The best response has been SD, except that one patient with ovarian cancer had a FDG-PET response (46% decrease) with improved symptoms. ^[225]

In human malignancies, there is an abundance of infiltrating innate immune cells (macrophages, mast cells, neutrophils) that correlate with increased immune evasion, angiogenesis, and poor prognosis. In contrast, an abundance of infiltrating lymphocytes correlates with a favorable prognosis. Early malignant tissue antigens are transported to lymphoid organs by dendritic cells (DCs) that activate adaptive immune responses, resulting in both tumor-promoting and anti-tumor effects. The pathways that regulate DC trafficking during early cancer development and the nature of the antigen(s) are tumor-type dependent.

Activation of B cells and the humoral immune response result in chronic activation of innate immune cells in neoplastic tissues. Inflammatory cells positively influence tissue remodeling and development of neoangiogenic vasculature by production of proangiogenic mediators and extracellular proteases. Tissues in which these pathways are chronically engaged exhibit an increased risk of tumor development. By contrast, activation of adaptive immunity elicits anti-tumor responses through T-cell-mediated cytotoxicity by induction of FAS, perforin, and/or cytokine pathways, in addition to antibody-dependent cell-mediated cytotoxicity (ADCC) and antibody-induced complement-mediated lysis (CDC).

Chronically activated innate immune cells can indirectly contribute to cancer development through suppression of anti-tumor adaptive immune responses, allowing tumor escape from immune surveillance. Myeloid suppressor cells, known to induce T-lymphocyte dysfunction by direct cell–cell contact and elaboration of immunosuppressive mediators, actively inhibit anti-tumor adaptive immunity. In addition, malignant lesions attract regulatory T cells (Treg), known to suppress cytotoxic T cell effector functions. Classic Treg cells are CD4+, CD25+, and FOXP3+, but different subtypes also exist. Initial investigations have revealed that in vivo depletion of Treg cells using antibodies against CD25 or denileukin diftitox enhances anti-tumor T-cell responses and induces regression of experimental tumors. Therefore, targeting defective immunity in cancer is an active area of research, which has also included vaccine-basedapproaches. ^{[226, 227]} Denileukin was discontinued from the market in January 2014.

Immunomodulatory drugs

The immunomodulatory drugs (IMiD) (thalidomide, lenalidomide, pomalidomide) appear to have multiple actions, including modulation of T cells and NK cells, adhesion, anti-angiogenesis, and direct anti-tumor activity. The agent has shown promising activity in several hematologic malignancies (MDS, MM, CLL, NHL) and is FDA approved for MDS (5q-) and MM. Combination strategies are being evaluated in MDS, AML, and MM that would further broaden the therapeutic potential of lenalidomide as a bone marrow microenvironment modulating agent. Lenalidomide is also being evaluated in refractory solid tumors (prostate cancer, NSCLC, melanoma, RCC, ovarian carcinoma).

A humanized Mab to CD200 (Alexion), an innate immune modulator, is currently in phase I studies for CLL and MM. Many vaccine trials have been conducted against targets in NHL, melanoma, prostate cancer, PDA, and ovarian cancer without a valid signal to pursue such strategies as anti-cancer agents. Despite these disappointments, targeting immune dysregulation in cancer is an active area of research, and future investigations should provide the necessary tools to fight cancer.

Bavituximab

Bavituximab (Peregrine Pharmaceuticals) is a Mab that binds to phosphatidylserine (PS), which is normally located in the inner membrane of cells but becomes exposed to the outside of the cells that line the blood vessels of tumors, creating a specific target for anti-cancer therapy. Binding of bavituximab to PS helps mobilize the body's immune system to destroy the tumor and associated blood vessels.

A phase I clinical trial of bavituximab in combination with chemotherapy in patients with refractory solid cancer indicated that about 50% of evaluable patients achieved an objective tumor response or SD after 8 weeks of treatment. The study showed that bavituximab was safe to be combined with chemotherapy and did not impact on PK properties. ^[228] Peregrine reported an update from its ongoing phase II trial of bavituximab plus docetaxel in advanced breast cancer. Enrollment of 15 patients in the 2-stage phase II study was recently completed. Of the 11 evaluable patients to date, none have experienced any measureable tumor growth or disease progression, with 5 of the 11 evaluable patients achieving a PR.

A phase I bavituximab monotherapy trial in refractory soild malignancies is now completed. Results for the first 20 patients (10 breast, 3 colorectal, 2 pancreatic, 1 each of hepatocellular carcinoma, head and neck cancer, melanoma, mesothelioma, and prostate cancer) treated at doses of 0.1 mg/kg, 0.3 mg/kg, and 1 mg/kg showed no DLTs. Common drug-related AEs were fatigue, nausea, dry skin, constipation, and dyspnea. The MTD has not been reached, and no responses have been reported. ^[229]

The genome and metabolism required to sustain cancer cell proliferation is under stress due to ongoing DNA damage and replication stress. Therefore, the stress response can be exploited to sensitize and/or overload cancer cells to propel them beyond a point of no return. ^[5] Several agents have entered clinical trials to utilize and exploit this concept.

Heat-shock response

The heat-shock pathway plays a major role in promoting protein folding and is activated in many cancer cells and is a nononcogene addicted (NOA) target. Pharmacologic evidence supports the notion that sensitizing tumor cells toward proteotoxic stress can inhibit tumorigenesis. HSP90, a chaperone involved in folding newly synthesized proteins and refolding misfolded proteins, has been targeted in cancer cells, as key client proteins (eg, CDK4, HER2, B-Raf, P53, c-Met) will remain in a destabilized form. Geldanamycin derivatives (17-AAG, 17-DMAG, IPI-504) have entered clinical trials. However, the benefit-to-toxicity ratio has precluded further development of these compounds. ^[230] Newer agents have entered early phase trials (SNX-5422, CNF2024). ^[231]

Ubiquitin-proteasome response

Unfolded or misfolded intracellular proteins (> 80%) are proteolyzed by the ubiquitin-proteasome pathway (UPP), controlling protein turnover essential to maintaining cellular homeostasis. The proteasome is a multicatalytic, macromolecular, cylindrical protease complex located within a 20S multisubunit structure consisting of 4 stacked rings arranged around an inner channel. When capped by the 19S regulatory complex at each end, 26S proteasome is formed. UPP possesses 3 enzyme functionalities (trypsin-like, chymotrypsin-like, caspase-like).

Normal cellular functions require an intact UPP that regulates the cell cycle, differentiation, angiogenesis, and DNA repair. Unfolded proteins are tagged with ubiquitin chains (ubiquitinated) that are recognized by the 19S regulatory cap, where they are de-ubiquitinated and unfolded in an ATP-dependent reaction. These unfolded proteins are then fed into the inner enzyme chamber for degradation, which generates peptides of 3 to 22 amino acids. ^[232] Since transformed cells were shown to be severalfold more susceptible to UPP inhibition than normal cells, enzyme inhibitors targeting this complex have been developed. ^{[232, 233]}

Bortezomib, a dipeptidyl boronic acid derivative, is the first proteasome inhibitor to be FDA approved for MM and mantle cell lymphoma. Bortezomib inhibits the degradation of Iκβα and downregulated NF-κβ, leading to reversal of chemoresistance and/or increasing chemotherapy sensitivity. ^[234] In fact, bortezomid is now FDA approved for combination with melphalan plus prednisone for newly diagnosed MM patients ineligible for stem cell transplantation. ^[235] Further, studies are ongoing in MCL and ABC-type DLBCL (NF-κβ is upregulated), where bortezomib is being combined with R-CHOP.

The main adverse events observed with bortezomib are activation of herpes zoster, reversible sensory and motor neuropathy, and cyclical reversible thrombocytopenia. ^[235] Other UPP targeting agents in early-phase clinical studies include CEP-18,770, RP-171, and NPI-0052.

RP-171 (PX-171, carfilzomib [CFZ]) is a novel irreversible peptide epoxy ketone inhibitor of the proteasome ^[236] that has undergone early-phase clinical trials in MM ^[237] and has now completed a phase I/II study in advanced refractory solid tumors with 2 or more prior therapies. ^[238] CFZ was administered to 14 patients IV on days 1, 2, 8, 9, 15, and 16 q28d for up to 12 cycles. Cycle 1 dosing in all cohorts was 20 mg/m², and subsequent dosing was 20 mg/m² or was escalated to 27 or 36 mg/m² in a stepped-up schedule. The 20/36 mg/m² was the MTD (grade 3 fatigue, DLT). In the phase-2 part, patients were split into 5 subgroups: SCLC, NSCLC, ovarian cancer, RCC, and other cancers. The most common AEs were fatigue, headache, diarrhea, nausea, and constipation. There were 2 PRs (RCC and SCLC), and SD was greater than 4 months in mesothelioma, ovarian cancer, RCC, andNSCLC. ^[238] An oral formulation of CFZ is also being developed.

Hypoxic and metabolic stress response

As tumors increase in size, they outstrip the vasculature and nutrient supply and are thus under stress to develop neo-angiogenesis. Hence, tumors upregulate hypoxia-inducible factor-1 (HIF-1), which induces a coordinated transcriptional program to stimulate vessel sprouting (VEGF-A, angiopoietin-2) and glycolysis (induction of glucose transporter 1, hexokinase, LDH, PDH kinase 1). HIF-1a is rapidly degraded under normoxic conditions but is upregulated in hypoxic tumors. Hence, targeting HIF-1a is thought to be an important NOA target for cancer therapy. ^[239] The success of mTOR (mammalian target of rapamycin) inhibitors in hypoxic HIF-1 driven RCC has validated HIF-1 as an anti-angiogenic therapy in cancer. However, there are no properly validated specific HIF-1a inhibitors in the clinic. Therefore, targeting mTOR should provide an alternative, as it is also dysregulated in human cancers. ^[240]

Mammalian TOR is a cytosolic S/T kinase that is the target of rapamycin (sirolimus), a macrolide immunosuppressant isolated from S. hygroscopicus. It is located downstream of Akt in the PI-3K cell survival pathway. In mammals, mTOR exists in 2 complexes: TORC1 (mTOR, raptor PRAS40, mLst8) and TORC2 (mTOR, rictor, Sin1, mLst8, mAvo3).

The TORC1 complex is regulated by cellular stress (hypoxia, metabolic, energy charge [ATP: AMP], growth factors) and controls the G1 to S phase transition of the cell cycle. It also regulates cap-dependent translation, membrane trafficking, protein degradation, ribosome biogenesis, transcription (eg, c-Myc, HIF-1a), proliferation, and survival.

TORC 2 regulates the actin cytoskeleton and activates Akt, as it is located upstream of Akt. ^[241] Rapamycin (Wyeth) is a potent mTORC1 inhibitor and is an approved drug for immune suppression in organ transplantation but not for cancer therapy, because of poor affinity for mTOR. Instead, several rapamycin ester analogs (rapalogs) with improved pharmaceutical properties have been designed (CCI-779 [temsirolimus]; RAD-001 [everolimus]; AP23573 [deforolimus]) that have been clinically evaluated. ^[242]

Temsirolimus, an mTOR inhibitor, is FDA approved for the treatment of advanced RCC. In the pivotal registration trial, the temsirolimus-alone arm achieved longer PFS and OS than the IFN-alone arm (10.9 mo vs 7.3 mo). The results of the combination arm were not different from those of the IFN-alone arm. AEs were asthenia, anemia, and dyspnea. Although temsirolimus is available in oral and IV formulation, the indication is 25 mg IV q1w. ^[243] Phase II studies, alone or in combinations, are being conducted for disease-specific sites (mantle cell lymphoma, endometrial cancer, MM), as there were responses observed in 2 phase I studies conducted in Europe (q1w) and the US (d 1-5, q2w).

Everolimus, an orally available mTOR inhibitor (10 mg qd), demonstrated antitumor activity in advanced RCC in patients who progressed on sunitinib, sorafenib, or both, in a randomized phase III trial. Based on the RECORD-1 trial, the FDA approved daily oral dosing in patients with advanced RCC. The common AEs were mucositis, lung inflammation, infection, fatigue, diarrhea, and dyspnea. ^[244] Studies are ongoing in NHL, BC, gastric cancer, NSCLC, neuroendocrine tumors, and tuberose sclerosis.

Deforolimus is an oral mTOR inhibitor that has completed phase I trials, with an MTD of 18.75 mg qd. The DLT was mucositis. The phase III study for patients with metastatic sarcoma is currently enrolling patients. A phase I study in refractory hematologic malignancies showed a few responses in MPDs. There are ongoing combination phase II studies of deforolimus plus bevacizumab targeting the vasculature, as well as deforolimus plus IGF-IR Mab inhibiting invasion in patients with refractory solid tumors.

The host microenvironment stromal response to an invasive malignancy is frequently observed in many tumor types, including CRC, BC, PDA, and hematologic malignancies (MDS, MM, CLL, MPD). The tumor-stroma interaction is characterized by a complex interplay between the normal host cells, invading tumor cells, stromal fibroblasts, inflammatory cells, proliferating endothelial cells, altered extracellular matrix (ECM), and growth factors that activate oncogenic signaling pathways by autocrine and paracrine mechanisms. Hence, the tumor microenvironment is a dynamic process promoting tumor growth, invasion through anoikis resistance, genomic instability, and drug resistance. Therefore, targeting the tumor microenvironment in the genetic context of the tumor is a potentially useful approach to cancer therapy. ^[13]

The sonic hedgehog (SHH) pathway normally regulates embryogenesis but is reactivated in a wide variety of human malignancies and appears to drive a tumor-stroma reaction that provides a growth advantage to an invasive tumor. In physiologic conditions, the HH pathway is actively repressed by the HH receptor, a patched homologue 1 (PTCH1) that inhibits smoothened (SMO) GPCR, a key activator of the pathway. However, binding of HH ligand (SHH) inhibits PTCH1, allowing SMO to become activated and drive GLI transcription factors to activate gene expression, thus influencing proliferation, survival, and differentiation of cells.

Normal HH signaling is reversible, so that in the absence of SHH, PTCH1 inhibits SMO and turns off the pathway. ^[245] In human malignancies, inhibition of the HH pathway would be of considerable therapeutic potential in cancer. The higher degree of selectivity would also mean fewer side effects chemother than with choemotherapy.

In a phase 1 study, GDC-0449, an HH pathway inhibitor, was orally administered to 33 patients with locally advanced or metastatic basal cell carcinoma, where the HH pathway is prominent. The agent showed favorable PK and PD profiles, with a median steady-state plasma concentration of 16.1 mM. The median time to reach this steady-state level was 14 days. Grade 3 AEs were fatigue, hyponatremia, weight loss, dyspnea, and QTc prolongation. The overall response rate was 55%, with a 50% response rate in patients with metastatic disease. ^[246]

The absence of significant hematologic adverse events with GDC-0449 should simplify its use in planned combination-therapy studies. The recommended phase II dose is 150 mg qd. Unfortunately, single-agent therapy with GDC-0449 results in resistance developing rapidly. The high stroma content and deficient vasculature in tumors such as PDA may lead to inefficient delivery of therapeutic agents. In a genetically engineered mouse model of PDA inhibition of the HH pathway by IPI-926 (Infinity), an oral HH inhibitor resulted in depletion of tumor stroma and increased the number of blood vessels in the tumor, thereby improving delivery of gemcitabine.

The combination of IPI-926 plus gemcitabine results in induction of apoptosis, decreased metastases, and a significant extension of survival. Unfortunately the responses are transient, and other compensatory mechanisms are elaborated with tumor progression. ^[247] Infinity has initiated a phase 1 study of IPI-926 in patients with advanced and/or metastatic solid tumors (ongoing) and is likely planning a study with gemcitabine for advanced PDA. A third HH pathway, SMI XL139 (BMS-833923/Exelexis), has entered a phase 1 trial in patients with advanced or metastatic solid tumors.

Oncoproteins have been identified that are involved in generating a malignant stromal response that promotes fibrosis or a desmoplastic reaction with an invasive-metastatic phenotype with consequent drug resistance. They include SHH, TGFb, HGF, FGF, and PDGF.

The tumor microenvironment is abundant in immune-cell-derived cytokines, chemokines, and pro-angiogenic proteins (eg, VEGF, FGF, IL-1, IL-6, tumor necrosis factor-α (TNFα), TGFβ, VEGF). ^[226] Production of VEGF and FGF is one mechanism by which tumor-infiltrating leukocytes increase angiogenesis and promote tumor development. ^[248]

Anti-TNF Mabs

TNFα is a key cytokine elaborated during acute inflammation that also mediates cancer development. As TNFα receptors are expressed on both epithelial and stromal cells, TNFα facilitates cancer development directly, by regulating the proliferation and survival of malignant cells, as well as indirectly, by exerting its effects on endothelial cells, fibroblasts, and immune cells in tumor microenvironments. ^[249] The first cytokine targeted in cancer was TNFα without much success. Mabs to TNF (etanercept or infliximab) are approved for rheumatoid arthritis. Their efficacy in human cancer has been disappointing, although there were 3 PRs seen in advanced RCC. ^[250] Clinical development of anti-TNF drugs as a cancer therapy appears to have been discontinued.

Anti-IL-6 chimeric Mab

IL-6, a glycoprotein, has a wide range of biological activity, including regulation of immune response (stimulates B cells to make antibodies, differentiation of naive CD4 T cells to Th17 cells), hematopoiesis, generation of acute-phase reactions, induction of inflammation, and oncogenesis (via the JAK-STAT pathway). ^[251]

In MM, IL-6 from bone marrow stroma binds the IL-6R on plasma cells, promoting tumor progression and drug resistance. Several approaches to inhibiting IL-6 function include inhibition of IL-6 production, inhibition of IL-6 binding to IL-6R, blocking IL-6/IL-6R complex binding to gp 130, and blocking the intracytoplasmic signal through gp 130. ^[251]

In several clinical trials of CNTO-328 (Centocor), an anti-IL-6 chimeric Mab showed efficacy in refractory MM, RCC, and B-lymphoproliferative disorders, with a decrease in C-reactive protein levels in all patients. The Mab was well tolerated, and no serious adverse effects were observed in the vast majority of studies. The fact that anti-IL-6 Mab therapy decreased the incidence of cancer-related anorexia and cachexia may also be useful in the treatment of cancer patients. ^[252] In prostate cancer, IL-6 helps tumor resist androgen deprivation therapy.

Therefore, a phase I study evaluated docetaxel (75 mg/m² q3w) in chemotherapy-naive patients with metastatic HRPC plus CNTO-328 (6 mg/kg q2w, and 9 and 12 mg/kg q3w) following an initial run-in cycle of docetaxel alone to examine the effect of CNTO-328 on PK properties. Of 29 CNTO-328 treated patients, 21 (72%) achieved 30% or greater PSA decrease within 3 months. Confirmed PSA responses of 50% or higher were observed in 16 (55%), with unconfirmed PSA declines of 75% or greater and 90% or greater in 13 (45%) and 8 (28%) patients, respectively. Two of 9 (22%) patients with measurable lesions had an unconfirmed PR. Median baseline CRP was 3.13 mg/L (range < 1 to 91.3). Of patients with at least one post-baseline value, CRP became undetectable in all 28 patients. ^[253]

A phase II study of CNTO-328 is being conducted through SWOG in patients with hormone-refractory prostate cancer. There are studies ongoing in MM combining CNTO-328 and bortezomib versus bortezomib alone. For ovarian cancer, a phase II trial of CNTO-328 alone showed 1 PR and 7 SDs in 18 evaluable patients. ^[254] Moreover, an open-label, non-randomized, dose-finding phase 1 study with CNTO-328 in patients with B-cell non-Hodgkin lymphoma, MM, or Castleman disease is ongoing. Tocilizumab (INN, or atlizumab, Roche and Chugai), a humanized Mab against the IL-6R used as an immunosuppressive drug, is likely to enter clinical trials in cancer.

Anti-RANK ligand Mab

The receptor activator of NF-κβ ligand (RANKL), its cognate receptor RANK, and its natural decoy receptor osteoprotegerin (OPG) have been identified as the final effector molecules of osteoclastic bone resorption.

RANKL signaling is required for osteoclast differentiation, activation, and survival. Furthermore, in vivo inhibition of RANKL leads to immediate osteoclast apoptosis. ^[255] Denosumab (AMG-162) is a fully human Mab directed against RANKL. The agent has been developed through a multitude of early-phase clinical trials, and has been approved by the FDA for use in women with postmenopausal osteoporosis and in patients with metastatic lytic bone lesions (MM, BC, and prostate cancer).

The broad clinical trial experience to date indicates that denosumab may be a biologic agent that can effectively inhibit bone resorption with minimal side effects. ^[256] The hope is that ongoing phase 3 clinical trials will demonstrate similar efficacy and safety profiles for denosumab therapy in metastatic cancer.

In November 2010, the FDA approved denosumab for prevention of skeletal-releated events (SREs) in patients with cancer metastases from solid tumors. SREs include bone fractures from cancer and bone pain requiring radiation.

CXCR4 antagonists

The chemokine receptor CXCR4 is widely expressed on many tumor types and is thought to be involved in cell migration and invasion. CXCR4 and its ligand CXCL12 are also important for maintaining tumor cells in close contact with the stroma. Therefore, specific targeting of the CXCR4-CXCL12 axis is considered an important therapeutic intervention in cancer. CXCR4 antagonists have the ability to release white blood cell progenitors from the bone marrow into the peripheral circulation.

Two CXCR4 antagonists are approved by the FDA for stem cell mobilization prior to auto-stem cell transplantation. Plerixafer (Mozobil) is indicated for patients with MM and NHL. ^[257]  Motixafortide (Aphexda) is indicated for patients with MM. 

Approval of motixafortide was supported by the phase 3 GENESIS trial. The primary endpoint was the proportion of patients collecting ≥6 × 10⁶ CD34+ cells/kg within 2 apheresis procedures; the secondary endpoint was to achieve this goal in 1 apheresis. The American Society for Transplantation and Cellular Therapy (ASTCT) guidelines recommend a collection target of 3-5 × 10⁶ CD34+ cells/kg. Motixafortide plus G-CSF enabled 92.5% to successfully meet the primary endpoint versus 26.2% with placebo plus G-CSF (P< .0001). Motixafortide plus G-CSF also enabled 88.8% to meet the secondary endpoint versus 9.5% with placebo plus G-CSF (P< .0001). ^[258]  

Anti-CCL2 Mab

CCL2 is a chemokine that attracts monocytes to the tumor microenvironment (tumor-associated macrophages) that promote angiogenesis, immune suppression, and metastasis through the PI3K and NF-κβ signaling. Moreover, CCL2 has direct effects on tumors by recruitment of myeloid-derived suppressor cells (MDSC) that helps tumors evade immune attack. Also, CCL2 acts as a maturation factor to osteoclasts, which promote tumor cells to metastasize to bone and continued growth advantage at these sites. ^[259] Therefore, targeting CCL2 or its receptor CCR2 (Millennium) would be a novel therapeutic approach.

CNTO-888 (Centocor) is a human IgG1κ Mab with high binding affinity for CCL2 with good preclinical antitumor activity. In a phase I dose-escalation study of refractory solid tumor patients, CNTO-888 was administered over 90 min IVI on day 1 and 28 and subsequently on a q14d schedule. Twenty-one patients received repeated CNTO-888 infusions at 5 dosage levels (0.3, 1, 3, 10, 15 mg/kg). In a 2-cohort expansion, 23 patients are being evaluated (10 mg/kg, 15 mg/kg). No DLTs were observed in dosages up to 15 mg/kg q14d. PK studies at dosages of 10 mg/kg or less showed linear kinetics and a t_½ of 4.4 to 8.7 days. Dose-dependent increases in bound CCL2 levels of greater than 1000-fold observed after treatment supports target modulation. In terms of response. 2 patients demonstrated SD greater than 6 months at 15 mg/kg (ocular melanoma and neuroendocrine tumor). An ovarian cancer patient on a dosage of 0.3mg/kg had a 50% decrease in CA125, with SD for 10 months. ^[260]

Anti-CCR4 Mab

Normal and malignant cutaneous T cells home in on the skin through interactions with dermal capillary endothelial cells. In patients with mycosis fungoides, a rare cutaneous T-cell lymphoma, cytokines released by keratinocytes promote the proclivity of the cutaneous T cell to home in on the skin, which infuse the dermis, coat the luminal surface of the dermal endothelial cells, and upregulate the adhesion molecules in the dermal capillary endothelial lumen. This action causes a reaction to CC chemokine receptor 4 (CCR4) found on cutaneous T cells.

The first CCR4-directed monoclonal antibody, mogamulizumab, was approved by the FDA in August 2018. It is indicated for adults with relapsed or refractory mycosis fungoides or Sézary syndrome after at least 1 prior systemic therapy. ^[261]

Over the last 3 decades-and more so in the last 10 years-remarkable advances have been achieved with the utilization of novel advances in technology. Myriad early-phase clinical trials with novel agents targeting oncogene and non-oncogene addicted signaling pathways have provided a wealth of biologic information. With the availability of prospective pretreatment and posttreatment tumor biopsies, gene expression profiling in combination with high-throughput DNA sequencing for detection of genetic defects, proteomics (body fluids), and imaging (FDG-PET, DCE-MRI), early-phase trials have begun to provide biologic insights into response and resistance to therapy. These studies have highlighted and continue to highlight the complexity of human tumor biology and the enormity of the task at hand, to achieve a cure for the most common types of cancer.

It is hoped that medical advances in the next 10-20 years will make great strides in achieving cures for common cancers, but there are monumental barriers to progress because of the cost of health care, insurance-company intrusion into many aspects of medical care, drug-company hidden agendas, and the unwillingness to collaborate with each other and academia because of legal barriers, lack of academic support for innovative science, and lack of funding of high-risk projects by the NIH. Moreover, a comprehensive public education of all aspects of cancer is missing, and patients and families are clearly misinformed of the benefits and pitfalls of clinical research.

Cancer can be depicted as a fatal wound, where genetic and epigenetic aberrations elaborate a medley of proteins in a tangled web of interactions that orchestrate the initiation and progression of cancer. The developmental pathways have become important targets; particularly, Notch, Hedgehog, and Wnt have been strongly implicated in cancer and stem cell progenitors. Hence, drug companies have targeted them, with inhibitors to Notch and Hedgehog already in clinical trials.

Inhibitors to Wnt have lagged behind, and no specific Wnt pathway inhibitor has entered clinical testing, even though about 80% of colorectal cancers are driven by Wnt pathway mutations. Recent studies have changed this view and have identified that CDK8 (cyclin C) orchestrates cross-talk between b-catenin and Rb signaling pathways that are frequently deregulated in human cancers. ^[262] A chemical genetic screen identified a small molecule, XAV939 (Novartis), which selectively inhibits b-catenin-mediated transcription. XAV939 stimulates b-catenin degradation by stabilizing axin by inhibiting the poly-ADP-ribosylating enzymes tankyrase 1 and tankyrase 2. Tankyrases interact with a highly conserved domain of axin and stimulate its degradation through the ubiquitin-proteasome pathway. ^[263] It is certain that Wnt inhibitors will enter clinical trials for both solid and hematologic malignancies in 2010.

It is impossible to provide a comprehensive review of the next generation of anti-cancer agents that target oncogenic (eg, Flt-3 in AML; JAK2 in MPD) and nononcogenic addiction (eg, STK33 in PDA) encompassing the 10 hallmarks of cancer within each genetic background of a heterogeneous tumor type. It is even more daunting to rationalize how these agents can or will be combined to eradicate cancer cells and their progenitors. The development of rational combinations in early-phase clinical trials is destined to be a slow process, despite the urgency to cure cancer. We are in the midst of a revolution in drug discovery and development that requires a collective effort from the science and medical community for the greater good of humanity.