Yttrium-90 Microspheres: Radiation Therapy for

Unresectable Liver Cancer

Riad Salem, MD, MBA, Kenneth G. Thurston, MA,1 Brian I. Carr, MD,1 James E. Goin, PhD,1 and
Jean-Francois H. Geschwind, MD1
Hepatocellular carcinoma (HCC) constitutes a difficult health challenge because of its poor prognosis and limited
treatment options. Most available therapies are used only for palliation. The use of yttrium-90 microspheres is a new
intraarterial therapy consisting of beta-irradiating microspheres measuring 20 –30 ␮m in diameter that can be deliv-
ered directly to the tumors. 90Y microspheres, which carry the radiation, are selectively taken up by the tumors, thus
preserving normal liver. In several studies to date, 90Y microspheres have proved to have a low toxicity profile and
have generally been well tolerated by patients. Other than transient elevation in liver enzyme levels and mild fatigue
and fever, no substantial treatment-related toxicities have been observed. Gastrointestinal toxicities occur in a limited
number of cases and are preventable with proper knowledge of visceral arterial anatomy. The effect on survival is also
promising, with median survival rates of 23 months (95% confidence interval ⴝ 14, 44) and 11 months (95% confidence
interval ⴝ 6, 26) for patients with Okuda stage I and stage II disease, respectively. On the basis of these data,
intraarterial delivery of 90Y microspheres offers a new alternative in the treatment of unresectable HCC.

Index terms: Liver neoplasms, therapeutic radiology • Microspheres • Yttrium, radioactive

J Vasc Interv Radiol 2002; 13:S223–S229

Abbreviation: HCC ⫽ hepatocellular carcinoma

HEPATOCELLULAR carcinoma into account hepatic tumor burden, as- survival rates for orthotopic liver
(HCC) represents a difficult clinical cites, jaundice, and serum albumin transplantation ranged from 20% to
challenge with regard to detection, level (1). These parameters are indirect 45%. In a recent series (12), liver trans-
treatment, and management. It usu- measures of functional hepatic reserve plantation was found to extend the
ally results in a life-limiting prognosis and the degree of cirrhosis. Depending median survival time to 59 months.
for the patient, with little possibility of on the extent to which liver function is Although surgical approaches offer a
cure. compromised, HCC is classified as potential cure for patients with HCC,
Median survival times for un- stage I, II, or III, with a worsening less than 15% are candidates for sur-
treated HCC range from less than 1 prognosis by stage. Patients present- gery because of multifocal disease, cir-
month to 18 months (1–12). Prognosis ing with Okuda stage III disease have rhosis with associated liver function
is highly dependent on the functional the worst survival outcome, with me- compromise, or chronic hepatitis (14).
status of the patient at diagnosis, as dian survival times of less than 1 Liver transplantation for HCC is lim-
indicated by Okuda stage, which takes month to 3 months reported in the ited because of the availability of do-
literature (1,5,6,8,9). nor organs, and a high recurrence rate
Treatment options for HCC include is experienced (15,16).
From the Department of Interventional Radiology, surgical and nonsurgical approaches. Nonsurgical treatments for HCC
William Beaumont Hospital, Royal Oak, Mich (R.S.); Surgical resection has resulted in me- have included both systemic and re-
DataMedix Corporation, Brookhaven, PA (K.G.T., dian survival times of 20 – 49 months gional therapies. Systemic therapies,
J.E.G.); Thomas E. Starzl Transplantation Institute,
Pittsburgh, Pa (B.I.C.); and the Russell H. Morgan
(1,4,6,10 –12). An extensive review of including chemotherapy, immuno-
Department of Radiology, the Johns Hopkins Uni- the surgical literature reported 5-year therapy, and hormonal therapy, have
versity School of Medicine, Baltimore, MD actuarial survival times of 20%–70%, met with limited success owing to
(J.F.H.G.). Received and revision requested May 5, with limited resection and noncir- compromised liver function at presen-
2002; revision received May 10; accepted May 17.
Address correspondence to J.F.H.G., Division of In- rhotic presentation resulting in longer tation, treatment-related toxicities,
terventional Radiology, The Johns Hopkins Hospi- survival (13). Patients with Okuda and general lack of therapeutic effect
tal, Blalock 545, 600 N. Wolfe St., Baltimore, MD stage I disease had a better surgical in randomized trials. Regional thera-
21287; E-mail: [email protected]
outcome than did those with stage II pies, including cryosurgery, micro-
1
These authors have identified the existence of a disease, most likely owing to inten- wave ablation, ethanol injection,
potential conflict of interest.
tional selection for lesser tumor bur- transarterial chemotherapy and embo-
© SIR, 2002 den (1,6,10 –12). Corresponding 5-year lization, transarterial chemoemboliza-

S223
S224 • Yttrium-90 Microspheres in Unresectable Liver Cancer September 2002 JVIR

tion, and conformal radiation therapy, months after treatment. Gastritis and ceiving a doses of 104 Gy or less (P ⫽
have shown varying degrees of suc- duodenal ulceration were observed in .06) and in patients with Okuda stage I
cess. Arterially directed chemotherapy several cases, with evidence of micro- disease compared with those with
delivered percutaneously via external sphere deposition in the stomach and Okuda stage II disease (P ⫽ .07). As
ports or surgically implanted pumps duodenum at histologic examination reported in the phase I studies, the
have yielded tumor response rates (biopsy) in one case, but they resolved most common side effects were transi-
ranging from 20% to 70%; however, promptly with medical therapy. Some tory nausea and fatigue. The most
these therapies are associated with of the gastrointestinal complications common serious adverse events were
substantial toxicity and morbidity ow- were attributed to erroneous catheter transitory elevation of liver enzymes
ing to the compromised liver function placement, which resulted in inadver- and bilirubin level (lasting for 1 week)
in patients with HCC. Moreover, no tent deposition of 90Y microspheres and gastric ulceration owing to inad-
survival benefit has been reported. into either the gastroduodenal or right vertent deposition of 90Y microspheres
Hepatic arterial embolization and che- gastric vessels (unappreciated at pre- into the gastric vascular bed. One pa-
moembolization have demonstrated treatment angiography). Prophylactic tient with known excessive lung
an objective tumor response ranging management of these events with an- shunting before the procedure, but for
from 17% to 58% in selected cohorts of tiulcer and protective agents was sug- whom there was no other therapeutic
patients and have been shown to pro- gested. Initial results were encourag- alternative, received 56 Gy to the
long survival only in nonrandomized ing, with most patients (60%–70%) lung and died of radiation-induced
studies (13). Promising results, with experiencing either stabilization or re- pneumonitis.
median survival times exceeding 17 duction of disease at 4 months (deter- The results of the phase II study,
months and low toxicity, have been mined with computed tomography therefore, helped confirm the findings
reported for conformal beam radia- [CT]), although disease progression of the phase I trial and showed en-
tion in combination with intraarte- was observed in some cases at the couraging results in terms of both ob-
rial hepatic 5-fluorodeoxyuridine. lower (50-Gy) dose. No patient deaths jective tumor response and patient
This focal radiation technique, in were found to be related to the treat- survival. The toxicities associated with
90
which a three-dimensional approach ment. In summary, this phase I trial Y microsphere therapy appeared to
is used instead of the broader axial showed 90Y microspheres to have a be less than those found after systemic
plane for external-beam photon tech- low toxicity profile and some thera- or intraarterial chemotherapy and
niques, may permit much higher lev- peutic benefit as an alternative to che- hepatic arterial chemoembolization
els of radiation to be delivered safely motherapy for patients with HCC. (25) (Goin et al, unpublished data).
than would be possible for whole In addition, given the favorable tox-
liver exposure (17–19). icity profile of 90Y microspheres, the
PHASE II FIXED-DOSE STUDY possibility of performing these pro-
cedures on an outpatient basis was
EARLY CLINICAL RESULTS: On the basis of the encouraging re- suggested, particularly because beta
PHASE I STUDIES sults of the initial phase I trial, a phase radiation does not require medical
II study was conducted in patients confinement of patients for radiation
90
Y microspheres (TheraSphere; with unresectable HCC (24). This time, protection.
MDS Nordion, Ottawa, Ontario, the therapeutic dose was set at 100 Gy
Canada)were first evaluated in a (administered to the whole liver in a DEVICE DESCRIPTION
phase I study to assess their safety and single infusion) because the 50-Gy
90
therapeutic benefit with a dose para- dose was associated with low toxicity Y microspheres (TheraSphere)
digm ranging from 50 to 150 Gy in during the phase I study. The purpose consist of insoluble glass microspheres
patients with unresectable HCC or of this phase II study was to determine in which the radionuclide 90Y is an
metastatic carcinoma to the liver (colo- the proportion and duration of treat- integral constituent. The microspheres
rectal, neuroendocrine, carcinoid, islet ment response and to evaluate patient have a mean diameter of 25 ␮m ⫾ 10
cell) (20 –23). In most patients enrolled survival, toxicity, and adverse effects (standard deviation), with less than
in the study, prior chemotherapy reg- after intrahepatic arterial injection of 5% smaller than 15 ␮m and less than
imens had been unsuccessful. Patients 90
Y microspheres. Of the 20 patients 10% larger than 35 ␮m (Fig 1a). Each
in whom excessive extrahepatic shunt- treated with 90Y microspheres who milligram contains between 22,000
ing (gastrointestinal or pulmonary) were evaluated for treatment efficacy, and 73,000 microspheres. The 90Y mi-
was seen on a technetium-99m macro- the median dose delivered was 104 Gy crosphere dose is supplied in 0.05 mL
aggregated albumin (MAA) study (range, 45–145 Gy). The tumor re- of sterile, pyrogen-free water con-
were excluded from treatment. Most sponse rate was 20% (four of 20 tu- tained in a 0.3-mL vial with a
patients were treated with a single in- mors), with one complete response. V-shaped bottom secured within a
fusion of 90Y microspheres to the The median duration of response was 12-mm clear polystyrene vial shield
whole liver via the proper hepatic ar- 127 weeks, the median time to pro- (Fig 1b). 90Y microspheres are avail-
tery. Other than a transient (few days gression was 44 weeks, and the me- able in six dose sizes: 3 GBq (81 mCi),
to weeks) elevation in liver enzyme dian survival was 54 weeks. There was 5 GBq (135 mCi), 7 GBq (189 mCi), 10
levels and mild fever and fatigue, no a trend for increased survival in pa- GBq (270 mCi), 15 GBq (405 mCi), and
hematologic, hepatic, or pulmonary tients receiving more than 104 Gy (the 20 GBq (540 mCi). Each dose of 90Y
toxicity was observed for up to 53 median dose) compared with those re- microspheres is supplied with an ad-
Volume 13 Number 9 Part 2 Salem et al • S225

Figure 1. (a) Micrograph (original magnification, ⫻500) shows the size of 90Y micro-
spheres compared with that of a human hair. Microsphere diameter was 15–35 ␮m. (b)
90
Y microsphere dose vial contained in a polystyrene shield.

ministration set that facilitates infu- CLINICAL AND Figure 2. Planar image obtained in the
sion of the microspheres from the dose METHODOLOGIC anterior projection after injection of 99mTc
vial (see below). CONSIDERATIONS MAA into the right hepatic artery illus-
trates pulmonary shunting. Arrows indi-
Patient Screening and Selection cate the areas of MAA accumulation within
the lungs.
RADIATION DOSIMETRY
Clinical and diagnostic evaluation.—
90
Y, a pure beta emitter, decays to Patients should be carefully evalu-
stable zirconium 90 with a physical ated before undergoing therapy with bed, three treatments, instead of the
90
half-life of 64.2 hours (2.68 days). The Y microspheres. First, the degree of usual two lobar infusions, may be
average energy of the beta emissions hepatic compromise (Okuda stage), necessary. In such cases, the activity
from 90Y is 0.9367 MeV. The average history of resection, performance sta- ordered must be reduced to account
range of the radiation in tissue is 2.5 tus (Eastern Cooperative Oncology for the relative contribution of each
mm, with a maximum range of less Group and Karnofsky scale), and im- vessel to the liver volume to be
than 1 cm. A 1-GBq (27-mCi) dose of aging findings should be reviewed. treated.
90
Y per kilogram of tissue gives an Once the patient is selected for 90Y 99mTc MAA study: pulmonary
initial radiation dose of 13 Gy (1,297 microsphere administration, physical shunt and gastrointestinal flow.—Be-
rad) per day. The mean life of 90Y is examination and laboratory studies cause arteriovenous shunting is com-
3.85 days. Thus, the radiation dose de- should be performed (complete mon in HCC, an excessive amount of
90
livered by 90Y over complete radioac- blood count with differential, platelet Y microspheres may shunt to the
tive decay starting at an activity level counts, prothrombin time and/or lungs (especially because the micro-
of 1 GBq (27 mCi) per kilogram of partial thromboplastin time, liver spheres measure only 15–35 ␮m),
tissue is 50 Gy (5,000 rad). function tests, albumin and tumor which results in radiation-induced
markers [␣-fetoprotein level]) and a pneumonitis. Therefore, assessment
medical history should be obtained. of possible shunting to the lungs is
PRINCIPLES OF OPERATION Finally, the volume of each liver lobe crucial (Fig 2). The catheter should
should be calculated because the ac- be placed in the right or left hepatic
As with other intraarterial treat- tivity of 90Y microspheres to be or- artery depending on which side will
ment modalities, 90Y microspheres are dered is determined by the volume be treated, after which 99mTc-labeled
delivered into the liver tumor via a of the liver rather than that of the MAA (2– 6 mCi [74 –222 MBq]) is ad-
catheter placed into the right or left tumor. ministered. A perfusion scan is then
hepatic artery because liver tumors are Pretreatment visceral angiography.— obtained with bedside planar or sin-
primarily fed by arterial rather than Because of the high incidence of arte- gle photon-emission CT gamma cam-
portal venous blood. 90Y microspheres rial variants in the liver, all candi- eras to reveal the degree of shunting
are unable to traverse the tumor vas- dates for intraarterial 90Y micro- to the lungs (ratio of total counts in
culature and are, therefore, embolized sphere administration must undergo both lungs divided by total hepatic
within the tumor, where they exert a pretreatment visceral angiography. plus pulmonary counts). This evalua-
local radiation therapy effect with rel- When anatomic variants are present, tion of lung shunting must be ob-
atively limited concurrent injury to treatment plans must be tailored to tained before each treatment. The
surrounding normal tissue. Because ensure safe and accurate delivery of lungs can tolerate a total cumulative
90
the microspheres are made of glass, Y microspheres. For example, if ac- dose of 30 Gy (16.5 mCi [610.5 GBq]).
they are not biodegradable and do not cessory branches (accessory right he- Thus, patients may be treated only if
redistribute to other organs of the patic branch off the superior mesen- their total pulmonary dose does not
body. teric artery) contribute to the tumor exceed 30 Gy.
S226 • Yttrium-90 Microspheres in Unresectable Liver Cancer September 2002 JVIR

Nontarget delivery of 90Y micro- spheres can safely administered. Suf-

Components of the Microsphere
spheres to the gastrointestinal tract ficient pressure should be applied to Administration System
can cause substantial morbidity, and, maintain the spheres in suspension,
therefore, patients should be care- allowing unimpeded passage Item
fully evaluated to avoid such pitfalls. through the catheter to the intended No. Item
Depending on the proximity of the target. If necessary, a 3-F coaxial sys-
1 Fluid source
left gastric artery to the other hepatic tem (0.0325 inches or larger) may be 2 Piercing pin
branches, coil embolization of this used. The use of smaller microcath- 3 Fluid line
vessel may be warranted, as this will eter systems, however, is not encour- 4 Red three-way stopcock
minimize the risk of reflux of 90Y mi- aged because there may be too much 5 Free port on the red three-way
crospheres into the gastric circula- outflow resistance to injection, pre- stopcock
tion. Every effort should also be venting adequate flow rates and par- 6 Monarch 25 syringe
made to identify the right gastric ar- ticle suspension. 7 Inlet line
tery. Although this vessel is often Administration set preparation.—The 8 Check valve
seen arising from the left hepatic ar- administration system for 90Y micro- 9 20-gauge needle at the free end of
the inlet line
tery, it can arise from the proper or spheres is comprised of a variety of
10 TheraSphere dose vial
right hepatic artery as well. Prophy- apyrogenic components: saline, 11 Acrylic vial shield
lactic coil embolization of the right sealed tubing, acrylic shields, two ra- 12 20-gauge needle at the free end of
gastric artery may be necessary to diation dosimeters and handling the outlet line
minimize inadvertent deposition of tools, three-way connection systems, 13 Outlet line
90
Y microspheres into the gastric and inlet and outlet lines. The essen- 14 Blue three-way stopcock
distribution. tial components of the administration 15 Free port on the blue three-way
system are listed in the Table. stopcock
The Monarch 25 syringe (Merit 16 Catheter
Treatment Planning Process 17 Vent line
Medical Systems, South Jordan, Vt)
18 Filter vent assembly
Dose calculation.—Cross-sectional and “red stopcock” control the flow 19 Sterile empty vial
imaging (magnetic resonance imag- of the fluid carrier to the dose vial, 20 Lead pot
ing or CT) is necessary to calculate causing the microspheres to be trans-
liver volumes, which is essential for ported through the outlet line to the
dosimetry. For maximal effect on the catheter. The “blue stopcock” permits
tumors, a dose of 100 –150 Gy should the delivery system to be purged of
air and primed with fluid in prepara- of 90Y microspheres is achieved. A
be delivered to either lobe. It is im- minimum saline flush of 100 –160 mL
portant to keep in mind that the dose tion for the infusion of microspheres.
Figure 3 illustrates the administra- is recommended with all catheter
is based on liver volume rather than systems (5 and 3 F).
tumor volume. The amount of radio- tion set configuration. It is crucial to
carefully assemble the delivery kit, as Radiation safety considerations.—Be-
activity required to deliver the de- cause of the potentially high radia-
sired dose to the liver is calculated any error during the set-up may lead
to misadministration. It is very im- tion exposure from 90Y, radiation
by using the following formula: safety is of the utmost importance.
portant to ensure that the connec-
tions to the saline bag, syringe, and Areas of increased vulnerability are
Treatment Activity 共GBq兲 ⫽
inlet and outlet lines are tight so that the eyes, skin, and hands. Beta radia-
关Desired Dose 共Gy兲兴 ⫻ the system is sealed and can main- tion from 90Y can typically travel
关Liver Volume 共mL兲兴 ⫻ tain a constant pressure. more than a meter in air but is sub-
关1.03 共 g/mL兲兴 Microsphere infusion technique.—Re- stantially reduced by less than a cen-
gardless of the catheter selection for timeter of acrylic. Although the dose
50 vial is shielded in acrylic, the outlet
the procedure, constant low pressure
The fraction of injected radioactiv- of 15–25 psi for 5-F catheters and line and catheter are not.
ity that is shunted to the lungs is not 40 – 60 psi for 3-F catheters must be Typical surface radiation dose
included in the dose-ordering calcula- maintained throughout the adminis- rates from the patient are 4 –12
tion unless this fraction is greater than tration. Actual radioactivity is moni- mrem/h. These dose ranges are well
0.1. Because 90Y microspheres are not tored as the spheres travel through within the accepted radiation levels
shipped as a unit dose, the available the blue stopcock. At times, it may for outpatient radiation treatments,
doses (3, 5, 7, 10, 15, and 20 GBq) must be necessary to slightly “tap” or “vi- and no special precaution is neces-
be decayed to the treatment activity. brate” both the inlet and outlet sides sary for either inpatients or
The maximum amount of time for de- of the blue stopcock while the ad- outpatients.
cay is 7 days from the dose calibration ministration is under way. Because of Calculation of residual activity and
date. potential spaces that exist at catheter proportional activity delivered.—After
Administration of 90Y microspheres.— connection sites, it is possible for 90Y the administration of 90Y micro-
After the catheter is accurately posi- microspheres to remain lodged, ne- spheres, the dose vial, outlet line,
tioned within the right or left hepatic cessitating a “tapping” process to re- catheter, and towels beneath the de-
arteries corresponding to the desired suspend the spheres. Flushing should livery line should be placed in a
treatment target area, the 90Y micro- be continued until optimal delivery plastic container. This container fits
Volume 13 Number 9 Part 2 Salem et al • S227

time. Long-term treatment response

(by means of imaging studies), he-
patic function, and tumor markers
are assessed at 3-month intervals.

RECENT CLINICAL
EXPERIENCE UNDER THE
HUMANITARIAN DEVICE
EXEMPTION
After the Humanitarian Device Ex-
emption was granted in 1999 for the
treatment of unresectable HCC with
90
Y microspheres, the clinical use of
90
Y microspheres was initiated at eight
institutions in the United States (un-
published data, 2002). To date, ap-
proximately 300 patients with liver
carcinoma have been treated. Each in-
stitution was required to obtain ap-
proval from its respective Institutional
Review Board before enrolling pa-
tients at the site. 90Y microspheres
were administered either as one or
more treatments to the same lobe or as
one or more treatments to both lobes.
Figure 3. Diagram illustrates the 90Y microsphere administration set. Numbers on Patients were followed up by using
diagram correspond to those in the Table. standard oncology clinical practice to
assess toxicities, adverse events, and
tumor response. So far, the data have
confirmed those reported earlier in the
into a cylindrical acrylic shield pro- tors are recommended to minimize literature. The toxicity profile remains
vided in the 90Y microsphere acces- radiation gastritis (40 mg of any anti- low, and the treatment is generally
sory kit. The residual activity in the gastric ulcer medication orally twice well tolerated. The most common side
assembly can be determined with use a day for 14 days with 1 g of antacid effects associated with 90Y micro-
of a portable ionization chamber. To orally four times a day for 14 days if sphere treatment were transient eleva-
determine the actual dose (in grays) necessary). tions in liver enzyme levels as a result
delivered to the liver after injection, After the procedure, patients are of tumor lysis, with return to baseline
the following formula is used: recovered for 6 hours and promptly within 2–3 weeks after treatment.
discharged. If the patient develops Other reported side effects included
Dose 共Gy兲 ⫽ transient, sudden onset chills, shakes, gastrointestinal symptoms—including
50 关Injected Activity 共GBq兲兴 and fever, meperidine hydrochloride abdominal pain, nausea, vomiting, an-
⫻ 关1 ⫺ F ⫺ R兴 and acetaminophen should be ad- orexia, and gastritis—and, occasion-
, ministered. Although poorly under- ally, gastric and duodenal ulcerations.
Liver Mass 共kg兲 stood, this phenomenon may be the Most of the more serious side effects
where F is the fraction of injected ra- result of tumor lysis causing the re- occurred after whole-liver administra-
dioactivity localizing in the lungs lease of various pyretics and acute tion of 90Y microspheres, which
(measured with Tc-99m MAA scintig- phase reactants, such as tumor necro- prompted a change in the protocol to
raphy) and R the fraction of injected sis factor and C-reactive protein. lobar rather than whole-liver treat-
radioactivity remaining in the dose Immediate and long-term follow- ment. Note that prophylactic therapy
vial, outlet line, and catheter (mea- up.—Follow-up consists of an initial with gastric acid inhibitors on the day
sured with the ionization chamber)— visit scheduled 14 days after treat- of treatment resulted in a substantial
typically .01–.05. ment to assess the patient’s tolerance reduction of associated gastrointesti-
to treatment, any adverse sequelae nal symptoms. The change to a lobar
Postprocedural Care and Follow-up (eg, fatigue, tumor lysis, gastrointes- treatment approach has also contrib-
tinal toxicity, or signs of radiation-in- uted to a reduction in complications
Postprocedural considerations.—The duced gastritis), and performance simply because the catheter could be
entire treatment process can be per- status. Hepatic function, hematology advanced farther into the left or right
formed on an outpatient basis. Due status, and treatment response (bio- hepatic branch, distal to the gastrodu-
to the proximity of the left lobe to logic and imaging) are evaluated at odenal or right gastric artery. This
the stomach, prophylactic antiulcer 30 days. Treatment of the second modification in the protocol has re-
medications and proton pump inhibi- lobe (if planned) is performed at that sulted in a lower rate of gastrointesti-
S228 • Yttrium-90 Microspheres in Unresectable Liver Cancer September 2002 JVIR

Figure 4. (a) Axial CT scan shows the right lobe lesion adjacent to the portal vein before 90Y microsphere administration. (b) Image
obtained 5 weeks after treatment shows that the lesion decreased significantly in size. Note the lobular retraction and central necrosis.

nal-related symptoms than that ob- decrease in ␣-fetoprotein level from for their assistance in preparing the figures
served in the previous phase I and/or 115.7 ng/mL (115.7 ␮g/L) before and reviewing the manuscript.
II studies, in which catheters were treatment to 35.3 ng/mL (35.3 ␮g/L)
placed in the common hepatic artery at 5 weeks. References
to deliver 90Y microspheres to the 1. Okuda K, Ohtsuki T, Osata H, et al.
Natural history of hepatocellular carci-
whole liver.
CONCLUSION noma and prognosis in relation to
The therapeutic benefit of 90Y mi- treatment: study of 850 patients. Can-
crospheres continues to show signs of 90
Y microsphere administration is a cer 1985; 56:918 –928.
promise and has, to date, confirmed promising new intraarterially deliv- 2. Stefanini GF, Amorati P, Biselli M, et al.
that of the two previous trials. Median ered therapy in the armamentarium Efficacy of transarterial targeted treat-
survival times for a cohort of 54 pa- against liver cancer. It can deliver up ments on survival of patients with hep-
tients with HCC were 23 months for to 15,000 rad (150 Gy) of beta radiation atocellular carcinoma: an Italian expe-
patients with Okuda stage I disease to liver tumors with minimal exposure rience. Cancer 1995; 75:2427–2434.
(95% confidence interval ⫽ 14, 44) and 3. Allgaier HP, Diebert P, Olchewski M,
to normal liver tissue. From 1986 to et al. Survival benefit of patients with
11 months for patients with Okuda 2002, more than 400 patients with liver inoperable hepatocellular carcinoma
stage II disease (95% confidence inter- carcinoma have been treated with 90Y treated by a combination of transarte-
val ⫽ 6, 26). Corresponding 1-year ac- microspheres. Results of phase I and II rial chemoembolization and percutane-
tuarial survival rates were 68% (⫾ 10% studies have demonstrated that 90Y ous ethanol injection: a single-center
[standard error]) and 37% (⫾ 11%), microspheres can be useful in stabiliz- analysis including 132 patients. Int J
respectively. These survival data are ing the disease process. 90Y micro- Cancer 1998; 79:601– 605.
consistent with those reported for pa- spheres also appear to provide an in- 4. Rose AT, Rose, DM, Pinson CW, et al.
tients with HCC undergoing surgical creased survival benefit, particularly Hepatocellular carcinoma outcomes
based on indicated treatment strategy.
resection (1,4,6,10 –12). Radiologic re- at a median dose of at least 100 Gy
Am Surg 1998; 64:1128 –1134.
sponse was more difficult to elicit, as (24). In addition, 90Y microspheres 5. Kouromalis E, Skordilis P, Thermos K,
most tumors treated local-regionally have a low toxicity profile and are Vasilaki A, Moschandrea J, Manousos
do not necessarily decrease in size. generally well tolerated by patients ON. Treatment of hepatocellular carci-
Most tumors either decreased without substantial side effects. noma with octreotide: a randomized
slightly in size or remained stable. Evidence from the collective clinical controlled study. Gut 1998; 42:442– 447.
Figure 4a and 4b illustrate such a experience provides compelling data 6. Markovic S, Gadzijev E, Stabuc B, et al.
case where the tumor size decreased that support the safety and therapeutic Treatment options in western hepato-
significantly after therapy. The pa- benefit of 90Y microspheres as an alter- cellular carcinoma: a prospective study
native treatment for unresectable liver of 224 patients. J Hepatol 1998; 29:650 –
tient presented with a diffuse lesion 659.
in the right hepatic lobe, which de- carcinoma. 7. Bruix J, Llovet JM, Castells A, et al.
creased significantly in size 5 weeks Transarterial embolization versus
after 90Y microsphere treatment. This Acknowledgments: The authors thank symptomatic treatment in patients
was correlated with a corresponding Michelle McErlane and Monica Ashbridge with advanced hepatocellular carcino-
Volume 13 Number 9 Part 2 Salem et al • S229

ma: results of a randomized, controlled eds. Cancer: principles & practice of artery floxuridine for unresectable
trial in a single institution. Hepatology oncology, 6th ed. Philadelphia, Pa: Lip- liver malignancies (abstr). Proc Am Soc
1998; 27:1578 –1583. pincott Williams & Wilkins, 2001; Clin Oncol 1999; 18(suppl);86.
8. Pawarode A, Voravud N, Sriuranpong 1162–1203. 20. Houle S, Yip TK, Shepherd FA, et al.
V, Kullavanijaya P, Patt YZ. Natural 14. Marcos-Alverez A, Jenkins RL, Wash- Hepatocellular carcinoma: pilot trial of
history of untreated primary hepato- burn K, et al. Multi-modality treatment treatment with Y-90 microspheres. Ra-
cellular carcinoma: a retrospective of hepatocellular carcinoma in a hepa- diology 1989; 172:857– 860.
study of 157 patients. Am J Clin Oncol tobiliary center. Arch Surg 1996; 131: 21. Herba MJ, Illescas FF, Thirlwell MP, et
1998; 21:386 –391. 292–298. al. Hepatic malignancies: improved
9. Sithinamsuwan P, Piratvisuth T, Ta- 15. Mazzaferro V, Regalla E, Doci R, et al. treatment with intraarterial Y-90. Radi-
nomkiat W, Apakupakul N, Tongyoo Liver transplantation for the treatment ology 1988; 169:311–314.
S. Review of 336 patients with hepato- of small hepatocellular carcinomas in 22. Andrews JC, Walker SC, Ackermann
cellular carcinoma at Songklanagarind patients with cirrhosis. N Engl J Med RJ, Cotton LA, Ensminger WD, Shapiro
Hospital. World J Gastroenterol 2000; 1996; 334:696 – 699. B. Hepatic radioembolization with
6:339 –343. 16. Selby R, Kadry Z, Carr B, Tzakis A, yttrium-90 containing glass micro-
10. Pawarode A, Tangkijvanich P, Vora- Madariaga JR, Iwatsuki S. Liver trans-
spheres: preliminary results and clini-
vud N. Outcomes of primary hepato- plantation for hepatocellular carci-
cal follow-up. J Nucl Med 1994;
cellular carcinoma treatment: an 8-year noma. World J Surg 1995; 19:53–58.
35:1637–1644.
experience with 368 patients in Thai- 17. Lawrence TS, Dworzanin LM, Walker-
land. J Gastroenterol Hepatol 2000; 15: Andrews SC, et al. Treatment of can- 23. Marn, CS, Andrews JC, Francis IR, Ho-
860 – 864. cers involving the liver and porta hepa- lett MD, Walker SC, Ensminger WD.
11. Hekele MS, Muller C, Kutilek M, Oes- tis with external beam irradiation and Hepatic parenchymal changes after in-
terreicher C, Ferenci P, Gangl A. Hep- intraarterial hepatic fluorodeoxyuri- traarterial Y-90 therapy: CT findings.
atocellular carcinoma in central Eu- dine. Int J Radiat Oncol Biol Phys 1991; Radiology 1993; 187:125–128.
rope: prognostic features and survival. 20:555–561. 24. Dancey JE, Shepherd FA, Paul K, et al.
Gut 2001; 48:103–109. 18. McGinn CJ, Ten Haken RK, Ensminger Treatment of nonresectable hepatocel-
12. Herold C, Reck T, Fischler P, et al. WD, Walker S, Wang S, Lawrence TS. lular carcinoma with intrahepatic 90Y
Prognosis of a large cohort of patients Treatment of intrahepatic cancers with microspheres. J Nucl Med 2000;
with hepatocellular carcinoma in a sin- radiation doses based on a normal tis- 41:1673–1681.
gle European centre. Liver 2002; 22:23– sue complication probability model. 25. Leung DA, Goin JE, Sickles C, Raskay
28. J Clin Oncol 1998; 16:2246 –2252. BJ, Soulen MC. Determinants of
13. Fong Y, Kemeny N, Lawrence TS. Can- 19. Dawson LA, McGinn NJ, Ensminger postembolization syndrome after he-
cer of the liver and biliary tree. In: De- W, et al. Preliminary results of esca- patic chemoembolization. J Vasc Inter-
Vita VT, Hellman S, Rosenberg SA, lated focal liver radiation and hepatic vent Radiol 2001; 12:321–326.