Therapeutic Radiology

Damon E. Smith, MD #{149}Kayur H. Shah, BS Aroor

#{149} R. Rao, MD #{149}
Daniel B. Frost, MD
Frank Latino, Paul J. Anderson,
MD #{149} MD Anuj
#{149} V. Peddada, MD #{149}A. Robert Kagan, MD

Cancer ofthe Anal Canal: Treatment

with Chemotherapy and Low-Dose
Radiation Therapy’

PURPOSE: To determine the efficacy E PIDERMOID cancer of the anal canal, were
whose
excluded
treatment
from
was
the study
planned
group-one
as preop-
of a standardized protocol of chemo- although a rare malignancy, is an
therapy and low-dose radiation example of success in the modern erative therapy and one who received
chemotherapy prior to being entered into
therapy in treatment of patients with multidisciplinary management of ma-
the protocol. All other patients were in-
anal canal cancer. lignancies. Until the midseventies,
cluded. To our knowledge, this group of
MATERIALS AND METHODS: treatment was almost entirely surgical 42 patients comprised all patients in our
Forty-two consecutive patients with and involved abdominoperineal me- medical plan in the greater Los Angeles
anal canal cancer were treated with section (APR) with 5-year survival area who had nonmetastatic anal canal
15 fractions of 30-Gy photon beam mates of 60%-70% in the most recent cancer during the study period. There
reports (1,2). Radiation therapy alone were 16 men and 26 women patients in
radiation therapy administered an-
had also been used with preservation the study group. Age at diagnosis of can-
teroposterior-posteroanterior in con-
of the sphincter muscle in a majority cer was 36-80 years with a mean of 54
junction with chemotherapy with years for the men and 63 years for the
of patients and survival approxi-
5-fluorouracil and mitomycin C. Sur- women. Thirty-four patients were Cauca-
mately equal to that with surgery (3).
viva! analysis was performed with sian, three were black, two were Hispanic,
the lifetest procedure. Late severe complications, however,
two were East Indian, and one was Amen-
were a relatively common concern. In
can Indian. Five patients, all men, had
RESULTS: In patients with stage Ti 1974, Nigmo et al (4) issued a prelimi- positive test results for human immunode-
and T2 tumors, 26 of 29 (90%) were nary report that noted excellent me- ficiency virus (HIV) at the time of anal
free of disease after chemotherapy sponse to a combination of chemo- cancer diagnosis. No patient had penianal
and radiation therapy and had no therapy and radiation therapy, and skin cancer (anal margin cancer).
recurrent tumors. In patients with subsequent reports (5-11) have con- Pre-enrollment work-up for all patients
stage T3 and T4 tumors, five of 13 firmed the efficacy of this approach, consisted of a biopsy of the anal lesion,
(38%) were free of disease after making it the treatment of choice in chest radiography, complete blood cell
therapy and had no recurrences. count, and blood urea nitrogen, creati-
cancers of this region. The optimal
nine, and liver function tests. Abdomino-
CONCLUSION: This therapy is effec- sequence, timing, dose of radiation,
pelvic computed tomography (CT) was
tive for epidermoid cancers of the and dose and types of chemotherapy, performed in all patients initially, but
anal canal that are smaller than 5 cm however, are all variables that are as since CT findings did not affect treatment
regardless of nodal status. Tumors yet undetermined. The current study in the first 20 patients, it was routinely
larger than this or that invade adja- consists of a series of patients treated performed only in those patients with ad-
cent structures are not adequately in a community setting by a variety of vanced disease. Cancer staging was deter-
controlled with this protocol. physicians in a single treatment cen- mined with the TNM system described by
tem during a relatively short period of the American Joint Committee on Cancer
(12). Ti tumors were defined as those 2 cm
Index terms: Anus, neoplasms, 757.321
time. All patients had their tumors
in diameter or smaller, T2 as those larger
Anus, therapeutic radiology, 757.1299 staged and were treated and followed
than 2 cm but smaller than or equal to 5
up after treatment in a single clinic cm, T3 as those larger than 5 cm, and T4 as
Radiology 1994; 191:569-572 with a standardized protocol. those with invasion of an adjacent organ.
As not all patients underwent CT scan-
ning of the abdomen and pelvis, and as
MATERIALS AND METHODS the finding of penirectah lymph node in-
volvement was believed to be excessively
Patient Population and examiner dependent, only palpable ingui-
Study Design nal nodes were staged. Thirty-three (79%)
of the tumors were squamous cell, while
From the Departments of Radiation Oncoh- Forty-four patients with epidermoid
the remainder were cloacogenic.
ogy (DES., K.H.S., A.R.R., P.J.A., A.V.P., ARK.), (squamous cell or cloacogenic) cancer of
Twenty-nine (69%) of the tumors were
Surgery (D.B.F), and Internal Medicine (FL.), the anal canal were treated between 1987
stage Ti or T2 (Fig 1). All of the tumors
Kaiser Permanente Medical Center, 4950 Sunset and 1991 at the Southern California Kaiser
Blvd. Station 2B, Los Angeles, CA 90027. From Permanente Radiation Oncology Depart-
the 1993 RSNA scientific assembly. Received
ment. All patients were seen before enroll-
September 7, 1993; revision requested October
ment in the treatment program by a surgi-
26; revision received November 22; accepted Abbreviations: AP-PA = anteroposterior-pos-
December 6. Supported by a grant from South-
cal oncologist and radiation oncologist at teroanterior, APR = abdominoperineal resec-
em California Permanente Medical Group. Ad- the combined rectal clinic. The protocol tion, 5-FU = 5-fluorouracil, HIV = human im-
dress reprint requests to A.R.R. was designed so all treatment would be munodeficiency virus, NED = no evidence of
i RSNA, 1994 given on an outpatient basis. Two patients disease.
staged as T4 were in women, and staging RESULTS
was based on the presence of vaginal in-
vasion. Six patients had palpable inguinal At this writing, with a mean fol-
lymph nodes at diagnosis; four had stage
low-up of 33.4 months (range, 18-61
T2 disease. A cancer grade was also as-
months), 36 (86%) patients have NED,
sessed in 26 patients; all had squamous
two (5%) are alive with disease, two
cell carcinoma, which was well differenti-
ated in five (19%), moderately well differ-
(5%) died of intercurrent disease, and
entiated in six (23%), and poorly differen- two (5%) died of their disease (Fig 2).
tiated in 15 (58%). Ten (24%) patients had a recurrence
of their tumor, an average of 18
months after the protocol was com-
Radiation Therapy pleted (range, 2-52 months). Nine T2 T3 T4
Radiation therapy was performed with (21%) patients had tumor recurrence Stage of Cancerous Tumors
either a 6-MV or 15-MV photon beam hin- in the anal canal (all at the primary
Figure 1. Anal cancer. Distribution of pa-
ear accelerator with opposed anteroposte- site), and they underwent exploratory
tients by size of tumor. Numbers above bars
nior-posteroantenior (AP-PA) fields. If pal- laparotomy followed by salvage APR represent tumors in each stage.
pabhe inguinal nodes were present or if a if no distant disease was found. Seven
15-MV accelerator was used, a 1-cm bohus of these patients currently have NED
was applied to the inguinal nodes. Pa-
with an average follow-up of 22 100%
tients were treated with a total dose of
months (mange, 1-54 months). The 90%
3,000 cGy in daily doses of 200 cGy dehiv-
patient who did not have locally me- 80%
ered 5 days a week for 3 weeks. The lateral
borders of the field were the inguinal current disease had recurrence in a 70%

lymph nodes, the superior border was the supraclaviculam lymph node and ex- 60%

bottom of the sacroiliac joints, and the in- tensive recurrence in the retmopemito- 50%
fenior border was placed 2 cm below the neum; at this writing, the patient is 40%
anal verge. alive with disease 2 months after the 30%
recurrence was diagnosed. 20%
For patients with stage Ti and T2
Chemotherapy 10%
tumors, including four who presented
0%
On day I of the protocol, 1,000 mg/rn2 with palpable inguinal adenopathy, NED DID AWD DOD
of 5-fluorouracil (5-FU) was administered 26 of 29 (90%) were free of disease Oveijil curs = 90%
Rat#{149}
by continuous infusion over 24 hours, and after combined chemotherapy and Figure 2. Current status, all patients. DID =
this was continued each of the first 4 days radiation therapy and had no recur- died of intercurrent disease, AWD = alive
of radiation therapy. Also on day 1, 10 with disease, DOD died of their disease.
mence of tumors. The local failure rate =

rng/m of rnitornycin C was administered LI = APR performed, U = APR not per-

was 7%. Two of the three recurrent
with bolus injection. After the radiation formed.
therapy portion of the protocol was corn-
tumors were in the anal canal, and
pleted, patients were given a 1-week the patients were treated with APR
break, then the 4-day course of 5-FU was and currently have NED (Fig 3).
repeated. The protocol was originally de- For patients with stage T3 and T4 tion, four with neutmopenic fevers
signed to give 15 mg/rn2 of rnitornycin C, tumors, five of 13 (38%) were free of and one with pulmonary edema. No
but the dose was decreased to 10 mg/rn2 disease after chemoradiation therapy deaths were attributed to the treat-
after four of the first five patients devel- and had no recurrence. The failure ment protocol, and all hospitaliza-
oped severe neutropenia with fever.
rate of the localized therapy was 62%. tions were for 1 week or less. All five
Five patients who had recurrent tu- HIV-positive patients completed the
Follow-up moms were treated with APR and cur- protocol as planned, although four of
rently have NED, one is alive with the five required hospitalization for
Patients were seen in the clinic I month disease, and two died of disease (Fig neutropenia; this group comprised
after the second course of 5-FU. At that
4). 38% of the patients requiring hospi-
time, a biopsy was performed in all pa-
Five patients were found to have talization.
tients. If a palpable or visible mass was
present, an incisional biopsy was per-
unilateral palpable inguinal nodes at Of all the biopsies performed 1
formed. In all other patients, a needle as- initial examination, and one was month after the end of treatment,
piration of the tumor bed was performed. found to have bilateral palpable in- only one showed positive results, and
If no evidence of disease (NED) was pre- guinal nodes (three nodes were T2N2, this was a punch biopsy performed in
sent, patients were then seen at 2-3- one was T2N3, one was T3N2, and the only patient who did not com-
month intervals for the first 2 years after one was T4N2). Four of the six pa- pletely respond to treatment. Most of
treatment and every 6 months thereafter. tients were free of disease after che- the needle aspirate results were
If the biopsy results were positive, the pa- moradiation therapy and had no me- found to contain only med blood cells.
tient was referred for APR. At each visit,
cummence, one currently has NED after Acute toxicity, besides the bone
body weight and toxic side effects were
recorded for each patient, and sigmoidos-
APR, and the other had extensive me- marrow suppression and occasional
copy was performed. A chest radiograph currence in the retmopemitoneum. neutmopenic fevers, resulted in diam-
was obtained each year. In 10 (24%) patients, the protocol rhea in almost all patients (none se-
was varied. In eight,
breaks from ma- veme enough to cause a break in treat-
diation treatment of 1 day to 2 weeks ment or hospitalization), frequent
Statistical Analysis were necessary because of bone mar- nausea and occasional vomiting, and
Differences among variables were as-
row suppression. In the remaining moist skin desquamation (17 pa-
sessed by using the Fisher exact test. Sun- two, only the second course of 5-FU tients). The last side effect was sex
vivah analysis was performed with the chemotherapy was delayed. Five of dependent, occurring in 58% of the
hifetest procedure. these 10 patients required hospitaliza- women and only 13% of the men

570 Radiology
#{149} May 1994
 Effects of Combine d Chemoth erapy and Ra diation Th erapy T reatment in Patients with Anal Canal Cancer in Selected Studies
No.of
Total No. Tumor No. Severe Local Control Survival
Radiation Chemo- of Size of Late
Reference Dose (Gy) therapy Patients (cm)* Patients Reactionst Percentage Follow-up Percentage Follow-up

Leichman et al (7) 30 5-FU, MMC 45 5 35 0 (0) 89 50 mot 76 50 mo*

>5 10 70
Sischyetal(10) 41 5-FU,MMC 79 <3 26 6(8) 84 3y 85
>3 50 62 68
Cummings et al (5) 48-60 5-FU, MMC 69 5 31 19 (28) 86 5 y 76 5 y*
Tanum et al (9) 50 5-FU, MMC 106 5 34 18 (17) 93 5 y 73 5 y
>5 58 75 ‘-70
Doci (6)1 54 5-FU, MMC 56 5 39 2 (4) 72 49 mo* 81 5 y’
x3cycles >5 17 59
Current study 30 5-FU, MMC 42 5 29 0 (0) 90 31 mo 100 31 mo
> 5 13 31 31 mo 87 31 mo*
Note.-MMC = mitomycin C.
* Stage 14 tumors are considered to be > 5 cm in diameter.
t Numbers in parentheses are percentages.
t Numbers represent the median.
§ Number represents cause spedfic survival.
Prescribed regimen was given to only 50% of patients.
- # Number represents actuarial survival.

100% 100%
80% 80%

60% 60%

40% 40%
20% 20%

0% 0%
NED DID AWD DOD NED DID AWD DOD
3. 4.
Figures 3, 4. (3) Current status of patients with primary tumors 5 cm or smaller (stage Ti and T2). (4) Current status of patients with primary
tumors larger than 5 cm (stage T3 and T4). DID = died of intercurrent disease, AWD = alive with disease, DOD = died of their disease. D =
APR performed, U = APR not performed. Numbers above bars represent patients in each category.

(P < .05), and is almost certainly be- primarily involve late damage to the In the current study, the protocol
cause of multiple skin folds in the rectal area. As neither Leichman et al was varied for 10 patients; all varia-
pemineal region in women. This com- (7) nor we have noted long-term tions were either treatment breaks
plication did not require a break in problems such as rectal stenosis or during the planned course of madia-
treatment of any patient, however, as radiation proctitis, even a moderate tion therapy or delay in the delivery
the desquamation usually became increase in the dose appears to in- of chemotherapy. The causes of de-
most severe shortly after the radiation crease late effects greatly. Sischy et al lays in radiation therapy included
treatments were finished. (10) noted severe late reactions in six leukopenia, nausea, fever, diarrhea,
(8%) patients, including anal stenosis and pulmonary edema (one patient).
in one. Cummings et al (5) reported a Thirteen (31%) patients were hospi-
DISCUSSION startling 28% occurrence of severe talized at some time during, or shortly
late reactions in the combined 5-FU, after, the treatment protocol was com-
Toxicity
mitomycin C, and radiation therapy pleted, although five of these were
As the Table illustrates, most cur- group, although this may have been the first patients enrolled in the study
rent studies use a higher dose of ma- due to a somewhat larger fraction size and were hospitalized as a precaution
diation than the 30 Gy originally used (250 cGy) than other researchers used. during chemotherapy. In recent
by Nigmo et al (4). As this dose was Tanun et al (9) reported serious late years, it has been mare for a patient to
chosen empirically, it is understand- morbidity, defined as the necessity of need hospitalization for any part of
able that other regimens have been surgical intervention or the inability the treatment.
researched. It is notable, however, to live a normal social life, in 17% of
that neither local control rates nor the patients, a result that is termed
HIV Status
survival rates appear to differ signifi- “unacceptable.” Of the studies cited,
cantly on the basis of dose of madia- only Doci et al (6) noted late compli- Five (12%) patients, all men, were
tion delivered; only in the current cations in less than the 5% of patients, documented as HIV-positive on en-
study is any local control number less which is generally considered desim- mollment in the study. All completed
than 50%, and this is seen only in able in the treatment of malignancies the protocol, and none had a recur-
stage T3 and T4 tumors. The in- with radiation. It is notable that only rence of the tumors, although one
creased dosage is delivered at the ex- half of these patients received the died of intercurrent disease. Three of
pense of increased side effects, which prescribed regimen, however. the five required delays in radiation

Vrlume 191 Number

#{149} 2 Radiology 571
#{149}
therapy or chemotherapy because of vage APR. Seven of these patients 2. Boman BM, Moertel CG, O’Connell MJ, et
al. Carcinoma of the anal canal: a clinical
hematologic toxicity. Four of the five currently have NED, which is in con-
and pathologic study of 188 cases. Cancer
required hospitalization at some time trast to results in other published me- 1984; 54:114-125.
during the protocol, usually during ports (7,13-15). Eight of nine patients 3. Papihlon J, Montbarbon JF. Epidermoid
the second course of chemotherapy. with recurrent anal carcinomas who carcinoma of the anal canal: a series of 276
were treated with APR in the Henry cases. Dis Colon Rectum 1987; 30:324-333.
None experienced late effects from
4. Nigro ND, Vaitkevicius VK, Considine B.
the treatment, although these pa- Ford Hospital series (13) died an aver- Combined therapy for cancer of the anal
tients, in general, tolerated the proto- age of 20 months after APR. Their canal: a prehiminary report. Dis Colon Rec-
col more poorly than the average analysis of five separate series showed him 1974; 17:354-356.
5. Cummings BJ, Keane TJ, O’Sullivan B,
man. that 71 % (18 of 25 patients) died of
Wong CS, Catton CN. Epidermoid anal
disease within 3 years. We see no rea-
cancer: treatment by radiation alone or by
son why our results are so different, radiation and 5-fluorouracil with and with-
Node Status
other than, perhaps, the relatively out mitomycin C. mt j Radiat Oncol Biol
Six (14%) patients had inguinal short follow-up in our series and the Phys 1991; 21:1115-1125.
6. Doci R, Zucahi R, Bombelhi L, Montalto F,
lymphadenopathy at initial examina- fairly small number of patients in all
Lamonica G. Combined chemoradiation
tion. No patient was noted to have the studies cited. therapy for anal cancer. Ann Surg 1992;
perirectal or pelvic node metastases, 215:150-156.
as CT of the abdomen was not per- 7. Leichman L, Nigro N, Vaitkevicius V, et al.
CONCLUSIONS Cancer of the anal canal: model for preop-
formed in all patients and no pelvic
erative adjuvant combined modality
adenopathy was described in the me- A dose of 30 Gy in 15 fractions over therapy. Am J Med 1985; 78:211-215.
ports of the patients who underwent 3 weeks with photon beam radiation 8. Tanun C, Tveit K, Karlsen K. Chemora-
CT. No patient, therefore, had stage therapy, in conjunction with standard diotherapy of anal carcinoma: tumour re-
sponse and acute toxicity. Oncology 1993;
Ni disease, as this included only pen- doses of 5-FU and mitomycin C che-
50:14-17.
rectal lymph nodes, and the patient mothemapy, appears to be a sufficient 9. Tanun G, Tveit K, Karlsen K, Hauer-Jensen
with stage N3 disease had bilateral dose of radiation for the control of M. Chemotherapy and radiation therapy
inguinal adenopathy. In those pa- anal epidermoid tumors up to and for anal carcinoma: survival and late mor-
tients with palpable adenopathy, the including 5 cm in size. Although fol- bidity. Cancer 1991; 67:2462-2466.
10. Sischy B, Doggett RLS, KrallJM, et al. De-
only variation in the radiation treat- low-up at this time is still relatively
finitive irradiation and chemotherapy for
ment was the placement of a bolus short, most of these tumors recurred radiosensitization in management of anal
over the inguinal area. No boost was within 2 years after treatment in most carcinoma: interim report on Radiation
applied in this area. studies (including this one) where the Therapy Oncology Group study no. 8314.
Natl Cancer Inst 1989; 81:850-856.
combined modality treatment is used.
11. Nigro ND, Vaitkevicius VK, Buroker 1,
At this writing, the median follow-up Bradley GT, Considine B. Combined ther-
Prognostic Factors
in the current study is 31 months with apy for cancer of the anal canal. Dis Colon
Only tumor size, 5 cm or smaller a minimum of 18 months. A 30-Gy Rectum 1981; 24:73-75.
12. Beahrs OH, Henson DE, Hutter RVP,
versus larger than 5 cm, was found to dose of radiation delivers high local
Kennedy BJ, eds. Manual for staging of
be important in the prognosis of the control (90%) and cause specific sum- cancer. 4th ed. Philadelphia, Pa: Lippin-
disease, and this was statistically sig- vival (100%) in small and medium- cott, 1993; 103-106.
nificant for survival (P < .05), as was sized tumors, which account for the 13. Zelnick RS, Haas PA, Ajlouni M, Szilagyi E,
Fox TA. Results of abdominoperineal re-
the patient being alive without dis- majority of tumors when first diag-
sections for failures after combination che-
ease and with preservation of sphinc- nosed. A somewhat higher, as yet un- motherapy and radiation therapy for anal
ter muscles (P < .01). In contrast to determined, dose appears necessary canal cancers. Dis Colon Rectum 1992; 35:
other studies, sex was not a factor for for control of larger stage T3 and T4 574-578.
local control of cancer or survival of tumors. It may also be that the current 14. Habr-Gama A, da Silva e Sousa AH Jr,
Nadahin W, Gansh R, da Sihva JH, Pinotte
the patient in this analysis, although results reflect the small number of
HW. Epidermoid carcinoma of the anal
sex was a factor in the moist desqua- patients studied. In view of the very canal: results of treatment by combined
mation response to the treatment. high occurrence of severe late compli- chemotherapy and radiation therapy. Dis
cations in most studies, we propose Colon Rectum 1989; 32:773-777.
15. Meeker WR Jr, Sickhe-Santanehlo BJ, Phil-
that future randomized trials incorpo-
Treatment of Recurrent Disease potts G, et ah. Combined chemotherapy,
mate a low-dose (30 Gy) arm for stage radiation and surgery for epithehial cancer
At recurrence, nine of the 10 pa- Ti and T2 tumors so that the optimal of the anal canal. Cancer 1986; 57:525-529.
tients were thought to have only local dose may be determined. U 16. Michaehson PA, Magihl GB, Quan SH, et al.
Preoperative chemotherapy and radiation
recurrence after repeat staging work-
therapy in the management of anal epider-
up. All underwent exploratory lapa- Acknowledgment: The authors thank Dennis
moid carcinoma. Cancer 1983; 51:390-395.
rotomy; one patient had palpable Aguinaga, RU, for his indispensable assistance
with this project.
liver metastases and the incision was
closed with no resection performed,
References
while the other eight underwent sal-
1. Frost DB, Richards PC, Montague ED, Gi-
acco GG, Martin RG. Epidermoid cancer
of the anorectum. Cancer i984; 53:1285-
1293.

572 Radiology
#{149} May 1994