Int. J. Radiation Oncology Biel. Phys., Vol. 29. No. 1, pp.

17-23, 1994
Copyright 0 1994 Elswier Science Ltd
Pergamon Printed in the USA. All rights rewved
0360-3016194 $6.00 + .OO

0360-3016(93)E0120-U

@ Clinical Original Contribution

RADIATION THERAPY IN THE CONSERVATIVE TREATMENT OF CARCINOMA

OF THE ANAL CANAL

JEAN-PHILIPPE WAGNER, M.D., MARC ANDRE MAHE, M.D., PASCALE ROMESTAING, M.D.,
FRANCOIS PIERRE ROCHER, M.D., CHRISTINE BERGER, M.D.,
VERONIQUE TRILLET-LENOIR, M.D. AND JEAN-PIERRE GERARD, M.D.
Servicede Radiothérapie
Cancérologie,
CentreHospitalierLyon Sud, 69495 Pierre Benite Cedex, France

Purpose:Radiotherapy
is thestandardtreatmentofanaleanalcarcinoma.We retrospeetively analyzedourexperience
with8 patients.Specialattentionwas given in evaluating 51 patients who received concomitant chemotherapy
with SFU-CDDP.
Methods and Materials: From January 1980 to December 1989, 108 patients with anal canal carcinoma were
treated with exclusive radiotherapy at the Centre Hospitalier Lyon Sud. There were 11 men and 97 women, mean
age was 65 years (30-86). Histologie types were 94 epidermoid carcinomas, 13 basaloid carcinomas, and one
adenocarcinoma. The TNM classification (UICC 87) was: 16 Tl (14.8%), 53 Tz (49%), 33 T3 (39,5%), six T4
(5.5%), 77 N. (71,3%), 20 Ni (18.5%), nine Ni (8.3%) and two NJ (1.8%). Papillon’s radiotherapy technique with
a Cobalt direct perineal field was used in 82 patients. Ninety-six patients were treated with an interstitial 19*Ir
implant with a mean delay of 55 days after the end of the radiotherapy. In 59 patients at least one course of either
5-FU-mitomycin (8) or 5-FU-CDDP was added with at least one course concomitantly to the radiotherapy in 53
patients.
Results: A complete response in 104/108 patients (96%) was obtained 2 months after the brachytherapy. A lo-
xnal relapse (local and/or pelvic failure) was seen in 18 patients (16.6%) and inguinal node relapse in nine
(8.3%). Eight patients with locoregional recurrence and five with inguinal relapse were salvaged. A systemic failure
occured in six (5.5%) patients. Twenty-nine patients died, 16 of progressive disease. One patient died of treatment
related toxicity. The overall 5-year survival was 64% f 6 and specific survival 72% + 8. None of the patient
parameters was found to be statistically significant but there was a trend toward longer 5-year survival in Ti-T2
patients and in those with wel1 or moderately differentiated tumors. Noteworthy are the same survival rates for N.
and Ni-N3 patients (65 vs. 62%). The objective response and the locoregional failure rates were similar in the
patients treated with or without chemotherapy. The differente did not reach statistical significante though it was
important for the following parameters: overall survival rates for Ti-T2 with and without chemotherapy (94 vs.
61%) and for N1_3 patients (73% vs. 27%). The main prognostic factors in this series were differentiation (5-year
overall survival with chemotherapy 95% vs. 27% without chemotherapy p = 0.02) and the response at 3 months
after treatment initiation, before brachytherapy implant (5-year overall survival for complete responders and “very
good responders” 71% vs. 34% in partial responders p = 0.002). The complications rate was acceptable (Grade
111 9%, Grade 11 14%). Anal preservation was possible in 85% of the patients (92/108). Nine abdominoperineal
resection were performed for recurrence and seven for severe necrosis. The T3-T4 group abdomino perineal resection
was 23% while it was 9.2% of the Ti-T2 group.
Conclusion: We confirm that exclusive radiotherapy is the treatment of choice for epidermoid carcinomas of the
anal canal. The role of chemotherapy is stil1 unclear.

Anal canal carcinoma, Radiatiotherapy, Chemotherapy, Brachytherapy, Conservative treatment.

INTRODUCTION with a permanent colostomy. It is only recently following

Anal carcinoma is a rare disease with a surprisingly unex-
plained female predominante which may be related to
the presence of HPV 16 in 80% of the tumours (2, 9).
Before 1980, the standard treatment was radical surgery
differentpublications (4-7, 10, 11, 13, 15 17) that radio-
therapy is considered in most institutes to be the first line
treatment of choice. It gives the same survival rates as
radical surgery with preservation of a normal anal sphinc-
ter in most of the cases. Some controversies stil1 remain.

Reprint request to : Jean-Pierre Gerard. Accepted for publication 16 December 1993.

Acknowledgement-We thank Doctor Charles Dumontet for
reviewing this paper. This work was supported by a clinical grant
from the Ligue contre le Cancer. Comité du Rhône et de Saône
et Loire.

17
18 1. J. Radiation Oncology 0 Biology 0 Physics Volume 29. Number 1, 1994

mainly the limits of the treatment with radiotherapy alone,
the technique of irradiation and the relevante of concom-
itant chemotherapy. This paper intends to present a series
of 108 patients treated according to Papillon’s technique
(14) and to discuss these different points of controversy.
METHODS AND MATERIALS
Patients
From January 1980 to December 1989, 135 patients
with primary anal canal carcinoma (ACC) were seen at
the radiotherapy department of the University Hospital
Lyon Sud. Out of these 135 patients with cancer of the
anal canal seen during this period of time, curative and
conservative treatment with radiotherapy alone or radio-
therapy and chemotherapy was administered to 108 pa-
tients. Twenty-seven patients were not included in this
retrospective analysis for the following reasons: 11 patients
were treated on a palliative basis either for very large T4
lesions and/or metastasis or very old age, 14 patients were
treated with preoperative irradiation and abdominoperi-
neal resection for T3-T4 tumors extending more than two-
thirds of the anal canal circumference and/or perirectal
nodes larger than 2 cm, two patients were referred to the
department after primary Mile’s abdominoperineal re-
section and received postoperative irradiation. According
to the international recommendations, the anal canal was
defined as the region extending from the anal margin (an-
ocutaneous junction) to the anorectal junction. The den-
tate line in such a definition lies in the middle of the anal
canal(1, 8).
From a pathologie point of view ACC was defined fol-
lowing the WHO classification and included: squamous
cel1 carcinoma from wel1 to poorly differentiated types,
basaloid (or transitional or cloacogenic) carcinoma, mu-
coepidermoid carcinoma, and adenocarcinoma ( 12). Ini-
tial clinical work-up comprised careful digital examination
of the anus and of the anovaginal wal1 with the patient
in the knee-chest position, examination of al1lymph-nodes
areas, anuscopy. Since 1987 an endorectal ultrasonogra-
phy was perforrned. In some cases, when this examination
was painful, it was performed under genera1 anesthesia.
Computerized tomography (CT) was performed only in
cases of clinical evidente of pathologie lymph nodes. If
clinically pathologie inguinal nodes were suspected, an
excision biopsy or a fine needle aspiration were performed.
The staging of al1 the tumors has been updated to fit
with the recommendations of the UICC criteria of 1987
(8) and is based on the maximum tumor dimension. The
definition of T4 tumors is very ambiguous and many tu-
mors involving just the rectovaginal septum were not
classified as T4 tumors. Only tumors with proven vagina1
mucosal invasion were classified as T4 in this series (UICC
definition). Pretreatment characteristics of the patients are
summarized in Table 1.
Thirteen patients were < 50 years old and 32 < 60.
About two-thirds were TI-T2 tumors (63.8%). Out of 108
patients, 3 1 (28.7%) patients had nodal involvment at di-

Table 1. Pretreatment patient characteristics

Radiation Radiation and

alone chemotherapy 5-PU-MMC 5-FU-CDDP Total

Patients
Total 49 (45.3) 59 (54.6) 8 51 108
Menwomen 7~42 4:55 1.7 3.48 11:97
Age
Range 39-85 30-86 32.81 30.86 30-86
Mean 68 (* 11.4) 63 (.f 12.7) 64 62 65 (-f 12.4) t = 0.019
Histologies
Squamous 42 (85.7) 52 (88.1) 7 45 94 (87)
Wel1 differentiated 19 (38.7) 21 (35.5) 2 19 40 (37)
Moderately diff. 11 (22.4) 13 (22) 2 11 24 (22)
Poorly diK 10 (20.4) 13 (22) 1 12 23 (21.3)
Unknown 2 (4.1) 5 (8.4) 1 4 7 (6.4)
Basaloid 7 (14.4) 6 (9.9) 1 5 13 (12)
Adenocarcinoma l(l.6) 1 0 1 (0.9)
UICC stage
Tl 13 (26.5) 3 (5) 3 0 16 (14.8)~ = 0.004
TZ 22 (44.9) 31 (52.5) 2 29 53 (49)
T3 13 (26.5) 20 (33.9) 2 18 33 (30.5)
T4 1 (2) 5 (8.47) 1 4 6 (5.5)
N0 37 (75.5) 40 (67.7) 7 33 77 (71.3)
NI 8 (16.3) 14 (23.7) 1 13 22 (20.3)
NZ 3 (6.1) 4 (6.7) 4 7 (6.5)
N3 0 2 (3.3) 2 2 (1.8)
NI-3 11 (22.4) 20 (33.9) 1 19 31 (28.7)

Note: Differences between the two subgroups are stated when significant. Figures within brackets are percentage.
 RT in treatment of anal canal carcinoma 0 J.-P. WAGNER et al.
agnosis (NI-N3). Seventy-three tumors (67.3%) extended
less than one-half of the canal anal circumference and 13
(12%) extended more than two-thirds of the circumfer-
ence. The longitudinal extension in the canal was less
than 40 mm for 54/108 (50%), 40-49 mm for 14 (12.9%)
tumors or more for 39 (36%) tumors.
An endorectal ultrasonography was performed in ten
patients. Concordante between ultrasonography and
clinical staging was quite good for the primary lesion. A
differente was seen in only one patient (ultrasound Ti
and clinical T2). For pararectal lymph node staging agree-
ment was seen in six patients (five NO and Ni). In one
case a clinical Ni was assessed as N0 by ultrasonography
and in three cases clinical N0 was staged as N, by endo-
recta1 ultrasonography.
Treatment
External beam radiotherapy. Al1 the patients but one
were treated with external beam radiotherapy (EBRT).
External beam radiotherapy was delivered following Pa-
pillon’s technique (14). The patient was treated with a
6oCo unit in the lithotomy position for the perineal field
(8 X 8 cm, 80 cm SSD) and the prone position for the
sacral field. The dose to the perineal field calculated at 6
cm is 3 Gy per fraction, total dose 30 Gy in 10 fractions
and 17 days. The dose to the sacral field is 18 Gy in 6
fractions of 3 Gy. This field was treated on days when the
perineal field was not treated. Overall treatment time was
24 days.
Some slight modifications are brought to the original
Papillon technique. The arc rotation for the sacral field
is 160” and the field size is 10 X 10 cm at 80 cm SAD to
enlarge the width of the treated volume. The wedge filter
is 30” to lower the hot spot toward the upper rectum.
This technique was used in 96 patients. Modifications in
the protocol were performed in 14 patients: in six N2 pa-
tients an inguinal field was added and a postoperative
dose of 50 Gyf25 F/5 weeks was given with a mixed beam
of Cobalt and electron, two NI patients received a boost
dose of 9 Gy by a fixed sacral field, the sacral dose was
reduced in six patients (3- 15 Gy) and one patient was
treated with only the direct perineal field because of ad-
vanced age. Eleven patients were treated either by a pen-
dular posterior field or by a three- or four-field box tech-
nique and one patient was treated by exclusive brachy-
therapy, because of previous irradiation of the pelvis.
Interstitial brachytherapy. Two months after EBRT,
an interstitial brachytherapy implant was performed ac-
cording to the Lyon technique (15). Ninety-five patients
received such an implant 23 to 91 (mean 55) days after
the end of the EBRT, 12 patients did not receive any
brachytherapy implant for the following reasons: two re-
fused, two died before the brachytherapy (one toxic death
related to chemotherapy, and one intercurrent disease),
in one patient the tumor response was not good and he
underwent APR, and seven patients were treated with
exclusive EBRT usually after a primary complete local
19
excision for smal1 tumors and refered to the department
after the results of histology. Usually, five needles (four
to seven) were implanted through a perineal template and
afierloaded with 1921rwires of 5 cm length (4-7 cm). Spac-
ing between the needles is 1 cm. The mean reference dose
rate (calculated according to the Paris system) was 108
cGy/h for a linear activity of 1.8-2 mCi/cm. Fifteen to
25 Gy (mean 20.2) were delivered and thus the usual
treatment time was 1 day.
Chemotherapy. During the period 1980-1982 radio-
therapy alone was used. Since 1983, 59 patients received
as part of their treatment at least one course of chemo-
therapy. At the beginning, chemotherapy was added only
in T3-T4 tumors and progressively most of the patients
with tumors larger than T1 and in good condition received
a course of chemotherapy. In 1983- 1984 the chemother-
apy regimen was 5-fluorouracil (5-FU) and mitomycin-
C (MMC), (5-FU: 1 g/m2 DI-D4, 96 hour continous in-
fusion, MMC: 10 mg/m2 D, bolus); this regimen was used
in eight patients. After 1985, the regimen was switched
for al1 patients to 5-FU-CDDP (5-FU idem and CDDP:
25 mg/m2 D2-D5, bolus after standard hydration), and
only Ti or old and frail patients were not given 5-FU-
CDDP concomitant with the radiotherapy. Fifty-three
patients received one course concomitanly to the begin-
ning of the radiotherapy and among these patients 12
received a course of 5-FU-CDDP 3 weeks before radio-
therapy and among these 12 patients receiving neoadju-
vant chemotherapy, four N2 patients were treated with
the following protocol: groin dissection followed by im-
mediate chemotherapy and 4 weeks later radiotherapy
alone in two cases or with a second cycle of concomitant
chemotherapy in two other cases.
Surgery. Thirteen local excisions were performed. These
were excisional biopsies for T, lesions. Six patients un-
derwent a limited groin dissection for N2 tumors. A di-
verting colostomy was performed in one T4 patient that
presented with a bowel occlusion before radiotherapy.
Fellow-up assessment
The first assessment of the tumor response was made
2 months after EBRT under general anesthesia just before
the brachytherapy implant. Complete response was de-
fined as the total regression of the tumor. Partial response
was defined as a regression of more than 50% of the initial
volume of the tumor as estimated by a reduction in the
product of the two largest tumor diameters. When only
a smal1 residual fibrous mass remained, the response was
described as a “very good response.” NO biopsy was per-
formed at this time to avoid risk of necrosis.
After the end of the treatment, follow-up examinations
were scheduled every 3 months for 2 years, 4 months for
one additional year and then every 6 months. Follow-up
was calculated from the date of the first treatment to the
date of last follow-up or June 1991. Recurrences were
defined as local if occuring at the primary site, regional
if occuring in perirectal or latero-pelvic areas. Inguinal
20 1. J. Radiation Oncology 0 Biology 0 Physics Volume 29, Numbex 1, 1994

Table 2. Tumor response 2 months after the end of external beam radiotherapy

CR and very
Objective responses Complete response good responses

Total population 103/108 (95.3) 79/108 (73.4) 94/108 (87)

Without chemo. 44/49 (9.8) 34149 (69.3) 40/49 (81.3)
With chemo. 59/59 (100) 45/59 (76.2) 54/59 (9 1.3)
With 5-FU-MMC 8/8 (100) 5/8 (62) 7/8 (87)
With 5-FU-CDDP 51/51 (100) 40/5 1 (78.4) 47/51 (92.1)

node recurrences and extrapelvic metastases were recorded
separately. A patient with locoregional control was defined
as a patient free of any pelvic disease. Overall survival
rate, cancer-specific survival rate were determined by the
Kaplan-Meier method and compared with Mantel-Han-
ze1 log rank test. Differences between proportions were
tested with the chi-square analysis and between means
with the t-test. NO patient was lost to follow-up. The mean
follow-up is 35.5 months for the chemotherapy group and
5 1.7 months for the nonchemotherapy group (t = 0.003).
Toxicity of radiotherapy was defined as Grade 1 if no
treatment was performed, Grade 2 if only medical treat-
ment was given and Grade 3 if a surgical treatment had
to be performed.
RESULTS
Response to chemotherapy
Twelve patients received one course of chemotherapy
with 5-FU-CDDP before the start of radiotherapy and
ten of them were evaluable for tumor response at the third
week. Eight patients had partial response after a single
course, thus showing that anal canal carcinoma is a very
chemosensitive disease with 5-FU-CDDP.
Tumor response
Two months after the end of the external beam radio-
therapy 93.5% (103/108) of the tumors had responded
more than 50%. The complete and complete + very good
response rates were, respectively, 73% (79/108) and 87%
(94/108) of the patients (Tables 2 and 3). Two months
after brachytherapy al1 the patients were considered in
complete response except four who presented with some
smal1 residual induration which was carefully followed
every 2 months.
Locoregional and inguinal node recurrence
A locoregional recurrence was seen in 18 patients
(16.6%) and an inguinal node relapse in nine patients
(8.3%). Out of these nine inguinal relapses, three occured
in patients with abnormal inguinal nodes at first presen-
tation which were treated by inguinal dissection and ir-
radiation. In six patients the inguinal recurrence occured
in an inguinal area previously clinically normal. The lo-
coregional and inguinal recurrences occured in 50% of
the cases within the first year and only in one case (later-
opelvic relapse) after the third year. Among the 18 patients
with locoregional relapse, ten (9.8%) occured in the anal
canal, two (1.8%) in the pararectum and the anal canal,
and six in the lateropelvis (5.5%) (two with synchronous
inguinal recurrences). Out of 18 patients with locoregional
relapse, eight patients were subsequently controled by
radical APR surgery. Al1 of them were NO at the time of
initial presentation. The overall locoregional control after
salvage surgery is 98/108 (90.7%). There is no significant
relation with TNM staging: overall locoregional control
for TI-T2 66/69 (95.6%), for T3-T4: 32/39 (82%), NO:70/
77 (90.9%), N, 20/22 (90.9%), N2-Ns 8/9. Out of nine
patients with inguinal node relapse, two occured with a
synchronous lateropelvic recurrence and were not sal-
vaged. Out of seven isolated inguinal node relapse five
were long-term disease-free after inguinal dissection fol-
lowed by EBRT.
Metastasis
Systemic failure occured in six (5.5%) patients; four
patients had liver metastasis and bone metastasis were
seen in two cases. Among these six patients, one patient
presented with local relapse, and two patients with pelvic
relapse. Only one patient, at the time of last follow-up,
was in clinical remission 6 years after irradiation and che-
motherapy for a metastasis of Lg.

Table 3. CR and very good response 2 months after the end of EBRT. Correlation with T stage and chemotherapy

Without With
Population chemotherapy chemotherapy With With
(S) @) (%) 5-FU-MMC 5-FU-CDDP

TI 16/16 (100) 13/13 (100) 3/3 (100) 313 Of0

T2 44153 (83) 17/22 (77) 27/31 (87) 112 26129
T3 30/33 (90) 11/13 (89) 19/20 (9 1) 212 17/18
T4 516 011 515 l/l 4/4
 RT in treatment of anal canal carcinoma 0 J.-P. WAGNER et nl.
Causes of death
Twenty-nine patients died. Sixteen died of progressive
disease, eight from intercurrent disease. Four patients died
at home from an unknown cause, they were al1 disease-
free at last follow-up. One patient, 72-years old died of
treatment related toxicity, 3 months after one course of
neoadjuvant chemotherapy and radiochemotherapy for
a T3 N2 tumor (prolonged bone marrow hypoplasia and
progressive cachexia).
Survival
Five-year survival is 64% + 6 for the whole group of
108 patients and cancer specific survival is 72 f 8%. Sur-
viva1 has been correlated with different parameters.
Patients parameters. Age, sex, and histology were not
significantly correlated with survival. There is a trend to-
ward statistical significante in 5-year overall actuarial
survival for TI-T2 vs. T3-T4 patients (71.4% vs. 52.3 p
= 0.059). In this series there is no differente in survival
for N0 and NI__3patients (65% vs. 62%). Moreover it is
quite interesting to note that 4/9 NZ_3patients with in-
guinal nodes at presentation are stil1 alive more than 5
years after treatment. Patients with wel1 or moderately
differentiated tumors have a 75% chance of surviving at
5 years compared to 57% for the patients with poorly
differenciated tumors. This differente is not statistically
significant.
Treatment. There are no significant differences between
the two treatment groups (chemotherapy vs. nonchemo-
therapy) in the objective response rates and in the lo-
coregional failures, and no obvious differences between
patients receiving 5-FU-MMC or 5-FU-CDDP. In TI-
T2 patients the differente is important but not significant
for survival rates at 5 years (94% with chemotherapy 5-
FU-MMC + 5-FU-CDDP vs. 61% without chemother-
apy) and less important for T3-T4 (57% vs. 46%,). In TI-
T2 out of 35 patients without chemotherapy, ten experi-
enced a relapse of their cancer, only one was salvaged by
surgery and four had metastases (liver: three, bone: one).
In the group of 34 patients with chemotherapy, six ex-
perienced a relapse of their cancer (two with 5-FU-MMC
and four with 5-FU-CDDP) and three were salvaged by
subsequent surgery. Two (one with 5-PU-MMC and one
with 5-FU-CDDP) developped distant metastases (liver:
one, bone: one). Death by intercurrent disease or unknow
Table 4. Complications
Grade 2
Population 15/108 (13.8)
Without chemo 6149 (12.2)
With chemo 9/59 (15.2)
TI-TZ without chemo 3/35 (8.5)
T3-T4 without chemo 3/14 (21.4)
TI-TI with chemo* 7134 (20.6)
Tj-T4 with chemo* 2125 (8)
* Out of 8 patients with 5-FU-MMC two presented complications:
21
reason was higher in the group without chemotherapy
(six vs. three).
For N1_3 patients, the 5-years survival rates are quite
different too (73% vs. 27%) but yet not significant. Of the
12 N1_3 patients without chemotherapy, five died of re-
current cancer, two with distant metastases (liver: one,
bone: one). In the group of 19 patients with chemotherapy,
three died of recurrent cancer, none of them developped
metastatis. NO recurrence was salvaged in this group of
patient. Locoregional and inguinal recurrences did not
differ significantly between the two groups. In this series,
the main prognostic factors are the tumor differenciation
with 95% of the patients with poorly differenciated tumors
treated with chemotherapy living at 5 years vs. 27% with-
out chemotherapy (p = 0.02) and the tumor response at
3 months after treatment initiation since 7 1% of the pa-
tients having a complete or “very goed” tumor response
are alive at 5 years compared to 34% in case of partial
response (p = 0.002).
Treatment related complications
Acute eficts. Acute effects are seen in al1 patients at
the end of the treatment with some degrees of anal pain,
perianal or vulvar skin reactions. In 3-4 weeks, these acute
effects disappear and several months later they are replaced
in 2/3 of the cases by smal1 and intermittent rectal bleeding
related to the development of radiation-induced anorectal
mucosal angiodysplasia which can be controlled by cryo-
therapy or laser therapy.
Chemotherapy toxicity was usually mild but there was
one toxic death with 5-FU-CDDP as already stated. Two
patients had platelet counts under 100,000/mm3 and two
patients were treated without problems for aplasia. Thirty-
seven percent of the patients had Grade 2 or 3 vomiting.
Diarrhea, Grade 2 and stomatitis, Grade 2 were seen in
two cases; alopecia was Grade 2 in four cases and Grade
3 in three cases. One patient presented thoracic pain dur-
ing 5-FU infusion but there was no electrocardiographic
or enzymatic modification. Another patient had an arterial
thrombosis in the left leg, 15 days after the end of the 5-
FU perfusion. NO neurological or renal toxicity were seen
with CDDP.
Late complications. Grade 2 or 3 complications occured
in 25/ 108 (23%) patients (Table 4). Grade 3 complications
were seen in 9% of the patients ( 1O/ 108) and Grade 2 in
Grade 3 Total
10/108 (9.25) 25/108 (23)
6/49 (12.2) 12/49 (24.5)
4159 (6.7) 15/59 (25.4)
5/35 (14.3) 8135 (22.8)
1/14 (7.14) 4/14 (28.5)
1/34 (2.9) 8/34 (23.5)
3/25 (12) 7/25 (28)
One Grade 2 (T2) and one Grade 3 (T3).
22 1. J. Radiation Oncology 0 Biology 0 Physics
14% of the patients (151108). Most of these complications
were limited but painful necrosis usually occurring lO-
18 months after the iridium implant. Treatment with an-
tibiotics, corticoids analgesics and sometimes hyperbaric
oxygen was always attempted and surgery was necessary
in nine patients with extensive necrosis which did not
heal with medical treatment (six APR, and three diverting
colostomies, which could be closed 3-6 months later).
Two patients presented with severe proctitis (one APR,
and one blood tranfusions) and four with incapacitating
recta1 bleeding which could be treated in al1 cases with a
local treatment (infrared, cryotherapy); there was a need
of blood transfusion in one case. There were no bowel
complications in this series.
Rates of complications were higher in the T3-T4 cate-
gory but chemotherapy had no significant incidence on
the complication rates and no obvious differente was seen
between the 5-FU-MMC and 5-FU-CDDP regime. NO
pathologie variable has been found to be predictive of a
higher risk of complications. The only predictive param-
eter is treatment-related. The number of iridium sources,
the total brachytherapy treatment time and the total
brachytherapy dose are statistically correlated with oc-
curence of necrosis; the mean dose in the group with
complication is 23 Gy vs. 19 Gy in the group without
complication and the number of lines is > 5 and I 5 in
the two groups, respectively.
Anal preservation. In 85% (92/108) patients, the anus
and its function have been preserved. Among the 16 APR,
nine had to be performed for anal recurrence and seven
for very severe necrosis. Five (10.2%) and 11 (18.6%) were
performed in the group without and with chemotherapy
(not statistically different). In T3-T4 patients the rate of
APR was 23% (9/39) compared to 9.2% (7/69) in the Tl-
T2 group. These were equally performed for recurrence
and necrosis.
DISCUSSION
The results of this series, using the same technique as
Papillon’s, are comparable to the results presented in 1989
(15): 65% overall 5-year survival, 85% of anal preservation
among surviving patients, 8.3% of Grade 3-4 complica-
tions and 17.3% of anopelvic relapses. These results com-
pare favorably to those published in the literature, though
comparisons of necrosis and local control rate are some-
times difficult. Most of the institutions having a large ex-
perience with the treatment of ACC agree that APR is no
longer the standard treatment for limited TI-T2 and some
T3-T4 tumors. In these cases, radiotherapy provides sim-
ilar results with the advantage of sphincter preservation
in the majority of the patients (4-7, 10, 11, 15-17).
Many problems are stil1 waiting a definitive answer.
What is the best radiotherapy technique? Cummings
(4) states that the acute toxicity of his chemotherapy pro-
tocol may be partly due to the radiotherapy technique.
The treated volume with Papillon’s technique is rather
Volume 29, Number 1, 1994
smal1 and one must keep in mind that anal canal carci-
noma is often a very superficial disease and is thus wel1
approached through a direct perineal field. With this per-
ineal field the dose to the primary tumor is inhomoge-
neous but clinical experience ( 14) has shown that adding
a sacral field (pendular and wedged) can decrease the
pararectal recurrences from 20-5%. It is also worth re-
membering that this technique needs careful positioning
under the 6oCo unit by the physician himself and expe-
rienced radiographers used to this perineal field. To reduce
the rate of complication the dose per fraction to the per-
ineal field could be lowered to 2 Gy. The overall treatment
time wil1be prolonged and the delay before brachytherapy
could be shortened. In tumors with large pelvic lymph
nodes a three or four field technique with the patient in
the prone position can provide better coverage of the target
volume.
As far as brachytherapy is concerned, the 2-month rest
between the end of external beam irradiation and brachy-
therapy is in apparent contradiction with the proliferation
and repopulation of the cancer cells, but this delay seems
to be a good compromise between local control and risk
of necrosis. The total dose of 20 Gy can also be discussed.
When complete response is reached before brachytherapy,
15 Gy may be enough and could reduce the risk of Grade
3 necrosis. If residual disease is present at the time of
brachytherapy, an abdominoperineal resection can be
proposed instead of an implant. If brachytherapy is pre-
fered the dose can be 20-25 Gy into the initial volume
or 15 Gy plus a booster dose of 10-15 Gy on 2 or 3 lines
centering the residual disease. Keeping the dose rate of
the implant below 80 cGy/h may also be beneficial.
What are the limits of irradiation alone? In our protocol,
T3 tumors involving more than two-thirds of the anal
circumference or tumors with pelvic nodes larger than 2
cm in diameter are often treated with preoperative radio-
therapy and abdominoperineal resection. In Papillon’s
(15) series, and in our series (data not shown), 50% of the
surgical specimens are sterilized after preoperative irra-
diation. On the other hand, patients with such T3 or NI
tumors are at a higher risk of local failure and/or necrosis
when treated with irradiation alone. Thus it is very difficult
to define the best treatment strategy. One attitude could
be to propose preoperative radiotherapy to every patient
and according to the tumor response either go on with
radiotherapy in case of complete response, or perform
early surgery in the absente of complete response. In our
series, patients with complete response after external beam
irradiation have a high probability of local control and
cure, whereas two-thirds of the patients with residual dis-
ease after external beam irradiation at 2 months wil1 die
within three years due to cancer progression with a high
rate of locoregional failure. Perhaps in this situation a 3-
field technique with more protracted treatment could give
a better therapeutic ratio.
What is the role of chemotherapy? Anal canal carci-
noma is a very chemosensitive disease (3) and concomi-
 RT in treatment of anal canal carcinoma 0 J.-P. WAGNER er al.
tant chemoradiotherapy has been discussed by many au-
thors for several years now. Since the first report by Nigro
(13), Cummings (4) and Papillon (15) have reported re-
sults tending to demonstrate a benefit for concomitant
radiochemotherapy vs. radiotherapy alone as far as local
control is concemed. In our series, local control is high
and identical in the two treatment groups. There is only
a trend toward longer survival in the chemotherapy group
and the only significant result is for the patients with un-
differentiated tumors.
What is the best chemotherapy protocol? Mitomycin
plus 5-FU is often considered as the standard regimen for
cancer of the anal canal but Cummings (4) has recently
REFERENCES
1. Beahrs, 0. H.; Henson, D. E.; Hutter, R. V. P.; Myers, M. H.,
eds. American Joint Comittee on Cancer. Manual for staging
of cancer, 3rd ed. Philadelphia: Lippincott; 1988:s 1-83.
2. Beckmann, A. M.; Daling, J. R.; Sherman, K. J.; Maden,
C.; Miller, B. A.; Coates, R. J.; Kiviat, N. B.; Myerson, D.;
Weiss, N. S.; Hislop, T. G.; Beagrie, M.; MC Dougall, J. K.
Human papilloma virus infection and anal cancer. Int. J.
Cancer 43:1042-1049; 1989.
3. Brunet, R.; Becouarn, V.; Pigneux, J.; Ravaud, A.; Faivre,
J.; Jeandroz, V. Cisplatine et fluorouracile en chimiothérapie
néoadjuvante des carcinomes épidermoïdes du canal anal.
Lyon Chirurgical. 87:77-78; 199 1.
4. Cummings, B. J.; Keane, T. J.; O’Sullivan, B.; Wong, C. S.;
Catton, C. N. Epidermoid anal cancer: Treatment by ra-
diation alone or by radiation and 5 fluorouracil with and
without Mitomycin C. Int. J. Radiat. Oncol. Biol. Phys. 2 1:
1115-1125; 1991.
5. Doggett, S. W.; Green, J. P.; Cantril, S. T. Efficacy of ra-
diation therapy alone for limited squamous cel1 carcinoma
of the anal canal. Int. J. Radiat. Oncol. Biol. Phys. 15: 1069-
1072; 1988.
6. Eschwege, P.; Lasser, P.; Chavy, A.; Wibault, P.; Kac, J.;
Rougier, P.; Bognel, C. Squamous cel1 carcinoma of the
anal canal: Treatment by extemal beam irradiation. Radioth.
Oncol. 3:145-150; 1987.
7. Glimelius, B.; Pahlmann, L. Radiation therapy of anal epi-
dermoid carcinoma. Int. J. Radiat. Oncol. Biol. Phys. 13 :
305-312; 1987.
8. Hermanek, P.; Sobin, L. H., eds. Union Internationale
Contre le Cancer. TNM classification of malignant tumors.
4th ed. Berlin: Springer-Verlag; 1987:77-8 1.
9. Hill, S. A.; Coghill, S. B. Human papillomavirus in squa-
mous carcinoma of anus. Lancet 6: 1333; 1986.
23
shown a high toxicity for this association in combination
with high dose irradiation. On the other hand 5-F’U alone
does not improve the results when compared with irra-
diation alone. Brunet (3) in Bordeaux, and our group
(present study) have shown that 5-F’U-CDDP gives a high
response rate in ACC as in the majority of squamous cel1
carcinomas.
It is our impression that 5-PU-MMC and 5-FU-CDDP
in association with irradiation give comparable results.
The role of CDDP alone is stil1 open to clinical research.
Data provided by the ongoing randomized trials (EORTC,
RTOG, UK) wil1 help to answer these questions in the
near future.
10. Hughes, L. L.: Rich, T. A.; Delclos, L.; Ajani, J. A.; Martin,
R. G. Radiotherapy for anal cancer: Experience from 1979-
1987. Int. J. Radiat. Oncol. Biol. Phys. 17:1153-1160; 1989.
ll. John, M. J.; Flam, M.; Lovalvo, L.; Mowry, P. A. Feasibility
of nonsurgical definitive management of anal canal carci-
noma. Int. J. Radiat. Oncol. Biol. Phys. 13:299-303; 1987.
12. Morson, B. C.; Sobin, J. H. Histological typing of intestinal
tumours. International histological classification of tumors,
No. 15. Geneva: World Health Organization; 1976:62-63.
13. Nigro, N. D.; Vaitkevicius, V. K.; Considine, B. Combined
therapy for cancer of the anal canal. Dis. Colon Rectum
27~763-766; 1974.
14. Papillon, J.; Mayer, M.; Montbarbon, J. F.; Gérard, J. P.;
Chassard, J. L.; Bailly, C. A new approach to the manage-
ment of epidermoid carcinoma of the anal canal. Cancer
51:1830-1837; 1983.
15. Papillon, J.; Montbarbon, J. F.; Gérard, J. P.; Chassard,
J. L.; Ardiet, J. M. Interstitial curietherapy in the conser-
vative treatment of anal and recta1 cancers. Int. J. Radiat.
Oncol. Biol. Phys. 17:1161-1169; 1989.
16. Ng, Y. K.; Pigneux, J.; Auvray, H.; Brunet, R.; Thomas,
L.; Denefoux, R. Our experience of conservative treatment
of anal canal carcinoma combining external irradiation and
interstitial implant: 32 cases treated between 1973 and 1982.
Int. J. Radiat. Oncol. Biol. Phys. 14:253-259; 1988.
17. Schlienger, M.; Krzisch, C.; Pene, F.; Marin, J. L.; Gindrey-
Vie, B.; Mauban, S. Barthélémy, N.; Habrand, J. L.; Socie,
G.; Part, R.; Gallot, D.; Malafosse, M.; Laugier, A. Epider-
moid carcinoma of the anal canal. Treatment results and
prognostic variables in a series of 242 cases. Int. J. Radiat.
Oncol. Biol. Phys. 17:1141-1151: 1989.