- Email: [email protected]
Postoperative radiation therapy of rectal cancer
Im. J. Radiafron Oncology Bid. Phys.. Vol. 13, pp. S-10 Printed in the U.S.A. All rights reserved.
Copyright
0360-3016/87 0 19R7 Pergamon
$3.00 + .OO Journals Ltd.
??Original Contribution
POSTOPERATIVE
RADIATION THERAPY OF RECTAL CANCER
JOEL E. TEPPER, M.D.,* ALFRED M. COHEN, M.D.,? WILLIAM C. WOOD, M.D.,? ERICA L. ORLOW* AND STEPHEN E. HEDBERG, M.D.? Massachusetts General Hospital, MGH Cancer Center, Boston, MA 02 114 Beginning in December 1975, at the Massachusetts General Hospital (MGH) patients with rectal carcinomas thought to be at high risk of local recurrence after potentially curative surgical resection, were entered on a treatment protocol of high dose postoperative radiation therapy. Treatment was given with X rays of 10 MeV, generally using a four-field box technique to a dose of 4500 cGy with a boost to 5040 cGy or higher when the small bowel could be excluded from the reduced field. One-hundred sixty-five patients who began their radiation therapy between December 1975 and December 1982 were entered into the study. The median age was 65 years. The median follow-up in the survivors was 56 months, with a minimum follow-up of 17 months. All but 10 patients were followed for more than 2 years. Of the entire group, the actuarial 5-year survival was 53%, with survival of 71% in patients with Stage B-2,39% in Stage C-2, and 17% in Stage C-3. Local failure was seen in 5/53 patients with Stage B-2 disease and O/7 of patients with Stage B-3 disease. In patients with positive lymph nodes, local failure occurred in 2/10 (20%) of patients with Stage C-l, 16/77 (21%) of Stage C-2, and 8/15 (53%) of patients with Stage C-3 disease. Compared to previous series of surgery alone, the local failure rate has been decreased by more than one-half in all patients, except those with Stage C-3. Efforts to maximize the radiation doses in all stages should be made to minimize local failure. For Stage C-3, newer strategies such as intraoperative radiation therapy should be employed to decrease the continuing high incidence of failures.
INTRODUCIION
positive lymph nodes, local failure is Seen in approximately 50%. Local failure produces major morbidity in a high percentage of these patients.6 In many patients, the local failure site is the major symptom-producing area after surgical resection. Because of the high incidence of local failure in certain patient subsets, a number of institutions have evaluated the usefulness of radiation therapy given as an adjuvant to surgical resection. Initial studies emphasized low dose preoperative radiation therapy,‘5,‘6,‘8 but now many institutions use radiation therapy delivered primarily postoperatively. The radiation therapy has been used with two objectives: to decrease the incidence of local failure, and to improve the long-term survival rate. Beginning in 1975, the MGH began a prospective evaluation of the use of high dose postoperative radiation therapy after either low-anterior resection or abdominalperineal resection in patients with rectal carcinoma. This report details the local control and survival results in a group of 165 patients treated with a consistent approach and technique with a median follow-up of almost 5 years.
For many years, single modality therapy by surgical resection was the treatment approach for rectal carcinoma. This has produced a high cure rate in patients with small localized tumors, but for patients with extensive local disease or with spread to regional lymph nodes, the cure rate has not been as high. Surgical series have shown 5 year disease-free survival rates of approximately 50% in patients with Stage B disease (invasive into or through the bowel wall, node negative) and approximately 30% in patients with Stage C disease (node positive). A number of series have evaluated recurrence patterns after potentially curative resections in patients with rectal cancer. These studies, which have been generated from institutions including the University of Florida,2 University of Minnesota,7 Maine Medical Center,6 and the Massachusetts General Hospital,‘4 indicate that for patients with Stage A or B, disease, the pelvic failure rate is less than 10%. For patients with Stage B2 disease, local failure is approximately 25%, and it is approximately 50% for patients with Stage B3 disease. For patients with
* Radiation Medicine Service. 7 Surgical Service.
Reprint requests to: Joel E. Tepper. Accepted for publication 13 August 1986. 5
I. J. Radiation Oncology 0 Biology 0 Physics
METHODS
AND MATERIALS
Patients included in the analysis were a consecutive group of 165 patients treated with a course of postoperative radiation after a potentially curative surgical resection. The resection was done at the MGH in slightly less than half of the patients (79 patients). Eighty-six patients had their primary operative procedure done at an outside hospital. Patients were treated with a variety of operative procedures: 93 with abdominal-perineal resection (APR), 66 with low-anterior resection, 4 with trans-sacral resection, and 2 with posterior exenteration. To be eligible for the study, the patients had no evidence of metastatic disease at the time of the operative procedure, and no evidence of residual disease after resection. During the same time period, other patients were treated with a high-dose preoperative radiation therapy regimen when their disease was locally advanced at presentation and were not thought to be amenable to primary surgical resection. These patients have not been included in this analysis. The radiation therapy technique was a high dose regimen and was relatively uniform during this time period. Treatments were generally given through a 4-field box technique although some patients were treated with a 3field technique including right and left laterals and a posterior field. The majority of patients were treated with a 10 MeV X ray beam from a linear accelerator*. Patients were treated in the prone position. The initial portion of the radiation treatment was given to a field encompassing the posterior portion of the entire pelvis. When an abdominal-perineal resection had been performed, the fields always included the perineal scar in the entirety and bolus was used on the perineal incision. For patients who had a low-anterior resection, the lower portion of the field extended approximately 4 to 5 cm below the level of the anastomosis. The superior border of the field generally extended to the mid portion of the L5 vertebral body. Laterally, the fields extended 1 to 1l/2 cm lateral to the true bony pelvis. To treat the entire pre-sacral space with adequate margins and full dose, the lateral fields were designed so that the posterior border encompassed the entire sacrum with a i/z to 1 cm margin posterior to the sacrum. This was necessary because the distance from the 50% isodose line to the 90% isodose line can be in the range of I’/2 centimeters when treating with lateral fields. Anteriorly, the field was not designed to include the external iliac lymph nodes but was designed to encompass the previous tumor bed and to encompass the tissues that might potentially have been infiltrated by tumor anteriorly. For females, the posterior wall of the vagina was always included in the radiation field and for males a large portion of the prostate was included. Bowel
* Clinac- 18.
January 1987. Volume 13, Number I
located anteriorly and a portion of the bladder were spared in the lateral ports. After a dose of 4,500 cGy had been given to the large field, the field size was reduced to encompass the areas thought to be at highest risk for residual disease. This area was boosted for an additional 540 cGy, but in patients who were thought to have very high risk for residual disease (B3, Cz and C3 tumors) a higher dose was used if small bowel could safely be moved from the field. In treatment planning for the boost field, patients had a special small bowel X ray series to determine the exact location of the small intestine and possible fixation of the small intestine in the region of the planned boost. Attempts were made to use bladder distention or other patient positioning devices to move the small bowel out of the radiation field. If this could not be done safely, the dose was limited to 5,040 cGy and for some patients with B2 tumors with a large amount of small bowel in the pelvis, the dose was stopped at 4,500 cGy. The surgeons at the MGH were encouraged to reperitonealize the pelvic floor so that a large amount of small bowel would not be located in the radiation boost field. Patients were not generally treated with adjuvant chemotherapy although 14 patients did have some additional therapy, generally with 5-fluorouracil alone. Patients were staged using the Gunderson-Sosin modification of the Dukes’ Staging System. Patients with Stage B disease had negative lymph nodes with Stage B, disease limited to the bowel wall, B2 with extension through the bowel wall into the perirectal fat, and Stage B3 having tumor adherence or fixation to adjacent normal structure. Stage C disease patients had positive lymph nodes with the subscripts defining the extent of local penetration as for Stage B. Data analysis was done based on patient status information available in March 1985. Survival curves were calculated using the actuarial method of KaplanMeier,3,9 unless 15 or more failures occurred in a group, when life table methods were used. Local control actuarial curves were plotted with patients dying of distant metastatic disease with no local failure censored at the time of death. Patients were evaluated for local failure by clinical criteria, with further work up done for symptoms or laboratory evidence of recurrence. Local failure was evaluated until the patients death (not just first site). RESULTS
A total of 165 patients were treated. The median patient age was 65 years. The median follow-up in the survivors was 56 months with a minimum follow-up of 17 months and with all but 10 patients having been followed for more than two years. Three patients with Stage B,
Postoperative radiation therapy of rectal cancer 0 J. E. TEPPER et
tumors were treated because of minimal surgical margins and a high risk of local failure at the anastomotic site. There were a total of 53 patients with Stage BZ disease, 7 with B3, 10 with Cr ,77 with Cz, and 15 with C3 disease. Five year survival rates and 5-year rates with no evidence of tumor are shown in Table 1. Actuarial 5-year survival was 53% in the entire group of patients. For patients with Stage B2 tumors, 11 patients have failed out of a total of 53 with an actuarial 5-year survival rate of 7 1%. For patients with Cz disease, 45/77 have had disease recurrence with a 5-year survival rate of 39%, and for Stage C3, 13/ 15 patients have failed for a 5-year survival rate of 17%. In the three smaller groups, there were O/3 deaths in Stage B1, 3/7 deaths in Stage Bj, and l/10 death in patients with Stage C, tumors. Because of the long median follow-up time, the survival rates closely follow the disease-free survival rates in this study. A detailed analysis of local failure was completed in the entire group of patients. A total of 3 l/ 165 ( 19%) patients have had evidence of local failure (Table 2). Local failure was seen in O/3 patients with Stage B, disease, in 5/53 (9%) patients with Stage B2 disease, and O/7 of patients with Stage B3 disease. In patients with positive lymph nodes, local failure occurred in 2/ 10 (20%) of patients with Stage C, tumors, 16/77 (2 1%) of C2 tumors, and 8/ 15 (53%) of patients with C3 disease the local failure rate of 53% in patients with Stage C3 disease was disappointing, but reflected the advanced stage of these tumors at presentation. In the patients with C3 disease, it was not possible to state the exact reason for the local failures. Five of the patients had large pelvic recurrences where it was not possible to determine the exact site from which the local recurrence arose. In at least two patients, the local failure may have represented widespread intraabdominal disease with pelvic extension rather than a primary local recurrence. To determine the relative role of the adjuvant radiation therapy in producing local control, a comparison was made to a group of 142 patients with carcinomas of the rectum and rectosigmoid who were treated with surgery alone at the MGH between 1968 and 1972, with a minimum 5 year follow-up.‘4 These results are shown in Table 2. In essentially every subset, the local control was substantially improved with the addition of adjuvant radiation therapy. The two subsets large enough for meaningful comparison are Stage B2 and Stage Cz . In the B2 tumors, local control was obtained in 34/44 patients (77%) with surgery alone compared to 48/53 (91%) with surgery plus radiation therapy. In patients with Stage C2 disease, local control was improved from 53% with surgery alone (18/34) to 79% (6 l/77) with the addition of postoperative adjuvant radiation therapy. In the same group of patients, a comparison was made of patients with no evidence of disease at 5 years (Table 3). Again, there appeared to be an improvement in the 5 year NED rate for the patients treated with adjuvant radiation therapy. The differences were most marked in
al.
Table 1. Actuarial 5-year survival and NED rates after
postoperative irradiation of rectal cancer 5-year survival %
5-year NED* %
3 53
100
B3
7
67
100 76 69
C,
10
c2
77 15
90 39 17
69 34 13
165
53
50
Stage
Number of patients
BI B2
c3
Total
71
* Patients dying of intercurrent
disease censored for NED
analysis at time of death. Stage B disease with a more modest improvement in the 5 year NED rate in patients with Stage C tumors. This undoubtedly reflects a greater propensity for metastatic spread in the patients with Stage C disease with a lower impact of local control on long-term survival. Complications were analyzed in the patients treated with adjuvant radiation therapy. Generally, the radiation therapy was well-tolerated. With the addition of bolus on the perineum in patients with an APR, a perineal skin reaction was common but this usually abated easily after the completion of the irradiation. The perineal incision was observed carefully and the bolus removed if the reaction was becoming severe. Breaks from treatment were needed in only a minority of patients because of the perineal skin reaction. For patients with a low anterior resection, frequent bowel movements were a common complaint preceding, during, and after the radiation treatment. Patients who have very low anterior resections often will have 4-6 bowel movements per day from the surgery alone. Generally these symptoms were adequately controlled with the administration of Lomotil, and very few patients needed a break because of a significant problem with diarrhea. Cystitis was not a significant problem when using the four-field approach. Table 2. Local failure after postoperative irradiation of rectal cancer compared to surgery alone
Stage
Local failure (percent) surgery + XRT
Local failure (percent) surgery alone
;z; (0;)
3/39 lo/44 8/15
(8%) (Stage A and B,) (23%) (53%)
c3
5163 2110 16177 8/15
(8%) (20%) (21%) (53%)
214 16/34 416
(50%) (47%) (67%)
Stage C Total Total
26/102 31/165
(25%) (19%)
BI B2 B3
Stage B Total ::
Of3 5153 O/7
I. J. Radiation Oncology 0 Biology 0 Physics
8
Table 3.5-year NED rates in rectal cancer surgery alone vs. surgery plus postoperative irradiation Surgery alone
Surgery + XRT
%
%
BI
-
B2
41
Stage
B3
27
100 76 69
Cl
25 27 0
69 34 13
c2 c3
Long-term problems were quite acceptable and are outlined in Table 4. The major severe complication was small bowel obstruction requiring abdominal surgery. This occurred in 9/ 165 patients (5%) and in one patient small bowel obstruction developed with subsequent aspiration pneumonia and death. This is the same incidence as the 7/142 (5%) of small bowel obstruction requiring surgery in the previous group of patients from the MGH treated with surgery alone. Other chronic complications included bowel frequency or proctitis in 4 patients, small bowel obstruction not requiring surgery in 2 patients, and self-limited perineal or scrotal tenderness in 3 patients. Two patients had delayed healing in the perineal wound, and 3 had urinary problems (incontinence 2, bladder atrophy 1). It is not clear that all these problems were related to the radiation since urinary incontinence is a well known complication of pelvic surgery. The total incidence of severe complications was 6%, and of mild to moderate complications 8%. Most of the latter group were self-limited and did not produce significant morbidity. DISCUSSION Because of the known high incidence of local failure in patients with Stage B2 and C rectal tumors, a number Table 4. Complications after postoperative irradiation of rectal cancer Severe complications Small bowel obstruction/ileitis-requiring surgery Small bowel obstruction producing aspiration pneumonia, death Total severe complications Mild to moderate chronic complications Bowel frequencylproctitis Small bowel obstruction-not requiring surgery Perineal/scrotal tenderness (resolved) Delayed perineal wound healing Urinary incontinence Bladder atrophy/bleeding (resolved)
9
10 (6%) 4
14 (8%)
January 1987, Volume 13, Number 1
of groups have shown an interest in evaluating adjuvant radiation therapy combined with surgical resection. A previous publication from the MGH8 presented some preliminary data on the survival rates and the failure patterns. This series extends these observations with a much longer follow-up period and with a substantially larger group of patients. A number of groups have analyzed failure patterns after either low anterior resection or abdomino-perineal resection for rectal cancer. Cass et al.’have reported a 30% local failure rate for all stages of rectal cancer (including A and B,), and a local failure rate of 52% for patients with CZ colorectal tumors. Gunderson and Sosin have reported on the failure patterns in the reoperation series from the University of Minnesota. There were 17 patients with Stage C, disease with 4 local failures (24%). Of 40 patients with Stage C2 or C3 disease, 33 (83%) had local failure. Interestingly, 44% of the patients with tumor recurrence had local failure as the only site of failure, whereas only 1 patient had only distant metastases. Gilbert,6 from the Maine Medical Center, has analyzed patient morbidity from tumor recurrence after surgery alone. Sixty-eight percent of the patients who failed had symptoms related almost entirely to the pelvic recurrence, an additional 1 I % of patients had symptoms from both local and distant disease, and 20% of patients had symptoms primarily from metastases. Thus, pelvic recurrence is the major cause of morbidity in patients treated only with surgery for rectal cancer, and a decrease in the incidence would be of great value. Compared to the previous series of patients from the MGH with rectal cancer treated with surgery alone, the addition of postoperative radiation therapy has produced a substantial decrease in the incidence of local failure. As shown in Table 2, local control was improved in Stage Bz patients from 77% with surgery alone to 9 1% with surgery plus radiation therapy. In patients with Stage C2 disease, local control was improved from 53% to 79% with the addition of adjuvant irradiation. This was associated with an improvement in survival with the 5 year NED rate improving from 47% to 76% in patients with Stage Bz disease and from 27% to 34% in patients with Stage C2 tumors. It is of note that this improvement in local control and survival was produced with no detectable increase in late severe complications. In both the patients treated with surgery and surgery plus postoperative radiation therapy the incidence of small bowel obstruction requiring laparotomy was 5%. Other toxicities occurring after the completion of the radiation were mild and generally self-limited. Other studies” have demonstrated major morbidity from the use of adjuvant irradiation, but this can be attributed to poor technique (APPA fields only, fields extending to cover the para-aortic chain, etc.) or high dose per fraction. In the present study, the patients were treated with a 4-field technique using carefully blocked fields extending only to the level of the L-5 vertebral body. Treatment was given at 180
Postoperative radiation therapy of rectal cancer 0 J. E. TEPPERet al.
cGy per day, and careful attention was paid to the location of the small bowel to minimize bowel complications. Because this study was not a prospective randomized trial, the value of the addition of radiation therapy must be determined by comparison to historical groups. A series has recently been reported by the Gastrointestinal Tumor Study Group evaluating the results of a 4-arm randomized study with surgery alone, surgery plus postoperative chemotherapy, surgery plus postoperative radiation therapy, and surgery plus radiation and chemotherapy.4 In their patients with Stage B2 disease treated with the most aggressive regimen of combined radiation and chemotherapy, 25% of patients had evidence of disease recurrence. This compares to a 24% incidence of disease recurrence in our Stage B2 patients treated with radiation therapy alone. The comparison in Stage C is more difficult because the Gastrointestinal Tumor Study Group used a different staging classification, where Stage C, tumors had 4 or fewer positive lymph nodes and Stage C2 disease had greater than 4 positive lymph nodes. Using that staging system, the actuarial 5-year survival in the MGH series was 52% in 67 patients with GITSG Stage C, and 2 1% in 35 patients with GITSG Stage CZ. This is somewhat inferior to the combined modality group in the GITSG study where 68% of patients with Stage Cr and 55% of those with Stage C2 disease treated with combined radiation therapy and chemotherapy had no evidence of disease recurrence. Our results are similar to the 47 and 33% survival for Stage C1 and CZ tumors, respectively, seen in the GITSG series of patients treated with postoperative radiation therapy alone. Note, that in the GITSG study, the number of patients in each subset was quite small (generally under 20) which makes a statistical evaluation difficult. It is not possible to make a direct comparison of local control between the present study and that from the GITSG because only first sites of relapse were scored in the GITSG study whereas all pelvic recurrences were scored as local failure in the present evaluation. For patients treated with postoperative radiation therapy alone in the GITSG study, lo/50 (20%) had initial local failure and 5/46 ( 11%) had initial local failure with combined radiation and chemotherapy delivered postoperatively. The total incidence of local failure in the present series with postoperative radiation was 18%. Early studies from the Veteran’s Administration,16 Memorial Hospital,‘* and subsequently by the Princess Margaret Hospital” indicated an advantage to low dose preoperative irradiation in improving survival in patients with rectal cancer (the Memorial results were not confirmed in a follow-up study). Subsequently, studies from Yale University” and the University of Oregon’ evaluated high dose preoperative irradiation, with data indicating a decreased local failure rate with this regimen. Mendenhall et al., l1 from the University of Florida, reported a 10% (3/30) local-regional failure rate after
9
high dose preoperative irradiation, compared to 29% (39/l 35) in patients treated with surgery alone. Since patients with early stage disease were also irradiated, a direct comparison with the present study cannot be made. Gerard et al.’ reported results of an EORTC trial of 34.5 Gy of preoperative irradiation compared to surgery alone, with an improvement in local control from 65 to 85% with the addition of the irradiation. No survival difference has yet been demonstrated, but disease-free survival has been improved from 52 to 66% at 5 years. There are other series in the literature that have evaluated the use of postoperative radiation therapy combined with surgical resection. Many of these have either limited follow-up or small patients numbers making a direct comparison impossible. Data from a number of groups, however, indicate that the addition of postoperative radiation therapy does decrease the overall incidence of local failure from that observed with surgery alone.” It is not clear whether the best method of delivering adjuvant radiation therapy is with high dose postoperative radiation therapy as used in the present study, high dose preoperative radiation therapy, or sandwich treatment with low dose preoperative radiation followed by high dose radiation in patients in whom there is a high risk of local failure. This latter approach has been used by Mohiuddin et al. I2913and appears feasible. There are no clinical data indicating that survival rates are improved with the sandwich therapy. The major advantage ofthe approach used at the MGH is that there is a substantial group of patients in whom the risk of local failure is quite low, and the costs, inconvenience, and potential morbidity of high dose radiation therapy can be avoided by avoiding radiation in those patients with Stage A, and B tumors. Of concern is the high incidence of local failure in patients presenting with locally advanced disease in the pelvis. In patients with Stage C3 tumors, 8/ 15 (53%) developed local failure. It is likely that the dose of radiation, which could safely be delivered by external beam techniques, was not sufficient to produce a high incidence of local control in this patient subset. It is likely that these patients would have been better treated with a course of high dose preoperative radiation therapy followed by attempted surgical resection. The high dose preoperative radiation therapy may have allowed a more complete surgical resection to have been performed with less risk of tumor spillage at surgery. At present, when we see patients of this sort preoperatively, we are recommending high dose preoperative radiation therapy and are adding intraoperative electron beam radiation therapy (IORT) at the time of the surgical resection. With the high incidence of local failure seen in this analysis even with doses of 5,000 cGy or greater delivered postoperatively, it seems appropriate to treat these patients aggressively as described above. At present, the local failure rate in a similar (or more advanced) group of patients treated aggressively at the MGH is less than 10%. Additional attempts to improve the local control rates in patients with
IO
I. J. Radiation Oncology 0 Biology 0 Physics
stage C 1 and C2 disease also seem warranted. Despite a substantial decrease in local failure compared to surgery alone (where local failure is in the range of 50 to 65%),
January 1987, Volume 13, Number I
specialized radiation techniques may be necessary to increase the dose still further and to decrease the incidence of local failure.
REFERENCES 1. Allen, C.V., Fletcher, W.S.: A pilot study on preoperative
2.
3. 4.
5.
6.
7.
8.
9.
10.
irradiation of rectosigmoid carcinoma. Am. J. Roentgenol. 114: 504-508, 1972. Cass, A. W., Million, R.R., Pfaff, W. W.: Patterns of recurrence following surgery alone for adenocarcinoma of the colon and rectum. Cancer 37: 286 l-2865, 1976. Cox, D.R.: Regression models and life tables. J. R. Stat. Sot. B. 34: 187-202, 1972. Gastrointestinal Tumor Study Group: Prolongation of the disease-free interval in surgically treated rectal carcinoma. New Engl. J. Med. 312: 23 1465-1472, 1985. Gerard, A., Berrod, J.L., Pene, F., Loygue, J., Laugier, A., Bruckner, R., Camelot, G., Amaud, J.P., Metzger, U., Buyse, M., Dalesio, O., Duez, N.: Interim analysis of phase III study on preoperative radiation therapy in resectable rectal carcinoma: Trial of the Gastrointestinal Cooperative Group of the European Organization for Research on Treatment of Cancer (EORTC). Cancer 55: 2373-2379, 1985. Gilbert, S.G.: Symptomatic local tumor failure following abdomino-perineal resection. Int. J. Radiat. Oncol. Biol. Phys. 4: 801-807, 1978. Gunderson, L.L., Sosin, H.: Areas of failure found at reoperation (second or symptomatic look) following “curative surgery” for adenocarcinoma of the rectum: Clinicopathologic correlation and implications for adjuvant therapy. Cancer 34: 1278- 1292, 1974. Hoskins, R.B., Gunderson, L.L., Dosoretz, D.E., Rich, T.A., Galdabini, J., Donaldson, G., Cohen, A.M.: Adjuvant postoperative radiotherapy in carcinoma of the rectum and rectosigmoid. Cancer 55: 6 1-7 1, 1985. Kaplan, E.L., Meier, P.: Non-parametric estimation for incomplete observations. J. Am. Stat. Assoc. 53: 457-48 1, 1958. Kligerman, M.M., Urdaneta, N., Knowlton, A., Vidone, R., Hartman, P.V., Vera, R.: Preoperative irradiation of rectosigmoid carcinoma including its regional lymph nodes. Am. J. Roentgenol. 114: 3 498-503, 1972.
11. Mendenhall,
12.
13.
14.
15.
16.
17.
18.
19.
W.M., Million, R.R., Pfaff, W. W., Copeland, E.M.: Preoperative radiation therapy for clinically resectable adenocarcinoma of the rectum. Int. J. Radiat. Oncol. Biol. Phys. 10: 90, Sup. 2, 1984. Mohiuddin, M., Kramer, S., Marks, G., Dobelbower, R.R.: Combined pre and postoperative radiation for carcinoma of the rectum. Int. J. Radiat. Oncol. Biol. Phys. 8: 133-136, 1982. Mohiuddin, M., Marks, G.: Preoperative radiation therapy as the key to extending sphincter preservation in rectal cancer. Int. J. Radiat. Oncol. Biol. Phys. lO(Suppl.2): 90, 1984. Rich, T., Gunderson, L.L., Lew, R., Galdibini, J. J., Cohen, A.M., Donaldson, G.: Patterns of recurrence of rectal cancer after potentially curative surgery. Cancer 52: 13171329,1983. Rider, W.D., Palmer, J.A., Mahoney, L.J., Robertson, C.T.: Preoperative irradiation in operable cancer of the rectum: Report ofthe Toronto trial. Can. J. Surg. 20: 335338, 1977. Roswit, B., Higgins, G.A., Keehn, R.J.: Preoperative irradiation for carcinoma of the rectum and rectosigmoid colon: Report of a National Veterans Administration randomized study. Cancer 35: 1597- 1602, 1975. Sparso, B.H., Von der Masse, H., Kristensen, D., Christiansen, J., Damgaard-Nielsen, S.A., Hebjom, M., Andersen, B.: Complications following postoperative combined radiation and chemotherapy in adenocarcinoma of the rectum and rectosigmoid: A randomized trial that failed. Cancer 54: 2363, 1984. Steams, M.W., Berg, J.W., Deddish, M.R.: Preoperative irradiation of cancer of the rectum. Dis. Col. Rec. 4: 403408, 1961. Withers, H.R., Romsdahl, M., Barkley, H.T., Saxton, J., McBride, C., McMurtrey M.: Postoperative radiotherapy radiotherapy for rectal cancer. In Adjuvant Therapy of Cancer ZZZ,Jones and Salmon, 1979, pp. 62 l-628.
Copyright
0360-3016/87 0 19R7 Pergamon
$3.00 + .OO Journals Ltd.
??Original Contribution
POSTOPERATIVE
RADIATION THERAPY OF RECTAL CANCER
JOEL E. TEPPER, M.D.,* ALFRED M. COHEN, M.D.,? WILLIAM C. WOOD, M.D.,? ERICA L. ORLOW* AND STEPHEN E. HEDBERG, M.D.? Massachusetts General Hospital, MGH Cancer Center, Boston, MA 02 114 Beginning in December 1975, at the Massachusetts General Hospital (MGH) patients with rectal carcinomas thought to be at high risk of local recurrence after potentially curative surgical resection, were entered on a treatment protocol of high dose postoperative radiation therapy. Treatment was given with X rays of 10 MeV, generally using a four-field box technique to a dose of 4500 cGy with a boost to 5040 cGy or higher when the small bowel could be excluded from the reduced field. One-hundred sixty-five patients who began their radiation therapy between December 1975 and December 1982 were entered into the study. The median age was 65 years. The median follow-up in the survivors was 56 months, with a minimum follow-up of 17 months. All but 10 patients were followed for more than 2 years. Of the entire group, the actuarial 5-year survival was 53%, with survival of 71% in patients with Stage B-2,39% in Stage C-2, and 17% in Stage C-3. Local failure was seen in 5/53 patients with Stage B-2 disease and O/7 of patients with Stage B-3 disease. In patients with positive lymph nodes, local failure occurred in 2/10 (20%) of patients with Stage C-l, 16/77 (21%) of Stage C-2, and 8/15 (53%) of patients with Stage C-3 disease. Compared to previous series of surgery alone, the local failure rate has been decreased by more than one-half in all patients, except those with Stage C-3. Efforts to maximize the radiation doses in all stages should be made to minimize local failure. For Stage C-3, newer strategies such as intraoperative radiation therapy should be employed to decrease the continuing high incidence of failures.
INTRODUCIION
positive lymph nodes, local failure is Seen in approximately 50%. Local failure produces major morbidity in a high percentage of these patients.6 In many patients, the local failure site is the major symptom-producing area after surgical resection. Because of the high incidence of local failure in certain patient subsets, a number of institutions have evaluated the usefulness of radiation therapy given as an adjuvant to surgical resection. Initial studies emphasized low dose preoperative radiation therapy,‘5,‘6,‘8 but now many institutions use radiation therapy delivered primarily postoperatively. The radiation therapy has been used with two objectives: to decrease the incidence of local failure, and to improve the long-term survival rate. Beginning in 1975, the MGH began a prospective evaluation of the use of high dose postoperative radiation therapy after either low-anterior resection or abdominalperineal resection in patients with rectal carcinoma. This report details the local control and survival results in a group of 165 patients treated with a consistent approach and technique with a median follow-up of almost 5 years.
For many years, single modality therapy by surgical resection was the treatment approach for rectal carcinoma. This has produced a high cure rate in patients with small localized tumors, but for patients with extensive local disease or with spread to regional lymph nodes, the cure rate has not been as high. Surgical series have shown 5 year disease-free survival rates of approximately 50% in patients with Stage B disease (invasive into or through the bowel wall, node negative) and approximately 30% in patients with Stage C disease (node positive). A number of series have evaluated recurrence patterns after potentially curative resections in patients with rectal cancer. These studies, which have been generated from institutions including the University of Florida,2 University of Minnesota,7 Maine Medical Center,6 and the Massachusetts General Hospital,‘4 indicate that for patients with Stage A or B, disease, the pelvic failure rate is less than 10%. For patients with Stage B2 disease, local failure is approximately 25%, and it is approximately 50% for patients with Stage B3 disease. For patients with
* Radiation Medicine Service. 7 Surgical Service.
Reprint requests to: Joel E. Tepper. Accepted for publication 13 August 1986. 5
I. J. Radiation Oncology 0 Biology 0 Physics
METHODS
AND MATERIALS
Patients included in the analysis were a consecutive group of 165 patients treated with a course of postoperative radiation after a potentially curative surgical resection. The resection was done at the MGH in slightly less than half of the patients (79 patients). Eighty-six patients had their primary operative procedure done at an outside hospital. Patients were treated with a variety of operative procedures: 93 with abdominal-perineal resection (APR), 66 with low-anterior resection, 4 with trans-sacral resection, and 2 with posterior exenteration. To be eligible for the study, the patients had no evidence of metastatic disease at the time of the operative procedure, and no evidence of residual disease after resection. During the same time period, other patients were treated with a high-dose preoperative radiation therapy regimen when their disease was locally advanced at presentation and were not thought to be amenable to primary surgical resection. These patients have not been included in this analysis. The radiation therapy technique was a high dose regimen and was relatively uniform during this time period. Treatments were generally given through a 4-field box technique although some patients were treated with a 3field technique including right and left laterals and a posterior field. The majority of patients were treated with a 10 MeV X ray beam from a linear accelerator*. Patients were treated in the prone position. The initial portion of the radiation treatment was given to a field encompassing the posterior portion of the entire pelvis. When an abdominal-perineal resection had been performed, the fields always included the perineal scar in the entirety and bolus was used on the perineal incision. For patients who had a low-anterior resection, the lower portion of the field extended approximately 4 to 5 cm below the level of the anastomosis. The superior border of the field generally extended to the mid portion of the L5 vertebral body. Laterally, the fields extended 1 to 1l/2 cm lateral to the true bony pelvis. To treat the entire pre-sacral space with adequate margins and full dose, the lateral fields were designed so that the posterior border encompassed the entire sacrum with a i/z to 1 cm margin posterior to the sacrum. This was necessary because the distance from the 50% isodose line to the 90% isodose line can be in the range of I’/2 centimeters when treating with lateral fields. Anteriorly, the field was not designed to include the external iliac lymph nodes but was designed to encompass the previous tumor bed and to encompass the tissues that might potentially have been infiltrated by tumor anteriorly. For females, the posterior wall of the vagina was always included in the radiation field and for males a large portion of the prostate was included. Bowel
* Clinac- 18.
January 1987. Volume 13, Number I
located anteriorly and a portion of the bladder were spared in the lateral ports. After a dose of 4,500 cGy had been given to the large field, the field size was reduced to encompass the areas thought to be at highest risk for residual disease. This area was boosted for an additional 540 cGy, but in patients who were thought to have very high risk for residual disease (B3, Cz and C3 tumors) a higher dose was used if small bowel could safely be moved from the field. In treatment planning for the boost field, patients had a special small bowel X ray series to determine the exact location of the small intestine and possible fixation of the small intestine in the region of the planned boost. Attempts were made to use bladder distention or other patient positioning devices to move the small bowel out of the radiation field. If this could not be done safely, the dose was limited to 5,040 cGy and for some patients with B2 tumors with a large amount of small bowel in the pelvis, the dose was stopped at 4,500 cGy. The surgeons at the MGH were encouraged to reperitonealize the pelvic floor so that a large amount of small bowel would not be located in the radiation boost field. Patients were not generally treated with adjuvant chemotherapy although 14 patients did have some additional therapy, generally with 5-fluorouracil alone. Patients were staged using the Gunderson-Sosin modification of the Dukes’ Staging System. Patients with Stage B disease had negative lymph nodes with Stage B, disease limited to the bowel wall, B2 with extension through the bowel wall into the perirectal fat, and Stage B3 having tumor adherence or fixation to adjacent normal structure. Stage C disease patients had positive lymph nodes with the subscripts defining the extent of local penetration as for Stage B. Data analysis was done based on patient status information available in March 1985. Survival curves were calculated using the actuarial method of KaplanMeier,3,9 unless 15 or more failures occurred in a group, when life table methods were used. Local control actuarial curves were plotted with patients dying of distant metastatic disease with no local failure censored at the time of death. Patients were evaluated for local failure by clinical criteria, with further work up done for symptoms or laboratory evidence of recurrence. Local failure was evaluated until the patients death (not just first site). RESULTS
A total of 165 patients were treated. The median patient age was 65 years. The median follow-up in the survivors was 56 months with a minimum follow-up of 17 months and with all but 10 patients having been followed for more than two years. Three patients with Stage B,
Postoperative radiation therapy of rectal cancer 0 J. E. TEPPER et
tumors were treated because of minimal surgical margins and a high risk of local failure at the anastomotic site. There were a total of 53 patients with Stage BZ disease, 7 with B3, 10 with Cr ,77 with Cz, and 15 with C3 disease. Five year survival rates and 5-year rates with no evidence of tumor are shown in Table 1. Actuarial 5-year survival was 53% in the entire group of patients. For patients with Stage B2 tumors, 11 patients have failed out of a total of 53 with an actuarial 5-year survival rate of 7 1%. For patients with Cz disease, 45/77 have had disease recurrence with a 5-year survival rate of 39%, and for Stage C3, 13/ 15 patients have failed for a 5-year survival rate of 17%. In the three smaller groups, there were O/3 deaths in Stage B1, 3/7 deaths in Stage Bj, and l/10 death in patients with Stage C, tumors. Because of the long median follow-up time, the survival rates closely follow the disease-free survival rates in this study. A detailed analysis of local failure was completed in the entire group of patients. A total of 3 l/ 165 ( 19%) patients have had evidence of local failure (Table 2). Local failure was seen in O/3 patients with Stage B, disease, in 5/53 (9%) patients with Stage B2 disease, and O/7 of patients with Stage B3 disease. In patients with positive lymph nodes, local failure occurred in 2/ 10 (20%) of patients with Stage C, tumors, 16/77 (2 1%) of C2 tumors, and 8/ 15 (53%) of patients with C3 disease the local failure rate of 53% in patients with Stage C3 disease was disappointing, but reflected the advanced stage of these tumors at presentation. In the patients with C3 disease, it was not possible to state the exact reason for the local failures. Five of the patients had large pelvic recurrences where it was not possible to determine the exact site from which the local recurrence arose. In at least two patients, the local failure may have represented widespread intraabdominal disease with pelvic extension rather than a primary local recurrence. To determine the relative role of the adjuvant radiation therapy in producing local control, a comparison was made to a group of 142 patients with carcinomas of the rectum and rectosigmoid who were treated with surgery alone at the MGH between 1968 and 1972, with a minimum 5 year follow-up.‘4 These results are shown in Table 2. In essentially every subset, the local control was substantially improved with the addition of adjuvant radiation therapy. The two subsets large enough for meaningful comparison are Stage B2 and Stage Cz . In the B2 tumors, local control was obtained in 34/44 patients (77%) with surgery alone compared to 48/53 (91%) with surgery plus radiation therapy. In patients with Stage C2 disease, local control was improved from 53% with surgery alone (18/34) to 79% (6 l/77) with the addition of postoperative adjuvant radiation therapy. In the same group of patients, a comparison was made of patients with no evidence of disease at 5 years (Table 3). Again, there appeared to be an improvement in the 5 year NED rate for the patients treated with adjuvant radiation therapy. The differences were most marked in
al.
Table 1. Actuarial 5-year survival and NED rates after
postoperative irradiation of rectal cancer 5-year survival %
5-year NED* %
3 53
100
B3
7
67
100 76 69
C,
10
c2
77 15
90 39 17
69 34 13
165
53
50
Stage
Number of patients
BI B2
c3
Total
71
* Patients dying of intercurrent
disease censored for NED
analysis at time of death. Stage B disease with a more modest improvement in the 5 year NED rate in patients with Stage C tumors. This undoubtedly reflects a greater propensity for metastatic spread in the patients with Stage C disease with a lower impact of local control on long-term survival. Complications were analyzed in the patients treated with adjuvant radiation therapy. Generally, the radiation therapy was well-tolerated. With the addition of bolus on the perineum in patients with an APR, a perineal skin reaction was common but this usually abated easily after the completion of the irradiation. The perineal incision was observed carefully and the bolus removed if the reaction was becoming severe. Breaks from treatment were needed in only a minority of patients because of the perineal skin reaction. For patients with a low anterior resection, frequent bowel movements were a common complaint preceding, during, and after the radiation treatment. Patients who have very low anterior resections often will have 4-6 bowel movements per day from the surgery alone. Generally these symptoms were adequately controlled with the administration of Lomotil, and very few patients needed a break because of a significant problem with diarrhea. Cystitis was not a significant problem when using the four-field approach. Table 2. Local failure after postoperative irradiation of rectal cancer compared to surgery alone
Stage
Local failure (percent) surgery + XRT
Local failure (percent) surgery alone
;z; (0;)
3/39 lo/44 8/15
(8%) (Stage A and B,) (23%) (53%)
c3
5163 2110 16177 8/15
(8%) (20%) (21%) (53%)
214 16/34 416
(50%) (47%) (67%)
Stage C Total Total
26/102 31/165
(25%) (19%)
BI B2 B3
Stage B Total ::
Of3 5153 O/7
I. J. Radiation Oncology 0 Biology 0 Physics
8
Table 3.5-year NED rates in rectal cancer surgery alone vs. surgery plus postoperative irradiation Surgery alone
Surgery + XRT
%
%
BI
-
B2
41
Stage
B3
27
100 76 69
Cl
25 27 0
69 34 13
c2 c3
Long-term problems were quite acceptable and are outlined in Table 4. The major severe complication was small bowel obstruction requiring abdominal surgery. This occurred in 9/ 165 patients (5%) and in one patient small bowel obstruction developed with subsequent aspiration pneumonia and death. This is the same incidence as the 7/142 (5%) of small bowel obstruction requiring surgery in the previous group of patients from the MGH treated with surgery alone. Other chronic complications included bowel frequency or proctitis in 4 patients, small bowel obstruction not requiring surgery in 2 patients, and self-limited perineal or scrotal tenderness in 3 patients. Two patients had delayed healing in the perineal wound, and 3 had urinary problems (incontinence 2, bladder atrophy 1). It is not clear that all these problems were related to the radiation since urinary incontinence is a well known complication of pelvic surgery. The total incidence of severe complications was 6%, and of mild to moderate complications 8%. Most of the latter group were self-limited and did not produce significant morbidity. DISCUSSION Because of the known high incidence of local failure in patients with Stage B2 and C rectal tumors, a number Table 4. Complications after postoperative irradiation of rectal cancer Severe complications Small bowel obstruction/ileitis-requiring surgery Small bowel obstruction producing aspiration pneumonia, death Total severe complications Mild to moderate chronic complications Bowel frequencylproctitis Small bowel obstruction-not requiring surgery Perineal/scrotal tenderness (resolved) Delayed perineal wound healing Urinary incontinence Bladder atrophy/bleeding (resolved)
9
10 (6%) 4
14 (8%)
January 1987, Volume 13, Number 1
of groups have shown an interest in evaluating adjuvant radiation therapy combined with surgical resection. A previous publication from the MGH8 presented some preliminary data on the survival rates and the failure patterns. This series extends these observations with a much longer follow-up period and with a substantially larger group of patients. A number of groups have analyzed failure patterns after either low anterior resection or abdomino-perineal resection for rectal cancer. Cass et al.’have reported a 30% local failure rate for all stages of rectal cancer (including A and B,), and a local failure rate of 52% for patients with CZ colorectal tumors. Gunderson and Sosin have reported on the failure patterns in the reoperation series from the University of Minnesota. There were 17 patients with Stage C, disease with 4 local failures (24%). Of 40 patients with Stage C2 or C3 disease, 33 (83%) had local failure. Interestingly, 44% of the patients with tumor recurrence had local failure as the only site of failure, whereas only 1 patient had only distant metastases. Gilbert,6 from the Maine Medical Center, has analyzed patient morbidity from tumor recurrence after surgery alone. Sixty-eight percent of the patients who failed had symptoms related almost entirely to the pelvic recurrence, an additional 1 I % of patients had symptoms from both local and distant disease, and 20% of patients had symptoms primarily from metastases. Thus, pelvic recurrence is the major cause of morbidity in patients treated only with surgery for rectal cancer, and a decrease in the incidence would be of great value. Compared to the previous series of patients from the MGH with rectal cancer treated with surgery alone, the addition of postoperative radiation therapy has produced a substantial decrease in the incidence of local failure. As shown in Table 2, local control was improved in Stage Bz patients from 77% with surgery alone to 9 1% with surgery plus radiation therapy. In patients with Stage C2 disease, local control was improved from 53% to 79% with the addition of adjuvant irradiation. This was associated with an improvement in survival with the 5 year NED rate improving from 47% to 76% in patients with Stage Bz disease and from 27% to 34% in patients with Stage C2 tumors. It is of note that this improvement in local control and survival was produced with no detectable increase in late severe complications. In both the patients treated with surgery and surgery plus postoperative radiation therapy the incidence of small bowel obstruction requiring laparotomy was 5%. Other toxicities occurring after the completion of the radiation were mild and generally self-limited. Other studies” have demonstrated major morbidity from the use of adjuvant irradiation, but this can be attributed to poor technique (APPA fields only, fields extending to cover the para-aortic chain, etc.) or high dose per fraction. In the present study, the patients were treated with a 4-field technique using carefully blocked fields extending only to the level of the L-5 vertebral body. Treatment was given at 180
Postoperative radiation therapy of rectal cancer 0 J. E. TEPPERet al.
cGy per day, and careful attention was paid to the location of the small bowel to minimize bowel complications. Because this study was not a prospective randomized trial, the value of the addition of radiation therapy must be determined by comparison to historical groups. A series has recently been reported by the Gastrointestinal Tumor Study Group evaluating the results of a 4-arm randomized study with surgery alone, surgery plus postoperative chemotherapy, surgery plus postoperative radiation therapy, and surgery plus radiation and chemotherapy.4 In their patients with Stage B2 disease treated with the most aggressive regimen of combined radiation and chemotherapy, 25% of patients had evidence of disease recurrence. This compares to a 24% incidence of disease recurrence in our Stage B2 patients treated with radiation therapy alone. The comparison in Stage C is more difficult because the Gastrointestinal Tumor Study Group used a different staging classification, where Stage C, tumors had 4 or fewer positive lymph nodes and Stage C2 disease had greater than 4 positive lymph nodes. Using that staging system, the actuarial 5-year survival in the MGH series was 52% in 67 patients with GITSG Stage C, and 2 1% in 35 patients with GITSG Stage CZ. This is somewhat inferior to the combined modality group in the GITSG study where 68% of patients with Stage Cr and 55% of those with Stage C2 disease treated with combined radiation therapy and chemotherapy had no evidence of disease recurrence. Our results are similar to the 47 and 33% survival for Stage C1 and CZ tumors, respectively, seen in the GITSG series of patients treated with postoperative radiation therapy alone. Note, that in the GITSG study, the number of patients in each subset was quite small (generally under 20) which makes a statistical evaluation difficult. It is not possible to make a direct comparison of local control between the present study and that from the GITSG because only first sites of relapse were scored in the GITSG study whereas all pelvic recurrences were scored as local failure in the present evaluation. For patients treated with postoperative radiation therapy alone in the GITSG study, lo/50 (20%) had initial local failure and 5/46 ( 11%) had initial local failure with combined radiation and chemotherapy delivered postoperatively. The total incidence of local failure in the present series with postoperative radiation was 18%. Early studies from the Veteran’s Administration,16 Memorial Hospital,‘* and subsequently by the Princess Margaret Hospital” indicated an advantage to low dose preoperative irradiation in improving survival in patients with rectal cancer (the Memorial results were not confirmed in a follow-up study). Subsequently, studies from Yale University” and the University of Oregon’ evaluated high dose preoperative irradiation, with data indicating a decreased local failure rate with this regimen. Mendenhall et al., l1 from the University of Florida, reported a 10% (3/30) local-regional failure rate after
9
high dose preoperative irradiation, compared to 29% (39/l 35) in patients treated with surgery alone. Since patients with early stage disease were also irradiated, a direct comparison with the present study cannot be made. Gerard et al.’ reported results of an EORTC trial of 34.5 Gy of preoperative irradiation compared to surgery alone, with an improvement in local control from 65 to 85% with the addition of the irradiation. No survival difference has yet been demonstrated, but disease-free survival has been improved from 52 to 66% at 5 years. There are other series in the literature that have evaluated the use of postoperative radiation therapy combined with surgical resection. Many of these have either limited follow-up or small patients numbers making a direct comparison impossible. Data from a number of groups, however, indicate that the addition of postoperative radiation therapy does decrease the overall incidence of local failure from that observed with surgery alone.” It is not clear whether the best method of delivering adjuvant radiation therapy is with high dose postoperative radiation therapy as used in the present study, high dose preoperative radiation therapy, or sandwich treatment with low dose preoperative radiation followed by high dose radiation in patients in whom there is a high risk of local failure. This latter approach has been used by Mohiuddin et al. I2913and appears feasible. There are no clinical data indicating that survival rates are improved with the sandwich therapy. The major advantage ofthe approach used at the MGH is that there is a substantial group of patients in whom the risk of local failure is quite low, and the costs, inconvenience, and potential morbidity of high dose radiation therapy can be avoided by avoiding radiation in those patients with Stage A, and B tumors. Of concern is the high incidence of local failure in patients presenting with locally advanced disease in the pelvis. In patients with Stage C3 tumors, 8/ 15 (53%) developed local failure. It is likely that the dose of radiation, which could safely be delivered by external beam techniques, was not sufficient to produce a high incidence of local control in this patient subset. It is likely that these patients would have been better treated with a course of high dose preoperative radiation therapy followed by attempted surgical resection. The high dose preoperative radiation therapy may have allowed a more complete surgical resection to have been performed with less risk of tumor spillage at surgery. At present, when we see patients of this sort preoperatively, we are recommending high dose preoperative radiation therapy and are adding intraoperative electron beam radiation therapy (IORT) at the time of the surgical resection. With the high incidence of local failure seen in this analysis even with doses of 5,000 cGy or greater delivered postoperatively, it seems appropriate to treat these patients aggressively as described above. At present, the local failure rate in a similar (or more advanced) group of patients treated aggressively at the MGH is less than 10%. Additional attempts to improve the local control rates in patients with
IO
I. J. Radiation Oncology 0 Biology 0 Physics
stage C 1 and C2 disease also seem warranted. Despite a substantial decrease in local failure compared to surgery alone (where local failure is in the range of 50 to 65%),
January 1987, Volume 13, Number I
specialized radiation techniques may be necessary to increase the dose still further and to decrease the incidence of local failure.
REFERENCES 1. Allen, C.V., Fletcher, W.S.: A pilot study on preoperative
2.
3. 4.
5.
6.
7.
8.
9.
10.
irradiation of rectosigmoid carcinoma. Am. J. Roentgenol. 114: 504-508, 1972. Cass, A. W., Million, R.R., Pfaff, W. W.: Patterns of recurrence following surgery alone for adenocarcinoma of the colon and rectum. Cancer 37: 286 l-2865, 1976. Cox, D.R.: Regression models and life tables. J. R. Stat. Sot. B. 34: 187-202, 1972. Gastrointestinal Tumor Study Group: Prolongation of the disease-free interval in surgically treated rectal carcinoma. New Engl. J. Med. 312: 23 1465-1472, 1985. Gerard, A., Berrod, J.L., Pene, F., Loygue, J., Laugier, A., Bruckner, R., Camelot, G., Amaud, J.P., Metzger, U., Buyse, M., Dalesio, O., Duez, N.: Interim analysis of phase III study on preoperative radiation therapy in resectable rectal carcinoma: Trial of the Gastrointestinal Cooperative Group of the European Organization for Research on Treatment of Cancer (EORTC). Cancer 55: 2373-2379, 1985. Gilbert, S.G.: Symptomatic local tumor failure following abdomino-perineal resection. Int. J. Radiat. Oncol. Biol. Phys. 4: 801-807, 1978. Gunderson, L.L., Sosin, H.: Areas of failure found at reoperation (second or symptomatic look) following “curative surgery” for adenocarcinoma of the rectum: Clinicopathologic correlation and implications for adjuvant therapy. Cancer 34: 1278- 1292, 1974. Hoskins, R.B., Gunderson, L.L., Dosoretz, D.E., Rich, T.A., Galdabini, J., Donaldson, G., Cohen, A.M.: Adjuvant postoperative radiotherapy in carcinoma of the rectum and rectosigmoid. Cancer 55: 6 1-7 1, 1985. Kaplan, E.L., Meier, P.: Non-parametric estimation for incomplete observations. J. Am. Stat. Assoc. 53: 457-48 1, 1958. Kligerman, M.M., Urdaneta, N., Knowlton, A., Vidone, R., Hartman, P.V., Vera, R.: Preoperative irradiation of rectosigmoid carcinoma including its regional lymph nodes. Am. J. Roentgenol. 114: 3 498-503, 1972.
11. Mendenhall,
12.
13.
14.
15.
16.
17.
18.
19.
W.M., Million, R.R., Pfaff, W. W., Copeland, E.M.: Preoperative radiation therapy for clinically resectable adenocarcinoma of the rectum. Int. J. Radiat. Oncol. Biol. Phys. 10: 90, Sup. 2, 1984. Mohiuddin, M., Kramer, S., Marks, G., Dobelbower, R.R.: Combined pre and postoperative radiation for carcinoma of the rectum. Int. J. Radiat. Oncol. Biol. Phys. 8: 133-136, 1982. Mohiuddin, M., Marks, G.: Preoperative radiation therapy as the key to extending sphincter preservation in rectal cancer. Int. J. Radiat. Oncol. Biol. Phys. lO(Suppl.2): 90, 1984. Rich, T., Gunderson, L.L., Lew, R., Galdibini, J. J., Cohen, A.M., Donaldson, G.: Patterns of recurrence of rectal cancer after potentially curative surgery. Cancer 52: 13171329,1983. Rider, W.D., Palmer, J.A., Mahoney, L.J., Robertson, C.T.: Preoperative irradiation in operable cancer of the rectum: Report ofthe Toronto trial. Can. J. Surg. 20: 335338, 1977. Roswit, B., Higgins, G.A., Keehn, R.J.: Preoperative irradiation for carcinoma of the rectum and rectosigmoid colon: Report of a National Veterans Administration randomized study. Cancer 35: 1597- 1602, 1975. Sparso, B.H., Von der Masse, H., Kristensen, D., Christiansen, J., Damgaard-Nielsen, S.A., Hebjom, M., Andersen, B.: Complications following postoperative combined radiation and chemotherapy in adenocarcinoma of the rectum and rectosigmoid: A randomized trial that failed. Cancer 54: 2363, 1984. Steams, M.W., Berg, J.W., Deddish, M.R.: Preoperative irradiation of cancer of the rectum. Dis. Col. Rec. 4: 403408, 1961. Withers, H.R., Romsdahl, M., Barkley, H.T., Saxton, J., McBride, C., McMurtrey M.: Postoperative radiotherapy radiotherapy for rectal cancer. In Adjuvant Therapy of Cancer ZZZ,Jones and Salmon, 1979, pp. 62 l-628.