External Radiation Therapy A shrinking-field technique should be used, with initial radiation therapy fields designed to treat the primary tumor volume and regional lymph nodes. Smaller fields then can be used to treat the primary tumor bed to higher doses, as clinically indicated. The width of posteroanterior (PA) portals (Fig. 54-3A and Fig. 54-3B) should cover the pelvic inlet with a 2-cm margin. The superior margin is usually 1.5 cm above the level of the sacral promontory. In patients who have had an anterior resection, the usual inferior margin is below the obturator foramina. If the pelvis is treated, lateral fields should be used for a portion of treatment to avoid as much small bowel as possible. Bladder distention and prone position are useful techniques for displacing the small bowel out of the pelvis. The posterior field margin for lateral fields is critical because the rectum and perirectal tissues lie just anterior to the sacrum and coccyx. Accordingly, the posterior field margin should be at least 1.5 to 2 cm behind the anterior bony sacral margin (Fig. 54-3A, Fig. 54-3B and Fig. 54-4). The entire sacral canal should be included in patients with locally advanced disease to avoid sacral recurrence from tumor spread along nerve roots (Fig. 54-5). In patients with rectal cancer, internal iliac and presacral nodes are at risk for metastatic involvement. These lymph node chains are not a standard part of the dissection for rectal cancer. Accordingly, they should be included in the initial radiation therapy volume treated to 45 Gy. External iliac nodes are not a primary lymph node drainage site and are not included unless pelvic organs with external iliac drainage (i.e., prostate, upper vagina, bladder, uterus) are involved by direct extension. Irradiation of the perineum after abdominoperineal resection appears to be effective in decreasing the rate of perineal recurrence. In surgical series, the risk of perineal recurrence after abdominoperineal resection ranges from 8% to 30%. At the Mayo Clinic, the perineal failure rate was 2% at 5 years for patients undergoing postoperative adjuvant radiation therapy for whom the perineum was included within the radiation therapy field for the initial 40 Gy. In contrast, if the entire perineum was not treated, the perineal failure rate was 23% at 5 years (P = 0.01). Although these results were not drawn from a randomized trial, multivariate analysis showed that perineal treatment with radiation was the only significant factor associated with perineal failure. Temporary, acute, moderate to moderately severe perineal discomfort occurs in all patients whose perineums are included within irradiation fields. This can be mitigated with the use of a three-field technique (PA and laterals, with wedges on the laterals, heels posterior). The incidence of chronic complications has not increased as a result of perineal inclusion. Radiopaque markers can be used to outline the extent of the perineal scar at the time of simulation for posterior and lateral fields (Fig. 54-6). Anteriorly, the lower third of the rectum abuts the posterior vaginal wall or prostate, and these structures should be included in patients with distal lesions. In female patients, this can be verified by placing a contrast-soaked tampon in the vagina during radiation therapy simulation. Bolus applied to the perineal scar during the PA treatment ensures adequate dosage to this site. The dose to the large fields, which include the tumor bed and regional lymph nodes, should be 45 Gy given over 5 weeks. After this, the use of a boost field to the primary tumor bed and immediately adjacent lymph nodes should be considered. Boost fields are defined by methods such as barium enema studies, CT scan, and clip placement. Doses greater than 50.4 Gy generally should not be administered unless there is complete shift of the small bowel out of the final boost field. If radiation therapy is used for locally advanced extrapelvic colon cancer, the tumor bed should be covered with a 3- to 5-cm margin. Adjuvant radiation therapy for colon cancer should not be given except in the context of a formal prospective clinical trial. Sphincter Preservation Endocavitary Radiation Therapy Endocavitary radiation therapy produces high rates of local control and long-term survival in appropriately selected patients with rectal cancer. The indications for this technique have been described by Papillon (Table 54-4). In assessing the suitability of a patient for this technique, intrarectal ultrasonography is helpful to determining tumor confinement to the rectal wall. The treatment is performed on an outpatient basis. Local anesthesia in the anal canal occasionally is necessary for introducing the 3-cm-diameter applicator into the rectum. The radiation oncologist verifies the position of the applicator and coverage of the lesion. A lead apron and gloves are worn by the radiation oncologist, who holds the applicator firmly in place during the x-ray exposure. The treatment regimen usually consists of four 30-Gy treatments separated by intervals of approximately 2 weeks. A short-focal-distance (contact) x-ray unit is used at 50 kVp at a dose rate of approximately 10 Gy per minute. If the size of the tumor exceeds the diameter of the applicator, several overlapping fields must be used. Most reports indicate that endocavitary radiation therapy for rectal cancer results in high rates of disease control. For 207 patients treated by Papillon [ref: 136] with 5-year follow-up, the locoregional failure rate was 11%. Overall, 11% of the patients died of cancer and 13% died of intercurrent disease. Distant metastasis was found in only 3% of all patients. In the largest experience in the United States, Sischy and colleagues [ref: 167] reported similar excellent results. At the Cleveland Clinic, 62 patients were treated between 1973 and 1984; the median follow-up was 30 months. [ref: 101] Local recurrence developed in 11 patients (18%). In 3 of these 11, local recurrence was associated with distant metastasis. The other 8 patients were rendered disease free by other treatment at a median of 20 months after recurrence. Only 3 patients in the entire series died of cancer. Similar favorable results with endocavitary irradiation have been reported from the Mayo Clinic, where the projected local control rate at 5 years was 89% among 20 patients treated with curative intent. Local Excision With or Without Postoperative Radiation Therapy In 1961, Jackman reported results in 211 patients with rectal cancer treated with sphincter-sparing procedures. With follow-up ranging from 8 to 18 years, 8 patients (4%) experienced local recurrence. A later report described 234 patients treated mostly with local excision. [ref: 13] Forty-nine (21%) subsequently had local recurrence. The 5-year probability of local recurrence was approximately 11% for patients with in situ disease and 27% for patients with invasive disease. Hager and associates described results of local excision of cancer of the rectum in 95 patients. Fifty-nine low-risk patients, who had a wide margin of healthy tissue at the time of local excision, were divided into those with invasion of tumor into the submucosa (group 1) and those with invasion into the muscularis propria (group 2). Local recurrence was observed in 3 (8%) of 39 patients in group 1 and 3 (15%) of 20 patients in group 2. The 5-year survival rate was 90% in group 1 and 58% in group 2. These data strongly suggest that patients with early lesions with minimal invasion (i.e., those similar to patients in group 1) may be treated successfully with surgery alone without the use of further adjuvant therapy. Thirty-six percent of patients were considered high risk, as defined by incomplete removal of tumor; at a median follow-up of 45 months, 24% had experienced local progression, and 39% had died of cancer. It is clear that there is a wide range in incidence of local recurrence after limited surgical procedures for rectal cancer. Interpretation of the data is made difficult by the variation in selection criteria. To minimize local recurrence, postoperative radiation therapy has been used at some institutions. Rich and colleagues described results of treating 17 patients by limited surgery and postoperative radiation therapy. With a short median follow-up of 26 months in surviving patients, 1 (6%) had local failure. In another study, the local failure rate was 21% after local excision and postoperative radiation therapy. These results are not clearly better than those that may have been expected with local excision alone. In the absence of randomized studies, it is not possible to clearly identify which patients, if any, may benefit from postoperative adjuvant treatment after local excision. Results of Therapy Rectal Cancer Preoperative Irradiation Low-dose preoperative irradiation using 5 Gy in 1 fraction to 25 Gy in 10 fractions has been compared with surgery alone in several randomized prospective trials. None of these studies showed improved survival with preoperative radiation therapy (Fig. 54-7). Retrospective subgroup analysis did suggest a possible effect of preoperative treatment in two of these trials. In the Princess Margaret trial, patients with rectal cancer were randomly assigned to receive surgery either alone or with preoperative radiation therapy of 5 Gy given in one fraction. Although survival for all patients in the trial was virtually identical (Fig. 54-7), retrospective subgroup analysis of patients with Dukes' stage C cancer suggested a survival advantage for preoperatively irradiated patients. Based on this observation, the authors recommended that "this form of preoperative irradiation become routine." This conclusion is not justified for several reasons. If, for the entire group of patients, survival was virtually identical in the two groups, an apparent positive impact of preoperative irradiation on survival in some subgroups must be balanced by an apparent negative impact in other groups. Unfortunately, analysis of subgroups complementary to the Dukes' stage C patients was not presented. Moreover, no reliable technique is currently available to identify Dukes' stage C patients preoperatively. Although useful for generating hypotheses, retrospective subgroup analysis is generally an invalid statistical technique for reaching conclusions about treatment efficacy. The Veterans Administration study group found better survival among preoperatively irradiated patients (20 to 25 Gy) who underwent abdominoperineal resection. The hazard of retrospective subgroup analysis is illustrated by a later Veterans Administration trial, which strongly suggested that the original evaluation favoring the preoperatively irradiated patients resulted from an imbalance in prognostic factors rather than from any effect of treatment. Analysis of a trial conducted by the Stockholm Rectal Cancer Study Group suggested a clinically measurable effect of low-dose preoperative radiation therapy at high dose per fraction, when compared with surgery alone. In this trial, the patients were randomly assigned to receive 25 Gy in five fractions preoperatively or to receive surgery alone. Survival rates through 5 years of follow-up were virtually identical. Postoperative mortality was 8% in the preoperatively irradiated patients and only 2% in patients treated with surgery alone (P < 0.01). Among the patients who underwent curative surgery, the incidence of pelvic recurrence at 5 years was approximately 18% in the preoperatively irradiated group and approximately 31% in the surgical control group (P < 0.01). Although preoperative radiation therapy, as used in this study, appeared to provide improved local control, the overall benefit to patients was questionable in view of the lack of survival benefit and increased morbidity seen with the preoperative regimen. The most recent test of the value of low-dose preoperative radiation therapy was a Radiation Therapy Oncology Group (RTOG) trial in which patients with rectal cancer initially were randomly assigned to receive either 5 Gy preoperatively or immediate surgery. After surgery, all patients in both arms who were found to have tumor penetration beyond the rectal wall or involved nodes received 45 Gy to the pelvis. No differences in local recurrence, survival, or freedom from distant metastases were found between the two groups. In view of these data, low-dose preoperative radiation therapy should no longer be used in patients with rectal cancer. Retrospective studies suggest that higher doses of preoperative radiation therapy may be associated with improved survival and decreased pelvic recurrence. Moderate doses (30 to 40 Gy) of preoperative radiation therapy have been formally tested in several randomized prospective trials. At the Rotterdamsch Radio-Therapeutisch, patients with rectal cancer were randomly assigned to preoperative irradiation (34.5 Gy in 15 fractions to a pelvic and paraaortic field) or to operation alone. Freedom from local recurrence and survival were not significantly different in patients with T2 lesions. For patients with clinical evidence of T3 or T4 disease, 97% of those in the preoperative radiation therapy group subsequently had potentially curative resections, compared with only 68% in the operation-only group (P < 0.05). Irradiated patients with T3 or T4 tumors also had better 5-year rates of survival (P< 0.005) and freedom from local recurrence (P = 0.08). A larger randomized trial, using the same preoperative dose-fractionation scheme and field design, was conducted by the European Organization for Research and Treatment of Cancer. Although local recurrence was significantly lower among patients given preoperative radiation therapy (Fig. 54-8, A), survival was not affected (Fig. 54-8, B). In the second randomized trial of neoadjuvant radiation therapy conducted by the Veterans Administration Surgical Oncology Group, a preoperative dose of 31.5 Gy in 18 fractions to a pelvic and paraaortic field was compared with surgery alone in patients with rectal cancer. The 5-year survival rate was 50% in both arms of the study. Overall recurrence (distant and local) was also virtually identical, although a detailed analysis of patterns of failure was not reported. No benefit from low-dose preoperative radiation therapy has been observed in randomized prospective studies. Most randomized studies with higher doses of preoperative radiation suggest better local tumor control but not improved survival. Postoperative Adjuvant Therapy The advantage of postoperative adjuvant therapy of rectal cancer is that it allows consideration of pathologic factors in the selection of patients for this treatment. In trials of preoperative radiation therapy, 22% to 37% of patients randomly assigned to surgery alone had tumors that were limited to the bowel wall and therefore were at low risk for recurrence. [ref: 24,127,153] An additional 8% to 14% were found to have distant metastasis at surgery. [ref: 24,53,127,153] A postoperative approach allows the physician to use pathologic information to exclude 30% to 50% of patients who are unlikely to benefit from adjuvant therapy. A retrospective comparison of adjuvant postoperative irradiation with surgery alone in patients at high risk for local recurrence was performed at Massachusetts General Hospital. [ref: 78] For patients with gross extension of tumor beyond the rectal wall, metastatically involved lymph nodes, or both, the incidence of local failure was lower after postoperative adjuvant radiation therapy. Other retrospective studies suggested that local failure could be decreased by as much as 20% with the use of postoperative radiation therapy after resection of modified Astler-Coller B2-B3 and C1-C3 (equivalent to T3-4 or node-positive) tumors. [ref: 19,55,78,161,176,192] However, results from these retrospective studies have not been consistently corroborated by prospective studies. In an RTOG study, for example, patients with tumors that penetrated the rectal wall and those with positive nodes received postoperative radiation therapy to a dose of 45 Gy. The rate of local recurrence was 31% among patients who received 5-Gy preoperative radiation therapy and 34% among those who did not. [ref: 159] The 31% to 34% rate of local recurrence is not different from what would have been expected without the use of adjuvant therapy. In a study from the National Surgical Adjuvant Bowel and Breast Project, there was a modest nonsignificant difference in patients who were treated with postoperative radiation therapy, compared with surgery alone (16% and 25% local recurrence, P = 0.06). [ref: 44] Two separate randomized studies of postoperative radiation therapy (50 Gy) versus no adjuvant treatment in patients at high risk for local recurrence also failed to demonstrate a significant benefit for either local control or survival among patients who received postoperative radiation therapy. [ref: 11,173] Although results of adjuvant postoperative radiation therapy without chemotherapy have been disappointing, randomized prospective clinical trials have provided scientific evidence that improved survival and local tumor control after combined adjuvant postoperative radiation therapy and chemotherapy can be achieved in patients with T3-4 or node-positive tumors. The Gastrointestinal Tumor Study Group randomly assigned patients postoperatively to four groups: no further therapy; methyl-CCNU and 5-FU; pelvic radiation therapy (40 to 48 Gy); or pelvic radiation therapy (40 to 44 Gy), methyl-CCNU, and 5-FU. [ref: 50] Overall survival was significantly better among patients receiving chemotherapy plus irradiation than among those who had no adjuvant therapy. [ref: 35] Local recurrence rates were 24% in control patients, 27% in chemotherapy patients, 20% in irradiated patients, and 11% in patients who received the combination therapy. The results of this study suggest that both chemotherapy and radiation therapy are required for adjuvant treatment to have a favorable impact on local control; neither modality alone had a significant impact on local control. The value of adjuvant postoperative radiation therapy and chemotherapy in T3-4 and node-positive patients was confirmed by a randomized North Central Cancer Treatment Group (NCCTG) study. After complete surgical resection, the patients were randomly assigned to receive either postoperative pelvic radiation therapy or sequential postoperative chemotherapy (methyl-CCNU and 5-FU) and pelvic radiation therapy. Disease-free survival, overall survival, freedom from local recurrence, and freedom from distant recurrence were significantly improved in patients who received radiation therapy and chemotherapy. [ref: 98] The results of the Gastrointestinal Study Group and the NCCTG trials provided clear evidence that a combination of adjuvant postoperative chemotherapy and radiation therapy improves local control and survival in high-risk rectal cancer patients who have undergone complete surgical resection. A subsequent randomized NCCTG trial was undertaken to assess the contribution of the relatively toxic drug methyl-CCNU [ref: 16] to adjuvant therapy and to determine whether continuous infusion of 5-FU during radiation therapy resulted in better outcome than bolus administration of 5-FU. An improvement in survival (Fig. 54-9) was observed in patients who received continuous-infusion 5-FU [ref: 133]; details of the regimen are shown in Figure 54-10. Methyl-CCNU did not improve survival. A Gastrointestinal Tumor Study Group randomized study also found that methyl-CCNU did not improve survival. [ref: 51] Methyl-CCNU is no longer used in the adjuvant treatment of rectal cancer. The National Surgical Adjuvant Bowel and Breast Project conducted a clinical trial to assess the contribution of radiation therapy to adjuvant combined-modality treatment. Patients were randomly assigned to receive one of two chemotherapy regimens and either pelvic irradiation or no irradiation. Initial results from this study showed no difference in survival for patients who did or did not receive radiation therapy. The local recurrence rates were 6.7% and 11.3%, respectively (P = 0.045). [ref: 154] Although this difference was particularly in view of the potential toxicity associated with pelvic irradiation. [ref: 94] Longer follow-up is needed to assess definitively the findings of this study. Further clinical trials may be needed to evaluate definitively the contribution of radiation therapy to adjuvant treatment when used in conjunction with modern chemotherapy. Several randomized studies support the use of combined adjuvant postoperative radiation therapy plus chemotherapy in patients with T3-4 and node-positive rectal cancer. Currently, postoperative adjuvant combined-modality therapy is preferable to preoperative radiation therapy on the basis of evidence from randomized prospective studies that have consistently demonstrated a survival benefit for this approach and because a postoperative approach to adjuvant therapy allows exclusion of patients who would achieve little benefit from such treatment. Preoperative radiation therapy should not be used in patients with resectable rectal cancer unless a clear advantage of this approach, relative to postoperative adjuvant treatment, is demonstrated in a randomized clinical trial. Locally Advanced Rectal Cancer External-beam radiation therapy alone or in combination with chemotherapy provides palliation and modest prolongation of life but has only minimal curative potential in patients with locally advanced rectal cancer. At the Mayo Clinic, 65 patients with locally unresectable carcinoma of the large bowel were treated with external-beam radiation therapy (35 to 40 Gy) with or without 5-FU in a randomized prospective study. [ref: 120] Survival free of progression, median duration of symptomatic control, and overall survival were better in patients who received 5-FU and radiation therapy. Several papers have described results of treating patients with postoperative radiation therapy after subtotal resection of large bowel cancer. [ref: 2,19,55,162,181] At the Mayo Clinic, 17 patients received postoperative radiation therapy with doses of 40 to 60 Gy. Local failure was observed in 76% of patients, and the 5-year survival rate was 24%. The minimum follow-up among surviving patients was 5 years. [ref: 162] Other investigators have reported lower local failure rates (15% to 32%) but similar overall survival. [ref: 2,19,55,181] The reason for the wide range of results for local control is not clear but may relate to the manner in which local failure was defined and to the short duration of follow-up in some series. [ref: 162] Preoperative irradiation with doses of 45 Gy or more has been used in patients presenting with unresectable colon and rectal cancer. [ref: 34,39,135,169] Resectability rates after preoperative radiation therapy vary widely, ranging from about 50% to 75%. After resection, local failure occurs in approximately 36% to 45% of patients, so long-term local control is achieved in only 25% to 35% of these patients. Intraoperative radiation therapy (IORT) may improve these results. In most cases, patients who are considered for combined radical operation and IORT are given preoperative chemotherapy and external-beam radiation therapy. Theoretically, preoperative irradiation may provide tumor shrinkage, improve resectability, and potentiate IORT effects. Typically, between 45 and 55 Gy in 1.8-Gy fractions over 5 to 6 weeks is given, often in conjunction with 5-FU-based chemotherapy. After completion of external-beam radiation therapy and a 3- to 5-week recovery period, patients are reevaluated for metastatic disease and prepared for operation. Usually, aggressive surgical resection and IORT are performed only in patients without distant metastases. At operation, exploration is performed initially to detect the presence or absence of metastatic disease. If no evidence of metastatic disease is detected, tumor resection is performed. An attempt is made to perform a complete resection. Areas of suspected or known residual disease are evaluated jointly by the surgeon and the radiation oncologist to determine the feasibility of IORT. Sites of adherence or residual disease are then fitted with a suitably sized translucent cone designed specifically for delivery of the electron beam. The IORT dose is selected according to the amount of disease subsequent to surgical resection and ranges from 10 to 20 Gy. The two largest groups of patients in whom IORT has been used are patients with locally recurrent rectal cancer and those with primary unresectable disease. Details of therapy with external-beam radiation therapy, surgical resection, and IORT at the Mayo Clinic have been reported [ref: 65] and recently have been updated for 116 patients with recurrent colorectal cancer. [ref: 66] At 5 years, the survival rate was 18%, and the local failure rate was 40%. In an analysis of 106 patients treated with palliative surgical resection alone, palliative resection with external radiation therapy, palliative resection with IORT with or without external-beam radiation therapy, or palliative resection with brachytherapy, the use of IORT was associated with significantly improved survival (Fig. 54-11). [ref: 171] These results must be interpreted with caution because they are not from a randomized clinical trial. Prognostic factors that were related significantly to survival included the amount of residual tumor after surgical resection, the use of IORT, the symptomatic status of the patient, the degree of fixation of tumor to surrounding structures, and the performance status of the patient (Table 54-5). Results from patients with a history of pelvic irradiation before local recurrence are less satisfactory. In a Mayo Clinic series, survival at 5 years was only 13%, and local control at 4 years was 34%. [ref: 69] Thirty-nine patients with locally recurrent rectal or rectosigmoid cancer were treated at Massachusetts General Hospital with preoperative external-beam irradiation followed by IORT. [ref: 187] Nine patients did not receive IORT because it was not technically feasible, the tumor was unresectable or metastatic, or the tumor was completely resected with negative margins. For all 39 patients, the 5-year survival rate was 29%, and the disease-free survival rate was 21%. Five-year local control and disease-free survival rates in the 30 patients who received IORT were 26% and 19%, respectively. Local control was related to the degree of surgical resection performed before IORT. Local control was 62% and 18% in patients who had complete and partial resections, respectively. Four patients received no or minimal preoperative irradiation because of a history of prior pelvic irradiation. Of these 4 patients, 3 had local failure; 3 died of cancer and 1 of intercurrent disease. Currently, patients with locally recurrent rectal cancer who have a history of pelvic irradiation are not considered candidates for IORT at the Massachusetts General Hospital. [ref: 187] Twenty-six patients received external-beam radiation therapy and IORT for localized pelvic recurrence of large bowel cancer at Rush-Presbyterian Hospital. [ref: 97] The 3-year relapse-free and overall survival rates were 15% and 25%, respectively. Local failure occurred in 11 of 18 patients with gross disease after surgery and in 4 of 8 patients with microscopic disease. The local failure rate was higher in patients who received external-beam radiation therapy doses of less than 40 Gy (11 of 15 patients, 73%) than in those who received more than 40 Gy (4 of 11 patients, 36%). Like the group at the Massachusetts General Hospital, these investigators no longer use IORT when a full course of external-beam pelvic radiation therapy is not possible. External-beam radiation therapy with IORT has also been used in the treatment of primary locally advanced rectal cancer. [ref: 65,66,186] At the Massachusetts General Hospital, the actuarial overall disease-free survival rate at 5 years was 32% for patients with incompletely resected tumors and 53% for patients treated with adjuvant IORT (i.e., patients with completely resected tumors). Local control at 5 years was 60% for patients who underwent partial surgical resection and 88% for patients who received adjuvant IORT. At the Mayo Clinic, local control and survival rates at 5 years for 56 patients with primary locally advanced disease were 82% and 42%, respectively. [ref: 66]