Selection factors for full-thickness local excision are the same as or similar to those used for endocavitary radiation therapy. Consequently, this decision is based largely on findings of a digital rectal examination with increasing integration of transrectal ultrasound or MRI with endorectal coil.20,23,32 Patients with T1 tumors without adverse pathological features have a low incidence of local failure (5% to 10%) or lymph node involvement (<10%). With unfavorable pathological features (lymphovascular invasion, high grade, deep submucosal invasion, signet ring cell, or colloid histology)^33–36 or evidence of tumor invasion into or through the muscularis propria,^35,37,38 the local recurrence rate is at least 17%, and the risk of regional lymph node involvement is at least 10% to 15%.³³ In an analysis by the Massachusetts General Hospital of 40 patients who underwent local excision only, patients were categorized according to unfavorable clinical or pathological features.³⁵ Among patients with T1 or T2 cancers following local excision, Blumberg and associates reported positive lymph nodes in 10% of T1 cancers and 17% of T2 cancers.³⁹ In addition, among the total group of 159 patients, the incidence increased with the presence of lymphatic and vascular invasion (14% without vs. 33% with). Even among the 42 patients with the most favorable features (negative lymphatic and vascular space involvement, well- or moderately differentiated T1 cancers), 7% were found to have lymph node involvement. The overall 5-year survival rate for the whole group was 65% with a locoregional recurrence rate of 27%. Hager and associates reported on a series of 20 patients with T2 rectal cancer for which local excision was performed and who were otherwise thought to be “low risk” (well to moderately differentiated, nonmucinous, no lymphovascular invasion, and negative margins), despite which the incidence of locoregional failure was still 17%.³⁷ Others have reported locoregional failure rates as high as 43% following either local or transanal excision in patients with T2 cancers.⁴⁰ There have been no randomized trials that compared transanal full-thickness local excision alone for T1/T2 with or without adjuvant chemoradiation with low anterior resection or abdominoperineal resection. Recent published results of local excision of T1/T2 rectal lesions without adjuvant therapy show local recurrence rates of 3.4% to 18% for T1 lesions and 27% to 67% for T2 lesions; Table 78-1 summarizes these results.

Surgical Issues in Radical Resections

Radial Margins and Total Mesorectal Excision

The ability to obtain a negative lateral circumferential margin is associated with a decreased risk of local recurrence.60–63 In a multivariate analysis, circumferential margin involvement was the most powerful predictor of local recurrence (hazard ratio: 12.2) and of overall cancer mortality (hazard ratio: 3.2; Fig. 78-3). Heald and colleagues have advocated total mesorectal excision (TME) in conjunction with low anterior resection or abdominoperineal resection (APR) as the optimal surgical treatment for rectal cancer.⁵⁶ This technique involves removal of the entire rectal mesentery, including that distal to the tumor, as an intact unit. Complete distal TME is essential for clearance of any tumor deposits, which occur in 50% of T3 tumors with a maximal distal spread of 4 to 5 cm,⁶⁴ and is associated with increased frequency of local recurrence and decreased overall survival.^65,66

Sharp mesorectal excision combined with TME provides the optimal surgical strategy. In contrast with conventional blunt dissection techniques, sharp mesorectal excision facilitates nerve preservation, enables complete hemostasis, and emphasizes gentle handling to avoid tearing or disruption of the smooth outer surface of the mesorectum. Sharp TME has been shown to achieve a negative circumferential margin in 93% of resected specimens. Although no randomized trial of TME has been performed, TME has been evaluated prospectively in Sweden, where it has been introduced via a formal preceptorship-based training program. A 5-year prospective audit reveals a local recurrence rate of 7% following the addition of TME compared with a historical control rate of 23%.

Distal Mucosal Margin

The ability to perform sphincter-preserving surgery in nonirradiated patients is dictated by the requirements of a 2-cm distal margin rather than the traditional 5-cm margin.67–71 Only 2.5% of patients (usually with poorly differentiated and node-positive rapidly disseminating disease) had disease spread greater than 2 cm.³¹ There is no correlation between risk of local recurrence and the extent of distal margin in excess of 2 cm.^72–77

Proximal Extent of Lymph Node Dissection

Proximal lymph node dissection should extend just distal to the origin of the left colic artery. No evidence indicates a relationship between local recurrence and survival and dissection of deep iliac lymph nodes⁷⁸ or high ligation of inferior mesenteric pedicle.^79,80 Patients with pathologically positive nodes along the inferior mesenteric artery have very low 5-year survival rates.^81,82

Patterns of Failure

Following potentially curative standard or conventional surgical resection for adenocarcinoma of the rectum, the incidence of locoregional or distant treatment failure is related to the extent of transmural disease and associated involvement of regional lymph nodes by metastases.89–96 The incidence of locoregional failure is 8% to 21% in American Joint Committee on Cancer (AJCC) stage I disease (modified Astler Coller [MAC] stages A/B1), 29% to 44% in AJCC stage II disease (MAC stages B2/B3), and 50% to 61% in AJCC stage III disease (MAC stage C). The incidence of distant failure (as a component of failure) is up to 28% in AJCC stage I disease, up to 47% in AJCC stage II, and up to 74% in AJCC stage III disease. It has been claimed that when one compares these patterns of failure between multiinstitutional trial settings versus single-institutional and predominantly single-operator (surgeon) series, great differences in results can be seen. Table 78-2 summarizes these patterns of failure according to multiinstitutional versus single-institutional/single-operator series.^89–96 Although distant metastasis is most likely to be attributed as the cause of death in rectal cancer patients, the potential influence of locoregional failure as an antecedent event to the development of distant metastases is clinically important. Hence, decreasing locoregional failure is an important end point of treatment in rectal cancer. These data and rationale serve as the basis for consideration of adjuvant chemoradiotherapy in the management of rectal cancer and, in particular, as a standard for AJCC stage II (MAC stages B2/B3) and stage III disease (MAC stage C). It is important to note that limited retrospective data identify subsets of patients with stage I disease who may be considered for adjuvant therapy, as well as subsets of patients with T3N0 disease who may not require adjuvant therapy.^97,98 Willett and associates identified a subset of patients with stage I disease who have an increased incidence of locoregional failure following APR.⁹⁷ In an additional review of 117 patients with T3N0 disease, Marks and associates identified a favorable group of patients with moderately or well-differentiated cancers invading less than 2 mm into perirectal fat who had a 10-year actuarial locoregional failure rate of only 5% following surgery alone, compared with 29% in T3N0 patients without these favorable features.⁹⁹

Optimal Timing of Surgery Following Preoperative Therapy

The most common interval between the end of radiation and surgery is 6 to 8 weeks. Until the publication of the Lyon R90–01 randomized trial,¹⁰⁵ the optimal timing of surgery following preoperative therapy in rectal cancer was based on hypothesis or retrospective data. This study randomly assigned 201 patients with stage cT2/T3, NX, M0 into two treatment groups: (a) the short-interval group, in whom surgery was performed within 2 weeks of completion of preoperative radiation therapy (39 Gy and 13 fractions) versus (b) the long-interval group, in whom surgery was performed within 6 to 8 weeks after completion of preoperative radiation therapy. At a median follow-up time of 33 months, there was no difference in morbidity, local recurrence, or short-term survival between the two groups. These findings, along with the previously demonstrated findings that rectal cancers undergo slow tumor shrinkage over several months after radiation,¹⁰⁶ lend further support to the rationale that a longer delay before surgery, particularly in locally advanced tumors, might be desirable to allow for maximal tumor regression prior to surgery. Current randomized trials are systematically investigating the optimal timing of surgery to as long as 12 weeks following completion of neoadjuvant therapy.

In the most extreme example of delayed surgical intervention, Habr-Gama and colleagues have reported their continuing series of patients in whom they have delayed surgical intervention indefinitely following a clinical complete response (cCR) that was assessed 8 weeks following neoadjuvant chemoradiation by clinical, endoscopic, and radiographic studies.107,108 This approach is considered highly investigational; however, it is thought-provoking in terms of a treatment paradigm mirroring that for patients with cancers of the anal canal, in whom surgery is reserved for salvage of chemoradiation failures in patients who achieve a cCR to initial treatment.

Optimal Timing of Adjuvant Radiation Following Surgery

Lee and associates reported the results of a phase III study of postoperative adjuvant therapy in stages II and III rectal cancer that was designed to define the optimal sequencing of chemotherapy and radiation.¹⁰⁹ In this study, 308 patients were randomly assigned to early versus late radiation. Patients received 45 Gy in 25 fractions with 8 monthly cycles of 5-fluorouracil (5-FU)+leucovorin chemotherapy. Radiation began with the start of chemotherapy in the early radiation group versus with the start of the third cycle of chemotherapy in the late radiation group. With a median active follow-up of 37 months, the disease-free survival rate was significantly improved in the early radiation group compared with the late radiation group (81% vs. 71% at 4 years; P = 0.043). This finding was associated with an increase in both distant and locoregional disease recurrence in the late radiation group and with an overall recurrence rate of 17% in the early radiation group versus 27% in the late radiation group (P = 0.047). Although overall survival was not significantly different between the treatment arms, these results suggest that the timing of adjuvant postoperative radiation can have a significant impact on the outcome of patients with rectal cancer.

Radiation Treatment Planning

The design of pelvic radiation therapy fields is mainly based on the incidence and pattern of local recurrence after surgery. For locally advanced disease, recurrences in the soft tissues may arise from tumor extension to the pelvic sidewall, the bladder, prostate in men, the vagina in women, and the presacral space in all patients. This is especially true for tumors penetrating the mesorectal fascia or those with involved or close (<1 mm) circumferential radial margins (CRMs). Incomplete mesorectal excision is also at higher risk to leave residual microscopic tumor cells behind.

The major lymphatic spread is in a cephalad direction contained within the perirectal fascia and along the mesenteric system, that is commonly dissected by standard TME surgery. Outside the mesorectum is a space containing vessels, nerves and lymphatics, which is not usually dissected. Tumors at or below the peritoneal reflection tend to spread laterally along the internal iliac and the obturator chains. The external iliac nodes may only become at risk with anterior tumor extension and adjacent organ involvement. Lesions that extend to the anal canal or the lower third of the vagina can spread to the inguinal nodes. In this setting, the bilateral inguinal nodes are usually included in the pelvic radiation fields.

The relative frequency and sites of pelvic failures were delineated by the seminal work of Gunderson and Sosin.¹¹⁰ In this reoperative series of 75 patients, 69% of locoregional failures occurred in the soft-tissue of the pelvis or the anastomotic site, 42% developed also or exclusively in pelvic lymph nodes, in 25% the perineum was involved. A more contemporary series of 269 patients by Hruby et al. confirmed that the majority of local failures occurred in the posterior central pelvis (47%) or at the anastomosis (21%), whereas anterior recurrences (11%) were mainly restricted to T4 tumors. Perineal recurrences occurred in 16% of patients who underwent APR.¹¹¹

Irradiation Field Design

The whole pelvic radiation field should adequately cover the primary tumor/tumor bed as well as the primary nodes at risk. The intent of the boost is to treat the primary tumor and not to include the nodes. The exact size is determined by the size and location of the primary tumor. Whole pelvic and boost fields are usually treated with a three-field (posteroanterior and lateral) technique for patients receiving conventional treatment. Field shaping by blocks is used to spare additional small intestine anteriorly and superiorly, the posterior muscle and soft tissue behind the sacrum, and inferior to the symphysis pubis. Standard treatment involves the use of 3D treatment planning which provides the ability to plan and localize the target and normal tissues at all levels of the treatment volume, and to obtain dose volume histogram data. Guidelines for the definition and delineation of the clinical target volumes,112,113 as well as examples of conventional radiation field design,¹¹⁴ are routinely available.

The clinical utility of intensity-modulated radiation therapy (IMRT) treatment planning techniques is being investigated.115,116 IMRT treatment planning techniques can further decrease the volume of small bowel in the field.¹¹⁷ A retrospective analysis of 92 patients treated with either IMRT (n = 31) or conventional chemoradiation (n = 61) at the Mayo Clinic revealed that IMRT was associated with significantly less grade II+ gastrointestinal (GI) toxicity (32% vs. 62%).¹¹⁸ There were no significant differences in hematologic or genitourinary toxicity or in the rate of pathological complete response (pCR). The ultimate clinical benefit of IMRT compared with 3D or conventional treatment delivery is still being investigated.^115,119 Figure 78-4 shows examples of 3D and IMRT treatment plans.

Some patients may benefit from being treated on a “belly board” to allow for displacement of the small bowel up and out of the pelvis (Fig. 78-5). Nijkamp et al. showed that in 11 volunteers simulated for IMRT with a full bladder, a belly board had added value in decreasing bowel volume.¹²⁰ A systematic review of 33 reports revealed that treatment in the prone position resulted in a lower volume of small bowel treated versus a supine position. However, a more significant reduction could be obtained with the use of the prone position plus a belly board.¹²¹ Using a 3D planning system, Koelbl and associates found that in patients receiving postoperative radiation, the use of the prone position plus a belly board decreased the small bowel volume treated versus the supine position.¹²²

Brierley et al. analyzed the variation of small bowel volume in the pelvis before and during adjuvant pelvic radiation therapy for rectal cancer and reported that the displacement of small bowel from the posterior pelvis by bladder distention was not reliably maintained throughout the treatment course.¹²³ Therefore, any physical maneuver beyond the prone position may not be beneficial in all patients and should be tailored to the individual patient.

The role of hyperfractionated radiation has been examined in phases I/II trials.¹²⁴ In general, the pCR rates may be improved but at the expense of increased acute toxicity. Coucke et al. treated 250 assessable patients with cT3 to cT4 and/or N+ disease with 41.6 Gy at 1.6 Gy BID and reported a 92% actuarial 5-year local control rate but survival of only 60%.¹²⁵ Acute grade 3+ toxicity in the 5-FU plus twice-a-day radiation arm of Radiation Therapy Oncology Group (RTOG) R-0012 (1.2 Gy to 45.6 Gy, with a boost of 9.6 Gy to 14.4 Gy) was 42%. Hyperfractionated and accelerated fractionated radiation, especially in combination with chemotherapy, remains investigational.

Other investigational approaches, such as neutrons,¹²⁶ carbon ions,¹²⁷ protons,¹²⁸ and hyperthermia,^131–131 have been examined. An analysis of 3D treatment planning techniques suggests that the volume of small bowel in the radiation field is decreased with protons as compared with photons.¹³² Although treatment plans show improved normal tissue sparing, clinical studies have not shown a clear advantage compared with conventional 3D photon-based pelvic radiation therapy.

Techniques to Reduce and Manage Toxicity

Randomized trials have investigated the use of sucralfate enemas to decrease acute radiation proctitis, olsalazine, and mesalazine to decrease acute enteritis, octreotide acetate to decrease diarrhea,¹⁶⁰ and butyric acid to decrease chronic radiation proctitis.¹⁶¹ All of these trials have been negative. The radioprotector WR-2721 did not reduce toxicity in early trials, but there is a suggestion of a benefit in a more recent study.¹⁶² Rectally administered amifostine is well tolerated; however, its efficacy remains to be determined.¹⁶³ Other trials of amifostine have not shown clear benefits.¹⁶⁴ The best management of side effects is to decrease their incidence by designing the optimal treatment. There are a number of radiation therapy techniques available to achieve this. Despite their use, toxicities will occur. With appropriate doses and techniques of radiation, 5% to 25% of patients will have acute grade 3+ toxicity and approximately 1% of will have long-term toxicity to pelvic organs.

Surgical techniques to minimize small bowel injury are limited to the postoperative setting. Dexon or Vicryl mesh helps remove the small bowel from the pelvis. However, because the radiation component of postoperative chemoradiation does not begin until at least 4 months after surgery, the mesh may have already resorbed. Other techniques, such as an inflatable pelvic small bowel displacement prosthesis,¹⁶⁵ reconstruction of the pelvic floor, construction of an omental pedicle flap, and retroversion of the uterus, have had variable success.

A number of simple treatments for acute grades I to II toxicity from chemoradiation are available. In patients with grade 3+ toxicity, these should be combined with a treatment break. If radiation cannot resume within 2 weeks after a treatment break, its biological effectiveness is questionable and the risks of restarting likely outweigh any benefit.

Common treatments by organ sites are as follows:

These treatments as well as careful attention to treatment details will help decrease the acute and long-term toxicity of chemoradiation.

Preoperative Chemoradiation

A series of phase III trials have addressed a number of controversies in the use of preoperative chemoradiation for rectal cancer (Table 78-5). These issues are discussed below. When 5-FU is used concurrently with radiation, CI is the conventional regimen.^167,176 The NSABP R-04 trial compared preoperative chemoradiation with CI 5-FU versus capecitabine (with or without oxaliplatin). Compared with CI 5-FU, capecitabine had similar rates of pCR (22% vs. 19%), sphincter-sparing surgery (63% vs. 61%), and grade 3 + diarrhea (11%).¹⁴³ Hofheinz et al. randomly assigned 401 patients with CI 5-FU–based chemoradiation versus capecitabine-based chemoradiation. Patients who received capecitabine had equivalent pCR rates (6% vs. 7%) and their 5-year survival was noninferior (76% vs. 66%, P = 0.0004) compared with CI 5-FU.¹⁷⁷ Therefore, CI 5-FU– and capecitabine-based chemoradiation regimens are equivalent.

In contrast with chemoradiation, the regimen used in the postoperative adjuvant chemotherapy component of treatment is different. Based on the efficacy demonstrated in patients with stage III colon cancer, the combination of CI 5-FU plus oxaliplatin (FOLFOX) has replaced CI 5-FU as a standard postoperative regimen.¹⁷⁸ Other agents, such as irinotecan¹⁷⁹ and bevacizumab,¹⁸⁰ have not improved survival in the adjuvant setting and, therefore, are not used in the adjuvant management of rectal cancer.

The two randomized trials of preoperative versus postoperative chemoradiation for clinically resectable T3-4Nany rectal cancer (NSABP R-03¹⁴⁰ and the German CAO/ARO/AIO 94)¹³⁹ reported opposite results. The German trial completed the planned accrual of more than 800 patients and randomly assigned patients with rectal cancers smaller than 16 cm from the anal verge to preoperative chemoradiation (with CI 5-FU weeks 1 and 5) versus postoperative chemoradiation.¹³⁹ Patients were stratified by surgeon and all underwent a TME. Compared with postoperative therapy, patients who received preoperative therapy had a significant decrease in local recurrence (6% vs. 15%, P = 0.006), acute toxicity (27% vs. 40%, P = 0.001), chronic toxicity (14% vs. 24%, P = 0.012), and in those 194 patients judged by the surgeon pretreatment to require an APR, a significant increase in sphincter preservation (39% vs. 20%, P = 0.004). With a median follow-up of 40 months, there was no difference in 5-year survival (74% vs. 76%). A separate analysis revealed that the treatment center, schedule, and gender were independent prognostic factors for 5-year local control.¹⁸¹ The results were updated with a median follow up of 11.2 years. At 10 years, the local control benefit of preoperative versus postoperative therapy was sustained (local failure: 7% vs. 10%, P = 0.048). There was no difference in the 10-year cumulative incidence of distant metastasis (29%) or overall survival (60%).¹⁷³ As seen with other cancers, both surgeons and hospitals with higher volumes of rectal cancer surgery have improved outcomes compared with those with lower volumes.¹⁷¹

The NSABP R-03 trial accrued only 267 of the planned 900 patients.¹⁴⁰ Patients received induction chemotherapy followed by conventional chemoradiation and were randomized to receive it either preoperatively or postoperatively. Some patients underwent a local excision and a TME was not required. Compared with postoperative therapy, patients who received preoperative therapy had a significant improvement in 5-year disease-free survival (65% vs. 53%, P = 0.011), and a borderline significant improvement in 5-year overall survival (75% vs. 66%, P = 0.065). There was no difference in 5-year local recurrence (11%). There was a corresponding higher incidence of grade 4+ toxicity (33% vs. 23%) but the incidence of grade 3+ toxicity was lower (41% vs. 50%). Lastly, based on a prospective office assessment by the operating surgeon, there was no significant improvement in sphincter preservation (48% vs. 39%).

The results of the NSABP R-03 trial should be interpreted with caution as only 267 of the 900 planned patients were accrued, thereby limiting the statistical power to detect differences. The German trial met its accrual goals and based on the positive results preoperative chemoradiation remains the standard of care.

Novel Chemoradiation Regimens

Recent trials have examined the role of different chemoradiation regimens. The NSABP R-04 trial reported equivalent rates of pCR (22% vs. 19%), sphincter-sparing surgery (63% vs. 61%), and grade 3+ diarrhea (11%) with CI 5-FU– and capecitabine-based chemoradiation.¹⁴³

Four randomized trials examined the role of adding oxaliplatin to 5-FU– or capecitabine-based preoperative chemoradiation (Table 78-6). These include three European trials (Studio Terapia, Adjuvante Retto [STAR]-01,¹⁴¹ Actions Concertees and les Cancers Colorectaux et Digestifs [ACCORD],¹⁴² and the German CAO/ARO/AIO-94),¹⁴⁴ and one from the U.S. (NSABP R-04).¹⁴³ Three of the four reported no significant improvement in the pCR rate and a corresponding increase in acute toxicity.^{141–143,182} Preliminary data from the ACCORD trial revealed no improvement in 3-year local control (4% vs. 5%) or survival (88% vs. 85%) with the addition of oxaliplatin.¹⁸² A fifth trial (Pan-European Trial in Adjuvant Colon Cancer [PETACC]-6) is asking a similar question; however, the results are pending.

The benefit of adding targeted biological agents, such as bevacizumab and cetuximab, is being tested. Initial phases I/II trials of preoperative chemoradiation with capecitabine and oxaliplatin (CAPOX) + bevacizumab revealed pCR rates of 18% to 24%.183,184 However, more recent trials report increased acute toxicity and have been closed early as a consequence of increased acute toxicity.^185,186 Furthermore, given the lack of a survival benefit in the NSABP C-08 adjuvant colon cancer trial, the ultimate role of bevacizumab in the adjuvant management of rectal cancer remains unclear.¹⁸⁰

Phases I/II trials of new preoperative chemoradiation programs have been explored. The STAR-02 trial reported a 21% cCR rate in 60 patients treated with oxaliplatin, 5-FU, panitumumab, and 50.4 Gy.¹⁸⁷ Only 8% achieved a pCR with irinotecan, capecitabine, and 50.4 Gy in the Mannheimer Arbeitsgruppe für Gastrointestinale Tumoren (MARGIT) phase II trial.¹⁸⁸ Two trials of S-1–based chemoradiation have reported different results. A phase I trial of S-1 plus oxaliplatin and 50.4 Gy had a 13% pCR rate whereas a phase II of S-1 plus irinotecan and 54 Gy (with a limited pelvic radiation field) achieved a pCR of 35%.

The RTOG (RTOG 0247) performed a randomized phase II trial of preoperative chemoradiation (50.4 Gy) with capecitabine plus irinotecan (CAPIRI) versus capecitabine plus oxaliplatin (CAPOX).¹⁸⁹ A total of 146 patients were randomly assigned and the pCR rate was higher in those receiving CAPOX (21% vs. 10%); there were no differences in acute toxicity. The CAPOX was the experimental arm of the NSABP R-04 phase III trial which subsequently did not confirm a benefit in pCR compared with capecitabine alone.

Therapy based on KRAS expression is useful in patients with metastatic disease.¹⁸⁴ In the adjuvant setting, preliminary results from the phase II EXPERT-C trial (50.4 Gy+CAPOX+cetuximab) revealed a significant increase in 3-year survival in the 90 patients whose tumors were KRAS wild-type and received cetuximab versus those who did not receive cetuximab (96% vs. 81%, P = 0.035).¹⁹⁰ The end point of the trial (pCR) was not impacted by the use of cetuximab (11% vs. 7%). Nonetheless, this is the first trial to report that treatment based on a molecular marker had a significant impact on survival in patients with rectal cancer who were treated in the adjuvant setting.

Given the improvements in systemic chemotherapy there may be an opportunity to use preoperative radiation more selectively. In a prospective trial reported in abstract form, Cercek et al. treated 32 selected patients with uT2N1 or uT3N0-1 rectal cancer with neoadjuvant FOLFOX + bevacizumab.¹⁹¹ Of note, patients who required an APR were excluded. Pelvic radiation was reserved for patients who progressed preoperatively or following surgery had pT4, pN2, or positive margins. Of the 30 patients who underwent surgery none required radiation, the pCR rate was 27%, and 2 required postoperative radiation. This approach remains investigational and is being prospectively tested in the phases II/III Alliance N1048 trial.

Preoperative Short-Course Radiation

There are 12 modern randomized trials of preoperative radiation therapy.¹⁹² The fractionation varied from 5 Gy in 1 fraction to the more standard 5 Gy in 5 fractions. Most of the trials showed a decrease in local recurrence, and in five of the trials this difference reached statistical significance. Although in some trials a subset analysis revealed a significant improvement in survival, the Swedish Rectal Cancer Trial is the only one that reported a survival advantage for the total treatment group. Two meta-analyses report conflicting results. Although both revealed a decrease in local recurrence, the analysis by Camma et al.¹⁹³ reported a survival advantage, whereas the analysis by the Colorectal Cancer Collaborative Group¹³³ did not.

In the Swedish Rectal Cancer Trial, patients with cT1-3 rectal cancer were randomly assigned to 25 Gy in 5 fractions followed by surgery 1 week later versus surgery alone.¹⁹⁴ With 13-year follow-up, survival was significantly improved (38% vs. 30%, P = 0.008).¹⁹⁵ Of note, the local recurrence rate in patients with lymph node-positive disease who underwent surgery alone was 46%, illustrating the inferior results of surgery before the adoption of TME. This trial and the other 10 that preceded it did not mandate TME surgery. Therefore, although interesting from a historical perspective, these trials are not discussed further.

The Dutch CKVO 95-04 trial randomly assigned 1805 patients with cT1-3 disease to TME or 25 Gy in 5 fractions followed by TME.¹⁷⁴ Radiation significantly decreased local recurrence (8% vs. 2%) but there was no difference in 2-year survival (82%). With a 12-year median follow-up, 5-year local failure was higher with TME (11%) however, was significantly decreased to 5% with preoperative radiation.¹⁹⁶ The acute toxicity in the Dutch CKVO 95-04 trial included 10% neurotoxicity, 29% perineal wound complications, and 12% postoperative leaks.¹⁷⁵ In the patients who experienced postoperative leaks, 80% required surgery resulting in 11% mortality. In contrast to the earlier randomized trials of short-course radiation, multiple-field radiation techniques were used. Therefore, whether the increases in morbidity and mortality were a result of the learning curve associated with a new surgical technique, the 1-week interval between the completion of radiation and surgery, or both, is not known.

Preoperative Therapy: 5 Gy × 5 or Chemoradiation?

Historically, the primary reasons for not using short-course radiation is the lack of sphincter preservation, the inability to safely combine it with adequate doses of systemic chemotherapy, and the acute toxicity.¹⁹⁷ However, these shortcomings may be mitigated by (a) increasing the interval between the completion of radiation and surgery and (b) delivering chemotherapy sequentially (after radiation) as opposed to concurrently with radiation (Table 78-7).

The Dutch CKVO trial compared short-course radiation (25 Gy in 5 fractions) versus surgery alone.174,175 Compared with surgery alone, preoperative short-course radiation significantly increased local control. The outcomes of these two trials are not comparable because patients selected for treatment with short-course radiation included stages cT1-3, whereas the German trial included stages T3 and/or N+. However, more recent trials of short-course radiation have included patients with stages cT3 and/or N+ as well delivered sequential or postoperative chemotherapy, thereby allowing a more relevant comparison with chemoradiation.

The U.K. Medical Research Trial MRC C07 randomized 1350 patients with clinical stages I to III rectal cancer to 5 Gy × 5 or selective postoperative chemoradiation (45 Gy with concurrent 5-FU), which was delivered only to patients with a histologic CRM of less than 1 mm (12% of all patients with immediate surgery).²⁰¹ It should be emphasized that this trial did not compare short-course radiation with chemoradiation as only patients with close/positive CRM were selected to receive postoperative treatment. With a median follow-up of 4 years, patients who received preoperative compared with selective postoperative treatment has significantly lower 3-year local recurrence rates (4.4% vs. 10.6%, P < 0.0001), and a higher 3-year disease-free survival (77.5% vs. 71.5%, P = 0.013). A separate quality-of-life analysis revealed that the main adverse effect of treatment was male sexual dysfunction and the primary cause was surgery.¹⁹⁷ However, preoperative radiation also adversely impacted sexual and some aspects of bowel function.

Randomized Trials of Short-Course Radiation Versus Chemoradiation

There are two randomized trials of short-course radiation versus chemoradiation. The Polish trial from Bujko et al. and the Intergroup Australian/New Zealand TROG, AGITG, CSSANZ, RACS trial reported by Ngan et al. (Table 78-8).

The Need for Long-Term Follow-up

Local recurrences can occur late in patients with rectal cancer. In contrast to the results reported in the trials of adjuvant treatment of colon cancer in which 3-year, and possibly 2-year, disease-free survival predicts for 5-year survival,²⁰⁵ the INT 0114 postoperative rectal adjuvant trial confirmed that local control and survival continue to decrease beyond 5 years.¹⁶⁹ At 7 years the local recurrence rate was 17% and the survival was 56% compared with 14% and 64%, respectively, at 5-years.

Limiting the analysis to trials where all patients underwent a TME, a similar detriment in outcomes was seen with long-term follow-up. In the German CAO/ARO/AIO-94 trial, patients who received preoperative chemoradiation had an increase in local recurrence (7% vs. 5%) and decrease in survival (60% vs. 74%) at 10 years versus 5 years, respectively.¹⁷³ The incidence of local recurrence for all patients in the preoperative radiation arm of the Dutch CKVO trial of 5 Gy × 5 increased from 3% at a median follow-up of 3.5 years to 6% at a median follow-up of 6 years.²⁰⁶ These data underscore the importance of long term follow-up, regardless of which preoperative approach is used.

The Role of Postoperative Adjuvant Chemotherapy

Two randomized trials address whether postoperative chemotherapy is beneficial following preoperative therapy. The EORTC 22921 was a four-arm randomized trial of 1011 patients who received preoperative 45 Gy with or without concurrent bolus 5-FU+LV followed by surgery with or without 4 cycles of postoperative 5-FU+LV.²⁰⁷ Only 37% had a TME. The FFCD 9203 was a 2-arm randomized trial of 742 patients randomized to preoperative 45 Gy with or without concurrent bolus 5-FU+LV.²⁰⁸ However, all patients were scheduled to receive postoperative chemotherapy, and 73% did receive it.

The EORTC trial patients who received preoperative chemoradiation versus radiation had a significant decrease in local recurrence (8% to 10% vs. 17%, P < 0.001) but no difference in 5-year survival (65%). However, only 43% received 95% or more of the planned postoperative chemotherapy, which may explain the negative results. Furthermore, a subset analysis of the 785 patients with cT3 to cT4 tumors who had a R0 (complete) resection revealed that those patients who responded to preoperative chemoradiation had a survival benefit from postoperative chemotherapy if they had been downstaged to ypT0-2.²⁰⁹ The FFCD trial reported a similar decrease in local recurrence (8% vs. 17%, P < 0.05), a corresponding increase in pCR (11% vs. 4%, P < 0.05) but no survival benefit (68% vs. 67%).²⁰⁸

A pooled analysis of the two trials with a median follow-up of 5.6 years confirmed that patients who received preoperative chemoradiation compared with radiation had a significant decrease in local recurrence (11% vs. 15%, P = 0.0001); however, there was no difference in 5-year overall survival (66%).²¹⁰

These results need to be examined in perspective. Given that most patients did not receive adequate doses of postoperative chemotherapy in the European Organization for Research and Treatment of Cancer (EORTC) trial and the FFCD trial tested the impact of only 6 weeks of chemotherapy concurrent with preoperative radiation, the standard practice in many centers remains preoperative chemoradiation followed by surgery and 4 months of postoperative adjuvant chemotherapy. An analysis of the National Comprehensive Cancer Network (NCCN) database revealed that of 810 patients who received preoperative chemoradiation at eight NCCN institutions, 20% did not receive postoperative adjuvant chemotherapy. The most frequent reason for physician refusal was comorbid illness (54%), and the most frequent reason chemotherapy was not received when it was recommended by the medical oncologist was patient refusal (73%).²¹¹ Whether the ability to deliver postoperative chemotherapy is more successful in patients who receive short course radiation compared with chemoradiation is unknown.

Because there is at least some reluctance to administer postoperative adjuvant chemotherapy for rectal cancer patients, there is now a trend in clinical trial design to incorporate sequences of combination chemotherapy with or without targeted therapeutic approaches plus chemoradiation as a neoadjuvant strategy. With such an effort to enhance the ability to deliver chemotherapy to this population, there is the potential to improve disease free and overall survival benefit.

Beets and associates performed a pooled analysis of 2724 patients who received preoperative chemoradiation at European centers.²¹² Overall, 41% received postoperative chemotherapy and there was no benefit in disease-free survival in the subsets of patients with ypT0N0 or ypT3-4Nany disease. Patients with ypT1-2N0 disease had the greatest benefit although the hazard ratio (and 95% CI) was 0.45 (0.27 to 0.75).

Two retrospective analyses from the MD Anderson Cancer Center evaluated patients who were treated with neoadjuvant chemoradiation and adjuvant chemotherapy.213,214 In the first series, 117 patients received neoadjuvant chemotherapy and 63% also were treated with 5-FU and LV adjuvant chemotherapy. There was a significant disease free survival advantage for the patients whose tumors were downstaged after neoadjuvant treatment and who received adjuvant chemotherapy compared with those who were downstaged but did not receive the adjuvant treatment. The patients who did not experience tumor downstaging did not appear to benefit from the adjuvant chemotherapy. The second series of 470 patients included 305 who received adjuvant chemotherapy after neoadjuvant chemoradiation and surgery. The pathological tumor stage significantly predicted both the local relapse-free and 5-year overall survival. Those individuals with ypT0-2 disease in particular were noted to have the best outcome with both 5-year disease free and overall survival of 89% and local relapse-free survival of 96%. Because most patients were also treated with adjuvant chemotherapy, it was not possible to determine the added benefit of this approach compared with those treated with neoadjuvant therapy alone. Another retrospective study included 156 cT3 to cT4 rectal cancer patients who received neoadjuvant chemoradiation or neoadjuvant and adjuvant radiation or neoadjuvant chemoradiation and adjuvant chemotherapy with 5-FU. Improvement in both 5-year progression-free and overall survival was noted for the 71 patients in this group who received adjuvant chemotherapy. The group of patients who had significant downstaging after neoadjuvant chemoradiation (ypT0-2) also had improvement in survival compared with those who were not downstaged.²¹⁵

The National Cancer Institute GI Intergroup developed E5204 as a randomized phase III prospective study of adjuvant FOLFOX with or without bevacizumab for stage II and III rectal cancer patients who completed neoadjuvant chemoradiation and underwent surgical resection. The trial was closed prematurely when two adjuvant colorectal cancer trials evaluating the survival benefit of FOLFOX with or without bevacizumab (NSABP C-08: AVANT) showed that the addition of bevacizumab provided no additional benefit compared with FOLFOX alone. E5204 did accrue 355 patients and will therefore provide important information in the future about toxicity and tolerability of adjuvant FOLFOX for rectal cancer patients, patterns of failure, and disease-free survival.

Nomograms may also be helpful decision making. Valentini et al. pooled data from five major European clinical trials for rectal cancer and developed multivariate nomograms based on Cox regression analysis.²¹⁶ The nomograms were able to predict events with a c-index for external validation of local recurrence, distant metastases, and overall survival. These may be used as decision support tools by using the three defined risk groups to select patients for postoperative chemotherapy versus close follow-up.

Most investigators feel it is reasonable and use the same adjuvant chemotherapy for adjuvant colon and rectal cancer. For patients selected to receive postoperative adjuvant chemotherapy, 4 months (8 cycles) of mFOLFOX6 is recommended. However, its benefit remains controversial.²¹⁷

Induction Chemotherapy

Given the difficulty of patients tolerating and/or accepting postoperative adjuvant chemotherapy, the use of neoadjuvant chemotherapy prior to preoperative chemoradiation has been investigated. The Spanish GCR-3 randomized phase II trial compared this approach with conventional preoperative chemoradiation followed by surgery and postoperative chemotherapy.²¹⁸ A total of 108 patients received preoperative 50.4 Gy plus CAPOX and were randomly assigned to receive 4 months of CAPOX either by induction or adjuvant (postoperative). With the induction approach there was no detriment in the pCR rate (14% vs. 13%), grade 3+ toxicity was lower (17% vs. 51%, P = 0.00004) and the ability to receive all 4 chemotherapy cycles was higher (93% vs. 51%, P = 0.0001).

Garcia-Aguilar and colleagues have tested a similar approach.²¹⁹ In their randomized phase II trial, 144 patients with cN+ rectal cancer received preoperative chemoradiation (50.4 Gy + CI 5-FU) followed by surgery 6 weeks later vs. chemoradiation, and if they had a clinical response, mFOLFOX6 followed by surgery 11 weeks later. The pCR rate was nonsignificantly higher in the mFOLFOX arm (25% vs. 18%) with no increase in postoperative complications (40% in each arm). The preliminary data suggest that delaying surgery in the patient who has a response to preoperative and delivering additional chemotherapy is not detrimental to the tumor response or surgical complication rates.

Does the Distance from the Anal Verge Impact the Treatment Approach?

The limited data examining the impact of the distance from the anal verge on local recurrence are subset analysis not stratified by distance. There are no prospective randomized data. Furthermore, there are additional variables which may have contributed to differences in local recurrence. For example, TME was standard in the Dutch CVKO and German trials and not in the Swedish trial. All three trials included patients with tumors more than 12 cm from the anal verge in the “upper or high” category. Because the peritoneal reflection varies from 12 to 16 cm, some patients with tumors above the peritoneal reflection (colon cancer) were included in the three trials. Most investigators now limit preoperative treatment to tumors less than 12 cm from the anal verge.²²⁰ Lastly, distance measurements with a flexible proctoscope are less accurate than a straight proctoscope. Flexible scopes were used in the Dutch CKVO trial. The German trial used a straight scope. In the Swedish trial proctoscopic information was not mentioned and eligibility was limited to tumors “below the promontory as identified by barium enema.” The Polish trial is not included because all tumors were within reach by digital examination.²⁰³

Tumors identified as “high” in both the Dutch CKVO and Swedish trials (defined as >10.1 cm and 11 cm, respectively) had a lower incidence of local recurrence compared with mid and lower tumors. Short-course radiation did not significantly decrease local recurrence. By multivariate analysis, tumor location was an independent prognostic variable in the Dutch CKVO trial. In the 12-year update, the impact of preoperative radiation significantly increased as the distance from the anal verge increased (P = 0.03). However, excluding patients with CRM+, the relationship between distance from the anal verge and the impact of radiation became nonsignificant (P = 0.62). It is interesting to note that radiation did significantly decrease local recurrence for mid tumors in both trials whereas for lower tumors it was helpful in the Swedish trial.

In contrast, there was no significant difference in local recurrence between mid and upper tumors in the German trial.²²¹ Nash and colleagues reported that in a retrospective analysis of 627 patients with stage I-IV rectal cancer treated with either surgery alone or chemoradiation, the pelvic recurrence rate was lower in tumors 7 to 12 cm from the anal verge versus 0 to 6 cm from the anal verge (3% vs. 7%, P = 0.009).²²² However, mucosal, distant, and overall recurrences were not significantly different.

Given the conflicting data combined with the report from Guillem et al. confirming that the incidence of positive nodes is the same from 0 to 12 cm from the anal verge,²²⁰ treatment decisions, whether with preoperative chemoradiation or short-course radiation, based on the current definitions of low versus mid versus high should not be used regardless if the patient receives preoperative chemoradiation or short course radiation.

Is Clinical Staging Reliable to Select Patients for Preoperative Therapy?

The most common techniques for predicting T stage are transrectal ultrasound and high-resolution MRI. Ultrasound is used most commonly in North America, whereas many European investigators prefer high-resolution MRI. The advantage of MRI is its ability to identify patients likely to have close or positive CRM margins if they underwent initial surgery and therefore, would be better treated with preoperative therapy.²²³

The overall accuracy in predicting T stage is approximately 50% to 90% with ultrasound²²⁴ or high-resolution MRI,²²⁵ and 50% to 70% with CT or conventional MRI.²²⁶ Although fludeoxyglucose (18F)(FDG)-PET may be more accurate compared with CT for identification of metastatic disease,²²⁷ its use to restage patients following preoperative chemoradiation remains controversial.^228–231

Overstaging is common, especially when there is a fibrotic thickening of the rectal wall following preoperative chemoradiation. A reasonably high level of accuracy has been observed by phased array MRI for differentiating ypT0-2 from ypT3 disease.²³² Both diffusion-weighted MRI and FDG-PET have been used to monitor therapy response and to predict outcome to preoperative therapy.²³³ With FDG-PET there is a decrease in standardized uptake value on postradiation in responders compared with nonresponders, but the clinical value of this information remains to be determined.²³⁴

Identification of positive lymph nodes is more difficult. Overall, the accuracy in detecting positive pelvic lymph nodes with the above techniques is approximately 50% to 75%.²³⁵ Although the accuracy of MRI is similar to CT, it is improved with the use of external and/or endorectal coils. Both CT and MRI can identify lymph nodes measuring 1 cm or larger, although enlarged lymph nodes are not pathognomonic of tumor involvement. The accuracy of ultrasound for the detection of involved perirectal lymph nodes may be augmented when combined with fine-needle aspiration.²³⁶ Despite these advances, the ability to accurately predict the pathological stage following preoperative chemoradiation with MRI,^237,238 ultrasound,²³⁹ FDG-PET,^228–231 or physical examination²⁴⁰ remains suboptimal.

Although some series show no correlation,²⁴¹ most series suggest that there is improved outcome with increasing pathological response to preoperative therapy.^244–244 A retrospective review of 566 patients who achieved a pCR after receiving a variety of preoperative chemoradiation regimens at multiple European centers was reported by Capirci and associates.²⁴³ With a median follow-up of 46 months the local recurrence rate was only 1.6% and the 5-year disease-free and overall survival rates were 85% and 90%, respectively. A pooled analysis 3105 patients from 14 studies confirmed a significant improvement in local recurrence, distant failure, and disease-free and overall survival for the 16% of patients who achieved a pCR (ypT0N0M0) compared to those without a pCR.²⁴⁵

A pooled analysis of 27 articles identified 484 of 3105 patients treated with preoperative chemoradiation who achieved a pCR.²⁴⁵ The overall incidence of pCR was 15% to 27% and the median follow-up was 4 years. Patients who achieved a pCR had a significantly higher 5-year disease survival (83% vs. 66%, P < 0.0001).

Although a number of molecular markers are predictive of outcome in colorectal cancer,²⁴⁶ they have had varying success in identifying patients with rectal cancer who may respond to preoperative therapy. Kuremsky et al. reviewed 1204 studies examining a total of 36 molecular biomarkers which may have predictive value.²⁴⁷ Restricting the analysis to patients treated with preoperative chemoradiation and to gene products examined by five or more studies, only p53, epidermal growth factor receptor (EGFR), thymidylate synthase, Ki-67, p21, and bax/bcl-2 met these criteria. Of these, quantitatively evaluated EGFR or EGFR polymorphisms, thymidylate synthase polymorphisms, and p21 have been identified as promising candidates that should be evaluated in larger prospective trials for their ability to guide preoperative therapy.

More recent trials of preoperative chemoradiation have identified prognostic significance of thymidylate synthase,250–250 nuclear factor kappa B (NF-κB) expression,²⁵¹ k-RAS, and p53.²⁵² Because the studies are limited retrospective trials and because most do not examine multiple markers, the need for adjuvant therapy should still be based on T and N stage. The presence of acellular mucin following preoperative chemoradiation is seen in approximately 15% of patients and has no prognostic significance.^253,254

Is Radical Surgery Necessary Following Preoperative Chemoradiation?

Patients who are downstaged to ypT0 following chemoradiation have a 5% incidence of positive nodes and, therefore, a low nodal recurrence rate.²⁷¹ Are there alternative surgical approaches for those patients?

Habr-Gama and colleagues have questioned the value of radical surgery in patients who had a biopsy proven cCR.²⁷² Of note, their series included patients with cT1 to cT3 disease and has not been reproduced by other investigators. In series limited to patients with cT3 disease who received preoperative chemoradiation, radical surgery is still necessary to fully evaluate the degree of pathological response. Neither posttreatment ultrasound^239,273 or physical examination²⁷⁴ are sufficient. Studies on the use of of PET scan^230,231,275 and diffusion MRI²⁷⁶ as noninvasive measures of response are being conducted and have reported mixed results. Although Kalff and associates reported FDG-PET identification of residual viable tumor in 63 patients following chemoradiation had a high positive predictive value (0.94; 95% confidence interval, 85% to 99%),²⁷⁷ other groups have reported opposite results.

Glynne-Jones and associates reviewed 218 phase II and 28 phase III trials of preoperative radiation or chemoradiation and confirmed that clinical and/or radiologic response does not sufficiently correlate with pathological response and do not recommend a “wait-and-see” approach to surgery following preoperative therapy.²⁷⁸

Standard therapy for resectable rectal cancer has been radical or conventional surgical resection (inclusive of the possibility of TME) with increasing integration of preoperative or postoperative adjuvant therapy. In highly select patients, this approach has often been challenged by the use of more conservative local measures. Table 78-1 provides a summary of selected series.

Local treatment for rectal cancer was first applied to patients with medical contraindications to radical surgery. For example, significant cardiopulmonary disease may preclude extensive surgery because of high surgical mortality. Patient blindness makes subsequent colostomy self-care difficult. In recent decades the indications for local excision have been broadened with the goal of sphincter preservation.

As discussed above, endocavitary irradiation alone has been used for early (T1) or noninvasive tumors. For more advanced tumors (cT2 to cT3 and/or lymph node-positive) it is usually combined with a temporary iridium-192 implant and/or pelvic radiation.

Early localized tumors (3% to 5% of rectal cancers) include small, exophytic, mobile tumors without adverse pathological factors (i.e., high-grade, blood vessel invasion, lymphatic vessel invasion, colloid histology, or the penetration of tumor into or through the bowel wall) and are adequately treated with a variety of local therapies such as endocavitary radiation or local excision.

Sphincter Function

Prospective assessment of functional results is limited. The groups from Memorial Sloan Kettering³⁰⁰ and Gemelli Hospital, Rome,²⁹⁷ report 94% and 100% good-to-excellent function, respectively. Using a different scale, investigators from Fox Chase Cancer Center²⁹⁸ reported 82% good-to-excellent function, the University of Pennsylvania²⁹⁵ reported 92% satisfactory function, and MD Anderson²⁹⁴ reported that all patients were continent. In the preoperative setting sphincter function was reported good to excellent in 88% to 91%.^306,307