Combined Modality Therapy
Patients with locally advanced rectal cancer are at risk for local recurrence in addition to distant metastases.10,13 Rectal cancer is particularly prone to local recurrence due to the absence of serosa surrounding the rectum, the proximity of the rectum to pelvic structures and other organs, and technical difficulties in obtaining wide circumferential margins in the pelvis. Accordingly, for the majority of patients with locally advanced rectal cancer (stage II and stage III), combined modality therapy consisting of (1) neoadjuvant chemoradiation, (2) surgery, and (3) adjuvant chemotherapy is recommended to decrease the risk of both local and distant recurrence. All patients with locally advanced rectal cancer should undergo preoperative consultation with medical oncology, radiation oncology, and the extirpative surgeon and cases should be presented at multidisciplinary cancer conference to determine the best course of treatment for each patient.
Obstruction as a Presenting Symptom
Up to 30% of patients with colorectal cancer present with acute intestinal obstruction.11 Most patients who present with obstructed rectal cancer have T3 or T4 disease, and thus neoadjuvant chemoradiotherapy is preferred prior to tumor resection. In this situation, both rectal stenting and emergency diversion have been used as a bridge to neoadjuvant chemoradiotherapy and subsequent extirpative surgery. Stenting of rectal tumors is frequently associated with pain and stent migration; consequently, fecal diversion is the recommended approach in patients who present with acute obstruction.12 Our preferred approach is a laparoscopic loop sigmoid colostomy (if there is sufficient colonic length) in patients who will require a permanent stoma and a laparoscopic loop ileostomy in patients in whom a sphincter-preserving procedure is planned. Patients with a competent ileocecal valve and a complete obstruction may require proximal transverse colostomy. Patients who present with partial obstruction, in our experience, will tolerate neoadjuvant therapy without fecal diversion if a pediatric esphagogastroduodenal scope can be passed beyond the tumor.
Neoadjuvant Chemoradiation and Radiation Therapy
Historically, radiation for rectal cancer has been delivered both alone and in conjunction with chemotherapy and has been administered in both the neoadjuvant and the adjuvant setting. The goals of neoadjuvant treatment for rectal cancer are to (1) decrease local recurrence, (2) downstage the tumor, (3) improve sphincter salvage, and (4) decrease the toxicity of radiotherapy (compared to adjuvant treatment).
Neoadjuvant chemoradiotherapy is considered the standard of care for patients with T3 and T4 rectal cancer based on the results of randomized controlled trials. Additional relative indications for neoadjuvant chemoradiotherapy include node positivity on MRI or TRUS in a patient with T1/T2 rectal cancer, distal tumors for which an abdominoperineal resection (APR) is thought to be required and tumors in which the circumferential margin is threatened.
The German rectal cancer established neoadjuvant chemoradiotherapy as the standard of care for patients with T3/T4 or node-positive rectal cancer.13 In this trial, 823 patients were randomized to preoperative chemoradiotherapy (total of 5040 CGy in 28 fractions plus infusional fluorouracil given during the first and fifth weeks of treatment) or postoperative chemoradiotherapy (the same treatment with the addition of a 540-cGy boost to the tumor bed). All patients in this study underwent total mesorectal excision (TME) (6 weeks after conclusion of chemoradiation for patients in the preoperative treatment arm) and also received four cycles of bolus adjuvant fluorouracil. In this trial, the incidence of local recurrence was significantly decreased in the patients assigned to preoperative chemoradiotherapy (5-year local recurrence 6% vs. 13%, p = 0.006). There was no difference in overall survival comparing patients assigned to preoperative treatment to those assigned to postoperative treatment (5-year overall survival 76% vs. 74%, p = 0.80). In addition, patients who received preoperative chemoradiotherapy had significant downstaging, improved sphincter salvage, and decreased acute and long-term toxic effects, compared to patients randomized to postoperative chemoradiotherapy.
Neoadjuvant chemoradiotherapy has been compared to neoadjuvant radiotherapy alone in several trials. A recent meta-analysis of six randomized controlled trials comparing neoadjuvant chemoradiotherapy to neoadjuvant radiotherapy in patients with T3/T4 rectal cancer demonstrated decreased local recurrence in patients receiving neoadjuvant chemoradiotherapy.14 There was no difference in overall survival or sphincter preservation; however, increased acute toxicity was seen in the chemoradiotherapy group.
A number of randomized controlled trials have been completed assessing various chemotherapy regimens used during neoadjuvant radiotherapy. The current standard of care is either infusional 5-FU or oral capecitabine.15 Multiple randomized controlled trials have demonstrated increased toxicity and no improvement in pathologic complete response rate, sphincter salvage, or overall survival with the addition of neoadjuvant oxaliplatin to long-course chemoradiotherapy with infusional 5-FU.16-19 Similarly, the addition of neoadjuvant irinotecan to chemoradiotherapy has not been associated with improvements in overall survival, disease-specific survival, or pathologic complete response rates.20
The optimal timing of surgery following long-course chemoradiotherapy has not been established. Traditionally, this interval has been 6 weeks, based on the German Rectal Cancer Trial.13 Because tumor regression takes time, it has been suggested that waiting a longer period between the conclusion of chemoradiation and surgery may result in higher pathologic complete response rates and better outcomes. A meta-analysis of 13 uncontrolled trials demonstrated that waiting beyond 8 weeks after the conclusion of radiotherapy results in a 6% increase in pathologic complete response rate; however, morbidity, sphincter preservation, and oncologic outcomes (overall survival, disease-free survival, R0 resection rate) were not improved with a longer waiting time.21 The clinical significance of an improved pathologic complete response rate in the absence of improved oncologic outcomes is debatable. Our practice is to wait 8 to 12 weeks between neoadjuvant therapy and surgery.
Short-course preoperative radiation represents an alternative to traditional long-course chemoradiotherapy. Short-course radiotherapy (25 Gy in 5 fractions) was compared to surgery alone in the Swedish Rectal Cancer Trial.22 In this trial, patients underwent surgery 1 week following completion of radiotherapy. In patients who underwent preoperative radiation, both local recurrence (5-year local recurrence 11% vs. 27%, p < 0.001) and overall survival (5-year overall survival 58% vs. 48%, p = 0.004) were improved compared to patients who underwent surgery alone. This is the only trial that has demonstrated a survival benefit to neoadjuvant radiotherapy. The Dutch rectal cancer trial compared short-course neoadjuvant radiotherapy with TME to TME alone.23,24 In the Dutch trial, patients undergoing short-course neoadjuvant radiotherapy had improved local recurrence (5-year local recurrence 5.6% vs. 10.9%, p < 0.001) compared to those undergoing TME alone; however, overall survival was not different (5-year overall survival 64.2% vs. 63.5%, p = 0.902) between the two groups. Two randomized controlled trials have compared short-course radiation to long-course chemoradiation. Both of these trials demonstrated improved pathologic complete response rates with long-course chemoradiotherapy. There were no differences seen in recurrence-free or overall survival comparing short-course radiotherapy to long-course chemoradiotherapy in these trials; however, both of these trials were underpowered to demonstrate equivalence of these treatments.25,26 In North America, long-term chemoradiotherapy is generally the preferred approach. In patients who present with metastatic disease, short-course radiotherapy is occasionally used to prevent delays in initiation of systemic therapy.
Adjuvant Chemoradiation and Radiation Therapy
Following the German Rectal Cancer Trial, which established neoadjuvant chemoradiotherapy as the standard of care, indications for adjuvant chemoradiation or radiation are limited. For patients with resected stage II and III rectal cancer who do not undergo neoadjuvant therapy, postoperative chemoradiotherapy is preferable to surgery alone.
A number of older clinical trials examined outcomes with adjuvant chemoradiotherapy or radiotherapy. A large meta-analysis including seven trials comparing surgery alone to surgery followed by radiotherapy demonstrated that adjuvant radiotherapy was associated with decreased local recurrence (5-year local recurrence 15.3% vs. 22.9%, p < 0.001), but did not result in improved overall survival.27 In addition, two randomized controlled trials comparing adjuvant chemoradiotherapy (using older chemotherapy regimens) to radiation alone demonstrated decreased local recurrence and improved survival.28-30
Intraoperative radiotherapy (IORT) has been used in conjunction with neoadjuvant chemoradiotherapy in a number of centers in the United States, Europe, and Asia, with the goal of further decreasing the risk of local recurrence in patients with locally advanced rectal cancer. IORT delivers a boost of radiotherapy to the operative bed at the time of surgery. The goal of IORT is to sterilize any remaining microscopic foci of tumor following surgical resection (Fig. 113-1A). The benefit of IORT on locoregional recurrence is thought to be additive to the benefit of neoadjuvant chemoradiotherapy.
Setup and delivery of intraoperative electron beam radiation therapy (IOERT): A. The operative field of a patient following resection of locally advanced rectal cancer abutting the sciatic notch and the left pelvic sidewall. The internal iliac artery and vein have been ligated and the tumor has been dissected off of the left pelvic sidewall. B. Intraoperative setup for IOERT. Cone is used to deliver focused intraoperative electron beam radiation therapy to the operative bed. C. Delivery of IOERT using a mobile linear accelerator.
Two types of IORT are commonly used in the treatment of locally advanced rectal cancer: intraoperative electron beam radiation therapy (IOERT) and high-dose rate intraoperative radiation therapy (HDR-IORT).31 IOERT is delivered using electrons using a fixed or mobile linear accelerator. IOERT can be delivered in any operating room (there is no need for a lead shielded operating room), because electrons are used which do not penetrate the tissue as deeply as conventional radiotherapy. Multiple cone shapes and sizes can be adapted to fit differing tumor bed volumes with IOERT (Fig. 113-1B, C). One limitation of IOERT is that overlapping cones may result in radiation “hot spots” with increased toxicity, particularly peripheral neuropathy. In contrast, HDR-IORT is delivered using high-dose rate brachytherapy administered by catheters placed in the tumor bed following en bloc resection of the tumor (Fig. 113-2A). Catheters are placed into a silicon flap that can be cut to fit the tumor bed (Fig. 113-2B). The entire dose of HDR-IORT is administered at the time of surgery via a wire with an iridium source at the end that passes through the catheters (Fig. 113-2C). In contrast to IOERT which often requires multiple applications to cover the treatment area, the full dose of HDR-RT can generally be administered with a single flap. HDR-RT also results in a higher surface dose and a lower dose at 2 cm than IOERT.
Setup and delivery of high-dose rate radiation therapy (HDR-RT): A. The silicon flap used to deliver HDR-RT. B. Placement of the silicon flap and brachytherapy catheters in the operative bed for HDR-RT. C. Delivery of HDR-RT.
One randomized controlled trial has compared surgery alone to surgery plus IORT in patients with stage II and III rectal cancer who had received neoadjuvant chemoradiotherapy.32 This trial did not demonstrate any difference in local recurrence or overall survival comparing patients who received IORT plus surgery to those who received surgery alone, although the trial was significantly underpowered. Observational studies have demonstrated conflicting results with respect to the efficacy of IORT: some studies have noted decreased local recurrence with IORT,33-37 while other studies have noted no benefit of IORT.38,39 A recent observational study reported that patients with microscopically involved circumferential resection margins who were treated with IORT appear to have significantly improved local recurrence-free survival (5-year local recurrence-free survival 84% vs. 41%, p = 0.01) when compared to patients treated with surgery alone.33 In this study, there was no significant improvement in local recurrence-free survival seen with IORT among patients who underwent an R0 resection. We recommend having IORT available intraoperatively for all patients who undergo surgery for locally advanced rectal cancer with close lateral margins.
INDICATIONS AND PRINCIPLES OF RESECTION
Surgical resection remains the cornerstone of treatment for patients with locally advanced rectal cancer. The primary goal of surgery for locally advanced rectal cancer is complete removal of the tumor with histologically negative margins. Secondary goals include preservation of bowel continuity and anorectal sphincter function and preservation of genitourinary function.
The choice of operative procedure in patients with locally advanced rectal cancer is based on the location of the tumor relative to the dentate line, involvement on the sphincters, presence of invasion into other intra-abdominal organs, size of the tumor, and accommodating features of the pelvis. Sphincter-sparing procedures (i.e., low or very low anterior resection) can be attempted in patients in whom there is no sphincter involvement and a histologically negative distal margin can be achieved. In contrast, patients in whom a histologically negative distal margin cannot be obtained or who have sphincter involvement require an APR. Patients who have tumors that invade other intra-abdominal organs require en bloc multivisceral resection of all involved organs along with the rectum—this situation frequently necessitates a pelvic exenteration.
The principles of resection of locally advanced rectal cancer are to (1) obtain histologically negative margins, (2) perform a TME, and (3) perform an adequate lymphadenectomy.
The proximal margin should be at least 5 cm from the tumor to ensure an adequate lymphadenectomy as well as a well-vascularized anastomosis. Often this margin is greater and determined by vascular ligation for proper lymphadenectomy. For tumors of the upper rectum, a distal margin of 5 cm with a tumor-specific TME is recommended.40 For tumors of the mid and lower rectum, a distal margin of at least 2 cm is recommended when feasible. A distal margin of 1 cm is acceptable for tumors located at or below the mesorectal margin. It is essential that the distal margin of resection is negative on the final pathology report; a positive distal margin is associated with increased local recurrence and decreased overall survival.41,42 In patients who undergo radiation and TME, there does not appear to be increased recurrence in patients who have a close, but negative margin. A meta-analysis of 13 studies demonstrated that in patients who undergo TME and radiotherapy, a distal margin of less than 1 cm was not associated with increased local recurrence as long as the final pathologic margin is negative.43 Conversely, a distal margin of less than 1 cm was associated with increased local recurrence and decreased survival in patients who did not undergo either TME or radiation. Generally, we accept a distal margin of 5 mm in patients who have undergone neoadjuvant chemoradiotherapy. A negative CRM is critical in minimizing local recurrence in patients undergoing surgery for locally advanced rectal cancer. Numerous studies have demonstrated that involvement of the CRM is associated with a poor prognosis including increased local recurrence and decreased overall survival.44-46
Total mesorectal excision facilitates en bloc removal of the rectum and the associated mesentery and lymphatics. TME is performed using sharp dissection between the visceral and parietal layers of the endopelvic fascia. Proper TME technique is important in ensuring optimal oncologic outcome, preservation of autonomic nerves, and good intraoperative hemostasis.47 Tumor deposits in the mesorectum may extend up to 4 cm distal to the primary cancer.48 Accordingly, excision of the mesorectum extending 5 cm distal to the lower margin tumor is recommended for patients with upper rectal tumors, while full TME is recommended for patients with mid or low rectal tumors.40 Local recurrence rates following TME range from 5% to 8%, while local recurrence rates from conventional surgery range from 14% to 45%.30,46,49–51
Mesorectal lymphadenectomy is important for staging and for local control in patients with locally advanced rectal cancer. Vascular ligation at the origin of the superior hemorrhoidal artery with resection of all associated lymphatic drainage and full TME generally ensures an adequate lymphadenectomy.40 The benchmark of removal of 12 lymph nodes has been established for colorectal cancer.52
PROCTECTOMY AND COLOANAL ANASTOMOSIS
The total proctectomy with coloanal anastomosis is a sphincter-sparing procedure, which involves removal of the sigmoid colon and the rectum to a level where the distal margin is cancer-free using the technique of total mesorectal excision. A number of maneuvers can be used to gain length to facilitate a tension-free anastomosis: (1) complete mobilization of the left colon including the splenic flexure, (2) high ligation of the inferior mesenteric artery at the level of the takeoff from the aorta, and (3) division of the inferior mesenteric vein at the inferior border of the pancreas.53 A primary anastomosis is then completed between the descending colon and the anus. Options for constructing the anastomosis include end-to-end, side-to-end, colonic J-pouch, or transverse coloplasty. Reconstruction with J-pouch provides superior short-term bowel function to straight coloanal anastomosis; however, at 2 years there are no significant functional differences.54,55 Randomized trials have demonstrated no difference between an end-to-side anastomosis and colonic J-pouch reconstruction. Functional results with transverse coloplasty, however, have been mixed;56-58 thus, J-pouch reconstruction or end-to-side reconstruction is preferred when technically possible following proctectomy with coloanal anastomosis.
Defunctioning stomas should be considered for patients undergoing low or very low anterior resection with TME for locally advanced rectal cancer. A large meta-analysis of randomized and nonrandomized studies including over 11,000 patients demonstrated significantly increased rates of clinical anastomotic leak and reoperation in patients who did not undergo diversion.59 The two most frequent factors associated with anastomotic leak in this study were male sex and low anastomosis. Diversion can be accomplished with either a diverting loop colostomy (generally transverse) or a loop ileostomy. Due to the ease of reversal, diverting loop ileostomy is generally preferred over diverting loop colostomy. Diverting loop ileostomy may, however, be associated with increased frequency of dehydration due to high stoma output.40
Abdominoperineal resection involves excision of the sigmoid colon, rectum, and anus with creation of a permanent colostomy. APR is necessary in patients with low rectal cancers involving the sphincter complex or when a negative distal margin cannot be obtained. Additionally, an APR generally results in improved quality of life and functional outcomes compared to low anterior resection in patients with poor preoperative continence. Generally, the abdominal portion of the APR is performed first using TME technique similar to a low anterior resection. The perineal portion of an APR can then be performed in either lithotomy or prone position. In histologic studies, APR specimens have higher rates of CRM involvement and tumor perforation than low anterior resection specimens.60 Intraoperative tumor perforation and CRM involvement are associated with increased local recurrence and decreased overall survival.45,46,61 To avoid CRM involvement and to decrease the tumor perforation rate, a cylindrical APR should be performed, resecting the levators widely en bloc with the rectum and anus.62
Pelvic exenteration, the en bloc resection of pelvic organs in addition to the rectum, is indicated for T4 tumors which invade adjacent organs or bony structures.63 Total pelvic exenteration refers to en bloc resection of the rectum, bladder, and internal reproductive organs (prostate and seminal vesicles or uterus, ovaries, and/or vagina). Modified pelvic exenterative procedures can often be performed for locally advanced rectal cancer depending on the extent of the tumor including posterior exenteration (removal of rectum, anus, uterus, ovaries, and posterior vaginal wall), supralevator exenteration (a modified exenteration in which a sphincter-saving procedure can be performed), and composite exenteration (en bloc resection includes bony structures such as sacrum or coccyx). Pelvic exenteration is associated with low rates of mortality (0% to 3%), but high rates of morbidity (up to 60%).64-66 In single institution studies, patients who undergo pelvic exenteration for locally advanced rectal cancer have 5-year local recurrence rates ranging from 11% to 22%.65,67 Similarly, in a large population-based analysis including 1741 patients with T4 locally advanced rectal cancer without metastases, patients who were treated with multivisceral resection had improved overall survival compared to patients undergoing standard resection (HR = 0.81, p = 0.007) with no increase in early mortality.68
Sacrectomy is required in conjunction with pelvic exenteration when tumor extends dorsally into the sacrum or when tumor extends laterally to the pelvic side walls or to the pelvic floor and wider dorsal access is needed to ensure complete resection of the tumor.69 Contraindications for pelvic exenteration with sacrectomy are unresectable extra-pelvic disease, S1/S2 involvement, and tumor infiltration through the sciatic notch. Pelvic exenteration with sacrectomy generally consists of four phases: the abdominal phase, the perineal phase, the sacral phase, and the reconstructive phase (Fig. 113-3).70 The patient is initially positioned in lithotomy position. During the anterior phase, a complete laparotomy is performed in order to rule out metastatic disease to assess the primary tumor and to determine the level of sacral transection needed. The anterior plane of dissection is established based on the extent of anterior tumor involvement and then the internal iliac artery and vein are divided (Fig. 113-3A). The sacral nerve plexus is identified and the S1 and S2 nerve roots are identified and protected. The perineal phase is then performed and the urogenital diaphragm and levator ani muscles are divided widely (Fig. 113-3B). Anterolaterally, the perineal dissection is connected with the abdominal dissection. The abdomen is then closed temporarily, the patient is positioned in the prone position, and the sacral phase of the operation is undertaken. The gluteus muscle is detached from the sacrum exposing the posterior sacrum (Fig. 113-3C). The sacrotuberous and sacrospinous ligaments and the piriformis muscles are then detached (Fig. 113-3D). The sciatic nerve and S1 and S2 are then identified and protected and the sacrum is amputated at the desired level (Fig. 113-3E). Reconstruction of the urinary system and pelvic floor and formation of colostomy are then undertaken.
Major operative steps of pelvic exenteration with sacrectomy: A. Abdominal phase: dissection and division of internal iliac artery and vein and ureter. B. Perineal phase: division of the levator ani muscle. C. Sacral phase: division of sacrospinalis, piriformis, and gluteus maximus muscles. D. Sacral phase: division of sacrospinous and sacrotuberous ligaments. E. Sacral phase: division of the coccyx. (Reproduced with permission from Khatri VP. Atlas of Advanced Operative Surgery. Philadelphia: Saunders/Elsevier;2013.)
Genitourinary and/or pelvic floor reconstruction is frequently required following resection for locally advanced rectal cancer. Preoperative multidisciplinary surgical planning is essential when undertaking surgery for these patients.
Patients who undergo resection of the genitourinary tract as part of a multivisceral resection for rectal cancer most often undergo total/partial cystectomy or distal ureteric resection.71 A number of reconstructive options exist; however, reconstruction with ileal conduit, ureterocystostomy, and colon conduit are the most common reconstructive procedures following urinary tract resection for locally advanced rectal cancer. The incidence of urologic complications requiring intervention is high in this population (approximately 25%); however, preoperative radiotherapy does not appear to increase the risk of urologic complications in this population.
Following pelvic exenteration or cylindrical APR, pelvic floor reconstruction using flaps is often utilized to fill dead space and allow for tension-free closure. Pelvic floor reconstruction with flaps appears to decrease wound complications following APR and pelvic exenteration, particularly in patients who have received preoperative radiotherapy.72,73 In patients who require vaginectomy, vaginal reconstruction should be considered in conjunction with pelvic floor reconstruction with the goal of improving body image and sexual function.74 A number of different types of flaps have been used for pelvic floor reconstruction, including the anterolateral thigh (ALT) flap and the transverse rectus abdominus myocutaneous (TRAM) flap. TRAM flaps are widely utilized and provide well-vascularized tissue for pelvic floor reconstruction, but are associated with increased risk of incisional hernia, when compared to ALT flaps. The particular choice of reconstruction depends on a number of factors including the size of the defect, prior irradiation, the goals of reconstruction, and the availability of donor tissue. Involvement of the multidisciplinary team including a plastic surgeon in preoperative planning is thus essential to optimize outcomes for these patients.
Adjuvant chemotherapy is recommended for all patients with stage II and III rectal cancer who were treated with neoadjuvant therapy.9,40 Following neoadjuvant therapy, the tumor may undergo downstaging which can lead to uncertainty regarding the initial stage of the tumor. The current recommendation is to base decisions on adjuvant chemotherapy on the preoperative locoregional staging rather than the final pathology. The data suggesting benefit for adjuvant chemotherapy for patients with locally advanced rectal cancer is largely extrapolated from trials of adjuvant chemotherapy for patients with colon cancer.
Based on the MOSAIC and the NSABP-C 07 trials which both demonstrated superiority of FOLFOX to 5-FU/leucovorin for colon cancer, it is recommended that patients with stage II and III rectal cancer undergo adjuvant therapy with FOLFOX.75-77 There is no direct evidence supporting the use of FOLFOX over 5-FU/leucovorin alone in patients with rectal cancer. Again, extrapolating from the colon cancer literature in which FOLFIRI did not improve outcomes compared to 5-FU/leucovorin alone, FOLFIRI is not recommended as adjuvant therapy for rectal cancer.78 Most clinical trials of adjuvant chemotherapy for colon cancer have required initiation of adjuvant chemotherapy within 6 to 8 weeks following surgery.75-77 As longer time to chemotherapy has been associated with decreased survival after resection for colorectal cancer, it is recommended that adjuvant chemotherapy be initiated as soon as the patient has recovered following surgery (generally 4 to 6 weeks).79 There is no literature that indicates the optimal duration of adjuvant chemotherapy in patients with rectal cancer. The NCCN recommends a total perioperative treatment duration of 6 months, which generally means 4 months of adjuvant chemotherapy following surgery.9