Adjuvant Chemotherapy for Locally Advanced Rectal Cancer: Is It a Given?

Abstract

Adjuvant therapy for solid tumors is designed to cure patients more often than surgery alone, and the first principles of adjuvant therapy include gaining both local and systemic control. Locally advanced rectal cancer isdefinedbyeitherT3ornodepositivity,andhistorically ithashad a high risk for both local recurrence (LR) and distant recurrence. Unlike other solid tumors, local control in rectal cancer is a worthy pursuit, even in the absence of systemic therapy, as a result of the profound morbidity associated with LR and the inability to salvage most LRs with surgery. Consequently, combined-modality therapy with chemotherapy, radiation,andsurgeryemergedinthe1980sand1990sastheoptimumtherapy for reducing the risk of both LR and distant recurrence. Adjuvant (postoperative) chemotherapy has been accepted as a component of multimodality therapy for locally advanced rectal cancer for two primary reasons. First, postoperative fluorouracil was associated with an improvement in overall survival (OS) in patients with Dukes’ B and C rectal cancer from the era preceding total mesorectal excision. Second, the relative 25% to 30% risk reduction in distant metastases and death in these studies is concordant with the benefits of adjuvant chemotherapyseenincoloncancer. Subsequently, randomizedtrials inpatients with locally advanced rectal cancer adopted as standard therapy that all patients receive fluorouracil after completion of surgery and radiation therapy.Accordingly, theNationalComprehensiveCancerNetworkVersion 2.2015 recommends the routine use of infusional fluorouracil, leucovorin, and oxaliplatin (FOLFOX) or capecitabine and oxaliplatin (CAPEOX) chemotherapy for a total perioperative therapeutic period (includingchemoradiotherapy)of6months.Inthemostcommonlyused treatment paradigm for patients with stage II or III rectal cancer, patients receive preoperative chemoradiotherapy or short-course radiation followed by total mesorectal excision (TME) which is then followed by 4 months of an adjuvant fluoropyrimidine with or without oxaliplatin. Yet the field of rectal cancer has changed substantially in three ways that make applying information from the early trials problematic. First, neoadjuvant chemoradiotherapy is routinely used now. Moreover, the downstaging effect of preoperative therapy leaves many clinicians afraid to omit postoperative chemotherapy “in case the nodes were originally positive.” Second, the adoption of TME has substantially decreased LR from 25% to 50% to less than 10%. Finally, imaging with chest-abdominal-pelvic computed tomography scans, rectal magnetic resonance imaging scans, positron emission tomography scans, and liver magnetic resonance imaging scans has caused stage migration. Many patients in the earlier studies undoubtedly had early stage IV disease and would have been excluded in modern trials. Are there any studies in the modern era that take into account these changes and support the routine use of postoperative chemotherapy in patients who received preoperative radiation therapy? The simple answer is no (Table 1). The study at the heart of this controversy is the European Organisation for Research and Treatment of Cancer (EORTC) 22921 (Radiation Therapy, Surgery, and Chemotherapy in Treating Patients With Rectal Cancer That Can Be Surgically Removed), which randomly assigned 1,011 patients with clinical stage II or III rectal cancer who were receiving preoperative radiation 45 Gy in a 2 2 factorial design to two separate chemotherapy regimens. In the first step, patients were randomly assigned to concurrent fluorouracil with preoperative radiotherapy or radiotherapy alone. In the second step, patients were randomly assigned to four cycles of adjuvant fluorouracil/leucovorin given every 3 weeks. The study was designed and powered to demonstrate an absolute 10% OS benefit. With long-term follow-up (median, 10.4 years), there was no differenceinOSat10years(51.8%[adjuvantchemotherapy]v48.4%[no chemotherapy]), disease-free survival (DFS) at 10 years (47.0% [adjuvant chemotherapy] v 43.7% [no chemotherapy]) or cumulative incidence of distant metastases (P .52). Interestingly, all three arms receiving any chemotherapy had decreased LR at 10 years (21.9% [radiotherapy alone] v11.8%to14.5%[threechemotherapyarms];P .0017),butthisdidnot translate into an OS advantage, perhaps because the trial was underpowered to determine an absolute survival advantage of less than 10%. Interestingly, the benefit of adjuvant chemotherapy in patients with ypT0-2 seen in the first analysis has disappeared. The unexpected outcome of this study led to controversy. First, the chemotherapy delivered was not the standard in the United States. The concurrent fluorouracil was given as a bolus regimen in the Mayo Clinic schedule(albeitatadosereductionof350mg/m asopposedto400to425 mg/m) rather than infusional. Adjuvant chemotherapy was given at the same dose-reduced schedule. Second, this study has been criticized for being underpowered for a four-arm comparison trial, with only 252 to 253patientsperarm.However, thestudywasdesignedasa2 2factorial trial, which allows for an appropriate comparison between the approximately500patientswhoreceivedadjuvantchemotherapycomparedwith the approximately 500 patients who did not. These numbers appropriately drove the power calculation seeking a 10% OS benefit at 5 years. Thus, although the comparison of adjuvant chemotherapy or not was appropriatelystratifiedbypreoperative treatment, thereporteddifference reflects the entire 1,011 patients randomly assigned in the intent-to-treat analysis. JOURNAL OF CLINICAL ONCOLOGY COMMENTS AND CONTROVERSIES VOLUME 33 NUMBER 17 JUNE 1