Phase II, Randomized Study of Concomitant Chemoradiotherapy Followed by Surgery and Adjuvant Capecitabine Plus Oxaliplatin (CAPOX) Compared With Induction CAPOX Followed by Concomitant Chemoradiotherapy and Surgery in Magnetic Resonance Imaging–Defined, Locally Advanced Rectal Cancer: Grupo Cáncer de Recto 3 Study

Eligibility criteria included age 18 to 75 years and histopathologically confirmed rectal adenocarcinoma, with the inferior margin within 12 cm from the anal verge. Patients were considered to have LAR disease on the basis of high-resolution, thin-slice (ie, 3 mm) magnetic resonance imaging (MRI) of the pelvis, which was performed as described in appendix. MRI criteria for LAR cancer were as follows: tumors extending to within 2 mm of, or beyond, the mesorectal fascia (ie, an involved or threatened circumferential resection margin); lower third (≤ 6 cm from the anal verge) cT3 tumors; resectable cT4 tumors; and any cT3N+ (Appendix). Additional inclusion criteria were an ECOG performance status ≤ 2 and adequate hematologic, liver, and renal function (ie, neutrophils ≥ 1.5 × 10⁹/L; platelet count ≥ 100 × 10⁹/L; creatinine clearance ≥ 30 mL/min; total bilirubin concentration ≥ two times the upper normal limit; and liver transaminase or alkaline phosphatase concentrations ≥ three times the upper normal limit). Exclusion criteria included M1 metastatic disease; previous radiotherapy to the pelvic region; previous CT; other cancers; clinically significant cardiovascular disease (active: stable or unstable angina, New York Heart Association class 2 or greater congestive heart failure, cardiac arrhythmias, acute coronary syndrome [even if adequately controlled by medication], uncontrolled hypertension, and myocardial infarction in the year before commencement of study treatment). Figure 1 shows the treatment schema of all enrolled patients.

Endorectal ultrasound was performed in case of doubtful distal T3 tumors after MRI. All patients were required to have a computed tomography scan of the chest, abdomen, and pelvis and a carcinoembryonic antigen (CEA) measurement.

In a randomized, open-label phase II trial, we compared arm A—which consisted of preoperative CRT (capecitabine plus oxaliplatin and concurrent radiotherapy) followed by TME and postoperative adjuvant capecitabine-oxaliplatin—with arm B—which consisted of induction capecitabine-oxaliplatin followed by the same preoperative CRT and TME (Fig 2). Random assignment was performed centrally, and patients were stratified by institution.

Radiotherapy was delivered by a linear accelerator with a minimum of 6 MV by using a three- or four-field technique. The treatment volume included the primary tumor and the mesorectal, presacral, and internal iliac lymph nodes up to the level of the bottom part of the fifth lumbar vertebra. All patients received a total dose of 50.4 Gy, and daily fractions of 1.8 Gy were received 5 days per week. During radiotherapy, capecitabine was delivered orally at a fixed dose of 825 mg/m² twice daily on days 1 to 5 for 5 weeks. For patients with moderate renal impairment (ie, baseline creatinine clearance of 30 to 50 mL/min), capecitabine was reduced by 25%. The first daily dose was administered approximately 2 hours before radiotherapy, and the second dose was administered 12 hours later. Oxaliplatin was administered as a 2-hour infusion on days 1, 8, 15, 22, and 29 at a dose of 50 mg/m² per day. Dose adjustment was made in the event of toxicity, which was assessed according to National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE) version 3.0. In arm B, CRT was commenced 3 weeks after the start of the fourth CAPOX cycle.

Patients treated on arm A who underwent an R0 to R1 (ie, complete or microscopic residual) resection received postoperative adjuvant CAPOX for four cycles at 4 to 8 weeks after surgery. Capecitabine was administered daily at a dose of 2,000 mg/m² for 14 days every 21 days. Oxaliplatin was administered on day 1 of each of the four cycles at a dose of 130 mg/m². Patients treated on arm B received four cycles of induction CAPOX at the same doses as patients in arm A.

Guidelines for CT dose reductions applied to both arms were as follows: for grade ≥ 2 toxicity (as defined by NCI CTCAE version 3.0),⁸ capecitabine was held, and appropriate symptomatic treatment was administered. Once toxicity resolved to grade 0 or 1, treatment was resumed at 75% of the original dose for the first or at 50% of the original dose for the second occurrence of toxicity. The oxaliplatin dose was reduced for grade 3 vomiting, grade 3 or 4 thrombocytopenia or neutropenia, and paresthesia with pain or functional impairment of more than 7 days.

TME was performed in both treatment arms at 5 to 6 weeks after completion of CRT. The final choice of surgical procedure (ie, abdominoperineal surgery or anterior resection) was at the surgeon's discretion. Standardized pathology examinations were performed by using the methodology of Quirke et al.⁹ The extent of residual tumor in the resected specimen was classified according to the TNM staging system of the American Joint Committee on Cancer version 6, with the prescript “y” used to indicate that the tumor had been treated before surgical resection.¹⁰ Residual tumor masses after preoperative CRT were semiquantitatively evaluated according to the 5-point r regression grading scale established by Dworak et al.¹¹ A pCR was defined as the absence of viable tumor cells in the primary tumor and in the lymph nodes (ypT0N0).

The primary end point was pCR. A two-stage design, as proposed by Simon,¹² was used to allow early termination of any ineffective arm early in the study. In the first stage, 24 response-assessable patients were entered into each arm. If one of the two arms had zero to one pCR, accrual to that arm would be terminated, and it would be concluded that that schedule should not be pursued. If two or more pCRs were seen in each arm, the trial would continue to the second stage with the accrual of an additional 25 response-assessable patients in both arms (n = 49 patients per arm total). For each arm, with a sample size of 49 patients, type I error amounted to 5%, and the type II error was set to 10%, which corresponded to a power of 90%, assuming P₀ = 10% and P₁ = 25%. The planned patient number was increased to 108 to allow a 10% drop-out rate. Because no treatment arm can be considered standard, we use a randomized selection design that was based on the approach proposed by Simon et al.¹³ With this sample size, there was a greater than 90% chance of selecting the more effective treatment if the difference in response rates was at least 15% and if the smaller response rate was assumed to be 10%. If both arms had similar responses, then other factors, such as toxicity and feasibility, would contribute to selecting a preferred schedule.

Differences between ratios were analyzed with the χ² or Fisher's exact test, as appropriate. Failure-free survival (FFS) was defined as the time from the date of trial entry until disease progression, relapse, or death from any cause. Overall survival (OS) was calculated from the date of trial entry until death from any cause or was censored at last follow-up. Both FFS and OS were estimated by using the Kaplan-Meier method and were compared between groups by the log-rank test, for which P < .05 was considered statistically significant. Data were stored in Oracle Clinical 4.5.3 (Oracle Corporation, Redwood Shores, CA), and statistical calculations were performed by SAS 9.1 for Windows (SAS Institute, Cary, NC). Secondary end points included toxicity and compliance with the regimen, downstaging and R0 resection rates, 30-day surgical complications, and relapses rates.

Between May 2006 and December 2007, a total of 108 patients from 14 centers in Spain were enrolled. The cutoff date for this report was May 30, 2009.

Demographic and tumor characteristics of the 108 patients (of which 65% were men) who underwent random assignment—52 to arm A (ie, postoperative adjuvant CT) and 56 to arm B (ie, induction CT)—are listed in Table 1. The predominant clinical disease stage was cT3N+. After random assignment, arm B included more patients with T4 lesions than arm A (13% v 6%) and also more grade 3 tumors (11% v 2%). Arm A included more patients with a threatened circumferential resection margin than arm B (10% v 0%). Otherwise, patient characteristics were well balanced between the two arms.

The number of eligible patients who commenced treatment was 49 (94%) of 52 in arm A and 54 (96%) of 56 in arm B. Overall, 46 patients (89%) in arm A and 54 (96%) in arm B underwent surgery. When patients who had the appropriate dose reductions were included, more patients in arm B than arm A completed the study as per protocol (91% v 54%; P < .0001). A total of nine patients (17%) in arm A and 1 patient (2%) in arm B discontinued study treatment because of adverse events (P = .006). The percentages of patients who discontinued treatment for other reasons were similar in the two groups (Table 2).

A total of 46 patients in arm A, after preoperative CRT, and 54 patients in arm B, after induction CT and preoperative CRT, underwent surgery. On the basis of an intent-to-treat analysis, a pCR (ypT0N0M0) was achieved in seven patients in arm A (13%; CI 95%, 5.6% to 25.8%) and in eight patients in arm B (14%; CI 95%, 6.4% to 26.2%). Downstaging (defined as lower pathologic T stage compared with the pretreatment clinical T stage) was observed in 30 patients (58%) in arm A and in 24 patients (43%) in arm B. An R0 resection was achieved in 45 patients (87%) in arm A and in 48 patients (86%) in arm B. In arm A, the tumor regression grade (TRG) of the primary tumor (ypT0) was TRG 4 in seven patients, and an additional 22 patients showed tumor regression of greater than 50% of the tumor mass (TRG 3). In contrast, in arm B, eight patients had a TRG of 4, and 20 patients had a TRG of 3 (Table 3).

We express special thanks to all members of the Grupo de Cancer de Recto group in alphabetical order: Medical Oncologist: C. Bosch (Hospital Dr. Peset, Valencia) and M. Guillot (Hospital Son Dureta, Palma de Mallorca); Radiation Oncologist: F. Andreu (Hospital General Universitario, Alicante), C. Conill (Hospital Clinic, Barcelona), F. Mestre (Hospital Son Dureta, Palma de Mallorca), V. Macías (Corporació Sanitaria Parc Taulí, Sabadell), L. Petriz (Hospital Sant Creu i Sant Pau, Barcelona), A. Reig (Hospital del Mar, Barcelona), A. Tormo (Hospital Universitario La Fe, Valencia) and J. Valencia (Hospital Clinico Lozano Blesa, Zaragoza); Radiologist: J.R. Ayuso (Hospital Clinic, Barcelona), M. Busto (Hospital del Mar, Barcelona), A. Darnell (Corporació Sanitaria Parc Taulí, Sabadell), J.A. Fernández (Hospial Clínico Lozano Blesa, Zaragoza), M. Gil (Hospital Arnau Vilanova, Lleida), F.L. Jiménez (Hospital de Navarra, Pamplona), V. Martinez (Hospital General Universitario Valencia), F. Mas (Hospital Universitario La Fe, Valencia), J. Pernas (Hospital Sant Creu i Sant Pau, Barcelona), C. Rosello (Hospital Son Dureta, Palma de Mallorca), J. Santos (Fundación Instituto Valenciano de Oncología, Valencia) and L. Sarriá (Hospital Miguel Servet, Zaragoza); Surgeons: P. Bombardó (Corporació Sanitaria Parc Taulí, Sabadell), J. Campos (Fundación Instituto Valenciano de Oncología, Valencia), S. Delgado (Hospital Clinic, Barcelona), M. Gamundi (Hospital Son Dureta, Palma de Mallorca), J. Garriga (Hospital Sant Creu i Sant Pau, Barcelona), F. Lluis (Hospital General Universitario, Alicante), A. Monzón (Hospital Miguel Servet, Zaragoza), M. Pera (Hospital del Mar, Barcelona), J. Sierra (Hospital Arnau Vilanova, Lleida), J. Suárez (Hospital de Navarra, Pamplona), E. Tejero (Hospial Clínico Lozano Blesa, Zaragoza) and F. Villalba (Hospital General Universitario Valencia); Pathologists: J. Artes (Hospital Universitario La Fe, Valencia), F. Aranda (Hospital General Universitario, Alicante), A. Calatrava (Fundación Instituto Valenciano de Oncología, Valencia), F. Canet (Hospital Son Dureta, Palma de Mallorca), A. Casalots (Corporació Sanitaria Parc Taulí, Sabadell), M. Gomez (Hospital de Navarra, Pamplona), C. Hörndler (Hospital Miguel Servet, Zaragoza), M. Iglesias (Hospital del Mar, Barcelona), E. Martínez (Hospital General Universitario, Valencia), R. Miquel (Hospital Clinic, Barcelona), J. Ortego (Hospial Clínico Lozano Blesa, Zaragoza), E. Rosello (Hospital Dr. Peset, Valencia), F. Sancho (Hospital Sant Creu i Sant Pau, Barcelona) and J. Tarragona (Hospital Arnau Vilanova, Lleida).

Scans were performed by using 1T to 1.5T MRI scanners. The protocol used a thin, 3-mm section, turbo spin-echo, T2-weighted technique with a surface pelvic phased-array coil. For all tumors, scans were performed perpendicular to the long axis of the tumor, and four sequences were used. Coronal imaging was performed for all tumors arising at or below the levator muscle origins. No bowel preparation, air insufflations, or intravenous antispasmodic agents were used. After a coronal localizer, sagittal scans were required from inner pelvic sidewall to sidewall by using a 24-cm field of view, 5-mm contiguous/interleaved slices (no gap), and repetition time more than 2,500 ms and less than 5,000 ms (ms = 85). These acquisitions were used to plan thin-section, oblique, axial images. Axial T2FSE acquisitions of the anatomic pelvis were made by using an 18- to 20-cm field of view; a 3-mm contiguous section thickness; 4,000/85; 320 × 320 matrix; an echo train length of eight; no fat saturation; a 32 kHz bandwidth; and two signals acquisitions (4 NEX). The sagittal, T2-weighted images obtained were then used to plan T2-weighted, thin-section, axial images through the rectal cancer and adjacent perirectal tissues. These images were performed perpendicular to the long axis of the rectum. These were obtained by using a 20- to 22-cm field of view; a 3-mm section thickness; no intersection gap; 4,000/85 TR 4,000/TE 85; a 256 × 256 matrix; an echo train length of eight; no fat saturation; a 32-kHz bandwidth; and four acquisitions (3-4 NEX).

Mesorectal nodes must fulfill one of the following criteria: irregularly bordered nodes and/or inhomogeneous signal intensity nodes or nodes that measured greater than 5 mm.