Clinical Outcome of the ACCORD 12/0405 PRODIGE 2 Randomized Trial in Rectal Cancer

Since the pivotal German¹ and Dutch² trials, preoperative chemoradiotherapy (CT-RT) and short-course preoperative RT³ have been introduced in T3-4 Nx M0 resectable rectal cancer in the routine practice of most institutions. The FFCD (Fédération Francophone de la Cancérologie Digestive) 9203⁴ and EORTC (European Organisation for Research and Treatment of Cancer) 22921⁵ trials demonstrated that the concurrent addition of fluorouracil (FU) and folinic acid to RT increased sterilization of the operative specimen (complete sterilization of operative specimen [ypCR], 11%) and reduced the risk of local recurrence (8% at 5 years) without survival benefit.

The ACCORD (Actions Concertées dans les Cancers Colorectaux et Digestifs) 12 trial was designed and initiated in 2005 using a pragmatic approach to test concurrently the relevance of radiation dose escalation (50 v 45 Gy over 5 weeks), addition of concurrent oxaliplatin, and use of capecitabine to replace FU (CAPOX50 [50-Gy RT for 5 weeks with concurrent capecitabine and oxaliplatin] group). The control group (CAP45) received 45-Gy RT for 5 weeks with concurrent capecitabine. The primary end point was ypCR, and the results (19.2% v 13.9%; P = .09) showed no significant benefit of the CAPOX50 regimen for this end point.⁶ The primary aim of this article is to report the clinical outcome of these patients (local and distant recurrence, overall and disease-free survival [DFS], toxicity, and bowel function), which was analyzed at 3 years as a secondary end point.

Between November 2005 and July 2008, 598 patients were randomly assigned to treatment and included in the present analysis (Fig 1). Patient characteristics were well balanced between both arms.⁶ As previously published, the rate of pathologic complete response according to Dworak score was 13.9% versus 19.2% in the CAP45 versus CAPOX50 arm (P = .09), and the rate of complete and nearly complete response (no cancer cells or few residual cancer cells detected in operative specimen) was 28.9% for CAP45 versus 39.4% for CAPOX50 (P = .008).⁶ Median follow-up was 36.8 months (range, 30.6 to 43.4 months). A total of 128 patients in the CAP45 group and 125 in the CAPOX50 group received adjuvant CT. The most frequent regimen was LV5FU2 (leucovorin [folinic acid] plus fluorouracil for 2 days), with a mean of five cycles. The clinical outcome and events at 3 years were as follows (Table 1):

The main result of this analysis is that intensification of neoadjuvant treatment with the CAPOX50 regimen did not translate at 3-year follow-up into any significant clinical benefit with regard to local recurrence, DM, or survival. No significant differences regarding late toxicity, bowel continence, ED, or social life disturbance were related to the intensified CAPOX50 regimen. This trial also confirmed the value of ypCR and CRM involvement in predicting local recurrence and DFS.

This trial had a number of limitations. Follow-up of 3 years was too short to capture all the clinical events occurring in these patients, and the number of patients (N = 598) was too small to detect differences of < 10%. Analyses of bowel function and sexual activity have a low level of confidence because they were performed with limited measurements at 3 years in a small number of patients. Finally, because the experimental arm (CAPOX50) of this trial tested two different parameters (radiation dose increase and addition of oxaliplatin) when compared with the reference arm (CAP45), it is difficult to extrapolate which of these contributed most to any benefit.

However, the results of the ACCORD 12 trial may also be analyzed in perspective with the results of three other recent trials testing the hypothesis of an optimization of neoadjuvant treatment in locally advanced resectable rectal cancer (Table 3). The STAR (Studio Terapia Adiuvante Retto) 01 trial⁹ was similar to the ACCORD 12 trial; and oxaliplatin was added to FU only in the experimental arm. With 720 randomly assigned patients, no difference was noted in the sterilization rate (16%) of the tumor, and grade 3 to 4 acute toxicity (diarrhea) was significantly increased with addition of oxaliplatin (24%), a percentage of acute toxicity identical to that observed in the ACCORD 12 trial. In the NSABP (National Surgical Adjuvant Breast and Bowel Project) R04 trial,¹⁰ 1,608 patients were randomly allocated to infusional FU or capecitabine, with or without oxaliplatin, and RT. Pathologic complete response rate showed no increase with the addition of oxaliplatin to FU or capecitabine, and an increase in grade 3 to 4 early toxicity was also reported (25%) in the oxaliplatin group.⁸ Conversely, the CAO/ARO (Working Group of Surgical Oncology/Working Group of Radiation Oncology) 04 trial,¹¹ which included 1,265 patients, did not show increased toxicity with the addition of oxaliplatin, which produced an increase in pathologic complete response (14% v 19%). However, in a subgroup analysis with 196 patients treated in Gottingen, a significantly higher proportion of grade 3 or 4 acute toxicity was found in the patients treated with FU and oxaliplatin (27.1% v 7% with FU alone, as in the three other trials).¹² Despite some differences in the results of these trials, we agree with the comments made by Weiser¹³ in his recent editorial: “Considering the currently available results … it is unlikely that oxaliplatin will improve tumor sterilization.”^13(p2747) This statement is in agreement with the experimental data on xenograft of colon adenocarcinoma in mice, showing that when added to capecitabine, oxaliplatin is not an efficient radiosensitizer.¹⁴ Given the results of both the ACCORD 12 and NSABP R04 trials, capecitabine seems to be equivalent to FU and could be proposed as an alternative concurrent CT with irradiation.

In the ACCORD 12 trial, the CAPOX50 regimen was associated with a significant increase in tumor response on the operative specimen, when Dworak tumor regression grade score 3 to 4 and negative CRM (2 mm from surgical resection margin) were taken as end points.⁶ It is possible to hypothesize that the radiation dose escalation (15% increase in biologic equivalent dose) was responsible for this increase. Two other randomized trials^15,16 have already shown a higher ypCR in the higher radiation dose group. Garcia-Aguilar et al¹⁷ recently reported a phase II trial involving 90 patients with T2N0 rectal cancer treated with capecitabine, oxaliplatin, and radiation followed by transanal local excision. After the first 44 patients, the radiation dose was reduced because of digestive toxicity. The ypCR achieved was 48% with a dose of 54 Gy (52 patients), identical to the result reported in the ACCORD 12 trial, and only 36% with 45 Gy (16 patients). Because these trials did not address exactly the same patients, and treatments were not exactly similar, comparison among them is problematic. Nevertheless, all point to the role of radiation dose escalation in the CAPOX50 regimen, which achieved ypCR of 48% in T2 and 19% in T3 tumors⁶ with no increase in late toxicity. Such a hypothesis could be further tested in randomized trials.

This radiation dose escalation was achieved without increase in 3-year grade 3 to 4 toxicity or bowel dysfunction, possibly because the planning target volume (PTV), as recommended in the trial protocol,⁶ was limited, not exceeding 1.5 L and with upper limit always below the promontorium. Despite this limited PTV, it was possible to achieve a low rate of local recurrence. This is in agreement with a randomized trial using small fields (10 × 10 cm) of preoperative RT¹⁸ and with the results of the Dutch trial showing that most local recurrences are located below the S2/S3 junction.¹⁹ The increase in acute grade 3 to 4 toxicity seen in all the four randomized trials (Table 3) is clearly related to the use of concurrent oxaliplatin. Therefore, it is probably important to keep the PTV as small as possible, with the upper border not above the S2/S3 junction, especially for distal and middle rectal tumors.^18,19

In light of the results of the ACCORD 12 trial, the CAP50 regimen (without oxaliplatin) has been adopted as the control arm in the ongoing French randomized trial involving patients age < 75 years. Moreover, it is the recommended neoadjuvant regimen in the updated French guidelines for rectal cancer.²⁰

In summary, surgery with total mesorectal excision remains the cornerstone of treatment of locally advanced rectal cancer. Neoadjuvant treatment with CT-RT is standard management, and various regimens have been tested in randomized studies. Because the ACCORD 12 trial showed no clinical benefit at 3-year follow-up with the addition of oxaliplatin, and taking into consideration other recent phase III trials, this CT regimen should not be recommended concurrently with irradiation.

Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a “U” are those for which no compensation was received; those relationships marked with a “C” were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.

Employment or Leadership Position: None Consultant or Advisory Role: Christophe Hennequin, Astellas Pharma (C), Ipsen (C), Roche (C), sanofi-aventis (C); Eric François, Merck (C), Roche (C); Olivier Bouché, Roche (C); Jean-François Seitz, Roche (C) Stock Ownership: None Honoraria: Christophe Hennequin, Janssen Pharmaceuticals; Eric François, Merck, Roche; Olivier Bouché, Roche Research Funding: Christophe Hennequin, AstraZeneca Expert Testimony: None Other Remuneration: None