Which is the best postoperative chemotherapy regimen in patients with rectal cancer after neoadjuvant therapy?

Rectal cancer has been defined as a cancerous lesion located within 12 cm of the anal verge [1]. At present, the main treatment for locally advanced rectal cancer is chemoradiotherapy plus total mesorectal excision (TME). Although it is debatable whether preoperative chemoradiotherapy improves long-term survival [2‐4], randomized clinical trials have shown better local control, lower toxicity, and higher compliance if preoperative chemoradiotherapy is administered rather than postoperative conventionally fractionated chemoradiotherapy [5, 6]. Thus, the current “gold standard” of treatment recommended by both the National Comprehensive Cancer Network (NCCN) [7] and the European Society for Medical Oncology (ESMO) [8] for locally advanced rectal cancers with invading through the muscularis propria into the pericolorectal tissues (cT3), penetrating to the surface of the visceral peritoneum (cT4a), invading or being adherent to other organs or structures (cT4b), or lymph nodal metastasis on imaging (cN1-2) is preoperative radiotherapy plus 5-fluorouracil (5-FU)-based chemotherapy.

Is adjuvant chemotherapy needed after curative surgery for rectal cancer patients who have received neoadjuvant chemoradiotherapy? The NCCN recommended postoperative chemotherapy for all patients undergoing preoperative chemoradiotherapy regardless of the pathological stage [7]. The ESMO guidelines state that “similar to the situation in colon cancer Stages III (and “high-risk” Stage II), adjuvant chemotherapy can be provided, even if the scientific support for sufficient effect is less” [8]. However, several studies questioned the use of adjuvant chemotherapy in patients with rectal cancer who underwent neoadjuvant chemoradiotherapy and curative surgery, especially in patients without pathological lymph node metastasis (ypN0) [2, 9‐14]. Results from all these studies showed that adding postoperative adjuvant chemotherapy did not significantly improve disease-free survival (DFS) or overall survival (OS) in patients who have already received neoadjuvant chemoradiotherapy. On the other hand, a unique randomized clinical trial suggested that good-prognosis patients (ypT0-2) benefit from postoperative chemotherapy [15]. Furthermore, it was not reported whether adding postoperative 5-FU-based chemotherapy could improve the survival of patients with pathological lymph node metastasis (ypN1-2). Hence, for patients who have received neoadjuvant therapy, the role of postoperative chemotherapy is still controversial.

Although lack of data from rigorous randomized clinical trials confirmed the effectiveness, oxaliplatin has been used in rectal cancer for several years based on the extrapolated data in colon cancer. Similarly, the role of postoperative oxaliplatin in patients with rectal cancer who have already received neoadjuvant chemoradiotherapy is still not yet defined. To the best of our knowledge, only two ongoing randomized clinical trials presented preliminary results, though contradictory, on this issue [16, 17].

The aim of the current study was to investigate whether postoperative 5-FU-based chemotherapy or 5-FU plus oxaliplatin provides a benefit for patients with resected rectal cancer who have already received 5-FU-based neoadjuvant chemoradiotherapy.

Preoperative chemoradiotherapy is the standard treatment for locally advanced rectal cancer. However, there is no general agreement about whether patients who have already received neoadjuvant chemoradiotherapy need further postoperative chemotherapy based on 5-FU. Janjan [25] proposed that there was significant improvement in CSS in response to preoperative chemoradiotherapy and the administration of adjuvant chemotherapy. Collette [15] analyzed a subset of data from the European Organization for Research and Treatment of Cancer (EORTC) Trial 22921, which revealed that the postoperative 5-FU-based chemotherapy prolonged survival in ypT0-2, but not in ypT3-4, patients. Hypothetically, distal micrometastasis may be cleared by chemotherapeutic drugs more effectively in patients who were good in response (in low ypStage) to preoperative chemoradiotherapy.

However, Das [26] proposed that postoperative chemotherapy may be of greater benefit for patients in a higher ypStage such as ypStage III (ypN1-2), and lower ypStage subgroups should receive a relatively conservative therapeutic regimen. Subsequently, a study by Huh [11] revealed that postoperative adjuvant chemotherapy for patients in ypT0-2 N0 classification after preoperative chemoradiation and curative surgery did not significantly improve the survival, which was consistent with the suggestion of Das [26], but these results contradictory to that of Collette [15]. Later the results of the study by Govindarajan [9] confirmed that there was no significant difference in the 5-year DFS between patients, in ypT0-2 N0 and ypT3-4 N0 classifications, who did and did not receive adjuvant treatment. In addition, both Fietkau [10] and Kiran [12] proposed that adding postoperative chemotherapy did not significantly improve the survival of patients in ypN0 classification. More recently, Bosset [14] completed the EORTC Trial 22921 and proposed that postoperative 5-FU-based chemotherapy after preoperative radiotherapy (with or without chemotherapy) did not affect DFS or OS. The result was confirmed in both ypT0-2 and ypT3-4 classifications; however, this analysis was not stratified on the basis of ypN classification.

Although the conclusions of the aforementioned studies were different, a consensus may be arrived at based on the postoperative pathologic stage of the patient in determining the need for adjuvant chemotherapy. Simultaneously, several studies proposed that the risk of distant metastases is directly proportional to the postoperative pathologic stage [9, 27‐29], and Quah [28] found that the outcome was most accurately estimated by the postoperative pathologic stage. Nevertheless, the NCCN recommended postoperative chemotherapy for all patients undergoing preoperative chemoradiotherapy regardless of the results of the surgical pathology tests [7].

Considering the importance of ypTNM stage in determining the need for adjuvant chemotherapy, all tests were made stratified by ypTNM stage. We compared the prognosis among patients in no-chemo, 5-FU, and oxaliplatin group in the unmatched univariable survival analysis. We found that postoperative 5-FU-based chemotherapy did not prolong the CSS in ypStage I (ypT1-2 N0) (Figure 1A) and ypStage II (ypT3-4 N0) (Figure 1B), which was similar to the study by Govindarajan [9]. On the contrary, adding postoperative 5-FU-based chemotherapy significantly improved survival of patients in ypStage III (ypN1-2) (Figure 1C). To the best of our knowledge, the outcome of postoperative 5-FU-based chemotherapy in ypStage III patients was never reported previously, although researchers stressed the need for a randomized clinical trial [30, 31]. To confirm our results, both the PS-matched analysis and the Cox proportional hazards model were used to make adjusted analysis, and the results were in accordance with the univariable survival analysis (Figure 2, Table 2).

Adding oxaliplatin to fluoropyrimidines in the adjuvant setting improved the OS in colon cancer compared with FU and leucovorin regimens [32]. According to the NCCN guidelines, for patients with resected rectal cancer who have already received 5-FU-based neoadjuvant chemoradiotherapy, 5-FU/leucovorin/oxaliplatin (FOLFOX) was an optional regimen for postoperative chemotherapy. Several ongoing randomized clinical trials in rectal cancer focused on improving 5-FU-based chemotherapy through the addition of oxaliplatin preoperatively, postoperatively, or both [16, 17, 33‐35]. Two study groups presented preliminary results. Hong [17] proposed that postoperative FOLFOX significantly improved the 2-year DFS relative to postoperative 5-FU-based chemotherapy for rectal cancer patients in ypStage II or III after 5-FU-based neoadjuvant chemoradiotherapy followed by TME. On the contrary, Nimeiri [16] discovered that there was no difference in the OS between patients who received 5-FU alone or FOLFOX as postoperative chemotherapy. In the current study, no significant differences were found in the survival between the two groups (5-FU vs. oxaliplatin) of patients with resected rectal cancer who have already received 5-FU-based neoadjuvant chemoradiotherapy (Figures 1 and 3; Table 2). Although this result need to be confirmed by further clinical trials, we argue that, for patients who received neoadjuvant chemoradiotherapy, adding oxaliplatin to fluoropyrimidines in the postoperative chemotherapy require serious consideration at present.

The current study has some limitations. First, as it was a retrospective exploratory study, the potential for confounding based on patient selection could not be eliminated. Both traditional Cox proportional hazards model and PS-matched techniques were used to account for known relevant confounders. Second, only patients aged ≥66 years at the time of diagnosis were included in this study, which may limit the applicability of the findings to younger patients with rectal cancer. Third, the role of several known prognostic features such as tumor regression grade, preoperative carcinoembryonic antigen, microsatellite instability, perineural invasion, and lymphovascular invasion could not be investigated, as these characteristics were not available within the SEER-Medicare database. Patients with pCR were excluded from analysis because the pCR status was not well supported by the SEER-medicare database. Fourth, this study retrospectively examined the use of chemotherapy as identified through the Medicare claims data using a “one-claim” algorithm [36, 37]. This created a heterogeneous population in which some patients received a substandard duration of therapy. However, O'Connor [38] proposed that the “none versus any” approach used to assign treatment status provided a window into the effectiveness of chemotherapy in real-world practice, in which an individual’s likelihood of completing the treatment course is not known at the outset of the study. Finally, the preoperative clinical stage as well as pathologic response to neoadjuvant therapy was not supported by the SEER-Medicare database. Both the preoperative clinical stage and pathologic response were related to outcome; however, Quah [28] found that the outcome was most accurately estimated by the postoperative pathologic stage and the clinical stage adds no predictive value to the prognosis. In addition, the 100% accuracy of clinical staging was untenable even in the best of centers [39]. Based on this, the decision of postoperative chemotherapy could be regarded mainly as based on postoperative pathologic stage.

It is concluded that, for patients with resected rectal cancer who have already received 5-FU-based neoadjuvant chemoradiotherapy, postoperative 5-FU-based chemotherapy prolongs the CSS of groups in ypStage III. Adding oxaliplatin to fluoropyrimidines in the postoperative chemotherapy did not improve the CSS for patients who received neoadjuvant chemoradiotherapy.