Therapeutic effects of oxaliplatin-based neoadjuvant chemotherapy and chemoradiotherapy in patients with locally advanced rectal cancer: a single-center, retrospective cohort study

Therapeutic strategies for locally advanced rectal cancer (LARC) have evolved significantly during the last three decades. In Western countries, neoadjuvant chemoradiotherapy (NACRT) followed by total mesorectal excision (TME) is currently the standard treatment option for patients with LARC [1]. However, it is well known that, although NACRT reduces the local recurrence to a rate less than 10%, distant relapse occurs in 25–40% of patients [2‐4]. This distant relapse rate is almost equal to the local relapse rate in patients who receive no neoadjuvant therapy (NAT) [5]. In addition, several patients who received NACRT had an unfavorable response, with unnecessary toxicity and a delay in surgery.

Early systemic treatment may be effective to reduce distant relapse and prevent dissemination of micrometastases. Neoadjuvant chemotherapy (NAC) without radiation for LARC has recently been explored to reduce distant relapse and avoid the toxicities of radiation without compromising local control [6‐8]. Although a few randomized, controlled studies are ongoing, there is no literature comparing NACRT and NAC performed at a single institution where the surgical technique for LARC is relatively homogeneous [9‐11].

In the present study, different neoadjuvant treatment strategies for LARC, short-term survival results, and oncological results at our institution were compared.

The present study compared the therapeutic effects of NAC and NACRT with regard to clinicopathological findings, therapeutic toxicities, relapses, and survival curves at a single institution. Only one study so far has been published on the comparison of these treatment strategies for LARC patients [14]. In the present study, although the numbers of patients treated either by NAC or NACRT were small, several interesting results were found with regard to perioperative factors. Anal sphincter preservation was more often possible after NAC than after NACRT, even though the distance between the distal margins of the tumor and the anal verge before treatment did not differ between patients treated with NAC and those treated with NACRT. SSI was less frequent after NAC than after NACRT. In addition, the distant relapse rate of patients who received NAC was low compared with that of NACRT (7.4 vs. 18%), and there was a significant difference in 3-year RFS excluding patients who received postoperative adjuvant chemotherapy (NAC vs. NACRT; 94.4 vs. 63.2%, p = 0.043).

Local relapse rates did not differ between NAC and NACRT, and they were almost equal to the local relapse rates reported in previous trials [1, 2]. Although local relapse has been the most significant concern in patients after curative resection of LARC, distant relapse has also attracted much attention in recent years [2‐4, 15]. Distant relapse was slightly less frequent after NAC than after NACRT. No significant differences in OS and RFS were found between NAC and NACRT.

In analyses excluding patients who received PAC for the purpose of eliminating the effect of PAC, no patients who received NAC had a distant relapse, and consequently, RFS was significantly longer after NAC than after NACRT. These results suggest that NAC may be at least equally useful to NACRT for the prophylaxis of not only local relapse, but also distant relapse in patients with LARC.

After both NAC and NACRT, lymph node metastasis was not found (ypN0) in approximately 75% of patients and an improvement of depth of invasion to ypT0–1 was achieved in approximately 40% of patients after both NAC and NACRT. Pathological lymph node metastasis is the strongest prognostic factor in patients with colorectal cancer. Recent studies have shown a reduction of number of LNY after NACRT because of the reduction of the size and number of lymph nodes (LNs) by NACRT, due to fibrosis or disappearance [16‐18]. Furthermore, LNY is an important prognostic factor because a smaller LNY may result in critical failure in identifying metastatic LNs, and several recent studies have also shown that LNY is also an important prognostic indicator in patients who received NACRT [19‐21]. Data from a Dutch nationwide trial on rectal cancer showed that 12 or more LNs were found in only 19% of patients who received NACRT [22]. In the present study, on routine pathological examination, the percentages of patients with 12 or more LNs in the NAC group and the NACRT group were 74 and 32% (p = 0.002), respectively. These values were less than those in previous studies, especially in the NACRT group [20, 23‐25]. Miller et al. reported that the mean LNY after NACRT was decreased by 7 to 53% compared with that without NACRT and that the mean value of the decrease in the studies used in their review was 28% [16]. In the present study, 3 of 14 (21%) patients after NACRT with less than 12 LNs developed relapses. Therefore, the required number of LNs for the precise staging of patients after NACRT needs to be revised by further analyses.

It has been generally acknowledged that NACRT increases the risk of treatment-related toxicities such as gastrointestinal, dermatologic, and neurologic toxicities in patients with LARC [1, 26]. However, there were no grade 3/4 acute treatment-related toxicities in either the NACRT or the NAC group in the present study. Several trials of NACRT suggested that NACRT increased the incidence of surgical morbidity, especially perineal wound healing after APR [27, 28]. In the present study, SSI was significantly more frequent in the NACRT group than in the NAC group (71 vs. 22%, p < 0.001). Therefore, neoadjuvant oxaliplatin-based chemotherapy appears to be a safe therapeutic option, and the risk of treatment-related toxicities and surgical complications does not exceed that of NACRT.

Currently, the PROSPECT study, comparing NAC alone and NAC plus NACRT (conventional CRT), is ongoing as a randomized phase II/III trial in North America with participation of all US and Canadian cooperative groups. More recently, the FOWARC study is also being carried out as a randomized phase III trial comparing NAC with or without radiation and NACRT (conventional CRT). Although disease-free survival (DFS) as the endpoint of the FOWARC study is still under evaluation, Deng et al. have concluded that NAC alone showed a lower pCR rate than NAC with radiation, but it had a similar downstaging rate to that of NACRT with less toxicity [9]. We should wait and evaluate their outcomes to improve the treatment cure for patents with LARC.

Studies comparing NAC and NACRT head-to-head for patients with LARC are very few. Cassidy et al. surveyed 21,707 patients who would meet the eligibility criteria of the PROSPECT trial from the National Cancer Data Base (NCDB) and compared NACRT and NAC (neoadjuvant multiagent chemotherapy) for clinical T2N1 (cT2N1), cT3N0, or cT3N1 rectal cancers [14]. Using propensity score-matched analysis to avoid selection bias, they observed a worse OS in the NAC group than in the NACRT group. They concluded that elimination of CRT for patients who have cT2N1, cT3N0, or cT3N1 rectal cancers should be avoided until the results of the PROSPECT trial are published. In Cassidy’s study, however, the number of patients who received NAC alone was much less than that who received NACRT. The regimens and duration of NAC were not described and might have differed by institution and the period when the patients were treated. The patterns of relapse are not specified either. Therefore, the evaluation of NAC and NACRT for patients with LARC requires the common regimen including contents, duration, and postoperative treatment.

To the best of our knowledge, there has been no report from a single institution comparing NAC and NACRT, though the numbers of patients of both the NAC and NACRT groups were too small and the regimens of NAC were not homogeneous in the present study. The selection of either NAC or NACRT was not randomized and was usually based on the patient’s preference. Surgeons’ preference may also influence the selection of either of NAC and NACRT, because the patient’s decision is often influenced by the surgeon’s explanation when obtaining informed consent. In addition, some of the patients refused PAC by themselves despite the surgeons’ recommendation based on the pathological findings. Therefore, no conclusion can be drawn from the present study alone. However, the results suggest that future prospective studies to explore the safety and efficacy of NAC for LARC would be valid.

In conclusion, the present study suggests that the therapeutic effect of oxaliplatin-based NAC appears equal to, or may be even greater than that of, NACRT in several clinicopathological factors such as reduced toxicity, increased number of LNY, and improved RFS in patients who did not received PAC. Although we are unable to specify the best regimen and duration of NAC from the present study, oxaliplatin-based NAC might be effective for both local and distant control of LARC. Further studies regarding the regimen and duration of NAC is warranted in order to identify the best method of NAC for LARC.