Background
Neoadjuvant chemoradiation is a considered standard treatment for locally advanced rectal cancer [
1]. Current guidelines for the treatment of colorectal cancer in Germany recommend the administration of adjuvant chemotherapy for all rectal cancer patients after neoadjuvant chemoradiation and total mesorectal excision (TME), regardless of the postoperative pathologic staging result [
2]. This recommendation is based on the CAO/ARO/AIO-94 and FFCD 9203 studies [
1,
3]. However, hard evidence is lacking, especially for patients staged ypT0–2 ypN0. While an exploratory analysis suggested that particular patients with good response (ypT0–2) benefit from adjuvant chemotherapy [
4] randomized controlled trials addressing the same question showed no benefit for adjuvant chemotherapy [
5,
6]. However, these trials have relevant methodological restrictions. While a recent pooled analysis showed positive effects for adjuvant chemotherapy, another recent meta-analysis failed to do so [
7,
8].
In adherence to the German national guidelines from before 2008, ypT0–2 ypN0 patients were then not treated with postoperative chemotherapy (NACT) at our institution. After the introduction of the amended guidelines in 2008, adjuvant treatment was routinely administered to the same group of patients (ACT). In the present study, we investigated patients with locally advanced rectal cancer in clinical staging (UICC stages II and III) treated with neoadjuvant chemoradiation and TME and then staged ypT0–2 ypN0. On the basis of a prospectively maintained database, the oncologic outcomes of these patients were analyzed.
Methods
Ethics approval
The institutional review board reviewed and approved the protocol; the study was conducted in accordance with the Declaration of Helsinki.
Patient selection
All surgically treated colorectal carcinomas at the Department of Surgery, University Hospital Mannheim, Germany, between 1999 and 2012 were retrospectively analyzed on the basis of prospective databases. Patients with locally advanced rectal cancer who underwent neoadjuvant chemoradiation and subsequent TME in curative intent were eligible for the study when diagnosed ypT0–2 ypN0 in postoperative pathological staging. Exclusion criteria were postoperative death (in-hospital mortality), UICC stage IV and recurrent disease, or missing information on whether adjuvant chemotherapy was administered. Primary outcome measure was disease-free survival (DFS). Disease was defined as the event of local and/or distant recurrence during follow-up. DFS was defined as absence of local and/or distant recurrence and death by any cause during follow-up after primary hospital stay.
Pre-treatment evaluation
The presence of adenocarcinoma was confirmed by pathological examination in all cases. Clinical staging was performed using rigid rectoscopy, endorectal ultrasound, radiographic imaging of the chest, and abdominal ultrasound. Routine performance of magnetic resonance imaging (MRI) of the pelvis was introduced in 2003. Computed tomography (CT) scans of the thorax and/or abdomen were obtained in the majority of cases.
Preoperative chemoradiation and surgery
Neoadjuvant chemoradiation was administered when locally advanced rectal cancer was diagnosed (uT3-4, uN+). As preoperative chemotherapy regimen, capecitabine, capecitabine + irinotecan (XELIRI), XELIRI + cetuximab, capecitabine + oxaliplatin (XELOX), intravenous 5-FU, or panitumumab were used. Radiation therapy was applied as external-beam radiation with a target dose of 50.4 Gy. TME was scheduled 4 to 5 weeks after completing neoadjuvant chemoradiation before 2008, and 8 to 12 weeks after completion of chemoradiation for patients from 2008 till 2012.
Pathology investigation
Resected specimens were fixated in formalin and pathological work-up was done according to published standards [
9]. If no residual tumor was apparent, the initial tumor-bearing area was sliced and embedded. Tumor regression grade was determined based on the classification proposed by the Japanese Society for Cancer of the Colon and Rectum (JSCCR) [
10].
Postoperative chemotherapy
According to national colorectal cancer guidelines before the year 2008, patients were not offered postoperative chemotherapy when diagnosed ypT0–2 ypN0 in final pathological staging. After 2008, adjuvant chemotherapy became the treatment of choice for those patients when no contraindications were present. For adjuvant chemotherapy capecitabine, XELOX or intravenous 5-FU was used.
Statistical analysis
Baseline characteristics of all patients together were evaluated with respect to their influence on outcome. The characteristics were then compared between patients who received adjuvant chemotherapy and those who did not. Comparisons of frequencies between the two groups were performed using the Student’s t test or the chi-square test. Differences of non-parametric quantitative data were analyzed using the Mann-Whitney U test. Kaplan-Meier estimates were computed for recurrence and survival and were compared between the two treatment groups using the log-rank test. A p value < 0.05 was considered statistically significant. All calculations were made with the SPSS version 22.0 (IBM© SPSS® Statistics).
Discussion
Introduction of neoadjuvant chemoradiation for rectal adenocarcinoma has in combination with TME surgery led to reduce rates of locoregional recurrence [
1]. This improvement of local control, however, did not result in prolonged overall survival [
11]. Our data show a significant benefit from adjuvant treatment for disease-free and overall survival but no benefits with respect to recurrence.
Adjuvant chemotherapy after neoadjuvant treatment and TME surgery is administered with the intention of reducing the incidence of distant metastasis and thereby improving survival. Although this has been prospectively investigated in several trials, controversy remains [
5,
6]. In the just recently published study by Breugom et al., patients with ypTNM stage 0 or I were explicitly excluded which was also criticized [
8,
12,
13]. A meta-analysis identified this subgroup to profit the most from adjuvant chemotherapy [
14]. Maas et al. found the most pronounced effect of adjuvant chemotherapy on disease-free survival in ypT1–2 patients both in comparison to higher stages but also to pCR patients [
7]. Our analysis found the most pronounced effect of adjuvant chemotherapy in ypT2 patients. The theoretical consideration that tumors with the combination of responsiveness (shown by downstaging) and continued considerable risk for local and distant recurrence (> ypT1) would profit from adjuvant treatment might in particular hold true for ypT2 [
4].
Another restriction in the analysis of Breugom et al. is that the majority of the patients received bolus 5-FU [
8]. However, an explanatory phase III trial showed better disease-free survival after perioperative treatment with capecitabine than with 5-FU [
15]. In our analysis, only two patients received 5-FU postoperatively; therefore, a comparison of the effect of the two agents cannot be undertaken.
Our results are in conflict with the EORTC 22921 study that previously reported ypT1–2 patients to benefit from postoperative chemotherapy after 5 years; however, recently published late results after 10 years showed no improvement in disease-free or overall survival [
4,
16]. Meta-analyses presented inherently contradictory results with respect to the positive effects of adjuvant chemotherapy on disease-free and overall survival [
7,
8,
17,
18]. However, it is difficult to draw conclusions from these meta-analyses as the included studies have relevant shortcomings. As mentioned above, TME was not mandatory in some of the trials; others revealed a questionable quality of surgery with a R1 rate above 10% and finally, but most important, many of the studies showed a high percentage of patients not undergoing any adjuvant chemotherapy or not the initially planned number of cycles [
5]. Low adherence to planned postoperative chemotherapy is one of the major problems of the available randomized trials, and it is a serious problem for interpretation of non-significant results as a proof for ineffectiveness of adjuvant chemotherapy [
6,
8,
19]. In the EORTC 22921 trial, only 41% of the patients received complete chemotherapy [
13,
16].
Anastomotic leakage is a major problem for patients who actually would have been eligible for adjuvant therapy. In our cohort, anastomotic leakage showed a trend towards negatively influencing application of adjuvant chemotherapy (p = 0.062). This is well in accord with general clinical experience that anastomotic leakage often prevents application of chemotherapy.
The results of the PROCTOR-SCRIPT trial challenge our study, as this is the first randomized trial on the application of adjuvant chemotherapy in neoadjuvantly treated patients with rectal cancer [
6]. In this study, no benefit from adjuvant chemotherapy could be detected. However, again, there are limitations in this study. The trial had to be closed earlier due to poor patient recruitment and survival was better than expected suggesting that the trial was probably underpowered. In the PROCTOR part of the trial, only 50% of the patients had a CT or MRI scan before treatment, so inaccuracy of staging is probably a major bias. Patients were preoperatively treated either with 5 × 5 Gy or with long-term chemoradiation; however, the longstanding oncological results of a randomized comparison of these two therapy schedules are still awaited [
20]. Furthermore, stagewise analysis was not performed in the PROCTOR-SCRIPT trial and the number of retrieved lymph nodes not reported [
6].
The question if the number of lymph nodes retrieved during surgery would influence long-term outcome respectively the application of chemotherapy and thereby the outcome has to be addressed. Most guidelines recommend investigation of at least 12 lymph nodes for determining final pathologic tumor stage [
21]. In several studies, the number of detected locoregional lymph nodes was decreased after neoadjuvant chemoradiation [
22,
23]. However, when < 12 lymph nodes are investigated, metastasis could be missed and histopathological stage underestimated. As performance of adjuvant treatment is stage-dependent also patients that would need therapy are then excluded. When intensified pathology work-up of the specimens is performed and more lymph nodes are evaluated, the number of metastatic lymph nodes may raise thereby possibly resulting in stage migration (“Will Rogers phenomenon”) [
24]. While some studies indicate an association between the number of harvested lymph nodes and oncologic outcome [
24], the same authors could not reproduce these results when neoadjuvant chemoradiation was administered [
25]. Recent studies on this topic continue to give conflicting results; therefore, the significance of retrieving more than 12 lymph nodes remains unclear [
26,
27]. In the present study, retrieval of less than 12 lymph nodes showed no influence of adjuvant chemotherapy with respect to disease-free survival, while more than 12 lymph nodes and adjuvant chemotherapy were correlated to a better disease-free survival.
There are several limitations to our study. First, this is a retrospective study. Patients were not prospectively randomized and selection bias cannot be excluded, even though the groups were well matched in size, age, gender, and tumor-specific parameters. As the indication for adjuvant therapy changed in 2008 in Germany, the comparison could be described as historical. Second, follow-up time was significantly longer in the NACT group. The difference is explainable by the consecutive change of guidelines. These two points in turn can be regarded as strength of this study, making it a “quasi-RCT.” Furthermore, follow-up in patients who received adjuvant chemotherapy still was 54 months in mean.
Third, a possible sign of selection bias is the higher proportion of ypT0 patients in the group without adjuvant chemotherapy. In fact, clinicians are often averse to the application of adjuvant chemotherapy in pCR patients. Breugom et al. criticized the retrospective character of the study by Maas et al. and the fact that the other study supporting adjuvant chemotherapy was a meta-analysis [
7,
14,
28]. However, a meta-analysis usually reduces the risk of confounding.
A more detailed analysis of surgical complications other than anastomotic leakage could not be performed. Even if the database was prospectively performed and updated, the number of parameters documenter increased over time, e.g., the Clavien-Dindo complication grading was only introduced at a later stage. However, anastomotic leakage, which was adequately documented in the database, is one of the most severe complications in rectal surgery and most often the reason why adjuvant chemotherapy is delayed or not started at all. Moreover, leakage has been shown to influence the oncological outcome, and as both groups demonstrated a comparable leakage rate, this factor can be ruled out as a biasing factor.
At last, the sample size is too small to be able to evaluate statistical significant difference in rare incidences such as local recurrence that occurred only once.
Regardless of these limitations, the results support current guideline recommendations that in patients with ypT0–2 tumors adjuvant chemotherapy should continue to be administered, especially in ypT2 stages.