Recent results from the randomized multicenter trial CAO/ARO/AIO-94 showed an enhanced local control and sphincter preservation with concurrently decreased toxicity after preoperative long-term RCT compared to postoperative RCT . These results led to the recommendation of preoperative RCT in locally advanced (stage II/III) rectal cancers . Preoperative RCT results in a very heterogeneous tumor response, which can be measured by various response parameters such as T-level downsizing, tumor downstaging, elimination of lymph node metastases, and pathomorphologic tumor regression.
Of 153 patients with stage II/III rectal cancer who received standardized preoperative RCT within randomized clinical trials, pCR as a major response criterion, was achieved in 16% (n = 10) of patients. pCR rates vary between 10 and 20% and were associated with a favorable outcome [8, 10]. Nevertheless the majority of rectal cancers (70% of the actual collective) show intermediate response with residual tumor either within (ypT2) or beyond (ypT3) the rectal wall (Figure 2).
It remains unclear which subgroup of patients with intermediate response can be considered as cured after preoperative RCT and subsequent TME surgery. Conversly, it is of enormous clinical interest to know which subgroup necessitates adjuvant systemic therapy.
Involvement of circumferential resection margins (CRM) has recently been described as a very strong prognostic factor after preoperative short term radiation . Although this is distinctly reasonable, fortunately only a considerable small group of patients is affected by positive CRM after preoperative long-term RCT. In our study, 7% of patients (n = 8) presented with cT4 status and potential CRM involvement in pretherapeuthical imaging. RCT-induced tumor downsizing was achieved in all cases, resulting in a maximal residual mesorectal infiltration of ≥ 1.5 cm (ypT3d) in 2 patients (2%). Pathologically confirmed complete (R0) resection with negative (>1 mm) CRM after RCT was accomplished in all patients including those previously classified as high-risk for positive CRM.
Prior to implementation of neoadjuvant strategies for rectal cancer, a tumor invasion of ≥ 5 mm into the mesorectal compartment, besides circumferential involvement, was described as a significant prognostic factor . The decision to apply postoperative radiation or radiochemotherapy, was based on tumor invasion as well as a positive nodal status, and led to reduced recurrence rates and prolonged survival .
We therefore evaluated the impact of intramural depth of tumor invasion (ypT2) together with minimal (<1 mm) transgression of the muscularis propria (ypT3a) compared to a distinct transmural tumor invasion into the mesorectum (>1 mm; ypT3b-d). Since patients with ypT3a status show only an extremely marginal infiltration of the mesorectal compartment (<1 mm) we consider them to prognostically belong to the ypT2 group rather than to the tumors with distict mesorectal infiltration. Our results underline this assumption showing an increased incidence of nodal metastases in ypT3b-d patients compared to ypT2/3a patients.
In the patients presenting with previous cT3/4 rectal cancers (only 1 patient had cT2 N+ status, according stage III) the RCT-induced regression of tumor invasion depth to ypT2/3a status had no impact on prolonged DFS and OS. Thus, residual tumor transgression into the mesorectum after preoperative RCT showed no significant influence on cancer recurrence, providing that complete resection with negative CRM is achieved by adequate TME surgery.
Tumor downsizing from the extramural mesorectal compartment into the actual rectal wall therefore seems to be of importance only when tumor-free CRM and R0-resection cannot be guaranteed (former T3d/4 status).
In contrast to ypT, nodal status after preoperative CRT (ypN) significantly influenced cancer recurrence and overall survival in stage II/III rectal cancer patients with intermediate response within our investigation. This finding coincides with previous results and supports recent investigations with considerable numbers of patients [25, 26] but it is based on a collective of patients with highly standardized diagnostic and treatment procedures according to the protocols of the respective clinical phase II and III trials of the German Rectal Cancer Study Group.
In agreement with other authors [25, 27], we observed that pretherapeutical nodal involvement (cN+) has no impact on the prognosis of patients, in which ypN0 status can be achieved. Patients with evidence of lymph node involvement in pretreatment staging can therefore not categorically be considered as high risk for cancer relapse.
Anyway, patients with ypN+ status should be considered for upcoming trials with intensified adjuvant CT regimes as this might be more efficient in preventing systemic tumor relapse. Nonetheless, mesorectal tumor invasion (ypT3b-d) was significantly associated with residual lymph node metastases after RCT in our study (p < 0.001). We interpret this finding with a generally lower response to RCT regarding both downsizing of the primary tumor and sterilization of lymph node metastases. This might be due to improved biological behavior and enhanced resistance to RCT in individual cancers. The prognostic impact of mesorectal tumor infiltration remains unclear. We could not show straight effects on tumor recurrence and survival but are well aware that this might be due to the relative small number of patients underlying this investigation.
Neoadjuvant RCT has repeatedly been accused of reducing lymph node yield in rectal cancer specimens [28–31]. It has also been reported that the number of detected nodes in stage II rectal cancer patients influences survival [32–34]. Within our investigation, we evaluated a median number of 21 lymph nodes per specimen. In contrast to other investigations , we found no significant difference in lymph node yield between ypN0 patients with and those without subsequent development of distant metastases and tumor-related death. This might be explained by the implementation of extensive lymph node recovery at our institution and a minor variance of evaluated lymph node numbers between both groups.
While histologic tumor grading in colorectal cancers after primary surgery has been ascertained as a prognostic factor , its prognostic relevance following preoperative RCT remains unclear and currently does not belong to standard pathologic staging in rectal cancer specimens. Our results show that histologic grading of residual tumor cells is a reliable parameter, which correlates with advanced tumor biology and has straight impact on DFS despite RCT-induced histomorphologic alterations of the tumor. Thus, histopathologiclogic grading of residual tumor cells should be considered within risk stratification in rectal cancers after RCT.
Not unexpectedly, lymph node status displays as the major criterion for therapy stratification after application of preoperative RCT within our study and several recent investigations and might subdivide patients with need of intensified adjuvant treatment from those who can be considered as cured after surgery. In contrast Collette et al. , who reported the results of the EORTC 22921 trial, underlined that only patients with RCT-induced tumor downsizing to ypT1/2N0 status benefited from adjuvant CT. They interpret their results with an increased sensitivity to preoperative RCT as well as postoperative CT in this subgroup. However, 5-FU monotherapy was used in both, neoadjuvant and adjuvant setting in this trial. This might explain the failure of adjuvant CT in patients with a minor response to preoperative RCT. Thus, in the adjuvant setting an intensified or combined CT should be applied with different anti-tumoral mechanisms (e.g. FOLFOX/FOLFIRI regime ± targeted therapy) in patients with minor response to neoadjuvant treatment.
To date, most patients with positive nodal status after preoperative RCT will intuitively get adjuvant CT. Prospective randomized clinical trials should therefore clarify the impact of adjuvant treatment in patients undergoing preoperative RCT and radical surgery. For ypN0 patients 5-FU based adjuvant CT was shown as a potential overtreatment and had no significant effect on survival [36, 37].
Nevertheless, in our actual study population, 7 patients with ypN0 status developed distant metastases during follow-up. All 7 had poorly differentiated residual tumors (low grade). Poor differentiation of residual tumor cell clusters after RCT and advanced invasion depth turned out to be predictors of lymph node metastases and may be indicators of occult nodal (micro-) metastases in patients classified as ypN0. Both parameters should thus be taken into account in ypN0 patients, particularly in cases of minor lymph node recovery, and might have influence on the decision for adjuvant CT.
Although this investigation is based on a homogeneous collective of patients treated within randomized clinical trials with replicable and standardized diagnostic and therapeutic procedures, its principal limitations are the retrospective character and the relatively small number of patients. Thus this study does not want to claim to ultimately answer the question which subgroup of patients need adjuvant CT after preoperative multimodal treatment and subsequent R0-resection. Prospective randomized trials will have to clarify the debatable role of postoperative CT in rectal cancer patients after preoperative RCT and radical TME surgery. The clinicopathologic parameters investigated in this study might give indications to stratify patient groups with lower and higher individual risk of tumor relapse and tumor-related death within future clinical trials.