Total neoadjuvant treatment of locally advanced rectal cancer with high risk factors in Slovenia

More than half of the patients with rectal cancer present with locally advanced stage of disease and are treated with combination of preoperative chemoradiotherapy (CRT) followed by total mesorectal excision (TME) and adjuvant chemotherapy with 5-fluorouracil or capecitabine with or without oxaliplatin. With this approach decreased 5 year local recurrence rates to approximately 5–10% has been observed. However, good local control did not result in better survival due to high, more than 30% rate of distant recurrence, which remains the leading cause of rectal cancer-related death.1,2 The reason probably lies in insufficient dose of chemotherapy (ChT) prior the operation and poor compliance of patients to recieve remaining postoperative ChT needed to influence on micrometastases and prevent distant spread of the disease. Randomised trials testing intensification of preoperative treatment by adding oxaliplatin to capecitabine based CRT failed to prove significant benefit over the gold standard. Oxaliplatin significantly decreased distant failure, but did not improve overall survival (OS), disease free survival (DFS) and local failure (LF) compared to 5-fluorouracyl CRT.3

In the light of these unfavorable facts shifting adjuvant ChT into preoperative setting, so called the total neoadjuvant treatment (TNT), was the next logical step. In comparison with standard treatment, TNT is more effective regarding tumor regression rate, the rate of radical resections, sphincter sparing proceadures, patological complete remissions (pCR) and offers a good platform for less radical surgery or potential non-operative management in selected patients. Further, more favorable compliance and lower toxicity rate if ChT is deliverd preoperatively, allows more patients to complete the treatment according to the protocol.4, 5, 6 Currently, there are two slightly different TNT approaches in the treatment of locally advanced rectal cancer (LARC) with high risk of local recurrence. While American National Comprehensive Cancer Network (NCCN) guidelines recommend induction 5-fluorouracyl, leucovorin and oxaliplatin (FOLFOX) or capecitabine and oxaliplatin (CAPOX) ChT followed by CRT and operation, in the northern Europe the same ChT given after short course RT as a consolidation therapy before surgery, is more prefered treatment option according to the latest guidelines from European Socienty for Medical Oncology (ESMO).7, 8, 9

In Slovenia, the TNT was introduced in 2016. The treatment scheme consists of four induction cycles of FOLFOX/CAPOX, capecitabine or 5-FU-based CRT and two additional cycles of FOLFOX/CAPOX as a consolidation therapy before the TME surgery. With this regimen the interval between conclusion of CRT and surgery at 8–10 weeks is preserved. At first, only the patients with T4 tumors or with the presence of EMVI or extramesorectal lymph nodes involvement on magnetic resonance imaging (MRI) were considered candidates for TNT. Later on two more indications for this treatment selection were added: N2 disease and the distance £ 1 mm of tumor or lymph nodes from mesorectal fascia (MRF).

The main objective of the present study is to evaluate efficiency and toxicity of TNT treatment in LARC with high risk factors for local or distant recurrence in Slovenia.

Optimal therapy for patients with LARC is still controversial, but TNT is gaining acceptance in the treatment of a high risk group. The goal of the present analysis was to evaluate this approach for patients with LARC-HR in Slovenia. Compared to our previous study with intensified neoadjuvant capecitabine based treatment with one induction cycle of capecitabine before CRT and two consolidation cycles before the operation, TNT achieved better pCR (17.5% vs 23.3%), T (55.5% vs 65%) , N (77.7% vs 96.7%) and stage (79.3% vs 83.4%) downstaging with comparable toxicity and compliance of patients.15 Direct comparison with the results of these near TNT study and other TNT studies is limited because in majority of them patients with stage II and III LARC were included. It is known that patients with clinical stage II tumors had higher response to treatment and pCR rate than patients with clinical stage III tumors.6 Taking into account only the high risk group of patients in the study of Golo et al, the greater efficacy of our TNT approach is even more prominent. The difference in pCR rate is 13.8% (10.5% vs 23.3%).15

Another difficulty in comparison with other studies presents our scheme of TNT approach which is rather unique with combination of induction ChT before CRT and consolidation ChT after it. We found only one Chinese study reporting results on TNT in LARC-HR only with similar approach: 3 cycles of induction and 3 cycles of consolidation CAPOX, but even more intensified by oxaliplatin added to CRT.16 Studies with different TNT schemes than our in LARC reported pCR rate ranging from 14% to 36%.4, 5, 6,16,17 Our pCR rate of 23.3% is consistent with most of them. We achieved similar pCR than study from Chau et al (24%) and slightly lower than Chinese study (31.7%).4,16 Still, our downstaging data appear to be encouraging. We observed slightly higher proportion of low pathological T stage (ypT0-2; 51.7%) than in Chinese study (42.6%).16 On the other side we observed slightly lower proportion of high pathological T stage (ypT3 and ypT4; 41.7% and 6.7%) than in Chinese study (ypT3 40.4%). In the only TNT-based randomized trial including both (II and III) stages of LARC T downstaging was observed in 43% of patients.5

We reported recently that preoperative IMRT-SIB can achieve a high rate of pCR and T or N downstaging.18 Radiotherapy techniques used in the Chinese study were IMRT or VMAT16 with standard fractionaction. The possible explanation for lower pCR rate in our study than in Chinese is that more than half of the patients were irradiated with 3D-CRT (54.5%) technique with standard fractionation. If we calculate CR (cCR+pCR) rate in subset of our patients (n = 30) who were irradiated with IMRT SIB or VMAT SIB with higher dose per fraction in shorter time (i.e hypofractionation) with the biological equivalent total dose as with standard fractionation, we get an excellent CR rate of 36.7% (1 patient with cCR and 10 patients with pCR). The result is as good as the result of the Chinese study even without intensification of CRT with oxaliplatin and is better also for the subset of patients with HR in the study of But et al in which pCR was 20%. Compared to this group we also achieved better N (85% vs 96. 7%) downstaging.18 Moreover, our results are comparable even to other studies involving also favorable stage II LARC.

Further, our interval from completion of CRT to surgery (mean and SD, 11.3 weeks ± 2.5 weeks) was shorter than in Chinese study (mean 20.1 weeks).16 Time from completing neoadjuvant therapy to surgery is one of important determinant for achieving complete response.19 Compared to the study from Cercek et al, we had similar interval between completion of CRT and surgery but shorter time from completing neoadjuvant therapy to surgery (most frequent 2–4 weeks versus 8–12 weeks) due to different TNT regimens.6 In the contrast with us, Cercek et al reported higher rate of pCR (32.8%), but we also have to take into consideration that they reported result for LARC stage III with or without risk factors. To date, there is no consensus about optimal time for surgery after CRT. A lot of studies reported that long interval between preoperative radiotherapy and surgery was associated with a significantly better clinical tumor response and pathologic downstaging.6,20,21 On the other side, longer interval to surgery was associated with increased risk of death and could have impact on surgical complication due to potential fibrosis development.21 However, it is difficult to determine the point where the benefit is greater than the risk because time to surgery is not the only factors affecting pCR.

We also evaluated neoadjuvant rectal score (NAR). NAR was developed as a composite short-term endpoint for clinical trials involving neoadjuvant therapy for rectal cancer.22 It’s calculation is based on downstaging data (cT, pT, pN) and has greater predictive validity for overall survival than does ypCR.23 In the NSABP R-04 randomised trial, the NAR score calculation was divided into three classes. Low (NAR < 8), intermediate (NAR = 8–16), and high score (NAR > 16) were associated with 92%, 89% and 68% 5 year OS, respectively.23 In our study 83.3% of patients had NAR in low and intermediate class. Direct comparison with other TNT studies was not possible as we did not find any reports on predictive NAR score. Taking into account the data from NSABP R-04 trial and from randomized trials with near TNT-based regimens for LARC reporting 5 year OS between 67–77%5,24,25, we can consider results of current study as promising.

Compliance within TNT protocol in our study was in the range of 83% to 100% and is consistent with others who have studied TNT approach.26 In the largest randomized study of the adjuvant ChT in rectal cancer poor adherence with all planned dose in only 43% of patients was reported.27 The rates of compliance in TNT-based regimen are promising including our compliance rate for all planned dose ChT (60.6%).

Toxicities were acceptable with minimal life-threatening side effects. Grade 3 adverse event developed only in 6% of patients who recevied TNT and 1 unexplained death occured. Postoperative morbidity rate was 25% which is comparable to postoperative morbidity of TNT-based studies in which ranged from 13 to 51%.26 In addition, most surgical complications were associated with operative wound healing and not with other serious complications. Compared to other TNT-based regimen in LARC-HR, no relevant differences in terms of treatment outcome and toxicity were observed.

One of the important factors in assessing the quality of life is also the time to temporary stoma closure and the presence of a temporary or permanent stoma due to negative impact on social functioning and gastrointestinal symptoms.28 There is a lack of studies reporting the time to stoma closure. Cercek et al reported that stoma closure was earlier in the TNT group (89 days in TNT group vs 192 days in group with standard therapy).6 There are two major reasons for prolonged interval in our study: first, complexity of surgery with prolonged recovery and second, too long waiting time to admission for stoma closure. As pointed out previously our result can not be compared with the control group or another group with similar characteristics, but in relation to the previously mentioned result we consider median time of 134 days in our study as acceptable.

Limitations of our study are lack of control arm and limited number of patients from a single institution. Further, long-term data for our TNT approach are not available yet. Moreover, it should be emphasized that the NAR calculation does not reflect direct clinical benefits but it predicts overall survival. All of the above mentioned facts will bet taken into consideration when designing future studies.

TNT of high risk LARC is well tolerated and highly effective with excellent tumor and node regression rate and with low toxicity rate. Treatment according to the protocol is achievable in a great proportion of patients. Regarding short term outcomes TNT seems to be better option for patients with LARC with high risk for local or systemic recurrence than standard preoperative CRT and adjuvant ChT. Longer follow up will show if this strategy will improve distant disease control and survival.