Timing of surgery after neoadjuvant chemoradiotherapy in the treatment of esophageal cancer

Esophageal cancer is a very serious disease with poor prognosis. Merely 20% are alive after 5 years according to our Swedish registers. Early stages of esophageal cancer can be treated with endoscopic resection. However, for most cases of advanced disease surgery is the main treatment. The procedure has over the years incrementally been improved and standardized, and is today usually performed with minimally invasive techniques and often robot-assisted. Nevertheless, this surgery is associated with severe morbidity and mortality.

Over the years neoadjuvant treatment has manifested its role as standard of care and new regimes are continuously being evaluated. The oncological treatment is physically demanding and time for recovery after completed treatment, before performing surgery, is required. The standard time to surgery (TTS) after completed neoadjuvant chemoradiotherapy (nCRT) varies somewhat in the world, but has traditionally been 4-6 weeks. Interestingly, several studies have indicated an increased rate of pathologic complete response (pCR) in the resected specimen when surgery was performed 6-12 weeks compared to <6 weeks after nCRT. In turn, pCR in the resected specimen is associated with improved survival. For these reasons, many centers have routinely started to delay surgery for up to 12 weeks after completed nCRT.

In this thesis, timing of surgery after nCRT is investigated in depth. In study I the existing data in NREV (national registry of esophageal and gastric cancer) was analyzed. In study II- IV results from the randomized controlled trial NeoRes II on timing of surgery after CROSS type nCRT was reported.

In Study I, 643 patients treated with nCRT and esophagectomy were identified in the NREV register and divided into two groups; TTS ≤7 weeks (344 patients, 53.5%) and >7 weeks (299 patients, 46.5%). No significant differences were found between ≤7 and >7 weeks TTS in terms of total postoperative complications, pCR or overall survival.

In the NeoRes II trial, on which studies II-IV are based, 249 patients were randomized after receiving CROSS type nCRT and of these 223 patients were resected; 117 allocated to 4-6 weeks and 106 to 10-12 weeks TTS. We found no significant difference concerning overall postoperative complications, severe complications or length of hospital stay. The trial's primary endpoint, complete histological response in the primary tumor, usually referred to as pathological complete response (pCR), for patients with adenocarcinoma (AC) did not differ significantly between allocation to standard versus prolonged TTS (21% versus 26%, P-value 0,429). Nor did it for squamous cell carcinoma (SCC), or for other histological outcomes such as overall TRG, tumor-free resection margins or number of resected or metastatic lymph nodes. With a median follow-up time of 36 months, no significant differences regarding overall survival or recurrence were found. However, the first quartile survival (time point when 25% mortality was reached) was significantly worse in the prolonged TTS group (difference 12.3 months, P=0.003, 95% CI 3.7-21.0). Furthermore, patients with TRG 4 (>50% remaining cancer cells) had significantly worse overall survival (HR 2.5, 95% CI 1.1- 5.8).

As the oncological treatment options overall improve over time, some patients may in the future be possible to cure without planned surgery. However, how to safely identify these patients and separate them from the majority of patients who currently still do need surgery, is not known. In study IV the accuracy of clinical response evaluation with endoscopy was investigated comparing findings just before surgery at 4-6 weeks versus at 10-12 weeks after completed nCRT. We found that the accuracy of endoscopic biopsies in predicting residual tumor (TRG 2-4), did not differ significantly between 4-6 and 10-12 weeks TTS.

In conclusion, TTS after completed nCRT does not seem to be of any major importance with regard to short-term postoperative morbidity or mortality. Moreover, prolonged TTS does not significantly improve cCR. In addition, our data suggest that routinely prolonging TTS from 4-6 to 10-12 weeks may be detrimental with regard to overall survival, especially for pathological non-responders (TRG 4). Detection of residual tumor by endoscopic biopsies did not improve with prolonged TTS.

List of scientific papers

I. Association between time interval from neoadjuvant chemoradiotherapy to surgery and complete histological tumor response in esophageal and gastroesophageal junction cancer: a national cohort study. F Klevebro, K Nilsson, M Lindblad, S Ekman, J Johansson, L Lundell, N Ndegwa, J Hedberg, M Nilsson. Diseases of the esophagus. 2020, 33(5). https://doi.org/10.1093/dote/doz078

II. Surgical morbidity and mortality from the multicenter randomized controlled NeoRes II trial. Standard versus prolonged time to surgery after neoadjuvant chemoradiotherapy for esophageal cancer. K Nilsson, F Klevebro, I Rouvelas, M Lindblad, E Szabo, I Halldestam, U Smedh, B Wallner, J Johansson, G Johnsen, E K Ahlin, H-O Johannessen, G O Hjortland, I Bartella, W Schröder, C Bruns, M Nilsson. Annals of surgery. 2020, 272(5): 684-689. https://doi.org/10.1097/SLA.0000000000004340

III. Oncological outcomes of standard versus prolonged time to surgery after neoadjuvant chemoradiotherapy for oesophageal cancer in the multicentre, randomised, controlled NeoRes II trial. K Nilsson, F Klevebro, B Sunde, I Rouvelas, M Lindblad, E Szabo, I Halldestam, U Smedh, B Wallner, J Johansson, G Johnsen, E K Aahlin, H-O Johannessen, G Alexandersson von Döbeln, G O Hjortland, N Wang, Y Shang, D Borg, A Quaas, I Bartella, C Bruns, W Schröder, M Nilsson. Annals of oncology. 34(11): 1015-1024. https://doi.org/10.1016/j.annonc.2023.08.010

IV. Early versus delayed clinical response evaluation after neoadjuvant chemoradiotherapy: Data from the randomized NeoRes II trial. K Nilsson, G Saliba, F Klevebro, B Sunde, I Rouvelas, M Lindblad, E Szabo, I Halldestam, U Smedh, B Wallner, J Johansson, G Johnsen, E K Aahlin, H-O Johannessen, G Alexandersson von Döbeln, G O Hjortland, N Wang, D Borg, A Quaas, I Bartella, C Bruns, W Schröder, M Nilsson. [Manuscript]