In this prospective, single-arm pilot study, neoadjuvant treatment with sintilimab and chemotherapy (Nab-paclitaxel 260 mg/m2, d1 and cisplatin 75 mg/m2, d1-3) for 2 cycles in the resectable ESCC resulted in a pCR rate of 28.6% and MPR rate of 42.9%. The radiographic complete response rate according to RECIST 1.1 was 16.7%. No surgical delays or mortality within 3 months after surgery occurred. Eight (38.1%) patients developed surgical complications and all recovered by non-surgical interventions. These results preliminarily disclosed the efficacy and safety of our treatment regimen in neoadjuvant treatment of resectable ESCC.
The prognostic significance of pathologic complete response (pCR) after induction therapy in patients with esophageal cancer has been demonstrated in several studies [9, 10]. The pCR and MPR rates of our study were 28.6% and 42.9%. Since two patients with radiographic complete response and one patient with radiographic partial response did not receive surgery and the data of their pathologic response was lost, the real pCR and MPR rate of our treatment regimen may further increased. As we expected, in addition of immune checkpoint inhibitor to chemotherapy obviously elevated the pCR rate compared with preoperative chemotherapy [11].
The CROSS study promote the concurrent chemoradiotherapy to become the standard neoadjuvant therapy for resectable ESCC. Neoadjuvant concurrent chemoradiotherapy can achieve higher pCR rate compared with neoadjuvant chemotherapy and was widely adopted in western countries. However, its accessibility and affordability is poor in same area of China. In our study, the pCR rate was comparable with CROSS study (28.6% vs. 29%), indicated neoadjuvant therapy with the immune checkpoint inhibitor and chemotherapy can be a possible alternative. Another advantage of neoadjuvant chemotherapy is tis favorable safety profile. Radiotherapy may increase the incidence and severity of some AEs such as leukopenia, neutropenia and radiation esophagitis. In NEOCRTEC5010 study, the incidence of grade 3 / 4 leukopenia and neutropenia was 31.8% / 17.0% and 23.3% / 22.4%, respectively, while in our study, the incidence of grade 3 leukopenia and neutropenia was 27.6% and 20.7%. and only 3 patients occurred grade 4 haematological AEs. In the aspect of postoperative complications, our study arose two cases (9.5%) of respiratory complication and 4 cases (19.0%) of anastomotic leakage occurred while all patients recovered after non-surgical intervention. No patient died in hospital or within 90 days after surgery. Perioperative complication and treatment-related death will impair the survival benefit of the neoadjuvant. A randomized clinical trial involved 181 patients showed no significant PFS and OS benefit when adding radiotherapy to neoadjuvant although obviously higher histological complete response rate and higher R0 resection rate were observed. This phenomenon was mainly attributed to the relatively higher incidence of postoperative complication in the group of neoadjuvant chemoradiotherapy, especially anastomotic leakage and respiratory and more postoperative mortality events [7].
Nab-paclitaxel and cisplatin was deemed as a preferred chemotherapy regimen and widely used in China. Compared with CROSS study [12], high dose carboplatin (AUC = 2 per weeks) was replaced with moderated dose cisplatin in our study. In a multicenter, randomized clinical trial involved 321 patients in China to compare the efficacy and toxicity of paclitaxel with fluorouracil, cisplatin or carboplatin in the definitive chemoradiotherapy against esophageal squamous cell carcinoma (ESCC), cisplatin combined with paclitaxel showed the best 3-year OS rate although the toxicity was relatively higher [13].
In accord with other studies [14], the pathologic response was positively correlated with radiographic response. In patients who achieved pCR, two were evaluated as radiographic CR and 4 reached PR. In the rest 8 patients who had radiographic PR, 3 had near pathologic complete response (TGS1).
In terms of surgery, All 21 patients successfully underwent McKeown MIE without open surgery, the R0 resection rate reached 100%. Our mean operative time were 230 minutes. The intraoperative blood loss were 125.0 ± 34.7 mL (mean ± SD), comparable with the esophageal cancer without neoadjuvant treatment. All of this demonstrates that neoadjuvant sintilimab combined with chemotherapy does not increase the difficulty of surgery. At the same time, we found that after this neoadjuvant treatment, esophageal tumors tended to loosely adhere, and easier removal from surrounding tissue during operating. This appears to be different from patients after radiotherapy or neoadjuvant therapy for lung cancer. NEOSTAR trial (NCT03158129) suggests that due to hilar fibrosis in some patients, it is more difficult to separate the blood vessels. This indicated that different cancer types may cause different response to ICIs [15]. It also indicated that neoadjuvant sintilimab combined with chemotherapy did not increase the difficulty of the surgery.
This study has some limitations. First, it was an exploratory single-arm study with a small sample size. A randomized controlled study is warranted, especially to compare this regimen with standard neoadjuvant chemoradiotheray. Second, the duration of follow-up was limited and the data of survival was not mature. In addition, the predictive biomarkers will be explored in the further studies.