Development of perioperative immune checkpoint inhibitor therapy for locally advanced esophageal squamous cell carcinoma

Abstract

The standard preoperative treatment for resectable locally advanced esophageal squamous cell carcinoma (ESCC) is chemoradiotherapy in western countries (based on the CROSS trial) and triplet chemotherapy in Japan (based on the JCOG1109 trial). Postoperative nivolumab has recently been shown to improve disease-free survival in resectable locally advanced esophageal cancer after preoperative chemoradiotherapy in patients who had residual pathological disease, based on the CheckMate 577 trial. Furthermore, preoperative immune checkpoint inhibitor-containing treatments have also been developed. The JCOG1804E trial is presently evaluating the safety and efficacy of preoperative nivolumab-containing chemotherapy for resectable locally advanced ESCC. This review discusses the treatment of resectable locally advanced ESCC and future perspectives on perioperative immune checkpoint inhibitor-containing treatments.

Article Highlights

• This review discussed the differences in treatment of resectable locally advanced esophageal cancer between western countries and Japan, summarized the results of trials that include immune checkpoint inhibitor (ICI) therapy, and discussed the prospects and problems of future development of ICI therapy for these cancers.

• Treatment strategies for resectable locally advanced esophageal cancer differ between Japan and western countries.

• The standard preoperative treatment is chemoradiotherapy containing carboplatin and paclitaxel based on the CROSS trial in western countries, and docetaxel, cisplatin, and fluorouracil chemotherapy in Japan.

• ICIs have demonstrated survival benefit in unresectable advanced, recurrent, or metastatic esophageal cancer, and postoperative nivolumab after preoperative chemoradiotherapy and esophagectomy was found to be superior in resectable locally advanced esophageal cancer in the CheckMate 577 trial. However, these findings cannot be extrapolated to Japan because of the difference in preoperative treatment.

• Several studies of preoperative ICI therapy combined with chemotherapy or chemoradiotherapy have been performed and showed promising efficacy and good tolerability.

• In Japan, the multicohort phase I JCOG1804E (FRONTiER) study is evaluating the safety and efficacy of preoperative nivolumab combined with either cisplatin and 5-fluorouracil, or docetaxel, cisplatin, and fluorouracil, or fluorouracil, leucovorin, oxaliplatin, and docetaxel for resectable locally advanced esophageal squamous cell carcinoma.

Keywords: :

• esophageal squamous cell carcinoma
• immune checkpoint inhibitors
• nivolumab
• pembrolizumab
• perioperative treatment
• postoperative treatment
• preoperative treatment

1. Introduction

Esophageal cancer had the seventh highest incidence and the sixth highest mortality among cancers worldwide in 2020 [1]. The major histological subtypes of esophageal cancer are adenocarcinoma and squamous cell carcinoma (SCC). The frequency of these histological subtypes is influenced by geographic region and culture. Esophageal adenocarcinoma (EAC) is common in high-income countries, including the USA and Europe; obesity, gastroesophageal reflux disease, and Barrett’s esophagus have been identified to be major risk factors for EAC. In contrast, esophageal squamous cell carcinoma (ESCC) is common in Asia and lower-income countries; ESCC is strongly associated with alcohol consumption and smoking [1]. ESCC accounts for about 87% of all esophageal cancers worldwide, while about 60% of esophageal cancers in the USA are adenocarcinoma [2]. Given that 90% of esophageal cancers in Japan are ESCC, most clinical trials in esophageal cancer in the Japanese population have focused on ESCC [^3-5]. EAC and ESCC differ not only according to region and culture but also in terms of their clinical features and prognosis. Patients with EAC have more advanced disease, a higher proportion of lymph node metastasis, a higher recurrence rate, and a lower incidence of respiratory and otolaryngological disease than those with ESCC [6]. Moreover, surgical treatment differs between western countries and Japan. In western countries, esophagectomy with less extensive lymph node dissection is commonly performed, whereas in Japan, esophagectomy and three-field lymph node dissection is the standard curative surgery [7]. Against this background, treatment strategies have differed between western countries and Japan.

2. Perioperative treatment for resectable locally advanced esophageal cancer

Surgery is the mainstay of treatment for resectable locally advanced esophageal cancer. However, the prognosis in patients undergoing surgery alone has been poor. Therefore, multimodal treatment that includes chemotherapy and/or radiotherapy has been developed to improve clinical outcomes. Preoperative chemoradiotherapy is established in western countries. Regardless of histological subtype, the standard preoperative treatment in the west is concurrent chemoradiotherapy with carboplatin and paclitaxel based on the findings of the CROSS trial, in which surgery alone was compared with preoperative chemoradiotherapy followed by surgery [8]. A total of 368 patients were enrolled, 75% of whom had adenocarcinoma. The primary end point was overall survival (OS), which was significantly improved in the preoperative chemoradiotherapy plus surgery group compared with the surgery alone group (49.4 vs 24.0 months; hazard ratio [HR]: 0.657; 95% CI: 0.495–0.871; p = 0.003). A subgroup analysis revealed that chemoradiotherapy tended to have a greater benefit in patients with SCC than in those with adenocarcinoma. A pathological complete response (pCR) was observed in 29% of cases. Hematological toxicity was mild, with 7% of patients experiencing grade 3 hematological toxicity and 1% experiencing grade 4 hematological toxicity and febrile neutropenia [8]. After a median follow-up of 147 months, OS was better in the preoperative chemoradiotherapy plus surgery group than in the surgery only group (HR: 0.70; 95% CI: 0.55–0.89; p = 0.004), with respective 10-year OS rates of 38 and 25%. The risk of death from a cause other than esophageal cancer was comparable between the two groups (HR: 1.17; 95% CI: 0.68–1.99). Although the local recurrence rate was lower in the preoperative chemoradiotherapy plus surgery group than in the surgery alone group (8 vs 18%), the distant recurrence rate was comparable between the groups (27 vs 28%) [9].

Meanwhile, perioperative chemotherapy has been developed in Japan. The randomized phase III JCOG9204 trial found that the 5-year disease-free survival (DFS) rate was better in patients with ESCC who underwent surgery followed by postoperative chemotherapy than in those who underwent surgery alone (55 vs 45%; HR: 0.73; 95% CI: 0.51–1.03; p = 0.037) [10]. Worldwide, several studies have demonstrated the effect of preoperative chemotherapy for esophageal cancer [^10-13]. However, it has been unclear whether preoperative or postoperative chemotherapy is superior. The phase III JCOG9907 trial compared preoperative and postoperative chemotherapy with cisplatin and 5-fluorouracil (CF) in Japanese patients with locally advanced ESCC. The 5-year OS was significantly better in the preoperative chemotherapy group than in the postoperative chemotherapy group (55 vs 43%; HR: 0.73; 95% CI: 0.54–0.99; p = 0.04) but the survival outcome was still poor [14]. Preoperative docetaxel, cisplatin, and fluorouracil (DCF) therapy has been developed to improve clinical outcomes and showed promising efficacy in a phase II study [15]. However, the clinical question of whether preoperative chemotherapy or preoperative chemoradiotherapy, which is the standard treatment in western countries, is superior is unresolved. The randomized three-arm phase III JCOG1109 trial was performed to validate the superiority of preoperative DCF therapy and preoperative chemoradiotherapy consisting of CF plus 41.4 Gy over preoperative CF therapy in patients with ESCC. The primary end point was OS, and it was expected that there would be a 10% improvement in the 3-year OS rate with preoperative DCF therapy or preoperative chemoradiotherapy over preoperative CF, which had a 3-year OS rate of 63% in the JCOG9907 trial [14]. A total of 601 patients were randomly assigned to the treatment arms. The 3-year OS rate was significantly superior after preoperative DCF therapy than after preoperative CF therapy (72.1 vs 62.6%; HR: 0.68; 95% CI: 0.50–0.92; p = 0.006). Furthermore, there was no significant difference in 3-year OS between patients who received preoperative chemoradiotherapy and those who received preoperative CF therapy (68.3 vs 62.6%; HR: 0.84; 95% CI: 0.63–1.12; p = 0.12). Median progression-free survival was 2.7 years after preoperative CF therapy, not reached after preoperative DCF therapy, and 5.3 years after preoperative chemoradiotherapy, with respective pCR rates of 2.2, 18.6, and 36.7%. Grade 3 or 4 adverse events were observed in at least 10% of patients after preoperative CF therapy, preoperative DCF therapy, and preoperative chemoradiotherapy, with the rates being: 6.7, 63.8, and 53.9%, respectively, for leukocytopenia; 23.4, 85.2, and 44.5%, respectively, for neutropenia; 1.0, 16.3, and 4.7%, respectively, for febrile neutropenia; 6.2, 26.0, and 11.0%, respectively, for hyponatremia; and 8.3, 21.4, and 14.7%, respectively, for loss of appetite [16]. Based on the results of the JCOG1109 trial, preoperative DCF therapy has become standard treatment for resectable locally advanced esophageal cancer in Japan [4,5]. However, although the preoperative triplet regimen has improved outcomes, the pathological response and prognosis are still limited.

3. Mechanism of action of immune checkpoint inhibitors

Immune checkpoint inhibitors (ICIs), which include anti-CTLA-4, anti-PD-1, and anti-PD-L1 antibodies, have been developed in recent years and shown clinical benefits in various types of cancer [^17-27]. The immune response to cancer entails several steps. Tumor-specific antigens are released from tumor cells damaged by necrosis and apoptosis as a result of genetic or epigenetic alterations. Antigen-presenting cells (e.g., dendritic cells) recognize tumor-specific antigens and present them to T cells. The activated T cells then infiltrate the tumor tissue and kill the tumor cells [28,29]. Immune checkpoints interfere with interactions between T cells and cancer cells and regulate T-cell activation. CTLA-4 expressed on T cells binds to CD80/86 expressed on antigen-presenting cells, preventing CD80/86–CD28 binding and suppressing T-cell activation [30]. The interaction between PD-1 expressed on T cells and PD-L1 expressed on tumor cells and immune cells also downregulates the immune response and leads to T-cell apoptosis [31,32]. Hence, blocking immune checkpoints activates the immune response and has an antitumor effect (Figure 1).

Figure 1. Mechanism of action of immune checkpoint inhibitor therapy.

4. Preoperative ICI therapy

Nivolumab, which is a humanized monoclonal IgG4 anti-PD-1 antibody, has shown efficacy in patients with advanced esophagogastric adenocarcinoma or SCC [24,33]. Several trials have also assessed the benefits of preoperative chemotherapy or chemoradiotherapy combined with an ICI. Two trials assessing preoperative chemoradiotherapy with ICIs and 13 trials evaluating preoperative chemotherapy with ICIs have been reported (Table 1) [^34-48]. Pembrolizumab, camrelizumab, toripalimab, sintilimab, and tislelizumab are anti-PD-1 antibodies, and atezolizumab is an anti-PD-L1 antibody. One study included patients with EAC and was conducted in western countries [35] and the others included patients with ESCC and were performed in China. One study administered ICIs after chemoradiotherapy in addition to concurrent chemoradiotherapy [35], and the others administered ICIs concurrently with chemoradiotherapy or chemotherapy.

Table 1. Reported studies of preoperative immune checkpoint inhibitor therapy for locally advanced esophageal cancer.

Trial	ICI therapy	Phase	Histology	Patients, n	Chemotherapy	Radiotherapy	Timing of ICI therapy	R0	pCR	ORR	mPFS	mOS	Ref.
PALACE-1	Pembrolizumab	Ib	SCC	20	Carboplatin + paclitaxel	41.4 Gy	Concurrent with CRT	94.4%	55.6%	NA	NA	NA	[34]
PERFECT	Atezolizumab	II	AC	40	Carboplatin + paclitaxel	41.4 Gy	Concurrent and after CRT	82.5%	25.0%	NA	19.4 months	29.7 months	[35]
Shen et al.	Nivolumab, pembrolizumab, camrelizumab	II	SCC	28	Carboplatin + nab-paclitaxel	None	Concurrent with CT	96.3%	33.3%	88.9%	NA	NA	[36]
PEN-ICE	Pembrolizumab	II	SCC	18	Nedaplatin + taxane	None	Concurrent with CT	84.6%	46.2%	50.0%	NA	Not reached	[37]
Liu et al.	Camrelizumab	II	SCC	60	Carboplatin + nab-paclitaxel	None	Concurrent with CT	98.0%	39.2%	49.0%	NA	NA	[38]
NIC-ESCC2019	Camrelizumab	II	SCC	56	Cisplatin + nab-paclitaxel	None	Concurrent with CT	100.0%	35.3%	66.7%	Not reached	Not reached	[39]
Wang et al.	Camrelizumab	Ib	SCC	30	Nedaplatin + nab-paclitaxel + apatinib	None	Concurrent with CT	96.6%	24.1%	93.3%	NA	NA	[40]
Yang et al.	Camrelizumab	Pilot study	SCC	23	Carboplatin + nab-paclitaxel	None	Concurrent with CT	100.0%	25.0%	90.5%	Not reached	NA	[41]
Yang et al.	Camrelizumab	Pilot study	SCC	16	Carboplatin + paclitaxel	None	Concurrent with CT	93.8%	31.3%	81.3%	18.3 months	19.2 months	[42]
Xing et al.	Toripalimab	II	SCC	30	Cisplatin + paclitaxel	None	Concurrent with CT	100.0%	20.8%	66.7%	NA	NA	[43]
He et al.	Toripalimab	II	SCC	20	Cisplatin + paclitaxel	None	Concurrent with CT	87.5%	18.8%	61.1%	NA	NA	[44]
ESONICT-2	Toripalimab	II	SCC	20	Cisplatin + docetaxel	None	Concurrent with CT	100.0%	16.7%	70.0%	NA	NA	[45]
ESONICT-1	Sintilimab	II	SCC	30	Cisplatin + nab-paclitaxel	None	Concurrent with CT	100.0%	17.3%	66.6%	NA	NA	[46]
SIN-ICE	Sintilimab	Pilot study	SCC	23	Nedaplatin + taxane	None	Concurrent with CT	94.1%	35.3%	66.7%	13.8 months	NA	[47]
TD-NICE	Tislelizumab	II	SCC	45	Carboplatin + nab-paclitaxel	None	Concurrent with CT	80.5%	50.0%	NA	NA	NA	[48]

AC: Adenocarcinoma; CRT: Chemoradiotherapy; CT: Chemotherapy; ICI: Immune checkpoint inhibitor; mOS: Median overall survival; mPFS: Median progression-free survival; NA: Not assessed; ORR: Objective response rate; pCR: Pathological complete response; R0: Complete tumor resection; SCC: Squamous cell carcinoma.

The complete tumor resection (R0) and pCR rates in these trials were 82.5–94.4% and 25.0–55.6%, respectively, in patients who received preoperative chemoradiotherapy plus ICIs, and 80.5–100.0% and 16.7–50.0%, respectively, in those who received preoperative chemotherapy plus ICIs. Objective response rates of 49.0–93.3% were reported for patients who received preoperative chemotherapy combined with ICIs. Grade 3 or worse adverse events occurred in 40.0–65.0% of patients who received preoperative chemoradiotherapy plus ICIs and in 3.0–56.7% of those who received preoperative chemotherapy plus ICIs. Most of the adverse events involved hematological toxicity. Rash, pneumonitis, diarrhea, hyperthyroidism, and increased transaminase levels occurred in up to 5% of patients and were considered immune-related adverse events.

Although these trials have shown promising efficacy, there are still some issues to be addressed. First, the preoperative chemotherapy combined with ICI therapy in these trials consisted of platinum agents and taxanes, which is different from the standard preoperative chemotherapy used in Japan. The safety and efficacy of preoperative DCF therapy combined with ICIs are still unknown. Second, in all studies with the exception of one, the ICI was administered concurrently with chemoradiotherapy or chemotherapy. The optimal dosing schedule for ICIs when combined with chemoradiotherapy or chemotherapy has not been established. Changes in the tumor immune environment caused by chemotherapy and radiotherapy might affect the efficacy of ICI therapy. ICIs combined with chemotherapy have shown remarkable efficacy in advanced esophageal cancer and gastric cancer [^49-52], and some studies have shown favorable outcomes after ICI therapy followed by chemotherapy in lung cancer [53,54]. In previous reports, the impact of adding ICI to chemoradiotherapy on the pCR rate was limited [34,35], but preoperative chemotherapy combined with ICI therapy has improved the pCR rate [^36-48]. Third, the impact of the addition of ICI on survival is unknown because of short follow-up durations and small numbers of patients.

5. Postoperative ICI therapy

Although preoperative chemoradiotherapy plus surgery was more effective than surgery alone in the CROSS trial, the risk of recurrence remained high, particularly in patients who did not achieve a pCR [55]. Therefore, the efficacy and safety of postoperative nivolumab monotherapy after preoperative chemoradiotherapy plus surgery for esophageal or esophagogastric junction cancer were evaluated in the randomized, double-blind, placebo-controlled, phase III CheckMate 577 trial. Patients in whom R0 resection, but not pCR, was achieved received nivolumab or placebo after surgery for 1 year. A total of 794 patients were enrolled, 71% of whom had adenocarcinoma and the remainder of whom had SCC; 13% of the patients were Asian, and 17% in the nivolumab group had PD-L1 expression ≥1%. The primary end point was DFS, which was significantly longer in the nivolumab group than in the placebo group (median: 22.4 vs 11.0 months; HR: 0.69; 96.4% CI: 0.56–0.86; p < 0.001). Nivolumab was observed to be effective regardless of histological subtype or PD-L1 expression status. Grade 3 or 4 adverse events of any cause occurred in 34% of patients in the nivolumab group and 32% in the placebo group. Most immune-related adverse events were grade 1 or 2, and grade 3 or 4 events were seen in no more than 1% of patients. Pneumonitis (<1%) and rash (<1%) were the most frequent grade 3 or 4 immune-related adverse events in the nivolumab group [56].

Based on these studies, preoperative chemoradiotherapy and postoperative nivolumab monotherapy have become the standard perioperative treatment for resectable locally advanced esophageal cancer in western countries.

A randomized, placebo-controlled, phase II study in Asia evaluated the efficacy of adjuvant durvalumab, an anti-PD-L1 antibody, in patients with advanced ESCC who also received preoperative chemoradiotherapy consisting of CF plus 41.1 Gy. Unlike CheckMate 577, this study included patients with pCR but not patients with EAC. The primary end point was DFS. Eighty-six patients were divided into a durvalumab group (n = 45) and a placebo group (n = 41). The median DFS was not reached in either group (HR: 1.18; 95% CI: 0.62–2.27; p = 0.61). Median OS was 50.6 months in the durvalumab group and not reached in the placebo group (HR: 1.08; 95% CI: 0.52–2.24; p = 0.85). Patients in the durvalumab group who achieved pCR tended to have shorter DFS (HR: 1.76; 95% CI: 0.42–7.40; p = 0.43) and OS (HR: 2.26; 95% CI: 0.41–12.34; p = 0.33) than their counterparts in the placebo group, but these differences were not statistically significant. In patients who did not achieve pCR, there was no significant between-group difference in DFS (HR: 1.02; 95% CI: 0.49–2.13; p = 0.95) or OS (HR: 0.79; 95% CI: 0.34–1.81; p = 0.43). The exploratory analysis showed that DFS after preoperative chemoradiotherapy was longer when PD-L1 expression was high than when it was low. Treatment-related adverse events of any grade occurred in 58% of patients in the durvalumab group and 32% in the placebo group; however, most were grade 1 or 2. Grade 3 or higher adverse events in the durvalumab group were pneumonitis (n = 1), fatigue (n = 1), adrenal insufficiency (n = 1), and bowel perforation (n = 1) [57]. Of note, in this trial, neither DFS nor OS was prolonged by postoperative durvalumab, which is in contrast with the results of CheckMate 577 [56]. Indeed, there are several differences between this study and CheckMate 577. As previously mentioned, all patients in this study were Asian and had ESCC, although subgroup analysis in CheckMate 577 showed benefits of nivolumab in patients with ESCC (HR: 0.61; 95% CI: 0.42–0.88) and in Asian patients (HR: 0.70; 95% CI: 0.41–1.22) [56]. Moreover, this trial was smaller than CheckMate 577, and the inclusion of pCR cases might have contributed to its negative results.

Although postoperative nivolumab monotherapy demonstrated efficacy in the CheckMate 577 trial, the efficacy of postoperative nivolumab is still unclear in Japan, where the standard preoperative treatment for resectable locally advanced esophageal cancer is triplet chemotherapy and not chemoradiotherapy. A phase II trial found postoperative tegafur/gimeracil/oteracil (S-1) after preoperative chemotherapy plus esophagectomy to have promising efficacy, with a 3-year relapse-free survival rate of 72.3% and a 3-year OS rate of 85.0% [58]. A phase III trial (JCOG2206, SUNRISE) is presently evaluating the efficacy of adjuvant nivolumab monotherapy or S-1 monotherapy with that of observation alone after neoadjuvant chemotherapy followed by surgery in Japan [59]. This trial includes patients who have received preoperative chemotherapy followed by esophagectomy with R0 resection, who are randomized to observation, postoperative nivolumab monotherapy, or postoperative S-1 therapy. The primary end point is relapse-free survival.

6. Development of ICIs in Japan

Chemotherapy has been reported to have immunosuppressive effects (e.g., leukopenia) and to activate the immune system [60,61]. Chemotherapy causes immunogenic cell death by promoting the release of tumor antigens and an immune response [62]. Therefore, chemotherapy is considered to amplify the effects of ICIs. A preclinical study found that ICI therapy was more effective when administered preoperatively rather than postoperatively [63]. The efficacy of a combination of nivolumab and cytotoxic chemotherapy was found to be superior to that of chemotherapy alone when used as a preoperative treatment in non-small-cell lung cancer [64].

Although preoperative DCF is the standard treatment for resectable locally advanced ESCC in Japan, the standard perioperative treatment for EAC has not been established. In western countries, perioperative fluorouracil and leucovorin, oxaliplatin, and docetaxel (FLOT) therapy has shown considerable efficacy in EAC and gastric cancer [65]. However, it is unclear whether FLOT or a triplet regimen like DCF is superior for ESCC. Furthermore, the optimal timing for ICI therapy when combined with preoperative chemotherapy has not been established. Myelosuppression and depletion of lymphocytes may diminish the effect of ICIs. Preoperative nivolumab monotherapy has shown efficacy in lung cancer, and administration of an ICI that does not induce myelosuppression might be effective [66]. In view of these outstanding questions, the multicohort phase I JCOG1804E (FRONTiER) study is now underway to assess the safety and efficacy of preoperative nivolumab plus CF, DCF, or FLOT administered by various protocols in patients with resectable locally advanced ESCC [67,68].

The accrual target is 36 patients, with six patients each enrolled in cohorts A–D and 12 enrolled in cohort E. Patients in cohort A receive two courses of cisplatin (80 mg/m²), nivolumab (360 mg) on day 1, and fluorouracil (800 mg/m²) on days 1–5 every 3 weeks. Patients in cohort B receive one dose of nivolumab (240 mg) 2 weeks before initiation of chemotherapy and then follow the same schedule as cohort A. Patients in cohort C receive three courses of docetaxel (70 mg/m²), cisplatin (70 mg/m²), and nivolumab (360 mg) on day 1, and fluorouracil (750 mg/m²) on days 1–5 every 3 weeks. Patients in cohort D receive one dose of nivolumab (240 mg) 2 weeks before the initiation of chemotherapy and then follow the same schedule as cohort C. Patients in cohort E receive four courses of docetaxel (50 mg/m²), oxaliplatin (85 mg/m²), leucovorin (200 mg/m²), fluorouracil (2600 mg/m²), and nivolumab (240 mg) on day 1 every 2 weeks. Esophagectomy with two- to three-field lymph node dissection is performed within 12 weeks after the final dosage of preoperative chemotherapy. The primary end point is the incidence of dose-limiting toxicities (DLTs) between the first dose and postoperative day 30, and the secondary end points are the objective response rate during preoperative chemotherapy, the pCR rate, the percentage of curative resections, the protocol treatment completion rate, progression-free survival and OS rates, and the frequency of adverse events.

Short-term results are now available for cohorts A and B. A total of 13 patients were registered (six in cohort A and seven in cohort B). Among these cohorts, 12 of the 13 patients did not experience DLTs; one patient (in cohort B) was excluded because a nonresidual resection could not be performed. Grade 3 or higher adverse events were: neutropenia during preoperative therapy (46.3%) and anastomotic leakage during the postoperative period (8.3%). One patient in cohort B experienced grade 2 adrenal insufficiency. There were no grade 4 adverse events or treatment-related deaths. In cohort A, the pCR rate was 33.3% (n = 2/6) and the R0 resection rate in cohorts A and B combined was 92.3% (n = 12/13) [67]. The short-term outcomes in cohorts C and D have now been reported as well. Twelve patients were enrolled, six each in cohorts C and D. No DLTs were seen in cohort C; however, one patient in cohort D experienced DLTs (grade 3 rash and dyspnea). Grade 3 or higher adverse events during preoperative therapy in cohorts C and D were: leukopenia (n = 7), neutropenia (n = 1), lymphopenia (n = 1), febrile neutropenia (n = 1), anorexia (n = 3), hyponatremia (n = 2), and nausea (n = 1). One patient developed intestinal pneumonitis during the postoperative period. The R0 resection rate in cohorts C and D combined was 91.7% (n = 11/12). The pCR rate was 16.7% in cohort C and 50.0% in cohort D [68]. Preoperative CF or DCF plus nivolumab therapy followed by esophagectomy has thus shown favorable tolerability and promising efficacy. Results from cohort E and the investigation of biomarkers that can predict efficacy are awaited. Tumor, blood, and stool samples are being collected in this study to investigate tumor PD-L1 expression, tumor mutation burden, tumor-infiltrating lymphocytes, and the transcriptome and gut microbiome.

7. Conclusion

Preoperative chemoradiotherapy is the standard treatment for resectable locally advanced esophageal cancer in western countries, while preoperative DCF therapy is the standard therapy in Japan. Although postoperative nivolumab therapy prolonged DFS in the CheckMate 577 trial, it has not become the standard treatment in Japan because of differences in preoperative treatment. Small prospective trials of preoperative ICIs combined with chemotherapy or chemoradiotherapy have shown promising efficacy and tolerability. A multicohort phase I study is underway in Japan to assess the safety and efficacy of preoperative nivolumab with CF, DCF or FLOT administered by various protocols for resectable locally advanced ESCC. Long-term survival outcomes and the results of ongoing large preoperative ICI trials are awaited, and studies of predictive biomarkers are warranted.

8. Future perspective

Phase III trials that are evaluating the efficacy of perioperative ICI therapy in patients with ESCC are shown in Table 2. The efficacy of postoperative sintilimab monotherapy was assessed in the NCT05495152 trial [56], which had a design similar to that of the CheckMate 577 trial [56]. The NCT04848753 and NCT05213312 trials are examining the efficacy of adding an anti-PD-1 inhibitor to preoperative chemotherapy, and the NCT04973306 and NCT05244798 trials are evaluating the efficacy of adding an anti-PD-1 inhibitor to preoperative chemoradiation therapy. NCT05244798 is a three-arm trial that is comparing preoperative chemotherapy plus an anti-PD-1 inhibitor, preoperative chemoradiotherapy plus an anti-PD-1 inhibitor, and preoperative chemoradiotherapy. The efficacy of adding an anti-PD-1 inhibitor to preoperative and postoperative treatment is being evaluated in NCT04280822 and NCT04807673.

Table 2. Ongoing phase III studies of perioperative immune checkpoint inhibitor therapy for resectable locally advanced esophageal cancer.

Trial	ICI therapy	Histology	Patients, n	Neoadjuvant therapy	Adjuvant therapy	Primary end point
Adjuvant
NCT05495152	Sintilimab	SCC	219	None	Sintilimab	DFS
					observation
Neoadjuvant
NCT04848753	Toripalimab	SCC	632	Cisplatin + paclitaxel + toripalimab	None	EFS
				Cisplatin + paclitaxel
NCT05213312	Nivolumab	SCC	90	Cisplatin + paclitaxel or cisplatin + 5-fluorouracil + nivolumab	None	pCR
				Cisplatin + paclitaxel or cisplatin + 5-fluorouracil
NCT04973306	Tislelizumab	SCC	176	Carboplatin + paclitaxel + tislelizumab + radiotherapy	None	pCR
				Carboplatin + paclitaxel + radiotherapy
NCT05357846	Sintilimab	SCC	422	Cisplatin + paclitaxel + sintilimab + radiotherapy	None	OS
				Cisplatin + paclitaxel + radiotherapy
NCT05244798	Sintilimab	SCC	420	Carboplatin + nab-paclitaxel + sintilimab	None	pCR
				Carboplatin + nab-paclitaxel + sintilimab + radiotherapy
				Carboplatin + nab-paclitaxel + radiotherapy
Perioperative therapy
NCT04280822	Toripalimab	SCC	400	Cisplatin + paclitaxel + tislelizumab	Tislelizumab	EFS
				Cisplatin + paclitaxel	None
NCT04807673	Pembrolizumab	SCC	342	Cisplatin + paclitaxel + pembrolizumab	Pembrolizumab	EFS
				Cisplatin + paclitaxel + radiotherapy	None

DFS: Disease-free survival; EFS: Event-free survival; OS: Overall survival; pCR: Pathological complete response; SCC: Squamous cell carcinoma.

Several problems remain regarding ICIs for resectable locally advanced esophageal cancer. First, although preoperative chemotherapy is the standard treatment in Japan, it is unclear whether preoperative chemotherapy or chemoradiotherapy is more effective when combined with ICI therapy. Radiotherapy enhances the immune response by damaging tumors and causing the release of tumor antigens [69,70]. Furthermore, radiation increases the expression of major histocompatibility complex class I on the surface of tumor cells, which present intracellular antigens [71]. Preclinical studies have shown that radiotherapy upregulates PD-L1 expression, and the combination of radiation and PD-1/PD-L1 inhibitors is expected to have a synergistic antitumor effect [72,73]. Second, the optimal timing for administration of ICI therapy is unknown. Whether ICIs should be administered preoperatively, postoperatively, or both preoperatively and postoperatively is controversial. Further research is needed to determine whether preoperative ICI therapy should be administered concurrently with, before, or after chemotherapy or chemoradiation. Third, there is no established predictive biomarker for perioperative ICI therapy. PD-L1 expression after but not before chemoradiotherapy was found to be associated with survival in patients who received postoperative durvalumab [57]. The reported studies of preoperative ICI therapy have investigated its efficacy in relation to CD8⁺, CD4⁺, TCF-1⁺, FOXP3⁺, CD56⁺, CD68⁺, and HLA-DR⁺ cells, tumor mutation burden, and cytokines such as IFN-γ [34,35,^37-39,41,44,48]. Furthermore, there have been several studies of the effect of ICIs on the gut microbiome [^74-77]. Considering that ICIs can cause permanent immune-related adverse events, including diabetes, thyroid dysfunction, and adrenal dysfunction, as well as fatal immune-related adverse events, such as pneumonitis, myositis, and encephalitis, there is a need to explore biomarkers that can predict adverse events as well as the response to ICIs.

Financial disclosure

The authors have no financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Writing disclosure

No writing assistance was utilized in the production of this manuscript.

Competing interests disclosure

T Kadono has no conflict of interest. S Yamamoto has received personal fees from ONO PHARMACEUTICAL and Bristol Myers Squibb, MSD, Taiho, M3, HOKUTO. K Kato has received research funds from Ono Pharmaceuticals, Bristol Myers Squibb, MSD, Merck Biopharma, Taiho Pharmaceutical, Bayer, AstraZeneca, Janssen, and Oncolys Biopharma; and honoraria from Ono Pharmaceuticals, Bristol Myers Squibb, MSD, and Taiho Pharmaceutical. The authors have no other competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript apart from those disclosed. This includes employment, consultancies, stock ownership or options and expert testimony.