Definitive concurrent chemoradiotherapy with paclitaxel plus carboplatin is superior to cisplatin plus 5‐fluorouracil in patients with inoperable esophageal squamous cell carcinoma using retrospective, real‐world evidence

Abstract Background The optimal definitive chemotherapy regimen during concurrent chemoradiotherapy (CRT) for patients with advanced esophageal squamous cell carcinoma (ESCC) remains unclear because of conflicting evidence. This study aimed to compare the effectiveness of taxane‐based chemotherapy with that of conventional cisplatin plus 5‐fluorouracil (PF) as the chemotherapy regimen in definitive CRT for ESCC. Patients and Methods This retrospective study included patients with ESCC who received paclitaxel plus carboplatin (PC) or PF during definitive CRT between May 2012 and February 2015 in a medical center in Taiwan. Survival outcomes were compared after adjustment for risk factors. Results Overall, 229 patients were evaluated. Patients in the PC group had an objective response rate of 71.1% compared with the 51.4% of the PF group (p = 0.016). The PC group showed a significantly longer progression‐free survival (PFS, p = 0.002) and overall survival (OS, p = 0.019) than the PF group. Salvage surgery also helped prolong both the PFS and OS (p < 0001). Sex (male vs. female, HR, 1.831; 95% CI, 1.016–3.303), clinical stage (HR, 1.282; 95% CI, 1.069–1.537), accumulative radiation dose (≥41.4 Gy vs. <41.4 Gy; HR, 0.640; 95% CI, 0.413–0.993), salvage surgery (yes vs. no, HR: 0.412, 95% CI: 0.298–0.570), and regimen (PF vs. PC; HR, 1.514; 95% CI, 1.109–2.067) were independent prognostic factors for cancer mortality. Conclusion Compared with the PF regimen, the PC regimen for definitive CRT yielded significantly increased response rates and longer survival times; therefore, the PC regimen may be preferable for chemotherapy for definitive CRT in patients with advanced ESCC.


| INTRODUCTION
Esophageal cancer (EC) is one of the 10 most common cancers in men worldwide. 1,2 It is highly common in Taiwan, with a crude incidence of 21.87 male cases per million and 1.73 female cases per million. 3 Histologically, EC can be classified into esophageal adenocarcinoma (EAC) and esophageal squamous cell carcinoma (ESCC). EAC is more prevalent in North America and Europe with gastroesophageal reflux disease and obesity being the main risk factors. Meanwhile, ESCC is the more predominant EC in Asia, Africa, and South America. It is also predominant in African Americans in North America. Tobacco use and alcohol consumption are the main risk factors and esophageal squamous dysplasia is the precursor lesion. 4 Surgery is the primary curative modality for EC, but inoperable patients are treated primarily with definitive concurrent CRT (dCRT). 5 Chemotherapy and CRT are used as neoadjuvant therapies to improve long-term survival. [6][7][8][9][10] Among the chemotherapy regimens in CRT, cisplatin plus 5-fluorouracil (PF regimen) remains the most common choice (Category 1) [9]. However, long PF infusions are associated with toxicity and intolerance, prompting the development of other regimens, in recent years. The National Comprehensive Cancer Network guideline also recommends paclitaxel and carboplatin (PC) for dCRT based on the findings of the Chemoradiotherapy for Oesophageal Cancer Followed by Surgery Study (CROSS) in 2012 that demonstrated not only a better 5-year overall survival (OS) but also lower toxicity of PC as the chemotherapy regimen in CRT. [11][12][13][14] Local recurrence and distant metastasis are common after primary treatment. However, prospective trials evaluating the efficacy of different CRT regimens are lacking, and thus, the optimal regimen for dCRT or nCRT is yet to be established. Therefore, this study aimed to evaluate and compare the efficacies of PC and PF as chemotherapy regimens in CRT for EC.

| Study design and patients
This retrospective study was approved by the institutional review board of Chang Gung Memorial Hospital (Approval number 201701030B0).
We evaluated the patients with ESCC who were administered PC or PF during CRT between May 2012 and February 2015 at Chang Gung Memorial Hospital, Linkou, Taiwan. The inclusion criteria were: (i) age ≥20 years; (ii) Eastern Cooperative Oncology Group (ECOG) performance status 0-2; (iii) histologically confirmed ESCC; (iv) locally advanced, inoperable, newly diagnosed disease; (v) adequate liver and renal functions for dCRT; and (iv) had received chemotherapy or radiotherapy for at least 4 weeks. All patients received standard examinations for staging at diagnosis and were restaged according to the 8th edition of the American Joint Cancer Committee (AJCC) guidelines. The patients were divided into two groups, the PF and PC groups, according to their chemotherapy regimen.

| Chemoradiotherapy protocol
After the medical oncologists discussed the differences between the regimens with the patients and their families, a clinical shared decision-making model was used to select the regimen to be followed.
For chemotherapy, the PF regimen involved cisplatin at a dose of 60-75 mg/m 2 on day 1 and 5-fluorouracil (5-FU) at a dose of 800-1000 mg/m 2 on days 1 to 4 every 4 weeks and the PC regimen involved paclitaxel at a dose of 50 mg/m 2 every week for six weeks. Carboplatin was delivered simultaneously at a dose calculated from an area under the curve of 2 mg/ml/min every week for 6 weeks. All the patients were premedicated intravenously with dexamethasone, cimetidine, diphenhydramine, and granisetron according to the standard protocol of the institution. All adverse effects were documented according to the National Cancer Institute's Common Terminology Criteria for Adverse Events, version 3.0.10.
Radiotherapy was performed with a total externalbeam radiation dose of 50.4-59.4 Gy in 28-30 fractions at 1.8 Gy per fraction. Radiotherapy commenced on the first day of the first chemotherapy cycle and was performed in five fractions per week. In this study, we defined the cumulative RT dose of more than 4.14 Gy as the completion threshold of CCRT.
The radiotherapy technique was based on the institution's guidelines.

| Surgery after definitive chemoradiation
All the patients received dCRT. Surgery was performed for those who had cancer down-staging. In these patients, surgery was performed within 4 to 6 weeks after completion of the CRT. All the pathological characteristics were documented.

| Follow-up
Patients who underwent surgery were followed up every 3 months in the first year, every 6 months in the second year, and then annually thereafter until 5 years. Late toxic effects, disease recurrence, and death were reviewed using data from the medical charts. Recurrence was evaluated at the time of the first recurrence. Follow-up diagnostic investigations were performed only when recurrence was suspected. Among patients who underwent dCRT alone, those with good responses were commonly followed up with only close surveillance.

| Statistical analysis
All the data, including the staging according to the AJCC guidelines, were revised and updated until February 2020, which guaranteed a follow-up duration of at least 2 years. All the patients intended for dCCRT were enrolled for analysis. Age, sex, histology, clinical stage, locations of the primary tumor, radiation dose, and the number of salvage surgeries were analyzed using a simple descriptive method, while the Chi-square or Fisher exact tests were used to compare the PF and PC groups. All the significant factors from the univariate analysis were analyzed in the multivariate Cox regression model. A forward LR model was used to evaluate statistical significance. Progression-free survival (PFS) and overall survival (OS) were calculated from the date of diagnosis (with tissue evidence) to the date of disease progression/ relapse and death, respectively. Survival curves were generated using the Kaplan-Meier method and compared using the log-rank test. All the statistical analyses were performed using SPSS software, version 17.0 (SPSS Inc.). A p-value of less than 0.05 was considered to be statistically significant.

| Patient characteristics
Of the 293 patients (including 22 patients with cervical EC) who received dCRT, 44 (15.0%) patients who had initial metastasis at diagnosis and 20 (6.8%) patients who had concurrent synchronous double cancer, were excluded. Some patients with relatively early-stage disease (Stages IIa-IIb) also refused surgery and received dCRT mainly because of (i) their medically inoperable status or (ii) the primary location being at the cervical esophagus. Consequently, 229 patients were included in the analyses. Among them, 146 patients received the PF regimen and 83 patients received the PC regimen ( Figure 1). The patients' characteristics are shown in Table 1. The median age was 60 (range, 39-92) years. All the patients were followed until March 2020, with a median follow-up time of 12.0 (range, 0.8-85.7) months. There were no significant differences in the clinicopathological features, including age, sex, histology, clinical stages, primary tumor locations, tumor grade, baseline ECOG performance status, cumulative radiation doses, and surgery rates, after concurrent CRT, between the PF and PC groups. In total, 78 of the 83 patients (94%) in the PC group and 133 of the 146 patients (91.1%) in the PF group were female. The majority of patients in both groups had ESCC (224 of 229, 97%). Most of the patients in both groups had stage IIIB (50.6% and 42.5% in the PC and PF groups, respectively) and stage IVA (40.9% and 49.3% in PC and PF groups, respectively) disease. The reasons for the enrolled patients being categorized as inoperable or unresectable are based primarily, on their initial stages (at least stage III, 91.5% and 91.8% in PC and PF, respectively, Table 1) and age/performance/ willing to undergo surgery (8.5% and 8.2% in PC and PF, respectively, With respect to treatment characteristics, there was no significant difference in the radiation dose between the two groups (accumulative doses: ≥41.4 Gy vs. <41.4 Gy, p = 0.367). There was also no significant difference in the proportion of patients who underwent salvage surgery between the PC and PF groups (30 (36.1%) vs. 54 (37.0%), p = 0.899). The reasons for the 28 dropouts in this study were (i) intolerance to CCRT (toxicity) (n = 18), (ii) economic issues (n = 3), and (iii) malnutrition or sepsis (n = 7).

| Survival outcomes
The PC group had a significantly longer PFS (p = 0.002; Figure

| Univariate and multivariate analyses of the prognostic factors
To confirm the independent prognostic factors in this cohort, we used a Cox regression model for the univariate

| Novelty of the present study with comparisons
The current study reported real-world evidence of the effectiveness of the PC regimen during dCRT in Asian patients with inoperable ESCC. We found that patients who received the PC regimen had significantly longer PFS and OS than those who received the PF regimen. In addition, the baseline AJCC stage, surgery after the CRT, and chemotherapy regimens were independent prognostic factors after adjusting for related risk factors. To the best of our knowledge, this is the largest retrospective study to report that PC had superior effectiveness to PF as the chemotherapy regimen in concurrent CRT for EC. Table 3 summarizes the previous studies on the chemotherapy regimens for concurrent CRT, using a two-arm comparative design. There were two main types of chemotherapy regimens in the literature, namely, the taxanebased regimens 11,12,14-21 (e.g., paclitaxel or docetaxel) and the platinum plus 5-fluorouracil regimen. [22][23][24][25][26][27] Our findings were in line with those of previous studies from Japan 20 and China 18 demonstrating that patients who received concurrent CRT with a taxane-platinum regimen have more prolonged survival than those who received the PF regimen. However, it should be noted these studies used docetaxel-cisplatin, whereas we used paclitaxelcarboplatin. In contrast to our findings, Qu et al. (2017) reported that PF (n = 34) was superior to PC (n = 26) in inoperable EC. 17 Moreover, some studies found similar outcomes between PF and PC in concurrent CRT. [14][15][16]19,21 Fang et al. (2017) compared paclitaxel plus cisplatin with TS-1 plus cisplatin and found that TS-1 plus cisplatin could yield statistically better compliance in inoperable patients with EC. 15 In their study, more than 30% (vs. the 95.6% in our cohort) of patients were inoperable. The population differences may be the reason for a non-significant difference in the OS and PFS. Münch et al. (2017 and) also reported no significant difference in the OS and relapsefree survival between PF and PC. 16,19 However, these two studies only enrolled 25 and 31 patients who received PC, which may have been the cause of the nonsignificant impact on the OS and DFS. Horning et al. (2014) and Blom et al. (2014) both concluded that although there were no differences in the OS and DFS, PC had a comparable efficacy but lesser toxicity than PF. 14,21 In our study, PC showed a high survival benefit, consistent with the findings of the CROSS study 12 and other Asian studies. 18,20 Other regimens using non-taxane agents, different schedules, or different intensities of radiotherapy have also been explored. [22][23][24][25][26][27] These reported regimens included the SOX (oxaliplatin-TS-1) regimen, 22 FOLFOX regimen, 28 ECX (epirubicin, cisplatin, and capecitabine), 29 pemetrexed plus cisplatin, 27 neoadjuvant CRT versus chemotherapy with PF regimen before surgery, 23 concurrent CRT and sequential CRT with capecitabine and cisplatin in elderly patients, 24,26 a higher dose or RT (>60 Gy) with the PF regimen, 25 and RT alone in the elderly. 26 However, there is still no evidence to support changing the current practice guidelines to concurrent CRT. The success of immunotherapy as the second-line therapy 30,31 in patients with recurrent or metastatic EC might someday contribute to the development of new, more combination regimens with radiotherapy.

| Impact of complete pathological remission
A study by Haisley et al. (2017) reported that the PF regimen is better than the PC regimen with respect to recurrence-free survival and the OS benefit which could be due to a higher pathological complete remission rate (pCR) (33% vs. 22%). 11 However, pCR was not analyzed in their multivariate regression model, which may have resulted in better survival outcomes in the PF group that had a higher pCR rate. In contrast, the pCR rates in the current study were 7.4% (n = 4/54) and 10% (3/30) in the PF and PC groups, respectively (p = 0.755), resulting in an overall pCR rate of 8.3% (7/84

| Limitations of the study
There were some limitations in the current study. First, its retrospective nature limited the power of the scientific conclusion. However, except for the neoadjuvant PC regimen from the CROSS study in 2012, to the best of our knowledge, no prospective phase III study evaluating the regimens in concurrent CRT has been reported. Further, our study involved a relatively large population of inoperable patients with EC who received definitive CRT. Second, owing to the chart-review design, the data on several toxicity profiles were incomplete, and thus, we could not evaluate the differences in toxicity between the PF and PC regimens. Third, we only compared the PC and PF regimens and did not enroll patients who received other regimens. Our findings need to be validated in prospective randomized phase III trials to establish the optimal chemotherapy regimen during dCRT for EC. Fourth, the patients did not choose their regimens randomly as is done in prospective trials. Thus, the final analysis may have been influenced by a selection bias. However, our comparison of the two groups in Table 1 T A B L E 3 Literature review of two-arm primary/definitive concurrent chemoradiotherapy studies in esophageal cancer

Author
Year Country Design Phase Resectability

RT dose (Gy)
Ref. indicated no apparent differences in the two groups' essential characteristics.
In conclusion, PC yielded superior OS and PFS benefits compared to PF as the chemotherapy regimen during dCRT for patients with advanced ESCC. On the basis of the findings from real-world evidence, we propose that the PC regimen may be a preferable chemotherapy choice for dCRT in patients with advanced inoperable EC.