Abstract
Purpose
The optimal mode of neoadjuvant treatment for esophageal squamous cell carcinoma (ESCC) has not been well characterized. Our study compared neoadjuvant chemotherapy (NCT) with neoadjuvant chemoradiotherapy (NCRT) for patients with ESCC.
Methods
Data from ESCC patients receiving NCRT or NCT combined with esophagectomy between 2010 and 2018 from the National Cancer Center in China were retrospectively collected. Long-term survival, pathological response, and perioperative mortality and morbidity were compared between the NCRT and NCT groups. A Cox proportional hazards model and propensity score matching (PSM) were used to minimize bias due to potential confounding.
Results
Out of 327 eligible patients with ESCC in our study, 90 patients were identified in each group by PSM. The complete pathologic response (pCR) rate in the NCRT group was markedly higher than that in the NCT group (before PSM: 35.1% vs. 6.0%; after PSM: 38.9% vs. 5.6%; both P < 0.001). The rates of 30-day or 90-day mortality were comparable between the two groups, but the NCRT group had a longer postoperative hospital stay (P < 0.001 before PSM and P = 0.012 after PSM) and more postoperative complications (P < 0.001 before PSM and P = 0.014 after PSM), especially, anastomotic leaks (P = 0.001 before PSM and P = 0.013 after PSM). No significant differences in 5-year overall survival (OS) (P = 0.439) or 5-year relapse-free survival (RFS) (P = 0.611) were noted between unmatched groups, but the trend favored NCRT in the propensity score-matched group (77.3% vs. 61.3%; hazard ratio [HR] 1.57; 95% confidence interval [CI] 0.86–2.87; P = 0.141 for OS, and 77.8% vs. 60.5%; HR 1.72; 95% CI 0.95–3.11; P = 0.073 for RFS). Multivariate analysis showed that only ypT and ypN stages were independent predictors of OS before and after PSM (both P < 0.05).
Conclusion
There was no difference in survival between the NCT and NCRT groups, although a trend favored NCRT related to the significantly higher pCR rates. Prospective head-to-head clinical trials to compare these two types of neoadjuvant therapies in ESCC are warranted.
Supplementary Information
The online version contains supplementary material available at 10.1007/s00432-021-03659-7.
Keywords: Esophageal squamous cell carcinoma, Neoadjuvant chemoradiotherapy, Neoadjuvant chemotherapy, Pathology, Prognosis
Introduction
Esophageal cancer remains one of the most common malignancies around the world, ranking eighth in global cancer incidence and sixth in cancer mortality (Bray et al. 2018). Notably, esophageal squamous cell carcinoma (ESCC) is responsible for over 90 percent of esophageal cancers in China (Abnet et al. 2018). Surgery remains the most important element of comprehensive treatment of ESCC (Ajani et al. 2019). Although unprecedented advances have been made in the field of esophageal cancer in recent decades, the prognosis for patients with ESCC remains extremely poor (Blom et al. 2013; Launoy et al. 2017; Gao et al. 2017).
In recent years, multimodal treatments have proven to be an appropriate therapeutic approach for locally advanced, operable ESCC (Watanabe et al. 2020). Two major neoadjuvant approaches—neoadjuvant chemotherapy (NCT) and neoadjuvant chemoradiotherapy (NCRT)—have been adopted. According to the CROSS trial and the NEOCRTEC5010 study, ESCC patients who received NCRT plus esophagectomy experienced a significantly greater benefit compared to those who received surgery alone (Shapiro et al. 2015; Yang et al. 2018), and NCT combined with surgery showed improvements in 5-year overall survival (OS) of 6% and 13% in comparison to surgery alone in the OEO2 or MAGIC studies, respectively (Cunningham et al. 2006; Allum et al. 2009).
To date, there have been three small randomized controlled trials conducted to directly compare NCRT and NCT. Two small, underpowered randomized trials wherein all enrolled patients had esophageal adenocarcinoma found no difference in OS benefit when comparing NCRT and NCT (Burmeister et al. 2011; Stahl et al. 2017). The NeoRES trial showed that NCRT led to higher rates of complete pathologic response (pCR) and R0 resection without significantly affecting prognosis (Döbeln et al. 2019). At least five meta-analyses have compared the efficacy of NCT and NCRT directly, but their conclusions were inconsistent (Sjoquist et al. 2011; Deng et al. 2017; Zhao et al. 2018; Petrelli et al. 2019; Zhou et al. 2020). Among these, some showed that patients with ESCC responded better to NCRT, while NCT may be preferable for patients with esophageal adenocarcinoma. Additionally, a plethora of retrospective studies have compared the therapeutic efficacy of NCRT with NCT, but those have mainly focused on esophageal adenocarcinoma (Petrelli et al. 2019; Markar et al. 2017).
Therefore, there is currently insufficient evidence in patients with ESCC to determine with certainty whether treatment with NCT or NCRT is more beneficial. Herein, we present a real-world study that aimed to investigate practice patterns of neoadjuvant therapy based on data from the National Cancer Center in China and to evaluate the differences in pathologic outcomes as well as short- and long-term survival rates for patients with ESCC who were treated with NCT vs. NCRT.
Patients and methods
Patient sample and data collection
Consecutive patients with ESCC who received NCRT or NCT plus esophagectomy between January 2010 and December 2018 at the National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences were retrospectively analyzed. At our institution, patients who meet the following inclusion criteria are recommended to receive neoadjuvant therapy: (I) aged 18–75 years; (II) histologically confirmed and potentially curable ESCC; (III) clinically staged as T1N + M0 or T2-4aN0-3M0; (IV) normal hematologic, renal, and hepatic function; and (IV) Karnofsky performance score ≥ 70. Patients who did not undergo esophagectomy or who had cervical esophageal cancer were excluded from the analysis.
Clinical characteristics (e.g., age, sex, comorbidities, history of smoking and alcohol, tumor localization), treatment details (e.g., NCRT vs. NCT regimens, surgical procedures, adjuvant therapy), pathological features (e.g., differentiation grade, pathological response, tumor-lymph node-metastasis (TNM) stage, neurovascular invasion) and perioperative outcomes (e.g., hospital stay, 30-day and 90-day morbidity, postoperative complications) were retrospectively collected from the database of patients’ medical records.
This research was carried out after approval by the Ethics Committee of Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College (Beijing, China). Written informed consent was waived because this was a retrospective study.
Pathology
All surgically resected specimens were assessed independently by two senior pathologists using a standard protocol. Diagnoses mainly depended on histomorphology with hematoxylin-and-eosin staining and immunohistochemistry. The pathological TNM stage was based on the latest edition of the American Joint Committee on Cancer/Union for International Cancer Control staging system (8th edition). In line with the criteria proposed by Mandard et al., tumor regression grade (TRG) was classified from grade 1 to 5. Complete pathologic response (pCR) was defined when there were no viable residual tumor cells (grade 1); otherwise, cases were defined as non-complete pathologic response (non-pCR) (grades 2–5) (Mandard et al. 1994).
Treatment protocols
All patients received preoperative chemotherapy or chemo-radiation. The radiation dose was 32.4–50.0 Gy (administered in 1.8 to 2.0 Gy fractions) on Monday through Friday with chemotherapy. Radiotherapy techniques included volumetric intensity modulated arc therapy (VMAT), intensity modulated radiation therapy (IMRT) or three-dimensional conformal radiation (3D-CRT) using 6 MV X-rays. Preoperative chemotherapy mainly consisted of simultaneous treatment with platinum-based drugs and paclitaxel (TP regimen) or platinum-based drugs and fluorouracil (PF regimen). Platinum-based drugs (nedaplatin at 25 mg/m2 or lobaplatin at 20 mg/m2) and paclitaxel (50 mg/m2) were injected intravenously (IV) every week for 3 cycles (days 1, 8, 15, and 22). The PF regimen consisted of fluorouracil at 50 mg/m2 via IV injection on days 1–5 along with platinum-based drugs (cisplatin or nedaplatin at 75 mg/m2, IV) on days 1, 22, and 28. The application of different chemotherapy and radiotherapy protocols is presented in Supplementary Table 1.
Patients usually rested for 4–8 weeks after neoadjuvant therapy before undergoing esophagectomy plus lymphadenectomy if there were no contraindications for surgery. All patients received different types of esophagectomy, including standard lymphadenectomy and open or minimally invasive transthoracic or transhiatal approaches with reconstruction by gastric pull-up. The three most common procedures were the Sweet procedure, the Ivor-Lewis procedure, and the McKeown procedure. Other procedures were also selected, according to the specific conditions of each patient and the experience of the surgeon. All intraoperative and postoperative complications were recorded for analysis of relative risk. Other parameters, including bleeding volume, operation time, postoperative hospital stay, postoperative 30- and 90-day mortality, were also compared between the two groups.
Clinical staging and follow-up
All patients had biopsy-proven resectable ESCC at initial staging. The examinations for clinical staging include computed tomography (CT), endoscopic biopsy, ultrasound endoscopy or 18F-fluorodeoxyglucose positron emission tomography (18F-FDGPET)/CT scanning. In our study, every patient underwent CT and endoscopic biopsy for diagnosis, 350 (72.7%) patients received ultrasound endoscopy for initial staging, and only 69 (20.1%) patients received 18F-FDGPET/CT scanning. The low popularization rate of PET may have contributed to the delayed clinical application of PET/CT and its high cost. As an increasing number of studies have shown that PET may minimize the risk of under-staging, PET/CT is being increasingly recommended for patients with ESCC treated with neoadjuvant therapy at our institution.
Postoperative follow-up was routinely conducted every 3 months during the first 2 years after surgery, then once every 6 months from 3 to 5 years, and then once a year after 5 years. The detection methods included physical examination, measurement of tumor biomarkers, and cervical, thoracic, and abdominal CT. If necessary, upper gastrointestinal radiography, bone emission CT, or brain magnetic resonance imaging were used. Local regional recurrence consisted of recurrence at the esophageal tumor bed, anastomotic site, mediastinal lymph node, supraclavicular, and abdominal lymph node, and distant recurrence was defined as lung, liver, bone, or pleural metastasis or distant lymph node metastasis. Relapse-free survival (RFS) was defined as the time from the first initiation of chemotherapy to the time of relapse or last follow-up, and OS was calculated from the first chemotherapy treatment to the time of death or last follow-up.
Statistical analysis
Categorical variables for patients in the NCRT and NCT groups were compared using Chi-square tests or Fisher’s exact tests, and parametric or nonparametric continuous data were compared using Student’s t tests and the Mann–Whitney U test. Between-group differences in OS and RFS were compared using Kaplan–Meier analyses with log-rank tests. We performed a Cox proportional hazards model to adjust for confounders and to identify independent factors affecting OS. P values < 0.05 were considered statistically significant.
We performed nearest-neighbor 1:1 propensity score matching (PSM) (caliper = 0.02) to minimize the bias of confounding factors on the assessed outcomes between the NCRT and NCT groups. The propensity score of every patient was calculated from a multivariable logistic regression model with the following parameters: age, sex, history of smoking, history of alcohol use, comorbidities, family history of any malignancy, differentiation grade, tumor localization and clinical T stage, and clinical N stage. PSM was carried out using SPSS software. All statistical analyses were performed, and figures were designed using SPSS 24.0 for Windows (IBM Corp, Chicago, IL), GraphPad Prism software (version 6.01), and R version 3.5.3 software (http://www.r-project.org).
Results
Patient characteristics
A total of 327 patients met the inclusion criteria and were reviewed in this study, with 194 patients receiving NCRT and 133 patients receiving NCT. The NCT group was significantly older (P = 0.045) and had more patients with clinical T2/3 stage (P < 0.001), located in the lower third of the esophagus (P < 0.001), than the NCRT group, before matching. Every patient eventually enrolled in our study received at least one cycle of chemotherapy and underwent surgery. There were 163 (84.0%) and 128 (96.2%) patients who received 2–4 cycles of chemotherapy in the NCRT and NCT groups, respectively. Five (2.6%) patients in the NCRT group and two (1.5%) patients in the NCT group did not complete the planned neoadjuvant scheme due to severe side effects or intention to undergo surgery directly. The administration of adjuvant therapy in the NCRT group was significantly less than that in the NCT group (22.7% vs. 49.6%, respectively, P < 0.001). To reduce the confounding bias, we produced a 1:1 propensity score-matched cohort between the NCRT group (n = 90) and NCT group (n = 90). After PSM, the clinical and demographic characteristics of the two groups were well balanced, including age, stage, and tumor location, etc. After matching, there was no difference in cycles of neoadjuvant chemotherapy and neoadjuvant chemotherapy regimens between the two groups (P = 0.058 and P = 0.056). The administration of adjuvant therapy in the NCT group was significantly more than that in the NCT group (47.8% vs. 25.6%, respectively, P = 0.002). The baseline clinicopathological characteristics between NCRT and NCT before and after PSM are shown in Table 1 and Supplementary Table 1.
Table 1.
Comparison of baseline clinicopathologic characteristics between NCRT and NCT before and after propensity score matching
| Clinicopathologic characteristics | Before matching | After matching | ||||
|---|---|---|---|---|---|---|
| NCRT (n = 194) |
NCT (n = 133) |
P value | NCRT (n = 90) |
NCT (n = 90) |
P value | |
| Sexa | ||||||
| Male | 180 (92.8%) | 116 (87.2%) | 0.091 | 80 (88.9%) | 75 (83.3%) | 0.281 |
| Female | 14 (7.2%) | 17 (12.8%) | 10 (11.1%) | 15 (16.7%) | ||
| Age(year)a | ||||||
| < 60 | 118 (60.8%) | 66 (49.6%) | 0.045 | 48 (53.3%) | 41 (45.6%) | 0.297 |
| ≥ 60 | 76 (39.2%) | 67 (50.4%) | 42 (46.7%) | 49 (54.4%) | ||
| BMI(kg/m2)b | 23.18 ± 2.88 | 23.11 ± 3.16 | 0.830 | 23.41 ± 2.87 | 23.05 ± 3.02 | 0.410 |
| History of smokinga | ||||||
| No | 58 (29.9%) | 39 (29.3%) | 0.911 | 25 (27.8%) | 28 (31.1%) | 0.624 |
| Yes | 136 (70.1%) | 94 (70.7%) | 65 (72.2%) | 62 (68.9%) | ||
| History of alcohola | ||||||
| No | 69 (35.6%) | 39 (29.3%) | 0.238 | 32 (35.6%) | 28 (31.1%) | 0.527 |
| Yes | 125 (64.4%) (71.0%) | 94 (70.7%) | 58 (64.4%) (71.0%) | 62 (68.9%) | ||
| Comorbiditya | ||||||
| No | 142 (73.2%) | 97 (72.9%) | 0.958 | 61 (67.8%) | 68 (75.6%) | 0.247 |
| Yes | 52 (26.8%) | 36 (27.1%) | 29 (32.2%) | 22 (24.4%) | ||
| Family historya | ||||||
| No | 138 (71.1%) | 103 (77.4%) | 0.203 | 68 (75.6%) | 69 (76.7%) | 0.861 |
| Yes | 56 (28.9%) | 30 (22.6%) | 22 (24.4%) | 21 (23.3%) | ||
| Differentiationa | ||||||
| Well | 22 (11.3%) | 23 (17.3%) | 0.210 | 7 (7.8%) | 13 (14.4%) | 0.340 |
| Moderately | 83 (42.8%) | 56 (42.1%) | 43 (47.8%) | 46 (51.1%) | ||
| Poorly | 63 (32.5%) | 44 (33.1%) | 30 (33.3%) | 25 (27.8%) | ||
| Unknown | 26 (13.4%) | 10 (7.5%) | 10 (11.1%) | 6 (6.7%) | ||
| Tumor localizationa | ||||||
| Upper | 32 (16.5%) | 8 (6.0%) | < 0.001 | 5 (5.6%) | 8 (8.9%) | 0.536 |
| Middle | 112 (57.7%) | 38 (28.6%) | 43 (47.8%) |
37 (41.1%) (0 () |
||
| Lower | 50 (25.8%) | 87 (65.4%) | 42 (46.7%) | 45 (50.0%) | ||
| Clinical T stagea | ||||||
| T2 | 19 (9.8%) | 20 (15.0%) | < 0.001 | 11 (12.2%) | 9 (10.0%) | 0.846 |
| T3 | 112 (57.7%) | 95 (71.4%) | 65 (72.2%) | 65 (72.2%) | ||
| T4 | 63 (32.5%) | 18 (13.5%) | 14 (15.6%) | 16 (17.8%) | ||
| Clinical N stagea | ||||||
| N0 | 29 (14.9%) | 32 (24.1%) | 0.038 | 20 (22.2%) | 17 (18.9%) | 0.580 |
| N + | 165 (85.1%) | 101 (75.9%) | 70 (77.8%) | 73 (81.1%) | ||
NCRT neoadjuvant chemoradiotherapy, NCT neoadjuvant chemotherapy, BMI body mass index
aVariables used for propensity score matching
bData presented as mean ± standard deviation
Surgery-related procedures
In the original cohort, before matching, there were more patients who underwent McKeown esophagectomy in the NCRT group than in the NCT group (70.1% vs. 54.1%; P = 0.006). The NCRT group had significantly fewer lymph nodes resected (median 20 [interquartile range (IQR): 14–26] vs. 28 [IQR: 19–37]; P < 0.001) and positive lymph nodes (median 0 [IQR: 0–1] vs. 1 [IQR: 0–2.5]; P < 0.001) than the NCT group. The distributions of open esophagectomy and minimally invasive esophagectomy were comparable between the two groups. There were 3 (1.5%) patients in the NCRT group and 4 (3.0%) patients in the NCT group who had planned to receive minimally invasive esophagectomy but converted to open esophagectomy. Notably, the operative time in the NCRT group was significantly longer than that in the NCT group (299.0 ± 95.6 min vs. 276.4 ± 87.9 min, P = 0.031), and there was more bleeding in the NCRT group than in the NCT group (218.4 ± 282.2 ml vs. 162.9 ± 183.6 ml, P = 0.047). However, analysis of surgery-related procedures suggested that there were no significant differences between the propensity score-matched groups after PSM (shown in Table 2).
Table 2.
Comparative analysis of surgical procedures between NCRT and NCT before and after propensity score matching
| Clinicopathologic characteristics | Before matching | After matching | ||||
|---|---|---|---|---|---|---|
| NCRT (n = 194) |
NCT (n = 133) |
P value | NCRT (n = 90) |
NCT (n = 90) |
P value | |
| Surgical types | ||||||
| Sweet | 36 (18.6%) | 44 (33.1%) | 0.006 | 16 (17.8%) | 29 (32.2%) | 0.078 |
| Ivor-lewis | 15 (7.7%) | 15 (11.3%) | 7 (7.8%) | 9 (10.0%) | ||
| McKeown | 136 (70.1%) | 72 (54.1%) | 66 (73.3%) | 51 (56.7%) | ||
| Others | 7 (3.6%) | 2 (1.5%) | 1 (1.1%) | 1 (1.1%) | ||
| Lymphadenectomy (n) | ||||||
| < 15 | 55 (28.4%) | 15 (11.3%) | < 0.001 | 21 (23.3%) | 10 (11.1%) | 0.030 |
| ≥ 15 | 139 (71.6%) | 118 (88.7%) | 69 (76.7%) | 80 (88.9%) | ||
| Surgical approach | ||||||
| MIE | 101 (52.1%) | 76 (57.1%) | 0.365 | 58 (64.4%) | 51 (56.7%) | 0.286 |
| OE | 93 (47.9%) | 57 (42.9%) | 32 (35.6%) | 39 (43.3%) | ||
| Conversion to OE | 3 (1.5%) | 4 (3.0%) | 0.448 | 3 (3.3%) | 3 (3.3%) | 1.000 |
| Operative time(minute)a | 299.00 ± 95.63 | 276.44 ± 87.88 | 0.031 | 295.28 ± 109.62 | 274.18 ± 83.90 | 0.149 |
| Bleeding volume(ml)a | 218.45 ± 282.17 | 162.93 ± 183.60 | 0.047 | 172.78 ± 180.85 | 159.89 ± 194.05 | 0.645 |
NCRT neoadjuvant chemoradiotherapy, NCT neoadjuvant chemotherapy, MIE minimally invasive esophagectomy, OE open esophagectomy
aData presented as mean ± standard deviation
Postoperative course and short-term outcomes
Fifty (25.8%) patients in the NCRT group had postoperative complications, significantly higher than in the NCT group, which had 13 (9.8%) patients in the NCT group (P < 0.001), before propensity score matching. The incidence of anastomotic leakage was markedly higher in the NCRT group than in the NCT group (15.5% vs. 3.8%; P = 0.001). There were no significant differences in the incidences of any other postoperative complications between the two groups prior to matching. Of note, the median lengths of postoperative hospital stay in the NCRT and NCT groups were 13 days and 11 days, respectively, (P < 0.001) and significantly related to the high incidence of anastomotic leaks in the NCRT group (P = 0.001). Postoperative 30- and 90-day mortality in the NCRT group did not differ significantly from those in the NCT group (1.0% vs. 0.8% for 30-day mortality, respectively, P = 1.000; 3.1% vs. 0.8% for 90-day mortality, respectively, P = 0.248), before propensity score matching. The results for postoperative course and short-term outcomes in the propensity score-matched groups were consistent with those of the original, unmatched cohort (shown in Table 3).
Table 3.
Comparative analysis of postoperative course and short-term outcomes between NCRT and NCT before and after propensity score matching
| Clinicopathologic characteristics | Before matching | After matching | ||||
|---|---|---|---|---|---|---|
| NCRT (n = 194) |
NCT (n = 133) |
P value | NCRT (n = 90) |
NCT (n = 90) |
P value | |
| Postoperative complications | 50 (25.8%) | 13 (9.8%) | < 0.001 | 24 (26.7%) | 11 (12.2%) | 0.014 |
| Anastomotic leakage | 30 (15.5%) | 5 (3.8%) | 0.001 | 14 (15.6%) | 4 (4.4%) | 0.013 |
| Pulmonary complications | 9 (4.6%) | 4 (3.0%) | 0.458 | 5 (5.6%) | 3 (3.3%) | 0.720 |
| Cardiac complications | 6 (3.1%) | 2 (1.5%) | 0.480 | 2 (2.2%) | 2 (2.2%) | 1.000 |
| Recurrent nerve paresis | 2 (1.0%) | 1 (0.8%) | 1.000 | 2 (2.2%) | 0 | 0.497 |
| Wound infection | 5 (2.6%) | 2 (1.5%) | 0.705 | 3 (3.3%) | 2 (2.2%) | 1.000 |
| Others | 6 (3.1%) | 3 (2.3%) | 0.743 | 2 (2.2%) | 3 (3.3%) | 1.000 |
| Postoperative hospital stay (days)a | 13 (0–98) | 11 (6–59) | < 0.001 | 12 (0–82) | 11 (6–59) | 0.012 |
| 30-day mortality | 2 (1.0%) | 1 (0.8%) | 1.000 | 1 (1.1%) | 1 (1.1%) | 1.000 |
| 90-day mortality | 6 (3.1%) | 1 (0.8%) | 0.248 | 1 (1.1%) | 1 (1.1%) | 1.000 |
NCRT neoadjuvant chemoradiotherapy, NCT neoadjuvant chemotharapy
aData presented as median (range)
Tumor pathology
In the original cohort, before matching, patients who received NCRT had a significantly higher pCR rate than those who received NCT (35.1% vs. 6.0%, P < 0.001). Correspondingly, a good tumor regression grade was observed more frequently in the NCRT group than in the NCT group (34.5% vs. 6.0% for TRG1, 34.0% vs. 7.5% for TRG2, respectively; P < 0.001). This is reflected in the fact that the NCRT group had a significantly increased incidence of ypT0/Tis (34.5% vs. 5.3%; P < 0.001) and ypN0 (73.7% vs. 48.1%; P < 0.001) compared with the NCT group. Moreover, patients who received NCRT had a significant reduction in the incidence of lymphovascular invasion (17.0% vs. 33.1%, P < 0.001) and neural invasion (24.7% vs. 42.1%, P = 0.001) relative to NCT. The rate of R0 resection in the NCRT group was comparable to that in the NCT group (99.5% vs. 99.2%, P = 1.000). The pathological outcomes of the propensity score-matched groups were in line with those of the original cohort before matching (shown in Table 4).
Table 4.
Comparative analysis of pathological outcomes between NCRT and NCT before and after propensity score matching
| Clinicopathologic characteristics | Before matching | After matching | ||||
|---|---|---|---|---|---|---|
| NCRT (n = 194) |
NCT (n = 133) |
P value | NCRT (n = 90) |
NCT (n = 90) |
P value | |
| Pathological response | ||||||
| pCR | 68 (35.1%) | 8 (6.0%) | < 0.001 | 35 (38.9%) | 5 (5.6%) | < 0.001 |
| Non-pCR | 126 (64.9%) | 125 (94.0%) | 55 (61.1%) | 85 (94.4%) | ||
| Tumor regression grade | ||||||
| 1 | 67 (34.5%) | 8 (6.0%) | < 0.001 | 35 (38.9%) | 5 (5.6%) | < 0.001 |
| 2 | 66 (34.0%) | 10 (7.5%) | 32 (35.6%) | 4 (4.4%) | ||
| 3 | 29 (14.9%) | 33 (24.8%) | 9 (10.0%) | 23 (25.6%) | ||
| 4 | 28 (14.4%) | 74 (55.6%) | 13 (14.4%) | 52 (57.8%) | ||
| 5 | 4 (2.1%) | 8 (6.0%) | 1 (1.1%) | 6 (6.7%) | ||
| ypT stage | ||||||
| T0/Tis | 67 (34.5%) | 7 (5.3%) | < 0.001 | 35 (38.9%) | 4 (4.4%) | < 0.001 |
| T1 | 33 (17.0%) | 20 (15.0%) | 14 (15.6%) | 10 (11.1%) | ||
| T2 | 33 (17.0%) | 29 (21.8%) | 15 (16.7%) | 18 (20.0%) | ||
| T3 | 56 (28.9%) | 66 (49.6%) | 24 (26.7%) | 48 (53.3%) | ||
| T4 | 5 (2.6%) | 11 (8.3%) | 2 (2.2%) | 10 (11.1%) | ||
| ypN stage | ||||||
| N0 | 143 (73.7%) | 64 (48.1%) | < 0.001 | 67 (74.4%) | 43 (47.8%) | 0.002 |
| N1 | 37 (19.1%) | 39 (29.3%) | 17 (18.9%) | 29 (32.2%) | ||
| N2 | 9 (4.6%) | 24 (18.0%) | 4 (4.4%) | 12 (13.3%) | ||
| N3 | 5 (2.6%) | 6 (4.5%) | 2 (2.2%) | 6 (6.7%) | ||
| R0 resection | ||||||
| Yes | 193 (99.5%) | 132 (99.2%) | 1.000 | 89 (98.9%) | 89 (98.9%) | 1.000 |
| No | 1 (0.5%) | 1 (0.8%) | 1 (1.1%) | 1 (1.1%) | ||
| LVSI | ||||||
| No | 161 (83.0%) | 89 (66.9%) | 0.001 | 77 (85.6%) | 56 (62.2%) | < 0.001 |
| Yes | 33 (17.0%) | 44 (33.1%) | 13 (14.4%) | 34 (37.8%) | ||
| Neural invasion | ||||||
| No | 146 (75.3%) | 77 (57.9%) | 0.001 | 69 (76.7%) | 45 (50.0%) | < 0.001 |
| Yes | 48 (24.7%) | 56 (42.1%) | 21 (23.3%) | 45 (50.0%) | ||
NCRT neoadjuvant chemoradiotherapy, NCT neoadjuvant chemotharapy, pCR complete pathologic response, LVSI lymphovascular invasion
Recurrence patterns and causes of death
In total, 75 (38.7%) patients in the NCRT group and 44 (33.1%) patients in the NCT group experienced recurrence (P = 0.303). In the NCRT group there were instances of loco-regional (n = 43 [22.2%]) and distant (n = 32 [16.5%]) recurrences, while in the NCT group, 17 (12.8%) patients exhibited loco-regional recurrence and 27 (20.3%) patients exhibited distant recurrence. The most common recurrence site was the mediastinal lymph nodes. In terms of the causes of death, 55 (28.4%) patients in the NCRT group and 33 (24.8%) patients in the NCT group died from tumor progression, respectively. Other causes of death included other non-tumorous diseases (NCRT n = 9 [4.6%]; NCT n = 3 [2.3%]) and treatment-related complications (NCRT n = 10 [5.2%]; NCT n = 4 [3.0%]). After PSM, there were no differences in the recurrence rate between the NCRT and NCT groups (20.0% vs. 31.1%, P = 0.087). The number of non-cancer-related deaths in the NCRT group was more than that in the NCT group; however, the difference was not statistically significant (P = 0.137 before PSM; P = 0.578 after PSM).
Overall survival and relapse-free survival
In the original, unmatched cohort, the median follow-up time was 48.20 months (range 0–125.0) in the NCRT group and 31.23 months (range 1–108.4) in the NCT group. The median OS time was 96.00 months after NCRT versus 89.97 months after NCT. The median RFS time was 88.00 months for the NCRT group and was not reached in the NCT group. There were no significant differences in 5-year OS (59.5% vs. 59.6%; HR 0.86; 95% CI 0.58–1.26; log-rank test P = 0.439) or 5-year RFS (60.1% vs. 58.8%; HR 0.91; 95% CI 0.63–1.32; log-rank test P = 0.611) between unmatched groups (Fig. 1a, b). Overall, patients with pCR had significantly better OS (HR 0.38; 95% CI 0.21–0.66; log-rank test P < 0.001) and RFS (HR 0.41; 95% CI 0.24–0.69; log-rank test P < 0.001) than those with residual tumors (Fig. 2a, b).
Fig. 1.
Unmatched Kaplan–Meier survival analysis of OS a and RFS b between the NCT and NCRT groups. Propensity-matched Kaplan–Meier survival analysis of OS c and RFS d between the NCT and NCRT groups. NCT neoadjuvant chemotherapy, NCRT neoadjuvant chemoradiotherapy, OS overall survival, RFS relapse-free survival
Fig. 2.
pCR was significantly associated with better OS a and RFS b in the original, unmatched cohort in the Kaplan–Meier analysis. pCR was significantly associated with better OS c and RFS d in the propensity-matched cohort in the Kaplan–Meier analysis. In the original, unmatched cohort, 76 patients achieved pCR (68 in the NCRT group and 8 in the NCT group). In the propensity score-matched cohort, 40 patients achieved pCR (35 in the NCRT group and 5 in the NCT group). pCR complete pathologic response, OS overall survival, RFS relapse-free survival
After matching, the median duration of follow-up was 34.50 months (range 0–125.0) in the NCRT group and 31.47 months (range 1–108.4) in the NCT group. The median OS and RFS times in the propensity score-matched groups were not reached. The log-rank test suggested a trend toward a survival benefit in 5-year OS (77.3% vs. 61.3%; HR 1.57; 95% CI 0.86–2.87; P = 0.141) and 5-year RFS (77.8% vs. 60.5%; HR 1.72; 95% CI 0.95–3.11; P = 0.073) (Fig. 1c, d). However, there were no significant differences. Patients with pCR had significantly better OS (HR 0.06; 95% CI 0.01–0.47; log-rank test P < 0.001) and RFS (HR 0.12; 95% CI 0.03–0.51; log-rank test P < 0.001) than those with residual tumors (Fig. 2c, d). Independent factors affecting OS were identified using a Cox proportional hazards model (Table 5). Multivariate analysis showed that advanced ypT stage (HR 1.90; 95% CI 1.33–2.71; P < 0.001) and ypN stage (HR 1.51; 95% CI 1.06–2.15; P = 0.024) were independent variables significantly associated with poor OS.
Table 5.
Multivariate Cox regression model analysis of clinicopathologic factors for OS before and after propensity score matching
| Clinicopathologic characteristics | Before matching | After matching | ||||
|---|---|---|---|---|---|---|
| HR | 95% CI | P value | HR | 95% CI | P value | |
| Age (< 60 vs ≥ 60) | 0.693 | 0.465–1.033 | 0.072 | 0.715 | 0.389–1.314 | 0.280 |
| Sex (M vs F) | 0.807 | 0.318–2.050 | 0.652 | 0.786 | 0.184–3.365 | 0.745 |
| Smoking (Yes vs No) | 0.997 | 0.609–1.631 | 0.990 | 0.882 | 0.401–1.940 | 0.756 |
| Alcohol (Yes vs No) | 1.232 | 0.773–1.963 | 0.380 | 1.596 | 0.670–3.799 | 0.291 |
| Localization (upper vs middle, lower) | 0.882 | 0.647–1.202 | 0.428 | 0.725 | 0.426–1.233 | 0.235 |
| Differentiation (well vs moderately, poorly) | 1.100 | 0.869–1.391 | 0.429 | 1.006 | 0.661–1.530 | 0.979 |
| Lymphadenectomy (< 15 vs ≥ 15) | 0.928 | 0.583–1.477 | 0.754 | 0.779 | 0.335–1.811 | 0.561 |
| Neoadjuvant therapy (NCRT vs NCT) | 0.640 | 0.398–1.030 | 0.066 | 0.668 | 0.338–1.321 | 0.246 |
| ypT stage (T0 vs T1,T2,T3,T4) | 1.599 | 1.299–1.870 | < 0.001 | 1.898 | 1.330–2.709 | < 0.001 |
| ypN stage (N0 vs N1,N2,N3) | 1.271 | 1.021–1.583 | 0.032 | 1.508 | 1.055–2.154 | 0.024 |
| clinical T stage (T2 vs T3,T4) | 1.332 | 0.952–1.863 | 0.094 | 0.917 | 0.496–1.694 | 0.782 |
| clinical N stage (N0 vs N +) | 1.937 | 0.984–3.813 | 0.056 | 1.123 | 0.422–2.898 | 0.816 |
OS overall survival, HR hazards ratio, CI confidence interval, M male, F female, NCRT neoadjuvant chemoradiotherapy, NCT neoadjuvant chemotharapy;
Discussion
The selection of the most appropriate preoperative therapy for patients with esophageal cancer has long been a topic of controversy, and there have been extremely limited studies providing comparisons of outcomes of patients with ESCC who were treated with NCRT or NCT. In this work, we compared the clinicopathological characteristics and prognoses of patients with ESCC who were treated with NCT and NCRT using data from the National Cancer Center in China. Our results demonstrated no significant differences in RFS or OS between the NCRT and NCT groups either before or after propensity matching analysis. However, patients treated with NCRT obtained significant pathological advantages, including tumor downstaging and a higher pCR rate. Furthermore, the duration of postoperative hospital stay in the NCRT group was markedly longer than that in the NCT group, because the likelihood of postoperative complications, especially of anastomotic leakage, was higher in the NCRT group. This is the first and largest available study to directly compare NCRT with NCT with a focus on patients with ESCC. To minimize the potential confounding bias caused by the non-randomly selected data, we compared the two groups before and after using a propensity score-matched analysis.
To date, at least three clinical trials have compared NCT with NCRT, and their conclusions were largely similar. A randomized trial conducted in Australia directly compared NCT with NCRT (Burmeister et al. 2011). Although the rates of R0 resection and pathological response were much higher in the NCRT group, the median OS was not significantly increased by NCRT. A multicenter phase III trial (POET) included 126 patients who were randomly enrolled into the NCT group or NCRT group after induction chemotherapy (Stahl et al. 2017). That study found that the pCR rate in the NCRT group after induction chemotherapy was obviously higher (16.0% vs 2.0%), and the 5-year OS rate in the NCRT group also trended toward superiority compared with the NCT group, but the difference was not statistically significant (39.5% vs. 24.4%, P = 0.055). Another multicenter phase III trial (NeoRes) demonstrated that patients with NCRT had a significantly higher rate of pCR and complete resection and a lower lymph node positive rate. However, these differences did not result in a significant improvement in the 5-year OS rate (42.2% vs 39.6%, P = 0.60) or RFS rate (38.9% vs 33.0%, P = 0.82) (Döbeln et al. 2019). Notably, the patients included in these clinical trials were either all or mostly patients with adenocarcinoma, and two of the three clinical trials were terminated prematurely because the test efficiency was insufficient to show a survival advantage. Several meta-analyses also included randomized clinical trials that compared NCRT with NCT. The subgroup analysis of ESCC showed that patients with NCRT had a higher 5-year OS rate, R0 resection rate and pCR rate (Zhao et al. 2018; Zhou et al. 2020). However, the results were limited by the high heterogeneity between studies or the small sample sizes of subgroups with ESCC. Hence, studies comparing NCRT head-to-head against NCT in ESCC are extremely limited. A randomized clinical trial to compare these two predominant paradigms of preoperative therapy in ESCC is warranted.
In our retrospective study, NCRT significantly increased the pCR rate relative to NCT, but these improvements in tumor control did not translate into improved survival. Previous published studies have demonstrated that patients who experience complete pathologic recovery after receiving neoadjuvant therapy had better OS and RFS compared with those patients who did not achieve pCR (Berger et al. 2005). In parallel, we also found that patients with pCR had an improved OS and RFS relative to those with non-pCR. However, in the subgroup analysis on patients who did not achieve pCR, the patients who received NCRT had significantly worse OS (P = 0.047) but not RFS (P = 0.095) in the non-matched cohort, and there was no difference in OS or RFS between the NCRT and NCT groups in the propensity score-matched cohort (shown in Supplementary Fig. 1A–D). This raises interesting questions regarding why the local tumor response did not translate into a survival benefit in our study. There may be several explanations for this phenomenon. First, the high number of non-cancer-related deaths in the short term in the NCRT group may have hindered our ability to detect a small potential survival advantage. Our results showed that the number of non-cancer-related deaths in the NCRT group were more than the NCT group, although the difference was not statistically significant, which may offset the survival advantage of the NCRT group to some extent. These results were consistent with the results in the POET and NeoRes trials, in which an increased trend in perioperative mortality in the NCRT group, as well as a significant increase in non-cancer-related deaths in the first year was also observed (Stahl et al. 2009; Klevebro et al. 2016). Second, some scholars hypothesized that it is possible that the radiotherapy may be poorly effective at controlling occult systemic metastasis in ESCC, although it increased the local pathological response (Daxuan et al. 2017; Nakashima et al. 2018). Another possibility is that the frequency of adjuvant therapy use in the NCRT group was significantly less than that in the NCT group, especially for the patients who did not achieve pCR. Leng et al. found that pCR did not guarantee a cure and indicated that patients with pCR need receive adjuvant therapy (Samson et al. 2018; Leng et al. 2019). Therefore, further prospective analyses to explore the long-term survival of patients who receive these two types of treatment strategies should consider the value of adjuvant therapy.
Currently, most clinical guidelines recommend neoadjuvant therapy combined with esophagectomy for locally advanced esophageal cancer (Ajani et al. 2019). Neoadjuvant therapy can reduce the difficulty of surgery by reducing the tumor stage, in addition to improving the prognosis of patients. However, NCRT can also have adverse effects on surgery because radiotherapy leads to increased adhesion and friability of tissue. Our findings showed that the number of patients who had less than 15 lymph nodes removed was higher in the NCRT group than that in the NCT group, which may be contributed to the increased complexity of lymphadenectomy by preoperative radiotherapy. Although the number of lymphadenectomies proved to be a negative prognostic factor, the number of lymph nodes dissected (< 15 or ≥ 15) was not associated with survival both in either the original cohort (OS: P = 0.417; RFS: P = 0.383, Supplementary Fig. 2A, B) or propensity score-matched cohort (OS: P = 0.977; RFS: P = 0.554, Supplementary Fig. 2C, D). Moreover, those treated with NCRT had a prolonged operative time and more bleeding than those treated with NCT, supporting the potential side effects caused by radiation therapy. Previous published works have shown that the dose and type of radiotherapy are directly related to the complexity of surgery and postoperative complications (Zhao et al. 2018; Vande Walle et al. 2012; Wang et al. 2013). An ongoing randomized clinical trial (NCT03381651) is evaluating different radiation doses of NCRT in patients with resectable ESCC. The findings of this trial will likely be helpful for determining the ideal radiation profile of NCRT. It is compelling to note that, in our study, the occurrence of postoperative complications in the NCRT group, especially of anastomotic leakage, was markedly more common than that in the NCT group. Moreover, 73.3% patients with NCRT underwent McKeown procedures versus 56.7% patients with NCT in the propensity score-matched cohort. To explore whether the incidence of anastomotic leakage was related to NCRT, we further performed a subgroup analysis between tumor localization and surgical approaches with anastomotic leakage in the propensity score-matched cohort. Our results did not observe that the anastomotic leak rate was significantly associated with tumor localization (P = 0.072) or with different surgical procedures (P = 0.650). Similar to our results, Markar et al. also found that anastomotic leakage was more common in the NCRT group (Markar et al. 2017). Chen et al. reported that preoperative radiochemotherapy was an independent factor predicting anastomotic leakage in esophageal cancer patients (Chen et al. 2015; Roh et al. 2019). A pooled analysis including six trials involving 408 squamous cell carcinomas also showed that the incidence of anastomotic leakage in the NCRT group was significantly higher than that in the NCT group (Zhao et al. 2018).
Furthermore, we found that there were more patients with advanced clinical TNM stage and more patients with tumors located in the upper esophagus in the NCRT group than in the NCT group before matching. Considering that there were no differences in long-term outcomes between NCRT and NCT, it seems more appropriate to choose a personally tailored therapeutic strategy of neoadjuvant therapy for patients. For patients with tumors located in the upper esophagus and advanced N stage, we prefer NCRT plus en bloc three-field McKeown esophagectomy. For patients with poor physical condition and treatment tolerance, NCT followed by esophagectomy should be considered first. Given that NCRT may increase the difficulty of surgery and postoperative complications, an increasing number of surgeons in our institution prefer to choose NCT.
In recent years, immunotherapy has facilitated unprecedented advances in the treatment of cancers. Given the promising efficacy and acceptable side effects of immunotherapy in ESCC, combining neoadjuvant immunotherapy with NCRT or NCT is an effective strategy to improve the outcomes of locally advanced ESCC (Huang et al. 2020; Ende et al. 2021; Sihag et al. 2020). Our results may provide guidance for the selection of NCRT combined with immunotherapy or NCT combined with immunotherapy when designing clinical trials. More research is warranted to develop noninvasive methods to identify radiosensitive individuals and to determine who may be at the greatest risk for side effects from radiation therapy.
There were several potential limitations to our study. First, this was a retrospective analysis that spanned a relatively long period, and the changes in cytotoxic drugs and variations in radiation techniques would potentially lead to confounding bias. Therefore, we tried our best to improve the comparability of the two groups using a PSM method. Second, the chemotherapy regimens in the NCRT and NCT groups were not exactly the same, and there were differences in the frequency of adjuvant therapy in the two groups, which may influence outcomes in locally advanced ESCC patients. Furthermore, there was limited information on dropout percentage and intent-to-treat results because only patients who underwent esophagectomy were enrolled in our study. Therefore, it is necessary to further compare the two types of preoperative treatment in patients in the future prospective trials.
Conclusion
In this work, we performed a head-to-head comparison of the current leading two paradigms of neoadjuvant therapy in ESCC patients—NCRT and NCT—for the first time. Notably, there were no statistically significant differences observed in either RFS or OS between the NCRT and NCT groups, despite a trend toward favoring NCRT. The postoperative complications and length of hospital stay in the NCT group were significantly improved relative to those in the NCRT group. Given the extra cost and potential side effects of NCRT, the significance of preoperative radiotherapy in the multimodal treatment of ESCC should be further evaluated in prospective studies.
Supplementary Information
Below is the link to the electronic supplementary material.
Authors’ contributions
JH supervised the project, design, interpretation, manuscript revision, and final approval of the version to be submitted. GCZ, CQZ, NS and QX were involved in concept, data acquisition, analysis and interpretation. GCZ and CQZ wrote the first draft, and revised it critically in light of comments from other authors. GCZ, ZHZ and YJL prepared all the figures and tables. LYX and ZYY reviewed all specimens enrolled in the study. LLF, JWM, YSG and FWT were involved in data acquisition and provided material support. SGG and QX were involved in the manuscript revision and discussion. All authors reviewed and approved the final manuscript.
Funding
This work was supported by funds from the National Key Research and Development Program of China [2016YFC1303200].
Data availability
The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.
Declarations
Conflict of interest
The authors declare no conflict of interest.
Ethical approval
The research was approved by the Ethics Committee of Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College (Beijing, China). Written informed consent was waived because this was a retrospective study.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- Abnet CC, Arnold M, Wei WQ (2018) Epidemiology of esophageal squamous cell carcinoma. Gastroenterology 154:360–373 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ajani JA, D’Amico TA, Bentrem DJ et al (2019) Esophageal and esophagogastric junction cancers, version 2.2019, NCCN clinical practice guidelines in oncology. J Natl Compr Canc Netw 17:855–883 [DOI] [PubMed] [Google Scholar]
- Allum WH, Stenning SP, Bancewicz J et al (2009) Long-term results of a randomized trial of surgery with or without preoperative chemotherapy in esophageal cancer. J Clin Oncol 27:5062–5067 [DOI] [PubMed] [Google Scholar]
- Berger AC, Farma J, Scott WJ et al (2005) Complete response to neoadjuvant chemoradiotherapy in esophageal carcinoma is associated with significantly improved survival. J Clin Oncol 23:4330–4337 [DOI] [PubMed] [Google Scholar]
- Blom RL, Lagarde SM, van Oudenaarde K et al (2013) Survival after recurrent esophageal carcinoma has not improved over the past 18 Years. Ann Surg Oncol 20:2693–2698 [DOI] [PubMed] [Google Scholar]
- Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A (2018) Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 68:394–424 [DOI] [PubMed] [Google Scholar]
- Burmeister BH, Thomas JM, Burmeister EA et al (2011) Is concurrent radiation therapy required in patients receiving preoperative chemotherapy for adenocarcinoma of the oesophagus? A randomised phase II trial. Eur J Cancer 47:354–360 [DOI] [PubMed] [Google Scholar]
- Chen C, Yu Z, Jin Q et al (2015) Clinical features and risk factors of anastomotic leakage after radical esophagectomy. Zhonghua Wai Ke Za Zhi 53:518–521 [PubMed] [Google Scholar]
- Cunningham D, Allum WH, Stenning SP et al (2006) Perioperative chemotherapy versus surgery alone for resectable gastroesophageal cancer. N Engl J Med 355:11–20 [DOI] [PubMed] [Google Scholar]
- Daxuan H, Xue L, Yuanyuan Y et al (2017) Neoadjuvant chemoradiotherapy versus chemotherapy and surgery for patients with locally advanced esophageal squamous cell carcinoma. Transl Cancer Res 6:346–353 [Google Scholar]
- Deng HY, Wang WP, Wang YC et al (2017) Neoadjuvant chemoradiotherapy or chemotherapy? A comprehensive systematic review and meta-analysis of the options for neoadjuvant therapy for treating oesophageal cancer. Eur J Cardiothorac Surg 51:421–431 [DOI] [PubMed] [Google Scholar]
- Gao X, Wang Z, Kong C et al (2017) Trends of esophageal cancer mortality in rural china from 1989 to 2013: an age-period-cohort analysis. Int J Environ Res Public Health 14:218 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Huang J, Xu J, Chen Y et al (2020) Camrelizumab versus investigator’s choice of chemotherapy as second-line therapy for advanced or metastatic oesophageal squamous cell carcinoma (ESCORT): a multicentre, randomised, open-label, phase 3 study. Lancet Oncol 21:832–842 [DOI] [PubMed] [Google Scholar]
- Klevebro F, Alexandersson von Döbeln G, Wang N et al (2016) A randomized clinical trial of neoadjuvant chemotherapy versus neoadjuvant chemoradiotherapy for cancer of the oesophagus or gastro-oesophageal junction. Ann Oncol 27:660–667 [DOI] [PubMed] [Google Scholar]
- Launoy G, Bossard N, Castro C et al (2017) Trends in net survival from esophageal cancer in six European Latin countries: results from the SUDCAN population-based study. Eur J Cancer Prev 26:S24–S31 [DOI] [PubMed] [Google Scholar]
- Leng X, He W, Yang H et al (2019) Prognostic impact of postoperative lymph node metastases after neoadjuvant chemoradiotherapy for locally advanced squamous cell carcinoma of esophagus: from the results of neocrtec5010, a randomized multicenter study. Ann Surg. 10.1097/SLA.0000000000003727 [DOI] [PubMed] [Google Scholar]
- Mandard AM, Dalibard F, Mandard JC et al (1994) Pathologic assessment of tumor regression after preoperative chemoradiotherapy of esophageal carcinoma. Clinicopathologic Correlations Cancer 73:2680–2686 [DOI] [PubMed] [Google Scholar]
- Markar SR, Noordman BJ, Mackenzie H et al (2017) Multimodality treatment for esophageal adenocarcinoma: multi-center propensity-score matched study. Ann Oncol 28:519–527 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nakashima Y, Saeki H, Hu Q et al (2018) Neoadjuvant chemotherapy versus chemoradiotherapy for patients with esophageal squamous cell carcinoma. Anticancer Res 38:6809–6814 [DOI] [PubMed] [Google Scholar]
- Petrelli F, Ghidini M, Barni S et al (2019) Neoadjuvant chemoradiotherapy or chemotherapy for gastroesophageal junction adenocarcinoma: a systematic review and meta-analysis. Gastric Cancer 22:245–254 [DOI] [PubMed] [Google Scholar]
- Roh S, Iannettoni MD, Keech J et al (2019) Timing of esophagectomy after neoadjuvant chemoradiation therapy affects the incidence of anastomotic leaks. Korean J Thorac Cardiovasc Surg 52:1–8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Samson P, Puri V, Lockhart AC et al (2018) Adjuvant chemotherapy for patients with pathologic node-positive esophageal cancer after induction chemotherapy is associated with improved survival. J Thorac Cardiovasc Surg 156:1725–1735 [DOI] [PubMed] [Google Scholar]
- Shapiro J, van Lanschot JJB, Hulshof MCCM et al (2015) Neoadjuvant chemoradiotherapy plus surgery versus surgery alone for oesophageal or junctional cancer (CROSS): long-term results of a randomised controlled trial. Lancet Oncol 16:1090–1098 [DOI] [PubMed] [Google Scholar]
- Sihag S, Ku GY, Tan KS et al (2020) (2020) Safety and feasibility of esophagectomy following combined immunotherapy and chemoradiotherapy for esophageal cancer. J Thorac Cardiovasc Surg. S0022–5223(20):33192–33195. 10.1016/j.jtcvs.2020.11.106 (Online ahead of print) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sjoquist KM, Burmeister BH, Smithers BM et al (2011) Survival after neoadjuvant chemotherapy or chemoradiotherapy for resectable oesophageal carcinoma: an updated meta-analysis. Lancet Oncol 12:681–692 [DOI] [PubMed] [Google Scholar]
- Stahl M, Walz MK, Stuschke M et al (2009) Phase III comparison of preoperative chemotherapy compared with chemoradiotherapy in patients with locally advanced adenocarcinoma of the esophagogastric junction. J Clin Oncol 27:851–856 [DOI] [PubMed] [Google Scholar]
- Stahl M, Walz MK, Riera-Knorrenschild J et al (2017) Preoperative chemotherapy versus chemoradiotherapy in locally advanced adenocarcinomas of the oesophagogastric junction (POET): long-term results of a controlled randomised trial. Eur J Cancer 81:183–190 [DOI] [PubMed] [Google Scholar]
- van den Ende T, Clercq N, van Berge Henegouwen MI et al (2021) Neoadjuvant chemoradiotherapy combined with atezolizumab for resectable esophageal adenocarcinoma: A Single Arm Phase II Feasibility Trial (PERFECT). Clin Cancer Res. 10.1158/1078-0432.CCR-20-4443 (Online ahead of print) [DOI] [PubMed] [Google Scholar]
- Vande Walle C, Ceelen WP, Boterberg T et al (2012) Anastomotic complications after Ivor Lewis esophagectomy in patients treated with neoadjuvant chemoradiation are related to radiation dose to the gastric fundus. Int J Radiat Oncol Biol Phys 82:e513–e519 [DOI] [PubMed] [Google Scholar]
- von Döbeln GA, Klevebro F, Jacobsen AB et al (2019) Neoadjuvant chemotherapy versus neoadjuvant chemoradiotherapy for cancer of the esophagus or gastroesophageal junction: long-term results of a randomized clinical trial. Dis Esophagus 32:1–11 [DOI] [PubMed] [Google Scholar]
- Wang J, Wei C, Tucker SL et al (2013) Predictors of postoperative complications after trimodality therapy for esophageal cancer. Int J Radiat Oncol Biol Phys 86:885–891 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Watanabe M, Otake R, Kozuki R et al (2020) Recent progress in multidisciplinary treatment for patients with esophageal cancer. Surg Today 50:12–20 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yang H, Liu H, Chen Y et al (2018) Neoadjuvant Chemoradiotherapy Followed by Surgery Versus Surgery Alone for Locally Advanced Squamous Cell Carcinoma of the Esophagus (NEOCRTEC5010): A Phase III Multicenter, Randomized, Open-Label Clinical Trial. J Clin Oncol 36:2796–2803 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhao X, Ren Y, Hu Y et al (2018) Neoadjuvant chemotherapy versus neoadjuvant chemoradiotherapy for cancer of the esophagus or the gastroesophageal junction: A meta-analysis based on clinical trials. PLoS ONE 13:e0202185 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhou HY, Zheng SP, Li AL et al (2020) Clinical evidence for association of neoadjuvant chemotherapy or chemoradiotherapy with efficacy and safety in patients with resectable esophageal carcinoma (NewEC study). EClinicalMedicine 24:100422 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.