Variables Associated to Pathologic Complete Response, Overall Survival and Disease-Free Survival in the Neoadjuvant Setting for Esophageal Cancer: A Retrospective Cohort Analysis

Adenocarcinoma and squamous cell carcinoma (SCC) comprise the main histologic types of esophageal cancer. Both histologic types usually evidence poor survival rates. About 80% of patients will not survive longer than 5 years.¹ Most of the patients are diagnosed at advanced clinical stages.^2,3

Concerning advanced stage neoplasms, neoplasm circumferential involvement or even complete esophageal lumen obstruction are common, leading to intense dysphagia, weight loss, and malnutrition.

In locally advanced tumors, neoadjuvant chemoradiotherapy facilitates tumor downstaging, increasing probability for curative intent surgery, and avoiding locoregional recurrence. Tumor shrinking after neoadjuvant therapy may relieve dysphagia and improve patients' nutritional status.^4,5

Neoadjuvant tumor response varies among patients, and the patient's response to chemoradiotherapy is difficult to evaluate and predict before esophagectomy. Consequently, variables that can predict tumor response to chemoradiotherapy and long-term survival rates after esophagectomy are required for esophageal cancer management.

Objectives

This study aimed to evaluate possible prognostic variables in neoadjuvant therapy that can predict pathologic complete response, overall survival, or disease-free survival.

Methods

A retrospective cohort of patients submitted to neoadjuvant therapy for esophageal carcinoma was performed in a single institution, between 2009 and 2017.

Clinical, demographic, endoscopic, and histopathologic variables were extracted and assessed. Improved nutritional status and dysphagia relief while on neoadjuvant therapy was also recorded. The outcomes evaluated were pathologic complete response (pCR), overall survival (OS), and disease-free survival (DFS).

Statistical Analysis

Quantitative variables were assessed by means and standard deviation. Qualitative variables association to pCR were assessed by Student's t, Mann-Whitney, and Kolmogorov-Smirnov tests. Qualitative variables were assessed by frequency and percentage. Quantitative variables' association with pCR was assessed by Chi-square and likelihood-ratio tests.⁶ Odds ratios (OR) using bivariate analysis and simple and multiple logistic regression were applied with 95% confidence interval (95% CI).⁷

Survival analysis was performed by Kaplan-Meier curves and log-rank test. Hazard ratio (HR) and bivariate Cox regression were applied with 95% CI.⁸

The significance level adopted was 0.05. IBM-SPSS 20.0 (Chicago, Illinois) software was used for statistical analysis.

Results

Baseline patient's characteristics

This study covered 140 esophageal cancer patients submitted to neoadjuvant therapy followed by esophagectomy. We observed a male predominance (74.3%), and the mean age was 59.5 (±8.1) years. Mean follow-up was 39.1 (±24.8) months. We identified 94 (67.1%) squamous cell carcinomas (SCC); 44 (31.4%) adenocarcinomas; 1 (0.7%) carcinosarcoma; and 1 (0.7%) mixed adenocarcinoma and neuroendocrine carcinoma. Oncologic stage before neoadjuvant therapy was 3.5% for I, 18% for II, 70% for III, 8.5% for IV, according to the 8^th edition of American Joint Committee on Cancer (AJCC).⁹ The mean percentage increase in body weight was 1.33% (SD: ±12.1) after neoadjuvant therapy. Baseline patient characteristics are reported in Table 1.

Table 1 Patients' baseline characteristics

View Fullscreen

Pathologic complete response (pCR)

Association between variables and probability for pCR by univariate analysis are reported in Table 2. SCC is more likely to achieve pCR (compared to adenocarcinoma, OR: 8.07, 95% CI: 2.91–22.38, P < 0.001). Gender, age, grade of cellular differentiation, chemotherapy regimen, and neoplasm circumferential involvement before neoadjuvant therapy are variables that are unrelated to pCR. Relief of dysphagia, weight gain, and increased in serum albumin levels after neoadjuvant therapy were also unrelated to pCR.

Table 2 Univariate analysis assessing the variables for the outcome pathologic complete response (pCR)

View Fullscreen

After multivariate analysis, only SCC was associated with pCR (compared to adenocarcinoma, OR: 9.51, 95% CI: 3.05–29.64, P < 0.001).

Overall survival (OS)

By univariate analysis, association between variables and probability for OS are reported in Table 3. SCC was associated with higher probability of OS than other esophageal carcinomas (compared to adenocarcinoma, HR for death was 0.59, 95% CI: 0.35–1, P = 0.036). Age was also associated with lower probability of survival (HR 1.05, 95% CI 1.02-1.09, P = 0.005). Gender, grade of cellular differentiation, chemotherapy regimen, and neoplasm circumferential involvement before neoadjuvant therapy are variables that are unrelated to OS. Relief of dysphagia, weight gain, and increased serum albumin levels after neoadjuvant therapy were also unrelated to OS. After multivariate analysis, HR for age at diagnosis was 1.05 (95% CI: 1.01–1.08, P = 0.021).

Table 3 Univariate analysis assessing the variables for the outcome overall survival (OS)

View Fullscreen

Disease-free survival (DFS)

By univariate analysis, association between variables and probability for DFS are reported in Table 4. SCC was associated to higher probability of DFS than other esophageal carcinomas (compared to adenocarcinoma, HR for death or recurrence was 0.6, 95% CI: 0.37–0.98, P = 0.025). Gender, age, grade of cellular differentiation, chemotherapy regimen, and neoplasm circumferential involvement before neoadjuvant therapy are variables that are unrelated to DFS. Relief of dysphagia, weight gain, and increased serum albumin levels after neoadjuvant therapy were also unrelated to DFS.

Table 4 Univariate analysis assessing the variables for the outcome disease-free survival (DFS)

View Fullscreen

After multivariate analysis, weight loss during neoadjuvant therapy was related to higher risk for recurrence or death (HR 1.02, 95% CI: 1–1.04, P = 0.047).

Discussion

Several variables associated with patients' demographics, surgery aspects, and neoplasm status have been used as prognostic factors for esophageal cancer. Large neoplasms and poor cellular differentiation are factors that can predict poor outcomes and, consequently, are usually related to advanced clinical stages.¹⁰

Regarding locally advanced esophageal cancer, neoadjuvant therapy has been proven to yield better results and is currently considered the standard therapy.^11,12

In the setting of neoadjuvant therapy, only a few studies examined the relationship between prognostic variables and the change of these variables during neoadjuvant therapy.

In this study, SCC was associated with better results. The probability for pCR was 8.07 (95% CI: 2.91–22.38) higher than adenocarcinoma; the incidence of death was 0.59 (95% CI: 0.35–1) compared to adenocarcinoma; and the incidence of death or recurrence was 0.6 (95% CI: 0.37–0.98) compared to adenocarcinoma.

Concerning the 5-year OS analysis, Lee et al (13) found that patients with SCC had better survival rates than those with adenocarcinomas after trimodal therapy (42% versus 14%; P = 0.009).

In their systematic review, Bollschweiler et al¹⁴ found that the median probability of pCR is about 24% in the SCC, and 19.5% for adenocarcinoma, with no statistically significant difference. At any rate, tumor response to neoadjuvant therapy has been identified as a prognostic factor. Therefore, the 8^th edition of the AJCC staging of epithelial esophageal cancer shows separate classifications for clinical (cTNM), pathologic (pTNM), and postneoadjuvant (ypTNM) stage groups.^9,14

Considering neoadjuvant radiation, no significant difference was noted among the different range of doses for the endpoints pCR, OS, and DFS.

The CROSS (ChemoRadiotherapy for Oesophageal cancer followed by Surgery Study) trial¹⁵ used a dose of 41.4 Gy in 23 fractions, yielding a pCR rate of 29%, with locoregional recurrence rate of 14%. These rates are comparable to doses of 50.4 or even higher.¹⁶ Consequently, 41.4 Gy dose has been increasingly being used, avoiding radiation toxicity, and CROSS trial is currently the basis for National Comprehensive Cancer Network (NCCN) guidelines for esophageal cancer management.¹⁷

None of the chemotherapy regimens was found to lead to better outcomes. Nevertheless, Li et al¹² reported that taxane-incorporated chemotherapy is associated to higher OS rate; and there was a nonsignificant trend toward increased progression-free survival rate for taxane-incorporated group compared to patients receiving traditional cisplatin and 5-fluorouracil.

Dysphagia relief after neoadjuvant therapy is not a valuable predictor of pCR, OS, and DFS. Strandby et al¹⁸ found similar results for adenocarcinoma of the gastroesophageal junction. Relief from dysphagia may not be linked only to the neoplasm shrinkage. Besides the neoplasm mechanical obstruction, neoplasm encircling lower esophageal sphincter or infiltration of myenteric plexus branches or vagus nerves may theoretically contribute to dysphagia symptoms.^18,19

Esophageal cancer patients usually evidence poor nutritional status due to intense dysphagia. Weight gain after neoadjuvant therapy was not significantly associated with pCR, OS, or DFS. After multivariate analysis, weight loss during neoadjuvant therapy was related to higher risk for recurrence or death (HR 1.02, 95% CI: 1–1.04). Forshaw et al²⁰ reported that weight gain and improved swallowing after chemotherapy are not sufficiently sensitive to identify pathologic responders from nonresponders.

Serum albumin level change during neoadjuvant therapy was unrelated to any of the outcomes. However, critical albumin levels before or after neoadjuvant therapy may impact survival.²¹

As limitations of this work, we would like to mention that this is retrospective, single institutional study, and data were missing in some of the studied analyses. Also, we only investigated a few groups of chemotherapy regimens, and some regimens consisted of only a few patients. A recent cohort reported promising results for biweekly docetaxel plus cisplatin and fluorouracil for advanced esophageal cancer, particularly for elderly patients and patients with moderate organ disorders. Biweekly docetaxel plus cisplatin and fluorouracil showed lower toxicity rates than traditional docetaxel plus cisplatin and fluorouracil regimen, with high clinical efficacy.²²

Conclusions

The SCC histologic type was associated with higher probability of pCR, and higher OS and DFS rates. Gender, grade of cellular differentiation, chemotherapy regimen, and neoplasm circumferential involvement before neoadjuvant therapy are variables that are unrelated to pCR, OS, and DFS. Relief of dysphagia and increased serum albumin levels after neoadjuvant therapy were also unrelated to the studied outcomes. Weight loss during neoadjuvant chemotherapy was associated with poor DFS rate in the multivariate analysis.