Abstract
Patients with clinical T4 esophageal squamous cell carcinoma (ESCC) have an unfavorable prognosis, mainly indicated by the response to chemoradiotherapy (CRT), crucial to estimating long-term survival. Other prognostic measures include systemic inflammatory or immunonutritional indices such as the Glasgow Prognostic Score (GPS) and Prognostic Nutritional Index (PNI) that have not been sufficiently documented. Patients and Methods: This study retrospectively evaluated 91 patients with T4 ESCC treated at our Hospital between 2000 and 2013. All patients initially received CRT, including 5-fluorouracil (5FU) and cisplatin or nedaplatin with concurrent 2-Gy/fraction radiation (total dose, 40-60 Gy). Curative tumor resection was undertaken in suitable patients on completing CRT. Patients were classified as GPS0, GPS1, or GPS2 based on C-reactive protein (CRP) ≤10 mg/l and albumin ≥35 g/l, CRP >10 mg/l or albumin <35 g/l, or CRP >10 mg/l and albumin <35 g/l, respectively. PNI was calculated as 10-times the serum albumin (g/dl)+0.005×total lymphocyte count (/mm3). The impact of the pre-treatment GPS and PNI on the prognosis of patients with T4 ESCC was investigated in univariate and multivariate analyses. Results: Sixty (67%) patients responded to CRT (9 complete responses and 51 partial responses). Forty-one (45%) patients also underwent surgical resection of the residual tumor. The overall 5-year survival rate and median survival time were 27.0% and 11.8 months, respectively. In the cohort of CRT-plus-surgical resection, the 5-year survival rate was significantly higher than in the groups treated with CRT-alone (51.1% vs. 6.5%; p<0.01). On multivariate analysis, good response to CRT [hazard ratio (HR) =0.449, p<0.01], GPS1/2 (HR=2.151, p=0.015), and surgical resection (HR=0.282, p<0.01) were significant prognostic factors, whereas PNI was not. Conclusion: The GPS is a useful, simple survival marker for patients with T4 ESCC undergoing multimodal therapy.
Esophageal carcinoma, one of the most aggressive malignant neoplasms, is the sixth leading cause of cancer death worldwide (1). In Japan, esophageal squamous cell carcinoma (ESCC), in contrast to adenocarcinoma in Western countries, is the main histopathological type of esophageal cancer that metastasizes directly with ease to adjacent vital organs, such as the tracheobronchial tree and aorta, and becomes unresectable (2, 3). Chemoradiotherapy (CRT) is considered the optimal treatment for patients with clinical T4 ESCC with a good response to CRT, crucial to estimating long-term survival and good prognosis, whereas a poor response indicates a dismal prognosis. However, other prognostic indicators have not been sufficiently investigated thus far.
Recently, pretreatment nutritional status, including serum albumin, inflammatory status, including serum C-reactive protein (CRP), and immune status, including lymphocyte counts, have been found to influence long-term prognosis post treatment in various types of cancer. Immunonutritional indexes such as the Glasgow Prognostic Score (GPS) and Prognostic Nutritional Index (PNI) have been reported to predict post-treatment survival in gastric (4), colorectal (5), and pancreatic (6) cancer. However, the utility of these indices in T4 ESCC have not been adequately documented. We hypothesized that the GPS and PNI scores before CRT are strongly indicative of long-term survival after CRT for patients with T4 ESCC.
This retrospective study aimed to elucidate prognostic factors, including the GPS and PNI, in patients with T4 ESCC who underwent multimodal treatment.
Patients and Methods
This study retrospectively evaluated 91 patients with T4 ESCC treated with CRT, with or without surgery, at the Department of Surgical Oncology, Osaka City University Hospital, between April 2000 and April 2013. The pathological diagnosis and classifications were made in accordance with the sixth edition of the Tumor Node Metastasis (TNM) Classification of Malignant Tumors (7). The staging work-up was performed using endoscopy, computed tomography (CT), positron-emission tomography, and bronchoscopy. The absence of a fatty layer between the tumor and the aorta and 90° or greater angles of tumor contact against the aorta on three or more slices (slice thickness, 5 mm) on CT was defined as direct involvement of the aorta. Tumor protrusion into the lumen of the trachea on CT and irregular appearance of the surface of the membranous portion of trachea on bronchoscopy were defined as indicating direct involvement of the trachea and bronchus.
CRT regimens varied across three major time periods. From April 2000 to April 2003, a low-dose 5-fluorouracil (5FU) and cisplatin regimen comprising 250 mg of 5FU plus 10 mg of cisplatin was administered on days 1-5, 8-12, 15-19, and 22-26 with concurrent external beam radiation of 2.0 Gy/fraction for a total dose of 40-60 Gy. Between May 2003 and April 2007, the CRT protocol was modified to replace cisplatin with nedaplatin, and surgical resection was undertaken on completing 40 Gy radiation in patients with resectable tumors, whereas 20 Gy radiation was added if the tumor remained unresectable or the patient refused surgery. From 2007 to the present, the CRT protocol comprised 700 mg/m2 5FU on days 1-4 and 29-32 and 60 mg/m2 nedaplatin administered on days 1 and 29 with concurrent 2 Gy/fraction radiation up to a total dose of 50 Gy followed by surgery for tumors curatively resected.
The following variables were evaluated as prognostic factors: age, sex, clinical nodal status, preoperative performance status (PS), response to CRT, PNI, and GPS. With regard to the GPS, patients were classified as GPS0, GPS1, or GPS2 based on CRP ≤10 mg/l and albumin ≥35 g/l, CRP >10 mg/l or albumin <35 g/l, or CRP >10 mg/l and albumin <35 g/l, respectively (8). The PNI was calculated as: 10×serum albumin (g/dl) + 0.005×total lymphocyte count (/mm3), with a cut-off value of 40, based on the original study (9). All blood samples analyzed were collected before CRT.
Adverse events of CRT were assessed and graded by the National Cancer Institute-Common Toxicity Criteria version 3 (10). Response to CRT was evaluated according to Response Evaluation Criteria in Solid Tumor (RECIST) (11) and Japanese Guidelines for the Clinical and Pathologic Studies on Carcinoma of the Esophagus (12) and classified into four categories: complete regression (CR) was defined as 100% tumor regression; partial response (PR) was a more than 30% reduction in the sum of the longest diameter of measurable lesions; stable disease (SD) ranged from 30% reduction to 20% tumor enlargement; and progressive disease (PD) involved more than 20% tumor enlargement or the occurrence of new lesions. All characteristic data were obtained from medical and nursing records. Informed consent in this study was obtained from all patients.
Statistical analysis. Demographic data are shown as means±standard deviations. Comparisons among patient cohorts were made with analyses of variance, followed by either Student's t-test or the chi-square test. The overall survival durations for various sub-cohorts were calculated by the Kaplan–Meier method. The prognoses were compared by log-rank test. Univariate and multivariate analyses of independent prognostic factors were conducted in a Cox proportional hazards model. Statistical significance was determined by p-values of less than 0.05. SPSS software version 22 (SPSS, Inc., Chicago, IL, USA) was used for data analysis.
Results
Patient demographics and oncological features are summarized in Table I. The male:female ratio was 74:17 and the mean age was 63.0 years in this study population. Organ infiltration by the esophageal tumor included tracheal involvement in 49 cases, aortic involvement in 37 cases, and both tracheal and aortic involvement in five cases. In total, 56 and 35 patients received cumulative doses of <50 Gy and ≥50 Gy radiation, respectively. Sixty (67%) cases responded to CRT and included nine (10%) CR and 51 (57%) PR, as per the RECIST evaluation system. The most frequent adverse event was leukopenia, with grade 3 or greater severity in 40% (37/91) of patients. CRT-related mortality occurred in one case due to radiation-induced pneumonitis, whereas surgery-associated mortality was observed in three cases. Forty out of the 91 patients (43.9%) underwent surgical resection of the residual tumor. Complete resection at surgery was achieved in 90.0% (36/40) of patients.
In this study population, the overall 5-year survival rate and median survival duration was 27.0% and 11.8 months, respectively (Figure 1). In the cohort ofCRT plus surgical resection, the 5-year survival was significantly higher than in the group treated with CRT alone (51.1% vs. 6.5%; p<0.01). Tables II and III show results of univariate and multivariate analyses, respectively, for prognostic factors predicting overall survival in the whole cohort of patients with T4 ESCC. On univariate analysis, GPS, PS, serum SCC level, and response to CRT were significant prognostic factors, whereas good response to CRT [hazard ratio (HR) =0.455; 95% confidence interval (CI)=0.264-0.783], GPS1/2 (HR=2.151; 95% CI=1.167-3.966), and surgical resection of tumor (HR=0.282; 95% CI=0.146-0.543) were independent prognostic factors on multivariate analysis. Figure 2 shows the overall survival curves for patients with T4 ESCC according to GPS status. GPS0 was associated with significantly better prognosis compared with GPS1/2 (p=0.001). However, PNI was not a significant prognostic factor in univariate (p=0.065) or multivariate (p=0.628) analyses. The association between clinico-oncological features and the GPS status of patients with T4 ESCC is shown in Table IV. Patients in GPS1/2 groups had a higher proportion of males, a higher rate of poor PS, and a higher value of serum SCC compared to those in the GPS0 group.
Discussion
The prognosis of ESCC involving adjacent organs is dismal because complete resection of T4 ESCC is technically challenging and difficult because of the difficulty in combining a resection of the tumor spread to adjacent organs including aorta and trachea. Response to CRT is reportedly one of most important prognostic factors in T4 ESCC. In the present study, PR and CR after CRT were independent prognostic factors with good correlation on multivariate analysis. Some reports show curative surgical resection post CRT for T4 ESCC is effective for long-term survival (13, 14); however, other reports suggest that post-CRT surgical tumor resection is not effective in patients who respond to CRT (15, 16). In this study, multivariate analysis revealed that post-CRT surgery was an independent prognostic factor. However, there could have been a significant selection bias wherein responders to CRT tended to be mainly chosen as the candidates for surgical tumor resection at our Institution. Therefore, the actual value of surgical resection may have been obscured.
Conventionally, SCC and cytokeratin 19 fragment antigen 21-1 (CYFRA21-1) (17) are well-known tumor markers for ESCC. However, the sensitivity of these tumor markers is relatively low. Molecular markers such as epidermal growth factor receptor (EGFR) (18) and p53 (19) in the resected specimen and biopsy samples obtained during endoscopy were reportedly useful as prognostic markers. However, determination of these specimen-based markers are costly and not clinically practical. Simple and inexpensive markers are desirable for establishing a prognosis in ESCC; therefore, we evaluated the utility of GPS and PNI for patients with T4 ESCC because these markers are easily obtained from blood samples. The GPS was allocated by serum CRP and albumin levels and the PNI was calculated from serum albumin and lymphocyte counts.
Recently, several articles have reported an association between systemic inflammation and poor long-term outcome in various cancer types. The GPS, consisting of the serum CRP and albumin levels, simply reflects the systemic inflammation status of patients with cancer. The GPS was initially introduced to determine its prognostic value in advanced non-small-cell lung cancer by Forrest et al. in 2003 (8). After that, many reports about a close association between GPS and prognosis in various cancer types have been reported in the literature. The present study showed that pre-treatment GPS was an independent prognostic factor in patients with T4 ESCC. However, most reports regarding GPS for ESCC investigated patients who underwent esophagectomy (20-22). Those studies also described GPS as reflecting deeper tumor involvement, higher numbers of nodal metastases, and more advanced cancer stage. On the other hand, in this study, most patients had the same cancer stage. In our study, GPS1/2 scoring was closely associated with the male sex, poorer PS, and higher value of serum SCC compared with patients having a GPS0 scoring. These findings suggest that the GPS for patients with T4 ESCC might be a synthetic marker for predicting overall survival because it reflects both tumor progression and patient background.
On the other hand, in this study, PNI was not associated with survival after CRT in patients with T4 ESCC. The PNI was initially reported by Onodera et al. as a useful tool for predicting prognosis in patients with cancer (9); it is calculated by using serum albumin levels and lymphocyte counts that reflect the nutritional and immunological status of patients with cancer. Certain reports have revealed that preoperative PNI is associated with survival after surgery in patients with gastric (4) and colorectal cancer (5). However, the association between PNI and long-term survival in patients with ESCC has not been completely evaluated. Feng et al. reported that PNI was closely associated with survival in patients who underwent esophagectomy and was also associated with tumor length, tumor grade, and nodal status (24). The utility of PNI for patients with T4 ESCC has remained controversial, and further studies are required to definitively evaluate this prognostic index.
This study has certain limitations: it was retrospective and conducted at a single center. In addition, the CRT regimen differed across different periods during the study duration. Moreover, nedaplatin, which we chose as a main chemotherapeutic agent for ESCC, is not the standard therapeutic agent used to treat for ESCC worldwide.
In conclusion, the GPS is a simple and practical survival marker for patients with T4 ESCC who undergo multimodal treatment. Aggressive treatment including surgery for patients with T4 ESCC accompanied by GPS1/2 may not be beneficial in terms of prognosis.
- Received April 28, 2015.
- Revision received May 30, 2015.
- Accepted June 3, 2015.
- Copyright© 2015 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved