Fuzzy Logic-Based Prognostic Score for Outcome ... - IEEE Xplore

4 downloads 0 Views 590KB Size Report
Nov 16, 2012 - the invasiveness of combined modality treatment, improved prog- ... logic-based approach also performed consistently better than the. Glasgow ...
1224

IEEE TRANSACTIONS ON INFORMATION TECHNOLOGY IN BIOMEDICINE, VOL. 16, NO. 6, NOVEMBER 2012

Fuzzy Logic-Based Prognostic Score for Outcome Prediction in Esophageal Cancer Chang-Yu Wang, Tsair-Fwu Lee, Member, IEEE, Chun-Hsiung Fang, Senior Member, IEEE, and Jyh-Horng Chou, Senior Member, IEEE

Abstract—Given the poor prognosis of esophageal cancer and the invasiveness of combined modality treatment, improved prognostic scoring systems are needed. We developed a fuzzy logicbased system to improve the predictive performance of a risk score based on the serum concentrations of C-reactive protein (CRP) and albumin in a cohort of 271 patients with esophageal cancer before radiotherapy. Univariate and multivariate survival analyses were employed to validate the independent prognostic value of the fuzzy risk score. To further compare the predictive performance of the fuzzy risk score with other prognostic scoring systems, time-dependent receiver operating characteristic curve analysis was used. Application of fuzzy logic to the serum values of CRP and albumin increased predictive performance for one-year overall survival (AUC = 0.773) compared with that of a single marker (AUC = 0.743 and 0.700 for CRP and albumin, respectively), where the AUC denotes the area under curve. This fuzzy logic-based approach also performed consistently better than the Glasgow prognostic score (AUC = 0.745). Thus, application of fuzzy logic to the analysis of serum markers can more accurately predict the outcome for patients with esophageal cancer. Index Terms—Esophageal cancer, fuzzy logic, radiotherapy (RT), receiver operating characteristic curve (ROC) analysis, survival analysis.

I. INTRODUCTION SOPHAGEAL cancer is one of the most lethal malignancies, and overall survival rates remain disappointing. The symptoms of esophageal cancer are often insidious at the onset, precluding early diagnosis. Therefore, the majority of patients present with locally advanced cancers and very few with early stage disease [1]. The only real prospect of cure for early stage disease lies in surgical resection. However, the optimal treatment strategy for locally advanced esophageal cancer has not

E

Manuscript received July 31, 2011; accepted July 28, 2012. Date of publication August 2, 2012; date of current version November 16, 2012. This work was supported in part by the National Science Council, Taiwan, under Grant NSC99-2221-E151-071-MY3. C.-Y. Wang is with the Department of Electrical Engineering, National Kaohsiung University of Applied Sciences, Kaohsiung 80778, Taiwan, and also with Yuan’s General Hospital, Kaohsiung 80249, Taiwan (e-mail: [email protected]). T.-F. Lee is with the Department of Electronics Engineering, National Kaohsiung University of Applied Sciences, Kaohsiung 80778, Taiwan (e-mail: [email protected]). C.-H. Fang is with the Department of Electrical Engineering, National Kaohsiung University of Applied Sciences, Kaohsiung 80778, Taiwan (e-mail: [email protected]). J.-H. Chou is with the Department of Electrical Engineering, National Kaohsiung University of Applied Sciences, Kaohsiung 80778, Taiwan, and also with the National Kaohsiung First University of Science and Technology, Kaohsiung 811, Taiwan (e-mail: [email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TITB.2012.2211374

been well established. Concurrent chemoradiotherapy (CCRT) followed by surgery (trimodality treatment) is a viable treatment option for patients with good performance status [2], but patients with major comorbidities would not benefit from this major operation that could be associated with a higher morbidity [3]. Definitive CCRT with their superiority over radiotherapy (RT) alone has been recognized as an alternative option for those who are not suitable surgical candidates [4]. Given the uncertain benefit and added morbidity and mortality of surgical resection after CCRT, and the high local failure rate after CCRT alone, there is growing interest in developing criteria to identify patients who may safely defer surgery after CCRT. Currently, staging of esophageal cancer requires the expertise of a multidisciplinary team and utilizes information gathered from computed topographic scans, endoscopic ultrasonography, bronchoscopy, or positron emission tomography. This information is used to determine the cancer stage as defined by American Joint Committee on Cancer (AJCC) [5]. However, AJCC stage may not be accurate in patients receiving multimodality therapy [5], [6]. It is of particular interest to identify patients who were unlikely to live more than 6–12 months following radical resection, thereby averting the morbidity of surgery and allowing for nonsurgical palliative therapies [6]. The prognosis of patients with esophageal cancer is poor even with active trimodality treatment, in part because patients exhibit impaired immunity, inflammation, and poor nutrition [7]. For example, many patients with esophageal cancer present with elevated serum C-reactive protein (CRP) concentration, a marker of systemic inflammation [8] and decreased albumin, an indicator of poor nutritional status [9]. We have previously examined the utility of these serum markers in an effort to more accurately predict patient outcomes [10]. The Glasgow prognostic score (GPS), based upon the systemic inflammatory response and nutritional decline, has been used to predict survival of patients with other cancers [11]–[16]. The GPS combines CRP and albumin into a score and is defined as follows: patients with both an elevated CRP (>5 mg/L) and hypoalbuminaemia (