A Machine Learning Decision Support System (DSS) for Neuroendocrine Tumor Patients Treated with Somatostatin Analog (SSA) Therapy.

Virginia Rossi,Daniele Lavacchi,Amedeo Amedei,Lorenzo Dreoni,Serena Pillozzi,Nicola Fazio,Francesca Spada,Alice Laffi,Jasminka Hasic Telalovic,Lorenzo Antonuzzo,Rachele Fabbri,Ernesto Iadanza

doi:10.3390/diagnostics11050804

Abstract

The application of machine learning (ML) techniques could facilitate the identification of predictive biomarkers of somatostatin analog (SSA) efficacy in patients with neuroendocrine tumors (NETs). We collected data from 74 patients with a pancreatic or gastrointestinal NET who received SSA as first-line therapy. We developed three classification models to predict whether the patient would experience a progressive disease (PD) after 12 or 18 months based on clinic-pathological factors at the baseline. The dataset included 70 samples and 15 features. We initially developed three classification models with accuracy ranging from 55% to 70%. We then compared ten different ML algorithms. In all but one case, the performance of the Multinomial Naïve Bayes algorithm (80%) was the highest. The support vector machine classifier (SVC) had a higher performance for the recall metric of the progression-free outcome (97% vs. 94%). Overall, for the first time, we documented that the factors that mainly influenced progression-free survival (PFS) included age, the number of metastatic sites and the primary site. In addition, the following factors were also isolated as important: adverse events G3–G4, sex, Ki67, metastatic site (liver), functioning NET, the primary site and the stage. In patients with advanced NETs, ML provides a predictive model that could potentially be used to differentiate prognostic groups and to identify patients for whom SSA therapy as a single agent may not be sufficient to achieve a long-lasting PFS.

Highlights

Neuroendocrine tumors (NETs) arise from neuroendocrine cells distributed throughout the body
Summarizing, we have documented for the first time that the machine learning (ML) techniques provided a predictive model, which could potentially be used to differentiate prognostic groups in patients with an advanced NET and treated with an SSA
Consistent with previous literature, the predictive factors identified in our study may be useful when stratifying patients with a NET in further studies

Summary

Introduction

Neuroendocrine tumors (NETs) arise from neuroendocrine cells distributed throughout the body. They consist of a wide family of tumors that includes the pancreatic NET (panNET) and gastrointestinal (GI) NET and neoplasms from other origin sites [1]. The incidence of NETs in the United States was 6.98 cases per 100,000 people in the year 2004 and it increased from 1975 to 2008. The reasons for this rise are unclear the improvement of diagnosis and classification seem to be two of the major factors [2,3]. In addition to information on the histologic classification and stage, other factors are under study to verify a potential prognostic role such as the margin status (positive or negative) and the presence of a vascular or perineural invasion [21,22], the serum level of Chromogranin

Objectives

Methods

Results

Discussion

Conclusion