On the prediction of Hodgkin lymphoma treatment response.

PURPOSE: The cure rate in Hodgkin lymphoma is high, but the response along with treatment is still unpredictable and highly variable among patients. Detecting those patients who do not respond to treatment at early stages could bring improvements in their treatment. This research tries to identify the main biological prognostic variables currently gathered at diagnosis and design a simple machine learning methodology to help physicians improve the treatment response assessment. METHODS: We carried out a retrospective analysis of the response to treatment of a cohort of 263 Caucasians who were diagnosed with Hodgkin lymphoma in Asturias (Spain). For that purpose, we used a list of 35 clinical and biological variables that are currently measured at diagnosis before any treatment begins. To establish the list of most discriminatory prognostic variables for treatment response, we designed a machine learning approach based on two different feature selection methods (Fisher's ratio and maximum percentile distance) and backwards recursive feature elimination using a nearest-neighbor classifier (k-NN). The weights of the k-NN classifier were optimized using different terms of the confusion matrix (true- and false-positive rates) to minimize risk in the decisions. RESULTS AND CONCLUSIONS: We found that the optimum strategy to predict treatment response in Hodgkin lymphoma consists in solving two different binary classification problems, discriminating first if the patient is in progressive disease; if not, then discerning among complete and partial remission. Serum ferritin turned to be the most discriminatory variable in predicting treatment response, followed by alanine aminotransferase and alkaline phosphatase. The importance of these prognostic variables suggests a close relationship between inflammation, iron overload, liver damage and the extension of the disease.

[A series of 618 cases of monoclonal gammopathies of undetermined significance: predictive factors of disappearance of monoclonal component or evolution to malignant gammopathies]. / Serie de 618 casos de gammapatías monoclonales de significado indeterminado (GMSI): factores predictivos de desaparición del componente monoclonal o de evolución a gammapatías malignas.

INTRODUCTION: How to identify monoclonal gammopathies of undeterminated significance (MGUS) at risk for progression has been studied for the last years. AIMS: To study the incidence of MGUS in a region with 300,000 inhabitants and factors which associate with a) monoclonal gammopathy disappearance (transient MGUS) b) evolution to malignant gammopathy. METHODS: Study of 618 MGUS. RESULTS: Incidence: 30/40 new cases a year with increase to 70 cases a years in the latest years of study. Age and gender: 71,4 y (32-100), male/female ratio 1.4. Associated pathology: infection 328, heart diseases 249, rheumatic diseases 211, liver diseases 108, cancer 80, neuropathy 43. Monoclonal proteins: IgG 407, IgM 78, IgD 2, biclonal 16, triclonal 1; no heavy chain 21, light chain Kappa 389. Variables (mean): monoclonal component: 14 g/l, ESR 32,5, bone marrow: 5,9% plasma cells beta2-microglobulin: 2,59 mg/l, albumin: 3,1g/l, bone survey: normal 39,5%. Evolution: transient MGUS 20 cases. Time to disappearance 2,6 months (1,4-4,6). Evolution to malignant gammopathy 24 cases, time to progression 3 years (IC 1,82-4,3). RESULTS: Several factors were associatedç with malignant transformation: heavy chain IgA (p < 0,002), ESR (p < 0,001), age < 70 (p < 0,05), bone marrow percentage of plasma cells (p < 0,002) y ostheoporosis (p < 0,005). A MGUS follow up model is suggested.