Informative Power Evaluation of Clinical Parameters to Predict Initial Therapeutic Response in Patients with Advanced Pleural Mesothelioma: A Machine Learning Approach

Our plain-English summary of Informative Power Evaluation of Clinical Parameters to Predict Initial Therapeutic Response in Patients with Advanced Pleural Mesothelioma: A Machine Learning Approach, by R. Massafra, A. Catino, P.M.S. Perrotti, et al (Journal of Clinical Medicine, 2022).

The study

• Authors: R. Massafra, A. Catino, P.M.S. Perrotti, P. Pizzutilo, A. Fanizzi, M. Montrone, D. Galetta
• Journal: Journal of Clinical Medicine
• Published: 2022;11(6)
• Cited by: 5 articles, as counted by NIH iCite
• Read it: PubMed · Publisher · Free full text

Abstract

Malignant pleural mesothelioma (MPM) is a rare neoplasm whose early diagnosis is challenging and systemic treatments are generally administered as first line in the advanced disease stage. The initial clinical response may represent a useful parameter in terms of identifying patients with a better long-term outcome. In this report, the initial therapeutical response in 46 patients affected with advanced/unresectable pleural mesothelioma was investigated. The initial therapeutic response was assessed by CT scan and clinical examination after 2-3 treatment cycles. Our preliminary evaluation shows that the group of patients treated with regimens including antiangiogenetics and/or immunotherapy had a significantly better initial response as compared to patients only treated with standard chemotherapy, exhibiting a disease control rate (DCR) of 100% (95% IC, 79.40-100%) and 80.0% (95% IC, 61.40-92.30%), respectively. Furthermore, the therapeutic response was correlated with the disease stage, blood leukocytes and neutrophils, high albumin serum levels, and basal body mass index (BMI). Specifically, the patients with disease stage III showed a DCR of 95.7% (95% IC, 78.1-99.9%), whereas for disease stage IV the DCR decreased to 66.7% (95% IC, 34.9-9.1%). Moreover, a better initial response was observed in patients with a higher BMI, who reached a DCR of 96.10% (95% IC, 80.36-99.90%). Furthermore, in order to evaluate in the predictive power of the collected features a multivariate way, we report the preliminary results of a machine learning model for predicting the initial therapeutic response. We trained a state-of-the-art algorithm combined to a sequential forward feature selection procedure. The model reached a median AUC value, accuracy, sensitivity, and specificity of 77.0%, 75%, 74.8%, and 83.3%, respectively. The features with greater informational power were gender, histotype, BMI, smoking habits, packs/year, and disease stage. Our preliminary data support the possible favorable correlation between innovative treatments and therapeutic response in patients with unresectable/advanced pleural mesothelioma. The small sample size does not allow concrete conclusions to be drawn; nevertheless, this work is the basis of an ongoing study that will also involve radiomics in a larger dataset.

In plain English

This retrospective study looked at 46 consecutive patients in Bari, Italy, with unresectable pleural mesothelioma who received first-line systemic treatment between July 2017 and October 2021. Asbestos exposure or occupational risk was recorded for 30 of them. Thirty received standard platinum plus pemetrexed chemotherapy; sixteen received newer treatments in clinical trials that included antiangiogenic drugs and/or immunotherapy. After two to three treatment cycles, 40 patients had a partial response or stable disease. Because the sample was small, the authors counted results as significant at p less than 0.10. Disease control was 100% (95% confidence interval 79.40–100%) with the newer regimens versus 80.0% (61.40–92.30%) with chemotherapy alone. It was 95.7% (78.1–99.9%) in stage III versus 66.7% in stage IV, whose printed interval appears misprinted. Overweight patients reached 96.10% disease control (80.36–99.9%). Better response was also seen with lower white cell and neutrophil counts and higher albumin. A random forest machine learning model reached a median AUC (how well it told responders from non-responders) of 77.0%. The authors called the model moderately accurate but concluded these preliminary data from a small number of patients do not allow concrete conclusions.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Retrospective observational study, single center, with a random forest machine learning model to predict initial response to treatment
• Who was studied: 46 consecutive patients with unresectable pleural mesothelioma who had undergone first-line systemic treatment, July 2017 to October 2021, at Istituto Tumori 'Giovanni Paolo II', Bari, Italy
• Exposure: The study compared types of first-line treatment and other clinical features, not asbestos. Asbestos exposure/occupational risk was recorded as one clinical feature: yes in 30 patients (65.22%), no in 13 (28.26%), unknown in 3. No fiber types are given.
• Main results:
  • After 2 to 3 treatment cycles, 40 of the 46 patients had a partial response or stable disease and 6 had progressive disease
  • 30 patients received standard platinum plus pemetrexed chemotherapy; 16 received innovative treatments in clinical trials (antiangiogenic drug bevacizumab and/or immunotherapy, with or without chemotherapy)
  • Patients on regimens including antiangiogenics and/or immunotherapy had a disease control rate of 100% (95% IC, 79.40–100%), compared with 80.0% (95% IC, 61.40–92.30%) for patients on standard chemotherapy only (p-value 0.049)
  • Response was associated with disease stage (p-value 0.049). Disease control rate was 95.7% (95% IC, 78.1–99.9%) in stage III patients and 66.7% in stage IV patients. The printed confidence interval for stage IV appears garbled.
  • Patients with lower leukocyte levels, lower neutrophil levels, and higher serum albumin each had a disease control rate of 100% (95% IC 83.90–100%, 83.16–100%, and 75.30–100%, respectively). These associations were significant only at the study's threshold of p less than 0.10.
  • Overweight patients had a better initial response (p-value 0.035), with a disease control rate of 96.10% (95% IC, 80.36–99.9%)
  • The machine learning model performed best with seven features, with a median AUC of 77.0% (interquartile range 73.3–78.8%). Median accuracy was 75% (IR 63.04–84.8%), sensitivity 74.8% (IR 57.5–87.5%), and specificity 83.3%.
  • The features most often selected by the model were gender (54%), histotype (52%), BMI (50%), smoking habits (44%), packs/year (41%) and disease stage (40%). Asbestos exposure/occupational risk was selected 23% of the time.
• The authors' conclusion: The authors concluded that their machine learning approach based on clinical data achieved a moderately accurate model. They stated that these preliminary data from a small number of patients do not allow concrete conclusions, and that clinical characteristics alone are not sufficient for a clinical support tool.
• Limitations the authors state:
  • The dataset was relatively small
  • Model performance was highly variable, possibly because of the small sample and the strong imbalance between responders and non-responders
  • The authors acknowledged possible validation bias from cross-validation

Topics

Medical terms in this study, explained in our Lexicon: Pleural Mesothelioma, Neoplasm, Unresectable Pleural Mesothelioma, CT Scan, Immunotherapy, Chemotherapy, Habit, Unresectable.

Related studies

• Transfer learning approach in pre-treatment CT images to predict therapeutic response in advanced malignant pleural mesothelioma — A. Fanizzi, A. Catino, et al. (2024)
• Comprehensive approach integrating remote sensing, machine learning, and physicochemical parameters to detect hydrodynamic conditions and groundwater quality deterioration in non-rechargeable aquifer systems — M.H. Eid, A. Shebl, et al. (2024)
• Evaluation and Exploration of Machine Learning and Convolutional Neural Network Classifiers in Detection of Lung Cancer from Microarray Gene-A Paradigm Shift — K. M S, H. Rajaguru, et al. (2023)
• Volume-based parameter of 18)F-FDG PET/CT in malignant pleural mesothelioma: prediction of therapeutic response and prognostic implications — H.Y. Lee, S.H. Hyun, et al. (2010)
• A radiotherapy community data-driven approach to determine which complexity metrics best predict the impact of atypical TPS beam modeling on clinical dose calculation accuracy — F.M.D. Brooks, M.C. Glenn, et al. (2024)
• The use of a next-generation sequencing-derived machine-learning risk-prediction model (OncoCast-MPM) for malignant pleural mesothelioma: a retrospective study — M.G. Zauderer, A. Martin, et al. (2021)

Browse more research in Journal of Clinical Medicine, or return to medical research.