Outcome Prediction at Patient Level Derived from Pre-Treatment 18F-FDG PET Due to Machine Learning in Metastatic Melanoma Treated with Anti-PD1 Treatment

Abstract

(1) Background: As outcome of patients with metastatic melanoma treated with anti-PD1 immunotherapy can vary in success, predictors are needed. We aimed to predict at the patients’ levels, overall survival (OS) and progression-free survival (PFS) after one year of immunotherapy, based on their pre-treatment 18F-FDG PET; (2) Methods: Fifty-six metastatic melanoma patients—without prior systemic treatment—were retrospectively included. Forty-five 18F-FDG PET-based radiomic features were computed and the top five features associated with the patient’s outcome were selected. The analyzed machine learning classifiers were random forest (RF), neural network, naive Bayes, logistic regression and support vector machine. The receiver operating characteristic curve was used to compare model performances, which were validated by cross-validation; (3) Results: The RF model obtained the best performance after validation to predict OS and PFS and presented AUC, sensitivities and specificities (IC95%) of 0.87 ± 0.1, 0.79 ± 0.11 and 0.95 ± 0.06 for OS and 0.9 ± 0.07, 0.88 ± 0.09 and 0.91 ± 0.08 for PFS, respectively. (4) Conclusion: A RF classifier, based on pretreatment 18F-FDG PET radiomic features may be useful for predicting the survival status for melanoma patients, after one year of a first line systemic treatment by immunotherapy.

The Role and Potential of 18F-FDG PET/CT in Malignant Melanoma: Prognostication, Monitoring Response to Targeted and Immunotherapy, and Radiomics

“…Finally, Flaus et al [78] aimed at predicting overall survival (OS) and progression-free survival (PFS) after one year of immunotherapy based on pre-treatment 18 F-FDG PET scan data. To this end the authors retrospectively examined 56 patients with metastatic melanoma without prior systemic treatment.…”

Imaging More than Skin-Deep: Radiologic and Dermatologic Presentations of Systemic Disorders

“…Metastatic melanomas most commonly present near the surgical site of previous lesions as macules, papules, or nodules, or may manifest as firm palpable nodules that may be subcutaneous ( Figure 20 A). Additionally, melanomas may present as angiomatoid metastasis (soft-tissue mass with hemorrhagic and necrotic components) ( Figure 20 B) or hematoma-like metastases with ecchymosis [ 87 ]. Melanoma often spreads to regional lymph nodes ( Figure 20 C) via lymphatics but can spread hematogenously to the liver, lung, and brain [ 85 , 87 ].…”

“…Additionally, melanomas may present as angiomatoid metastasis (soft-tissue mass with hemorrhagic and necrotic components) ( Figure 20 B) or hematoma-like metastases with ecchymosis [ 87 ]. Melanoma often spreads to regional lymph nodes ( Figure 20 C) via lymphatics but can spread hematogenously to the liver, lung, and brain [ 85 , 87 ]. Brain metastases are best depicted on post-contrast MRI; as melanin reduces T1WI relaxation time, lesions with an adequate amount of melanin represent high SI on T1WI ( Figure 20 D), which is an uncommon finding in other cancers.…”

“…Fluorodeoxyglucose–positron emission tomography (FDG-PET) is another modality commonly used in detecting metastatic foci, revealing high FDG uptake due to increased metabolic activity [ 1 , 84 , 85 , 86 , 87 ]. A novel machine-learning-based model known as radiomics allows for the translation of the medical images to quantitative data.…”

“…A novel machine-learning-based model known as radiomics allows for the translation of the medical images to quantitative data. Radiomic features of 18F-FDG-PET have been investigated in recent studies to predict prognosis in patient with metastatic melanoma before immunotherapy treatment [ 1 , 84 , 85 , 86 , 87 , 88 ].…”