[68Ga]DOTATOC PET/CT Radiomics to Predict the Response in GEP-NETs Undergoing [177Lu]DOTATOC PRRT: The Theragnomics Concept

Simple Summary The radiological response assessment of neuroendocrine tumors (NET) to peptide receptor radionuclide therapy (PRRT) using [Lu-177]DOTATOC is still suboptimal due to the high variability in targeted somatostatin receptor 2 (SSTR-2) expression and histological heterogeneity among patients with well-differentiated NET. Promising and innovative laboratory assays have been proposed, but they are highly costly and not easily accessible. Machine learning offers new opportunities to provide quantitative characteristics from molecular images that cannot be appreciated by the human eye. We therefore retrospectively analyzed [Ga-68]DOTATOC PET/CT images before and after complete [Lu-177]DOTATOC PRRT in well-differentiated progressive, metastatic gastroenteropancreatic NET and obtained radiomics features as new and reliable imaging parameters that correlate to the response to PRRT and might be used for improved patient selection in the future. Despite impressive results, almost 30% of NET do not respond to PRRT and no well-established criteria are suitable to predict response. Therefore, we assessed the predictive value of radiomics [Ga-68]DOTATOC PET/CT images pre-PRRT in metastatic GEP NET. We retrospectively analyzed the predictive value of radiomics in 324 SSTR-2-positive lesions from 38 metastatic GEP-NET patients (nine G1, 27 G2, and two G3) who underwent restaging [Ga-68]DOTATOC PET/CT before complete PRRT with [Lu-177]DOTATOC. Clinical, laboratory, and radiological follow-up data were collected for at least six months after the last cycle. Through LifeX, we extracted 65 PET features for each lesion. Grading, PRRT number of cycles, and cumulative activity, pre- and post-PRRT CgA values were also considered as additional clinical features. [Ga-68]DOTATOC PET/CT follow-up with the same scanner for each patient determined the disease status (progression vs. response in terms of stability/reduction/disappearance) for each lesion. All features (PET and clinical) were also correlated with follow-up data in a per-site analysis (liver, lymph nodes, and bone), and for features significantly associated with response, the Delta radiomics for each lesion was assessed on follow-up [Ga-68]DOTATOC PET/CT performed until nine months post-PRRT. A statistical system based on the point-biserial correlation and logistic regression analysis was used for the reduction and selection of the features. Discriminant analysis was used, instead, to obtain the predictive model using the k-fold strategy to split data into training and validation sets. From the reduction and selection process, HISTO_Skewness and HISTO_Kurtosis were able to predict response with an area under the receiver operating characteristics curve (AUC ROC), sensitivity, and specificity of 0.745, 80.6%, 67.2% and 0.722, 61.2%, 75.9%, respectively. Moreover, a combination of three features (HISTO_Skewness; HISTO_Kurtosis, and Grading) did not improve the AUC significantly with 0.744. SUVmax, however, could not predict the response to PRRT (p = 0.49, AUC 0.523). The presented preliminary theragnomics model proved to be superior to conventional quantitative parameters to predict the response of GEP-NET lesions in patients treated with complete [Lu-177]DOTATOC PRRT, regardless of the lesion site.

Automated summary

Machine learning techniques offer new opportunities for radiomics analysis to predict treatment response in patients with neuroendocrine tumors.