ABSTRACT
OBJECTIVES: Cancer patients have worse outcomes from the COVID-19 infection and greater need for ventilator support and elevated mortality rates than the general population. However, previous artificial intelligence (AI) studies focused on patients without cancer to develop diagnosis and severity prediction models. Little is known about how the AI models perform in cancer patients. In this study, we aim to develop a computational framework for COVID-19 diagnosis and severity prediction particularly in a cancer population and further compare it head-to-head to a general population. METHODS: We have enrolled multi-center international cohorts with 531 CT scans from 502 general patients and 420 CT scans from 414 cancer patients. In particular, the habitat imaging pipeline was developed to quantify the complex infection patterns by partitioning the whole lung regions into phenotypically different subregions. Subsequently, various machine learning models nested with feature selection were built for COVID-19 detection and severity prediction. RESULTS: These models showed almost perfect performance in COVID-19 infection diagnosis and predicting its severity during cross validation. Our analysis revealed that models built separately on the cancer population performed significantly better than those built on the general population and locked to test on the cancer population. This may be because of the significant difference among the habitat features across the two different cohorts. CONCLUSIONS: Taken together, our habitat imaging analysis as a proof-of-concept study has highlighted the unique radiologic features of cancer patients and demonstrated effectiveness of CT-based machine learning model in informing COVID-19 management in the cancer population.
ABSTRACT
PURPOSE: Tumor hypoxia is associated with poor response to radiation (RT). We previously discovered a novel mechanism of metformin: enhancing tumor RT response by decreasing tumor hypoxia. We hypothesized that metformin would decrease tumor hypoxia and improve cervical cancer response to RT. EXPERIMENTAL DESIGN: A window-of-opportunity, phase II randomized trial was performed in stage IB-IVA cervical cancer. Patients underwent screening positron emission tomography (PET) imaging with hypoxia tracer fluoroazomycin arabinoside (FAZA). Only patients with FAZA uptake (hypoxic tumor) were included and randomized 2:1 to receive metformin in combination with chemoRT or chemoRT alone. A second FAZA-PET/CT scan was performed after 1 week of metformin or no intervention (control). The primary endpoint was change in fractional hypoxic volume (FHV) between FAZA-PET scans, compared using the Wilcoxon signed-rank test. The study was closed early due to FAZA availability and the COVID-19 pandemic. RESULTS: Of the 20 consented patients, 6 were excluded due to no FAZA uptake and 1 withdrew. FHV of 10 patients in the metformin arm decreased by an average of 10.2% (44.4% to 34.2%) ± SD 16.9% after 1 week of metformin, compared to an average increase of 4.7% (29.1% to 33.8%) ±11.5% for the 3 controls (p = 0.027). Those with FHV reduction after metformin had significantly lower MATE2 expression. With a median follow-up of 2.8 years, the 2-year disease-free survival (DFS) was 67% for the metformin arm vs 33% for controls (p=0.09). CONCLUSIONS: Metformin decreased cervical tumor hypoxia in this trial that selected for patients with hypoxic tumor.