Jonathan Simons, MD, President, and CEO of the Prostate Cancer Foundation speaks about Study Training Dogs to Detect Prostate Cancer Gets One Paw Closer to a ‘Robotic Nose’ to Diagnose the Disease, Including Most Lethal Form.
Link to Study:
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0245530
Summary
In the developed world, prostate cancer is the second leading cause of cancer death in men. Beyond serum prostate-specific antigen (PSA) population screening, a more sensitive and specific detection strategy for lethal prostate cancer is urgently required. Canine olfaction diagnosis has been shown to be both specific and sensitive by using dogs trained to detect cancer by smell. Although dogs as diagnostic sensors are impractical, computer olfaction for cancer detection can be checked. However, due to the considerable divide between these disciplines, studies bridging the gap between clinical diagnostic methods, artificial intelligence, and molecular analysis remain difficult. In a double-blinded pilot study, we examined the clinical feasibility of a cross-disciplinary, integrative approach to early prostate cancer biosensing in urine using qualified canine olfaction, volatile organic compound (VOC) analysis through gas chromatography-mass spectroscopy (GC-MS), ANN-assisted testing, and microbial profiling. Two dogs were qualified to sniff out Gleason 9 prostate cancer in urine samples taken from biopsy-confirmed men. Canine specificity as prostate cancer biodetectors was tested using biopsy-negative controls. Urine samples were tested for VOC content in headspace using GC-MS and urinary microbiota content using 16S rDNA Illumina sequencing at the same time. Furthermore, the dogs’ diagnoses were used to train an artificial neural network (ANN) to detect significant peaks in the GC-MS data. When it came to detecting Gleason 9 prostate cancer, the canine olfaction system was 71% sensitive and 70–76% precise. We have used GC-MS to confirm VOC differences and 16S rDNA sequencing to confirm microbiota differences between cancer-positive and biopsy-negative controls. Furthermore, using the canine diagnoses as input, the trained ANN identified regions of interest in the GC-MS results. To improve machine olfaction diagnostic tools, the methodology and feasibility are developed to inform larger-scale studies using canine olfaction, urinary VOCs, and urinary microbiota profiling. Scalable multidisciplinary methods can then be compared to PSA screening for detecting clinically aggressive prostate cancers in urine samples earlier, non-invasively, more specifically, and sensitively.