Artificial intelligence / machine-learning tool for post-market surveillance of in vitro diagnostic assays.

The study compares an artificial intelligence technology with traditional manual search of literature databases to assess the accuracy and efficiency of retrieving relevant articles for post-market surveillance of in vitro diagnostic and medical devices under the Medical Device Regulation and In Vitro Diagnostic Medical Device Regulation. Over a 3-year period, literature searches and technical assessment searches were performed manually or using the Huma.AI platform to retrieve relevant articles related to the safety and performance of selected in vitro diagnostic and medical devices. The manual search involved refined keyword searches, screening of titles/abstracts / full text, and extraction of relevant information. The Huma.AI search utilized advanced caching techniques and a natural language processing system to identify relevant reports. Searches were conducted on PubMed and PubMed Central. The number of identified relevant reports, precision rates, and time requirements for each approach were analyzed. The Huma.AI system outperformed the manual search in terms of the number of identified relevant articles in almost all cases. The average precision rates per year were significantly higher and more consistent with the Huma.AI search compared with the manual search. The Huma.AI system also took significantly less time to perform the searches and analyze the outputs than the manual search. The study demonstrated that the Huma.AI platform was more effective and efficient in identifying relevant articles compared with the manual approach.

Paired-like homeobox gene (PHOX2B) nonpolyalanine repeat expansion mutations (NPARMs): genotype-phenotype correlation in congenital central hypoventilation syndrome (CCHS).

PURPOSE: CCHS is an extremely rare congenital disorder requiring artificial ventilation as life support. Typically caused by heterozygous polyalanine repeat expansion mutations (PARMs) in the PHOX2B gene, identification of a relationship between PARM length and phenotype severity has enabled anticipatory management. However, for patients with non-PARMs in PHOX2B (NPARMs, ~10% of CCHS patients), a genotype-phenotype correlation has not been established. This comprehensive report of PHOX2B NPARMs and associated phenotypes, aims at elucidating potential genotype-phenotype correlations that will guide anticipatory management. METHODS: An international collaboration (clinical, commercial, and research laboratories) was established to collect/share information on novel and previously published PHOX2B NPARM cases. Variants were categorized by type and gene location. Categorical data were analyzed with chi-square and Fisher's exact test; further pairwise comparisons were made on significant results. RESULTS: Three hundred two individuals with PHOX2B NPARMs were identified, including 139 previously unreported cases. Findings demonstrate significant associations between key phenotypic manifestations of CCHS and variant type, location, and predicted effect on protein function. CONCLUSION: This study presents the largest cohort of PHOX2B NPARMs and associated phenotype data to date, enabling genotype-phenotype studies that will advance personalized, anticipatory management and help elucidate pathological mechanisms. Further characterization of PHOX2B NPARMs demands longitudinal clinical follow-up through international registries.

Diagnosing hereditary cancer predisposition in men with prostate cancer.

PURPOSE: We describe the pathogenic variant spectrum and identify predictors of positive results among men referred for clinical genetic testing for prostate cancer. METHODS: One thousand eight hundred twelve men with prostate cancer underwent clinical multigene panel testing between April 2012 and September 2017. Stepwise logistic regression determined the most reliable predictors of positive results among clinical variables reported on test requisition forms. RESULTS: A yield of 9.4-12.1% was observed among men with no prior genetic testing. In this group, the positive rate of BRCA1 and BRCA2 was 4.6%; the positive rate for the mismatch repair genes was 2.8%. Increasing Gleason score (odds ratio [OR] 1.19; 95% confidence interval [CI] 0.97-1.45); personal history of breast or pancreatic cancer (OR 3.62; 95% CI 1.37-9.46); family history of breast, ovarian, or pancreatic cancer (OR 2.32 95% CI 1.48-3.65); and family history of Lynch syndrome-associated cancers (OR 1.97; 95% CI 1.23-3.15) were predictors of positive results. CONCLUSION: These results support multigene panel testing as the primary genetic testing approach for hereditary prostate cancer and are supportive of recommendations for consideration of germline testing in men with prostate cancer. Expanding the criteria for genetic testing should be considered as many pathogenic variants are actionable for treatment of advanced prostate cancer.

Assessment of Diagnostic Outcomes of RNA Genetic Testing for Hereditary Cancer.

Importance: Performing DNA genetic testing (DGT) for hereditary cancer genes is now a well-accepted clinical practice; however, the interpretation of DNA variation remains a challenge for laboratories and clinicians. Adding RNA genetic testing (RGT) enhances DGT by clarifying the clinical actionability of hereditary cancer gene variants, thus improving clinicians' ability to accurately apply strategies for cancer risk reduction and treatment. Objective: To evaluate whether RGT is associated with improvement in the diagnostic outcome of DGT and in the delivery of personalized cancer risk management for patients with hereditary cancer predisposition. Design, Setting, and Participants: Diagnostic study in which patients and/or families with inconclusive variants detected by DGT in genes associated with hereditary breast and ovarian cancer, Lynch syndrome, and hereditary diffuse gastric cancer sent blood samples for RGT from March 2016 to April 2018. Clinicians who ordered genetic testing and received a reclassification report for these variants were surveyed to assess whether RGT-related variant reclassifications changed clinical management of these patients. To quantify the potential number of tested individuals who could benefit from RGT, a cohort of 307 812 patients who underwent DGT for hereditary cancer were separately queried to identify variants predicted to affect splicing. Data analysis was conducted from March 2016 and September 2018. Main Outcomes and Measures: Variant reclassification outcomes following RGT, clinical management changes associated with RGT-related variant reclassifications, and the proportion of patients who would likely be affected by a concurrent DGT and RGT multigene panel testing approach. Results: In total, 93 if 909 eligible families (10.2%) submitted samples for RGT. Evidence from RGT clarified the interpretation of 49 of 56 inconclusive cases (88%) studied; 26 (47%) were reclassified as clinically actionable and 23 (41%) were clarified as benign. Variant reclassifications based on RGT results changed clinical management recommendations for 8 of 18 patients (44%) and 14 of 18 families (78%), based on responses from 18 of 45 clinicians (40%) surveyed. A total of 7265 of 307 812 patients who underwent DGT had likely pathogenic variants or variants of uncertain significance potentially affecting splicing, indicating that approximately 1 in 43 individuals could benefit from RGT. Conclusions and Relevance: In this diagnostic study, conducting RNA testing resolved a substantial proportion of variants of uncertain significance in a cohort of individuals previously tested for cancer predisposition by DGT. Performing RGT might change the diagnostic outcome of at least 1 in 43 patients if performed in all individuals undergoing genetic evaluation for hereditary cancer.

False-positive results released by direct-to-consumer genetic tests highlight the importance of clinical confirmation testing for appropriate patient care.

PURPOSE: There is increasing demand from the public for direct-to-consumer (DTC) genetic tests, and the US Food and Drug Administration limits the type of health-related claims DTC tests can market. Some DTC companies provide raw genotyping data to customers if requested, and these raw data may include variants occurring in genes recommended by the American College of Medical Genetics and Genomics to be reported as incidental/secondary findings. The purpose of this study was to review the outcome of requests for clinical confirmation of DTC results that were received by our laboratory and to analyze variant classification concordance. METHODS: We identified 49 patient samples received for further testing that had previously identified genetic variants reported in DTC raw data. For each case identified, information pertaining to the outcome of clinical confirmation testing as well as classification of the DTC variant was collected and analyzed. RESULTS: Our analyses indicated that 40% of variants in a variety of genes reported in DTC raw data were false positives. In addition, some variants designated with the "increased risk" classification in DTC raw data or by a third-party interpretation service were classified as benign at Ambry Genetics as well as several other clinical laboratories, and are noted to be common variants in publicly available population frequency databases. CONCLUSION: Our results demonstrate the importance of confirming DTC raw data variants in a clinical laboratory that is well versed in both complex variant detection and classification.

Quality of Clinician-Reported Cancer History When Ordering Genetic Testing.

PURPOSE: Clinical history data reported on test requisition forms (TRFs) for hereditary cancer multigene panel testing (MGPT) are routinely used by genetic testing laboratories. More recently, publications have incorporated TRF-based clinical data into studies exploring yield of testing by phenotype and estimating cancer risks for mutation carriers. We aimed to assess the quality of TRF data for patients undergoing MGPT. PATIENTS AND METHODS: Ten percent of patients who underwent hereditary cancer MGPT between January and June 2015 at a clinical laboratory were randomly selected. TRF-reported cancer diagnoses were evaluated for completeness and accuracy for probands and relatives using clinical documents such as pedigrees and chart notes as the comparison standard in cases where these documents were submitted after the time of test order. RESULTS: TRF-reported cancer sites and ages at diagnosis were complete for > 90.0% of proband cancer diagnoses overall, and the completion rate was even higher (> 96.0%) for breast, ovarian, colorectal, and uterine cancers. When reported, these data were accurate on TRFs for > 99.5% of proband cancer sites and > 97.5% of proband ages at diagnosis. Cancer site and age at diagnosis data were also complete on the TRF for the majority of cancers among first- and second-degree relatives. Completeness decreased as relation to the proband became more distant, whereas accuracy remained high across all degrees of relation. CONCLUSION: Data collected as part of cancer genetic risk assessment is completely and accurately reported on TRFs for the majority of probands and their close relatives and is comparable to information directly obtained from clinic notes, particularly for breast and other cancers commonly associated with hereditary cancer syndromes.

Male breast cancer in a multi-gene panel testing cohort: insights and unexpected results.

PURPOSE: Genetic predisposition to male breast cancer (MBC) is not well understood. The aim of this study was to better define the predisposition genes contributing to MBC and the utility of germline multi-gene panel testing (MGPT) for explaining the etiology of MBCs. METHODS: Clinical histories and molecular results were retrospectively reviewed for 715 MBC patients who underwent MGPT from March 2012 to June 2016. RESULTS: The detection rate of MGPT was 18.1% for patients tested for variants in 16 breast cancer susceptibility genes and with no prior BRCA1/2 testing. BRCA2 and CHEK2 were the most frequently mutated genes (11.0 and 4.1% of patients with no prior BRCA1/2 testing, respectively). Pathogenic variants in BRCA2 [odds ratio (OR) = 13.9; p = 1.92 × 10-16], CHEK2 (OR = 3.7; p = 6.24 × 10-24), and PALB2 (OR = 6.6, p = 0.01) were associated with significantly increased risks of MBC. The average age at diagnosis of MBC was similar for patients with (64 years) and without (62 years) pathogenic variants. CHEK2 1100delC carriers had a significantly lower average age of diagnosis (n = 7; 54 years) than all others with pathogenic variants (p = 0.03). No significant differences were observed between history of additional primary cancers (non-breast) and family history of male breast cancer for patients with and without pathogenic variants. However, patients with pathogenic variants in BRCA2 were more likely to have a history of multiple primary breast cancers. CONCLUSION: These data suggest that all MBC patients regardless of age of diagnosis, history of multiple primary cancers, or family history of MBC should be offered MGPT.