Genome sequencing in the Parkinson disease clinic

Background and Objectives: Genetic variants impact both Parkinson disease (PD) risk and manifestations. While genetic information is of potential interest to patients and clinicians, genetic testing is rarely performed during routine PD clinical care. The goal of this study was to perform genome sequencing and examine patient interest in comprehensive genetic testing for PD in 2 academic movement disorder clinics. Methods: In 208 subjects with PD (age=63 years, 67% male), genome sequencing was performed and filtered using a custom panel, including 49 genes associated with PD, parkinsonism, or related disorders, as well as a 90-variant PD genetic risk score. Separately, 231 patients (age=67 years, 63% male) were surveyed on interest in genetic testing at baseline and in response to vignettes covering (i) familial risk of PD (LRRK2); (ii) risk of PD dementia (GBA); (iii) PD genetic risk score; and (iv) secondary, medically-actionable variants (BRCA1). Results: Genome sequencing revealed a LRRK2 variant in 3.4% and a GBA risk variant in 10.1% of our clinical sample. The genetic risk score was normally distributed, identifying 42 subjects with high risk of PD. Medically-actionable findings were discovered in 2 subjects (1%). In our survey, the majority (82%) responded they would share a LRRK2 variant with relatives. Most registered unchanged or increased interest in testing when confronted with potential risk for dementia or medically- actionable findings, and most (75%) expressed interest in learning their PD genetic risk score. Discussion: Our results highlight broad interest in comprehensive genetic testing among patients with PD and may facilitate integration of genome sequencing in clinical practice.

Automatic detection of actionable radiology reports using bidirectional encoder representations from transformers

Background It is essential for radiologists to communicate actionable findings to the referring clinicians reliably. Natural language processing (NLP) has been shown to help identify free-text radiology reports including actionable findings. However, the application of recent deep learning techniques to radiology reports, which can improve the detection performance, has not been thoroughly examined. Moreover, free-text that clinicians input in the ordering form (order information) has seldom been used to identify actionable reports. This study aims to evaluate the benefits of two new approaches: (1) bidirectional encoder representations from transformers (BERT), a recent deep learning architecture in NLP, and (2) using order information in addition to radiology reports. Methods We performed a binary classification to distinguish actionable reports (i.e., radiology reports tagged as actionable in actual radiological practice) from non-actionable ones (those without an actionable tag). 90,923 Japanese radiology reports in our hospital were used, of which 788 (0.87%) were actionable. We evaluated four methods, statistical machine learning with logistic regression (LR) and with gradient boosting decision tree (GBDT), and deep learning with a bidirectional long short-term memory (LSTM) model and a publicly available Japanese BERT model. Each method was used with two different inputs, radiology reports alone and pairs of order information and radiology reports. Thus, eight experiments were conducted to examine the performance. Results Without order information, BERT achieved the highest area under the precision-recall curve (AUPRC) of 0.5138, which showed a statistically significant improvement over LR, GBDT, and LSTM, and the highest area under the receiver operating characteristic curve (AUROC) of 0.9516. Simply coupling the order information with the radiology reports slightly increased the AUPRC of BERT but did not lead to a statistically significant improvement. This may be due to the complexity of clinical decisions made by radiologists. Conclusions BERT was assumed to be useful to detect actionable reports. More sophisticated methods are required to use order information effectively.

The value of long-term angiographic follow-up following Pipeline embolization of intracranial aneurysms

BackgroundFlow diversion of intracranial aneurysms with the Pipeline Embolization Device (PED) is commonly performed, but the value of long-term angiographic follow-up has not been rigorously evaluated. Here we examine the prevalence of actionable findings of aneurysm recurrence and development of in-stent stenosis in a cohort of patients that underwent long-term angiographic follow-up at multiple time points.MethodsAngiographic data from eligible patients were retrospectively assessed for aneurysm occlusion, in-stent stenosis, and aneurysm regrowth or recurrence. Patients were included in this study if they underwent angiographic imaging at 6 months post-treatment and at least one later time point.Results100% (132/132) of aneurysms occluded at 6 months remained occluded at final follow-up. 85.7% (6/7), 56.3% (27/48), and 25% (6/24) of aneurysms with entry remnant, subtotal filling, and total filling, respectively, at 6 months were completely occluded at final follow-up. 98.7% (147/149) of PED constructs that demonstrated no stenosis at 6 months demonstrated no stenosis at final angiography, while 44.4% (8/18) of PED constructs demonstrating in-stent stenosis at 6 months had resolution of stenosis on final angiography.ConclusionsAmong patients who undergo treatment of intracranial aneurysms with PED, the value of long-term angiography in patients demonstrating complete aneurysm occlusion and no in-stent stenosis on 6 month post-treatment angiography is low.

The marginal diagnostic benefit of pancreatic cancer molecular profiling after germline testing.

10513 Background: Germline genetic testing is now universally recommended for patients (pts) with pancreatic ductal adenocarcinoma (PDAC) for purposes of both familial screening and therapeutic guidance. Treatment selection can be further informed by tumor molecular profiling (TMP) to identify targetable somatic alterations in pts with advanced disease, but this is inconsistently applied. Determination of the rate of actionable findings identified with TMP after germline testing, which we term marginal diagnostic benefit, may inform practice patterns and workflow in this patient population. Methods: This retrospective analysis included all pts with PDAC who underwent germline testing and TMP at UCSF over a 4-yr period. Medical records were reviewed for demographics, disease-specific data, and germline testing/TMP clinical reports. Alterations classified as ‘pathogenic’ or ‘likely pathogenic’ were included, and were deemed ‘actionable’ if there was clinical or preclinical evidence of benefit from targeted therapy in any cancer, as previously described. Results: From 1/2016-1/2020, 144/738 (20%) UCSF pts with PDAC completed both germline testing and TMP. Germline testing identified actionable pathogenic alterations in 10 (7%). TMP confirmed 8/10 of these alterations and identified 3 additional therapeutic targets. Among the 134 pts without actionable germline findings, TMP identified 45 new therapeutic targets in 41 (31%) pts, increasing the overall rate of actionable findings from 7% to 35%. Most (35/58, 60%) actionable alterations involved genes associated with the Homologous Recombination DNA Damage Repair (HR-DDR) pathway (Table). 80% of pts with HR-DDR pathway alterations (9/10 germline, 19/25 somatic) received platinum-based chemotherapy. Four pts were treated with targeted therapy based on test results: PARP-inhibitor (n = 2, germline BRCA1 and PALB2 mutations), PARP-inhibitor + ATR inhibitor (n = 1, somatic ARID1A mutation) and mTOR inhibitor (n = 1, somatic STK11 deletion). Conclusions: In this analysis, PDAC TMP after germline testing increased the detection of actionable alterations (the marginal diagnostic benefit) by five-fold. As more somatic tumor alterations become actionable with the development of targeted therapeutics, TMP is a necessary complement to germline testing to fully inform personalized treatment decisions for all pts with PDAC.[Table: see text]

Scientific Citizenship’s Youngest Domain: Function Creep in Norway’s Newborn Screening Programme

Newborn screening (NBS) for inborn errors of metabolism and other serious conditions with onset during infancy is a widespread public health initiative. Like other screening programmes, it aims to discover and treat a disease before effects manifest themselves. Recently, there have been two prominent changes in NBS: a substantial increase in the number of conditions screened for and growing attention to secondary use of residual newborn blood spots. Here, we analyse how this latter change has transpired in Norway. In 2018, Norway’s parliament sanctioned the secondary use of NBS samples for epidemiological research unrelated to NBS. This broadened the programme’s scope, co-opting it for research purposes, making samples available for inclusion in Norway’s biobanking strategy. We argue that this transformation is a case of function creep, whereby the function of screening samples is expanded to serve purposes other than helping newborns. The process provided only minimal involvement from ordinary citizens, but it transformed screened infants into potential scientific citizens. Henceforth, all future generations of Norwegians must choose to stay in or opt out of biobank research when they turn sixteen. Additionally, consenting to this research may occasion a second form of function creep, as ‘actionable findings’ are fed back to participants.