Literature Review: January 2023
AI powered variant assessment, novel predictive biomarkers for cancer, quantum computing, microbiome dietary interventions, disruptiveness in science, and health-related social needs
Artificial intelligence-based recognition for variant pathogenicity of BRCA1 using AlphaFold2-predicted structures
In this study, the researchers developed a deep learning model called variant effect recognition network for BRCA1 (vERnet-B) to predict the clinical significance of genetic variants called variants of uncertain significance (VUS) in the BRCT domain of the BRCA1 gene. Leveraging AlphaFold2, the vERnet-B model incorporated protein tertiary structures to consider features associated with the pathogenicity of VUS. The vERnet-B web server is available for use at the provided URL.
Importance: Previous methods for predicting the effects of VUS only used primary amino acid sequences and ignored the important three-dimensional structure of proteins, which is a more comprehensive approach closer to protein function.
Prognostic and Predictive Value of Immune-Related Gene Expression Signatures vs Tumor-Infiltrating Lymphocytes in Early-Stage ERBB2/HER2-Positive Breast Cancer: A Correlative Analysis of the CALGB 40601 and PAMELA Trials
This study examined whether tumor-infiltrating lymphocytes (TILs) and immune-related gene expression signatures can predict the pathologic complete response (pCR) and event-free survival (EFS) in patients with early-stage ERBB2/HER2-positive breast cancer. The study analyzed data from two clinical trials involving 305 and 151 patients and found that TILs were associated with pCR. In addition, 36 immune signatures were significantly associated with higher pCR rates, and seven of these signatures outperformed TILs for predicting pCR. When both TILs and gene expression were available, the prognostic value of immune-related signatures appeared to be superior to TILs in predicting EFS.
Importance: This study identified potential predictive markers for response to neoadjuvant therapy in patients with breast cancer, providing clues for the targeting of treatment to those most likely to benefit and sparing others from the potential adverse effects of treatment.
Factoring integers with sublinear resources on a superconducting quantum processor
Shor's algorithm is a method for breaking information security systems that rely on public key cryptography. It has been difficult to use Shor's algorithm to break the widely used RSA-2048 scheme because it requires millions of physical qubits, which are not currently available. However, this research presents a new quantum algorithm for integer factorization that only requires a sublinear number of qubits O(logN/loglogN) and can successfully factor integers up to 48 bits using only a few qubits.
Importance: The study is sound in theory but likely out of practical reach. However, if validated, the methods proposed by this study can theoretically break the most commonly used online encryptions utilizing only 372 qubits, which is not much in the quantum computing world. Last year, IBM unveiled its quantum processor with 433 qubits and detailed its Quantum System Two, which will integrate multiple quantum processors into a single system, expected to launch by the end of 2023.
Microbiota-dependent proteolysis of gluten subverts diet-mediated protection against type 1 diabetes
This study suggests that a diet that includes hydrolyzed casein as a protein source can protect against the development of type 1 diabetes in mice. The addition of gluten to the diet appears to facilitate the development of autoimmune disease, but this effect is dependent on the presence of certain bacteria that can digest gluten and generate immune-activating peptides. Mice that were colonized with these bacteria on a diet that included both hydrolyzed casein and gluten were resistant to the development of type 1 diabetes.
Importance: These findings provide insights into the potential for dietary interventions to protect against autoimmune diseases in humans and highlight the importance of studying the interactions between diet, the microbiome, and autoimmune diseases.
Analysis of Tumor Mutational Burden, Progression-Free Survival, and Local-Regional Control in Patents with Locally Advanced Non–Small Cell Lung Cancer Treated With Chemoradiation and Durvalumab
This study evaluated the effectiveness of the addition of the drug durvalumab to chemotherapy and radiation therapy for the treatment of locally advanced non-small cell lung cancer (NSCLC). The study included 81 patients with NSCLC who were treated with this combination therapy between 2013 and 2020. The primary outcomes measured were disease control and survival rates. The patients were divided into two groups based on the level of a biomarker called tumor mutational burden (TMB) in their tumors. The group with higher levels of TMB had better disease control and survival rates compared to the group with lower levels of TMB. More specifically, patients with high levels oTMB appeared to have a 5-times lower risk of local recurrence (LRF) and improved progression-free survival (PFS) compared to patients with low levels of TMB. The researchers also found that tumors with certain genetic alterations in the KEAP1/NFE2L2 pathway had an increased risk of LRF.
Importance: The results of this study suggest that measuring TMB may help tailor treatment and improve outcomes for patients with locally advanced, inoperable NSCLC. TMB is a “complex” biomarker, which has supported tissue-agonstic indications in the past in advanced stage disease. TMB has previously been shown to be a useful biomarker for guiding treatment decisions in advanced stages of cancer. In 2020, the US Food and Drug Administration approved the use of the immune checkpoint inhibitor pembrolizumab (Keytruda) in conjunction with the companion diagnostic assay FoundationOneCDx for the treatment of patients with advanced solid tumors with high TMB levels that have not responded to other treatments. It will be interesting to see if TMB can be effectively used as a biomarker in earlier stages of various cancers in the future.
Papers and patents are becoming less disruptive over time
The number of science and technology research papers published in recent decades has increased, but the 'disruptiveness' of those papers has decreased, according to an analysis published in Nature. The analysis examined 45 million manuscripts and 3.9 million patents, measuring the “CD index” of disruptiveness and found that the average CD index declined by more than 90% between 1945 and 2010 for research manuscripts and by 78% from 1980 to 2010 for patents. The decline in disruptiveness was seen across all fields and patent types, with research in the 2010s more likely to focus on incremental progress, using terms such as 'improve' or 'enhance' rather than 'create' or 'discover'.
Importance: Although this study has a few limitations, including the use of the CD index as an appropriate measure for disruptiveness, it does appear that the scientific enterprise has been prioritizing incremental progress over disruptive or groundbreaking discoveries. This trend seems to have particularly affected the field of biomedicine, where a shift towards incremental safe bets have been the norm. However, I believe there is currently a renewed sense of disruptive experimentalism in biomedicine, driven by the convergence of biology, technology, and data science. This convergence is paving the way for exciting developments and discoveries in the future. It is worth noting that incremental progress and disruptive experimentation are not necessarily mutually exclusive, and it’s striking the right balance between the two that can be most beneficial for the advancement of science.
Hospitals and Health Equity — Translating Measurement into Action
This perspective suggests that the US healthcare system may be lacking in equity compared to other high-income countries. Whereas US hospitals have the potential to address health-related social needs (HRSNs), such as food insecurity and housing instability, only a quarter currently screen for these needs. The Centers for Medicare and Medicaid Services (CMS) has recently introduced three measures to address this issue in the Hospital Inpatient Quality Reporting program. The measures require hospitals to report on their commitment to health equity and their efforts to address health disparities, and to screen adult patients for HRSNs at the time of admission and report the percentage who screen positive for these needs. Hospitals will need to decide on a screening tool, consider how to collect data, integrate screening results into electronic health records, and identify resources and partners to help address identified needs.
Importance: CMS’s new framework provides potentially new use cases for health technology solutions including digital health tools. For example, digital health tools can be used to administer HRSN screening tools to patients using tablet-based screening approaches, which could reduce staff burden and potentially elicit higher rates of disclosure than face-to-face approaches. In addition, electronic health records (EHRs) can be used to store and analyze data on HRSNs, allowing hospitals to track the prevalence of these needs among their patient population and identify patterns and trends. This information can be used to inform the development and implementation of targeted interventions to address HRSNs. Digital health tools can also be used to connect patients with resources and services to address their identified HRSNs. For example, hospitals could use telemedicine to connect patients with social workers or other specialists who can help them access resources such as housing assistance or transportation.