Citations

Abstract TP53 aberrations, including mutations and deletion of 17p13, are important adverse prognostic markers in chronic lymphocytic leukemia (CLL) but are less studied in high count monoclonal B-cell lymphocytosis (HCMBL), an asymptomatic pre-malignant stage of CLL. Here we estimated the prevalence and impact of TP53 aberrations in 1,230 newly diagnosed treatment-naïve individuals (849 CLL, 381 HCMBL). We defined TP53 state as: wild-type (no TP53 mutations and normal 17p), single-hit (del(17p) or one TP53 mutation), or multi-hit (TP53 mutation and del(17p), TP53 mutation and loss of heterozygosity, or multiple TP53 mutations). Cox regression was used to estimate hazard ratios (HR) and 95% confidence intervals (CI) for time to first treatment and overall survival by TP53 state. We found 64 (7.5%) CLL patients and 17 (4.5%) HCMBL individuals had TP53 mutations with variant allele fraction >10%. Del(17p) was present in 58 (6.8%) of CLL and 11 (2.9%) of HCMBL cases. Most individuals had wild-type (N=1,128, 91.7%) TP53 state, followed by multi-hit (N=55, 4.5%) and then single-hit (N=47, 3.8%) TP53 state. The risk of shorter time to therapy and death increased with the number of TP53 abnormalities. Compared to wild-type patients, multi-hit patients had 3-fold and single-hit patients had 1.5-fold increased risk of requiring therapy. Multi-hit patients also had 2.9-fold increased risk of death compared to wild-type. These results remained stable after accounting for other known poor prognostic factors. Both TP53 mutations and del(17p) may provide important prognostic information for HCMBL and CLL that would be missed if only one were measured.

Advances in computer science and data-analytic methods are driving a new era in mental health research and application. Artificial intelligence (AI) technologies hold the potential to enhance the assessment, diagnosis, and treatment of people experiencing mental health problems and to increase the reach and impact of mental health care. However, AI applications will not mitigate mental health disparities if they are built from historical data that reflect underlying social biases and inequities. AI models biased against sensitive classes could reinforce and even perpetuate existing inequities if these models create legacies that differentially impact who is diagnosed and treated, and how effectively. The current article reviews the health-equity implications of applying AI to mental health problems, outlines state-of-the-art methods for assessing and mitigating algorithmic bias, and presents a call to action to guide the development of fair-aware AI in psychological science.