Evaluating the performance of large language models in haematopoietic stem cell transplantation decision-making.

In a first-of-its-kind study, we assessed the capabilities of large language models (LLMs) in making complex decisions in haematopoietic stem cell transplantation. The evaluation was conducted not only for Generative Pre-trained Transformer 4 (GPT-4) but also conducted on other artificial intelligence models: PaLm 2 and Llama-2. Using detailed haematological histories that include both clinical, molecular and donor data, we conducted a triple-blind survey to compare LLMs to haematology residents. We found that residents significantly outperformed LLMs (p = 0.02), particularly in transplant eligibility assessment (p = 0.01). Our triple-blind methodology aimed to mitigate potential biases in evaluating LLMs and revealed both their promise and limitations in deciphering complex haematological clinical scenarios.

Humoral and cellular immune response in patients with hematological disorders after two doses of BNT162b2 mRNA COVID-19 vaccine: A single-center prospective observational study (NCT05074706).

Hematological patients at higher risk of severe COVID-19 were excluded from the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) vaccine trials. In this single-center observational prospective study (NCT05074706), we evaluate immune response in the hematological patients followed at the Hematological Division of San Gerardo Hospital, Monza (Italy) deemed to be severely immunosuppressed after vaccination with two doses of the BNT162b2 vaccine. Anti-SARS-CoV-2 immunoglobulin G titers above the cutoff value of 33.8 BAU/ml were detected in 303 (80.2%) out of the 378 patients enrolled. Patients with lymphoproliferative disorders had a significant lower probability of immunization (43.2% vs. 88.4%, p < 0.001). Patients treated with anti-CD20 showed a significantly lower probability of immunization compared to all other treatments (21.4%, p < 0.0001). Among 69 patients who failed seroconversion, 15 patients (22.7%) showed a positive T-cell response. Patients previously treated with anti-CD20 were 2.4 times more likely to test positive for T-cell responses (p = 0.014). Within a follow-up of 9 months from the second COVID-19 vaccination, symptomatic SARS-CoV-2 infections were reported by 20 patients (5.3%) and four of them required hospitalization. Successful serological or T-cell-mediated immunization conferred protection from symptomatic COVID-19. Patients treated with anti-CD20 who were not seroconverted after vaccination might still be protected from COVID-19 due to the T-cell immune response.