The future of research in hematology: Integration of conventional studies with real-world data and artificial intelligence

Transformative Role of Artificial Intelligence in Reporting Haematology Cases: A Case Report

Artificial intelligence (AI) is transforming haematology reporting by improving accuracy, standardisation, and speed, addressing the need for timely and precise diagnostics. This study explores the use of the AI100 (SigTuple Technologies Private Limited, Bangalore, India) automated machine, a smart robotic microscope designed to automate the microscopic analysis of peripheral blood smears. Through the analysis of four haematology cases, this study demonstrates how AI technology facilitates efficient cell identification, enhances risk stratification, enables early detection of abnormalities, and accelerates diagnostic turnaround times. These advancements support pathologists in delivering improved patient care by augmenting traditional diagnostic methods. While AI can streamline processes and increase diagnostic accuracy, it is intended to complement, rather than replace, the expertise and judgement of skilled pathologists.

Prognostic and Predictive Models in Myelofibrosis

PURPOSE OF REVIEW

Myelofibrosis (MF) includes prefibrotic primary MF (pre-PMF), overt-PMF and secondary MF (SMF). Median overall survival (OS) of pre-PMF, overt-PMF and SMF patients is around 14 years, seven and nine years, respectively. Main causes of mortality are non-clonal progression and transformation into blast phase.

RECENT FINDINGS

Discoveries on the impact of the biological architecture on OS have led to the design of integrated scores to predict survival in PMF. For SMF, OS estimates should be calculated by the specific MYSEC-PM (MYelofibrosis SECondary-prognostic model). Information on the prognostic role of the molecular landscape in SMF is accumulating. Crucial treatment decisions for MF patients could be now supported by multivariable predictive algorithms. OS should become a relevant endpoint of clinical trials. Prognostic models guide prediction of OS and treatment planning in MF, therefore, their timely application is critical in the personalized approach of MF patients.

Artificial Intelligence in Medical Affairs: A New Paradigm with Novel Opportunities

The advent of artificial intelligence (AI) revolutionizes the ways of working in many areas of business and life science. In Medical Affairs (MA) departments of the pharmaceutical industry AI holds great potential for positively influencing the medical mission of identifying and addressing unmet medical needs and care gaps, and fostering solutions that improve the egalitarian and unbiased access of patients to treatments worldwide. Given the essential position of MA in corporate interactions with various healthcare stakeholders, AI offers broad possibilities to support strategic decision-making and to pioneer novel approaches in medical stakeholder interactions. By analyzing data derived from the healthcare environment and by streamlining operations in medical content generation, AI advances data-based prioritization and strategy execution. In this review, we discuss promising AI-based solutions in MA that support the effective use of heterogenous information from observations of the healthcare environment, the enhancement of medical education, and the analysis of real-world data. For a successful implementation of such solutions, specific considerations partly unique to healthcare must be taken care of, for example, transparency, data privacy, healthcare regulations, and in predictive applications, explainability.

SOHO State of the Art Updates and Next Questions | Diagnosis, Outcomes, and Management of Prefibrotic Myelofibrosis

Prefibrotic primary myelofibrosis (prefibrotic PMF) is a myeloproliferative neoplasm with distinct characteristics comprising histopathological and clinico-biological parameters. It is classified as a subtype of primary myelofibrosis. In clinical practice, it is essential to correctly distinguish prefibrotic PMF from essential thrombocythemia especially but also overt PMF besides other myeloid neoplasms. Risk stratification and survival outcomes for prefibrotic PMF are worse than that of ET but better than that of overt PMF. Rates of progression to overt PMF and blast phase disease are also higher for prefibrotic PMF than ET. In this review we first discuss the historical context to the evolution of prefibrotic PMF as an entity, its presenting features and diagnostic criteria. We emphasize the differences between prefibrotic PMF, ET, and overt PMF with regards to presenting features and disease outcomes including thrombohemorrhagic events and progression to fibrotic and blast phase disease. Next, we discuss the risk stratification models and contextualize these in the setting of clinical management. We share our view of personalizing treatment to address unique patient needs in the context of currently available management options. Lastly, we discuss areas of critical need in clinical research and speculate on the possibility of future disease course modifying therapies in prefibrotic PMF.

Opioids and Immunosuppression: Clinical Evidence, Mechanisms of Action, and Potential Therapies

Background

In addition to the more well-known adverse effects of opioids, such as constipation, mounting evidence supports underlying immunosuppressive effects as well.

Methods

In this study, we provide a narrative review of preclinical and clinical evidence of opioid suppression of the immune system as well as possible considerations for therapies.

Results

In vitro and animal studies have shown clear effects of opioids on inflammatory cytokine expression, immune cell activity, and pathogen susceptibility. Observational data in humans have so far supported preclinical findings, with multiple reports of increased rates of infections in various settings of opioid use. However, the extent to which this risk is due to the impact of opioids on the immune system compared with other risk factors associated with opioid use remains uncertain. Considering the data showing immunosuppression and increased risk of infection with opioid use, measures are needed to mitigate this risk in patients who require ongoing treatment with opioids. In preclinical studies, administration of opioid receptor antagonists blocked the immunomodulatory effects of opioids.

Conclusions

As selective antagonists of peripheral opioid receptors, peripherally acting mu-opioid receptor (MOR) antagonists may be able to protect against immune impairment while still allowing for opioid analgesia. Future research is warranted to further investigate the relationship between opioids and infection risk as well as the potential application of peripherally acting MOR antagonists to counteract these risks.

The role of randomized controlled trials, registries, observational databases in evaluating new interventions

Approaches to comparing safety and efficacy of interventions include analyzing data from randomized controlled trials (RCTs), registries and observational databases (ODBs). RCTs are regarded as the gold standard but data from such trials are sometimes unavailable because a disease is uncommon, because the intervention is uncommon, because of structural limitations or because randomization cannot be done for practical or (seemingly) ethical reasons. There are many examples of an unproved intervention being so widely-believed to be effective that clinical trialists and potential subjects decline randomization. Often, when a RCT is finally done the intervention is proved ineffective or even harmful. These situations are termed medical reversals and are not uncommon [1,2]. There is also the dilemma of when seemingly similar RCTs report discordant conclisions Data from high-quality registries, especially ODBs can be used when data from RCTs are unavailable but also have limitations. Biases and confounding co-variates may be unknown, difficult or impossible to identify and/or difficult to adjust for adequately. However, ODBs sometimes have large numbers of diverse subjects and often give answers more useful to clinicians than RCTs. Side-by-side comparisons suggest analyses from high-quality ODBs often give similar conclusions from high quality RCTs. Meta-analyses combining data from RCTs, registries and ODBs are sometimes appropriate. We suggest increased use of registries and ODBs to compare efficacy of interventions.

Surveying Hematologists' Perceptions and Readiness to Embrace Artificial Intelligence in Diagnosis and Treatment Decision-Making

AIM

This study aims to explore the critical dimension of assessing the perceptions and readiness of hematologists to embrace artificial intelligence (AI) technologies in their diagnostic and treatment decision-making processes.

METHODS

This study used a cross-sectional design for collecting data related to the perceptions and readiness of hematologists using a validated online questionnaire-based survey. Both hematologists (MD) and postgraduate MD students in hematology were included in the study. A total of 188 participants, including 35 hematologists (MD) and 153 MD hematology students, completed the survey.

RESULTS

Major challenges include "AI's level of autonomy" and "the complexity in the field of medicine." Major barriers and risks identified include "lack of trust," "management's level of understanding," "dehumanization of healthcare," and "reduction in physicians' skills." Statistically significant differences in perceptions of benefits including resources (p=0.0326, p<0.05) and knowledge (p=0.0262, p<0.05) were observed between genders. Older physicians were observed to be more concerned about the use of AI compared to younger physicians (p<0.05).

CONCLUSION

While AI use in hematology diagnosis and treatment decision-making is positively perceived, issues such as lack of trust, transparency, regulations, and poor AI awareness can affect the adoption of AI.

A multimodal analysis of genomic and RNA splicing features in myeloid malignancies

RNA splicing dysfunctions are more widespread than what is believed by only estimating the effects resulting by splicing factor mutations (SF^MT) in myeloid neoplasia (MN). The genetic complexity of MN is amenable to machine learning (ML) strategies. We applied an integrative ML approach to identify co-varying features by combining genomic lesions (mutations, deletions, and copy number), exon-inclusion ratio as measure of RNA splicing (percent spliced in, PSI), and gene expression (GE) of 1,258 MN and 63 normal controls. We identified 15 clusters based on mutations, GE, and PSI. Different PSI levels were present at various extents regardless of SF^MT suggesting that changes in RNA splicing were not strictly related to SF^MT. Combination of PSI and GE further distinguished the features and identified PSI similarities and differences, common pathways, and expression signatures across clusters. Thus, multimodal features can resolve the complex architecture of MN and help identifying convergent molecular and transcriptomic pathways amenable to therapies.

Agents contributing to secondary immunodeficiency development in patients with multiple myeloma, chronic lymphocytic leukemia and non-Hodgkin lymphoma: A systematic literature review

Introduction

Patients with hematological malignancies (HMs), like chronic lymphocytic leukemia (CLL), multiple myeloma (MM), and non-Hodgkin lymphoma (NHL), have a high risk of secondary immunodeficiency (SID), SID-related infections, and mortality. Here, we report the results of a systematic literature review on the potential association of various cancer regimens with infection rates, neutropenia, lymphocytopenia, or hypogammaglobulinemia, indicative of SID.

Methods

A systematic literature search was performed in 03/2022 using PubMed to search for clinical trials that mentioned in the title and/or abstract selected cancer (CLL, MM, or NHL) treatments covering 12 classes of drugs, including B-lineage monoclonal antibodies, CAR T therapies, proteasome inhibitors, kinase inhibitors, immunomodulators, antimetabolites, anti-tumor antibiotics, alkylating agents, Bcl-2 antagonists, histone deacetylase inhibitors, vinca alkaloids, and selective inhibitors of nuclear export. To be included, a publication had to report at least one of the following: percentages of patients with any grade and/or grade ≥3 infections, any grade and/or grade ≥3 neutropenia, or hypogammaglobulinemia. From the relevant publications, the percentages of patients with lymphocytopenia and specific types of infection (fungal, viral, bacterial, respiratory [upper or lower respiratory tract], bronchitis, pneumonia, urinary tract infection, skin, gastrointestinal, and sepsis) were collected.

Results

Of 89 relevant studies, 17, 38, and 34 included patients with CLL, MM, and NHL, respectively. In CLL, MM, and NHL, any grade infections were seen in 51.3%, 35.9% and 31.1% of patients, and any grade neutropenia in 36.3%, 36.4%, and 35.4% of patients, respectively. The highest proportion of patients with grade ≥3 infections across classes of drugs were: 41.0% in patients with MM treated with a B-lineage monoclonal antibody combination; and 29.9% and 38.0% of patients with CLL and NHL treated with a kinase inhibitor combination, respectively. In the limited studies, the mean percentage of patients with lymphocytopenia was 1.9%, 11.9%, and 38.6% in CLL, MM, and NHL, respectively. Two studies reported the proportion of patients with hypogammaglobulinemia: 0-15.3% in CLL and 5.9% in NHL (no studies reported hypogammaglobulinemia in MM).

Conclusion

This review highlights cancer treatments contributing to infections and neutropenia, potentially related to SID, and shows underreporting of hypogammaglobulinemia and lymphocytopenia before and during HM therapies.

Towards a Personalized Definition of Prognosis in Philadelphia-Negative Myeloproliferative Neoplasms

Purpose of Review

Philadelphia-negative myeloproliferative neoplasms (MPNs) include polycythemia vera (PV), essential thrombocythemia (ET), prefibrotic (pre-), and overt-primary myelofibrosis (primary MF, PMF). PV and ET could evolve into secondary MF (SMF), whose early diagnosis relies on monitoring signs of possible progression. All MPNs have a risk of blast phase (BP), that is associated with a very dismal outcome. Overall survival (OS) is different among MPNs, and disease-specific prognostic scores should be applied for a correct clinical management. In this review, an overview of current prognostic scores in MPNs will be provided.

Recent Findings

The biological complexity of MPNs and its role on the trajectory of disease outcome have led to the design of integrated prognostic models that are nowadays of common use in PMF patients. As for PV and ET, splicing gene mutations could have a detrimental role, but with the limit of the not routinary recommended application of extensive molecular analysis in these diseases. SMF is recognized as a distinct entity compared to PMF, and OS estimates should be calculated by the MYSEC-PM (Myelofibrosis SECondary-prognostic model). Both in PMF and SMF, decisions as selection of patients potentially candidates to allogenic stem cell transplant or that could benefit from an early shift from standard treatment are based not only on conventional prognostic scores, but also on multivariable algorithms.

Summary

The expanding landscape of risk prediction for OS, evolution to BP, and SMF progression from PV/ET informs personalized approach to the management of patients affected by MPNs.