ReFlections: RGA’s Global Medical Newsletter – October 2026 ReFlections: RGA’s Global Medical Newsletter

Also in this issue

Letter from the Editor

What's New? AI and heart disease, cell therapy for GVHD, and the latest research updates

Underwriting Updates GUM’s guidelines and information on a variety of disorders

Case ReView A case on leukopenia and neutropenia

Claims Updates Discover new Pathfinder training and tools to elevate claims excellence

Expert Q and A

RGA's Global Medical Newsletter

The Box Is Open: AI in medicine, insurance, and the case for informed optimism by Dr. Georgiana Willwerth The Next Era of Cancer Care: Earlier detection, more precise treatment, better outcomes by Dr. Stephen Falk Older-Age Underwriting: Rethinking mortality risk by Hilary Henly and Steve Simkin

Featured Articles

BACK

Jennie Calder Brown Executive Director of Global Claims Philosophy and Education jcalderbrown@rgare.com

Hilary Henly Global Medical Researcher hhenly@rgare.com

Assistant Editors

The Box Is Open: AI in medicine, insurance, and the case for informed optimism

Longer Life Foundation

The Box Is Open: AI in medicine, insurance, and the case for informed optimism

AI’s clearest value today is practical augmentation: Reducing administrative burden and helping clinicians and underwriters focus more on judgment. Its most meaningful medical promise is earlier insight: AI-powered triage tools, AI-supported imaging, and AI-ECG can surface risk before it becomes obvious to the human eye. A key emerging risk is false confidence: Polished, persuasive outputs may still be incomplete, biased, manipulated, or wrong. AI can scale human bias as well as data bias: Motivated reasoning, imperfect science, and flawed interpretation do not disappear when an algorithm is involved. Responsible optimism depends on governance: Validation, transparency, privacy, and human accountability must remain central as adoption expands.

Key takeaways

AI in medicine has transitioned from possibility to practice. Three themes define this moment: Where real value is emerging The developing risks that warrant attention The need to keep AI’s promise anchored in human judgment as the technology evolves Artificial intelligence is no longer a future concept in medicine. It is already embedded in workflows, beginning to reshape how disease is detected, how clinicians spend their time, and how medical information is interpreted and managed in underwriting and claims. That makes this more than a technology story. It is about evidence, trust, accountability, and professional judgment. Insurance medicine is a well-established practice, and these two domains – insurance and medical science – are converging around the same central question: “How do we allow AI to inform outcomes without quietly allowing it to make decisions for us?” To make sense of this moment, it helps to recall an old story. In the myth of Pandora, her curiosity leads her to open the box, releasing hardship into the world; but one thing remains inside – hope. That image is useful for AI. The box is already open. The technology is moving quickly, and some of its consequences are still uncertain. But uncertainty does not require pessimism. It necessitates disciplined optimism: the ability to recognize what AI is genuinely achieving, to name its most serious risks clearly, and to build the guardrails that keep hope inside.

Introduction

References

About the author

Georgiana Willwerth, MD, DBIM, is Vice President and Medical Director at RGA and a member of RGA’s Global Medical team. Dr. Willwerth is board certified in Insurance Medicine by the American Academy of Insurance Medicine (AAIM) and specialized in internal medicine, nephrology, and ultrasonography. She is a past president and scientific chair of the Canadian Life Insurance Medical Officers Association (CLIMOA) and chair of the Board of Insurance Medicine. She is a frequent industry presenter and contributor to industry publications. Her particular interests are global morbidity and mortality trends arising from infectious disease threats, and the applications of mortality analytics and generative AI in underwriting.

Dr. Georgiana Willwerth

Vice President and Medical Director, Global Medical

Georgiana.Willwerth@rgare.com

AI can surface signals the human eye may miss. But earlier detection is not diagnostic certainty.

Rate this article

To share your thoughts about the article, or to suggest topics we could cover in future issues, please click here to send us an email.

AI is changing the medical landscape that underwriting and claims depend on. New AI-derived signals may appear earlier than symptoms of disease and sometimes before clinical consensus is mature. This creates opportunity, but also ambiguity. For insurers, the near-term value of AI is clearest in workflow efficiency, evidence summarization, and decision support – not in replacing expert judgment. The priority must be disciplined adoption: validate the tools, understand the data, monitor for bias, and keep humans accountable for the decisions that follow.

Abstract: Why this matters for insurers

1. https://www.utoronto.ca/news/ai-tool-reduces-risk-unexpected-hospital-deaths-26-cent-study2. https://temertymedicine.utoronto.ca/news/study-chartwatch-ai-tool-reduces-risk-unanticipated-death-hospitalized-patients-26-cent3. https://www.michiganmedicine.org/health-lab/ai-model-can-read-and-diagnose-brain-mri-seconds4. https://emedicine.medscape.com/article/346529-overview5. https://www.nature.com/articles/s41591-026-04454-y6. https://pmc.ncbi.nlm.nih.gov/articles/PMC12808761/7. https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2842987

BACK

1. The biggest success so far: Enhancing care through earlier risk stratification and improved efficiency in healthcare delivery

If the hype is removed, AI’s clearest success in medicine so far is twofold. First, it is reducing administrative burden through transcription, summarization, and workflow support. Those may sound like modest wins, but they matter because time is one of healthcare’s scarcest resources. Second, and more important, AI is beginning to detect disease patterns earlier than human eyes. This is where the impact becomes tangible. A particularly powerful example is CHARTWatch, an AI early-warning system developed at Unity Health Toronto.1 CHARTWatch illustrates one of AI's most clinically meaningful roles: earlier detection of patient deterioration. In a real-world evaluation on the general internal medicine ward at St. Michael’s Hospital, 482 patients became high-risk during the 19-month CHARTWatch intervention period, compared with 1,656 patients during the 43-month pre-intervention period.2 The system monitored patients in real time, flagged those at high risk of unexpected death or ICU transfer, and alerted clinicians to intervene sooner. Its implementation was associated with a 26% reduction in unanticipated mortality, with fewer non-palliative deaths in the CHARTWatch group than in the pre-intervention group (1.6% versus 2.1%). That matters because it moves AI beyond theory into real-world clinical benefits: helping teams recognize deterioration earlier, escalate care sooner, and intervene in ways that appear to save lives. The same pattern emerges across imaging and data-driven, signal-based medicine. A recent neuroimaging example makes that promise especially concrete. As the familiar medical phrase reminds us, “Time is brain” – during a stroke, an estimated 1.9 million neurons are lost with each minute that passes without treatment. Researchers at the University of Michigan developed Prima, an AI system that can interpret brain MRI scans in seconds and rapidly flag urgent findings such as stroke and brain hemorrhage. In testing, it achieved up to 97.5% accuracy across more than 50 radiologic diagnoses. In neuroimaging – where time is critical – that combination of speed, accuracy, and triage support illustrates one of AI’s most meaningful successes so far: helping clinicians recognize life-threatening disease earlier, streamline workflow, and potentially improve outcomes.3 AI-supported mammography has improved cancer detection while reducing radiologist workload.4In Sweden’s MASAI trial of more than 105,000 women, AI-supported mammography detected 338 breast cancers versus 262 with standard double-reading, the conventional approach in which two radiologists independently review each mammogram – a 29% increase – while reducing radiologist workload by 44%, offering a strong real-world example of how AI can improve both early cancer detection and efficiency. In cardiology, AI is extending the diagnostic reach of routine testing: AI-ECG can detect early signatures of atrial fibrillation, ventricular dysfunction, and other cardiac diseases, while AI-enhanced echocardiography is improving recognition of structural and functional abnormalities.5,6 Much of this work is being led by highly reputable institutions, including major academic centers such as the Mayo Clinic, which lends credibility to the field’s rapid progress. Even though some AI-cardiology tools have entered routine clinical use, broader implementation remains uneven. These developments suggest that AI’s greatest medical contribution may be its ability to surface weak signals early, before they become visible to the naked eye or actionable under traditional thresholds. Taken together, these examples show where AI’s biggest medical success lies today – not as a universal remedy for a strained healthcare system but as tools that ease administrative burden, improve efficiency, and surface disease earlier in areas such as radiology, diagnostics, and cardiology. The momentum is real and increasingly international. More than 1,250 AI-enabled medical devices have been authorized by the FDA, Health Canada has approved a growing number of mainly imaging- and diagnostics-focused systems, and Asia Pacific regulators are increasingly applying digital-health and software-as-medical-device frameworks to AI-enabled tools, although approval data remain country-specific, rather than regionally consolidated. Adoption remains more limited than headlines suggest. Many tools are still concentrated in pilot programs, validation studies, and selected academic centers, with barriers including high implementation cost, limited reimbursement, uneven adoption, and gaps in prospective clinical outcome data. For now, these systems are best understood as promising and complementary, rather than stand-alone decision tools. In insurance settings especially, their outputs remain one piece of the puzzle, rather than a basis for independent underwriting or claims decisions. Insurance has used advanced analytics for decades, with modern AI expanding its role across the value chain. For underwriting, AI tools create both opportunity and complexity. AI can digest fragmented evidence, summarize records, and extract relevant findings, reducing manual review burden and allowing underwriters to focus more on judgment versus data assembly. Predictive AI adds value by improving risk stratification, helping to identify patterns that may not be apparent through traditional review. Models introduce new signals – sometimes emerging earlier than symptoms of disease or before clear clinical consensus exists – which should be interpreted cautiously. When available, AI-derived insights represent one additional input alongside clinical information and human expertise, not a stand-alone basis for decision-making. Their safe use depends on appropriate governance, training, and clear guidelines for interpretation and follow-up, ensuring that these signals are treated as prompts for further evaluation, rather than automatic rating allocation.

2. A key risk: False confidence in systems that sound convincing but remain vulnerable, biased, or opaque

Many risks around AI deserve attention, among them privacy, bias, liability, data governance, and gradual erosion of independent professional judgment. But a key risk is subtler and potentially more dangerous: false confidence. AI outputs often sound polished, efficient, and authoritative. That style can create a powerful illusion of reliability, especially when the system is flawed in ways that are difficult to detect. In medicine and underwriting alike, that is hazardous because these are fields where authoritative errors can cause harm. This concern appears in several forms. a) The “black box” dilemma High-performing models can produce useful outputs without offering a defensible explanation of how they were derived. Clinicians and underwriters are trained to ask why, but many deep-learning models offer limited answers – an important limitation in domains where decisions require explainability, defensibility, and accountability. In a profession grounded in evidence, models that learn autonomously yet offer limited insight into their reasoning can generate mistrust. b) Hallucination or manipulation in generative systems A 2024 quality-improvement study in The Lancet Digital Health illustrates this risk. In a simulated clinical messaging setting, medical large language models could be highly vulnerable to prompt-injection attacks, including fabricated evidence designed to induce unsafe recommendations, some carrying a risk of serious harm.7 c) Bias compounds the problem Bias is not only in data – it is in people. Science itself is rarely perfect or final; evidence evolves, datasets are incomplete, and human interpretation is never entirely free of cognitive bias. AI learns from historical data, and history is not neutral – but bias is not only in the data; it is also in how humans interpret it. Even when data is intended to be accurate, we still bring motivated reasoning and other universal human biases to how evidence is collected, interpreted, and applied. Motivated reasoning is the tendency to interpret evidence in a way that aligns with prior beliefs, rather than objectively evaluating it. AI does not erase those distortions – it can perpetuate and proliferate them at scale. That is why expertise remains essential: Professionals must be able to challenge conclusions, judge the quality of the underlying data, and resist treating AI outputs as absolute truth. In underwriting and claims, the danger is not only inaccurate prediction. It is also the temptation to convert a statistically interesting signal into a pricing or underwriting action before the evidence is mature and externally validated, calibrated, and understood across relevant populations. That is why technical fluency matters. The real risk is not AI alone; it is using AI without the expertise to challenge it. Professionals must be able to ask basic but essential questions: What was the model trained on? In whom has it been validated? Does performance hold across subgroups? Is this tool diagnostic, predictive, or merely assistive? What happens when it fails? Without that discipline, AI can quietly displace judgment, rather than strengthen it.

3. Keeping hope inside the box

Pandora’s story offers a useful perspective acknowledging that transformative discoveries can introduce risk and uncertainty while still leaving space for something meaningful and enduring. In the case of AI, hope lies in the possibility that these tools will help clinicians detect disease earlier, accelerate drug discovery, improve healthcare access and delivery, and allow experts to spend more time where human presence matters most. In underwriting and claims, hope lies in augmentation, rather than replacement. AI can assemble evidence, identify patterns, improve workflow efficiency, and support consistency. But it should not become an invisible decision-maker. The right model is supervised collaboration: AI acting within guardrails, under human authority, and with transparent escalation when complexity, ambiguity, novelty, or ethical sensitivity arise. Maintaining optimism requires a few practical commitments: start with use cases that solve real problems; demand validation before escalation; distinguish assistive tools from predictive tools; preserve privacy, transparency, and accountability; and continuously educate the professionals expected to use these systems. Above all, humans must remain responsible for the outputs, decisions, and consequences. Hope survives not because the technology is harmless but because we choose to govern it wisely.

ConclusionThe central AI story in medicine and underwriting can be distilled into three messages. First, one of AI’s clearest successes so far is practical augmentation: improving efficiency, supporting earlier risk recognition, and helping clinicians focus attention where it is most needed. Artificial intelligence early-warning systems suggest that AI can support risk triage and contribute to better outcomes, while deep learning models in medical imaging are already helping detect subtle patterns in areas such as radiology, mammography, and cardiology – supporting earlier diagnosis and risk prediction. Second, an emerging key risk is false confidence in outputs that are persuasive but may be biased, manipulated, opaque, or insufficiently validated. Third, hope remains available if we choose responsible optimism: curiosity paired with healthy caution, innovation paired with governance, and AI paired with human judgment. The box is open. Whether hope stays inside depends on us.

The Next Era of Cancer Care: Earlier detection, more precise treatment, better outcomes

Precision oncology is transforming cancer care, with therapies increasingly targeted to tumor biology, rather than cancer type. Earlier and more personalized intervention is improving outcomes while helping some patients avoid unnecessary treatment. Emerging technologies – from multi-cancer early detection (MCED) screening to GLP-1 therapies – show promise but require further evidence to define their long-term impact.

Key takeaways

References

1. Daily pill can double survival time for world’s deadliest cancer, trial shows; The Guardian, May 31, 20262. Conroy T, Desseigne F, Ychou et al. FOLFIRINOX versus Gemcitabine for Metastatic Pancreatic Cancer, N Engl J Med 2011;364:1817-18253. Wolpin BM, Wainberg ZA, Hendifar A et al. Daraxonrasib, a RAS(ON) multi-selective inhibitor vs chemotherapy in previously treated metastatic pancreatic adenocarcinoma (mPDAC): Primary and final analysis from the phase 3 RASolute 302 study. J Clin Oncol 44, LBA5(2026) Volume 44, Number 17_suppl4. Burtness B, Harrington KJ, Ji D et al. Amivantamab in HPV-unrelated recurrent/metastatic head and neck squamous cell cancer after disease progression on immune checkpoint inhibitor and chemotherapy: Pivotal results from the phase 1b/2 OrigAMI-4 study J Clin Oncol 44, 6008(2026) Volume 44, Number 16_suppl5. Taplin M-E, Gleave M, Shore ND et al. Perioperative (neoadjuvant and adjuvant) apalutamide (APA) + androgen deprivation therapy (ADT) vs placebo (PBO) + ADT with radical prostatectomy (RP) in high-risk localized or locally advanced prostate cancer (HR LPC/LAPC): Final analysis of the PROTEUS phase 3 study J Clin Oncol 44, LBA1(2026) Volume 44, Number 17_suppl6. Stein RC, Makris A, Macpherson IR. First results from the OPTIMA phase III randomized non-inferiority trial of test-directed chemotherapy in patients with high clinical risk ER-positive HER2-negative early breast cancer. J Clin Oncol 44, 500(2026) Volume 44, Number 16_suppl7. Swanton RC, Johnson P, Round T et al. NHS-Galleri: Primary results from a randomized, controlled trial to assess the clinical utility of a multi-cancer early detection (MCED) test in population screening J Clin Oncol 44, LBA100(2026) Volume 44, Number 17_suppl8. Orland MD, Mandala A, Unlu S et al. Can GLP-1 receptor agonists mitigate cancer progression? A propensity-matched analysis across seven solid tumors. J Clin Oncol 44, 3143(2026) Volume 44, Number 16_suppl 9. Four in ten cancer cases could be prevented globally, World Health Organization - https://www.who.int/news/item/03-02-2026-four-in-ten-cancer-cases-could-be-prevented-globally

The common theme at ASCO was a move toward more precise, effective, and patient-centered care.

Part II: Lewy Body Dementias

To share your thoughts about the article, or to suggest topics we could cover in future issues, please click here to send us an email.

Rate this article

Every year, the world’s leading cancer researchers, clinicians, and industry experts gather in Chicago for the American Society of Clinical Oncology (ASCO) Annual Meeting – the premier event on the global oncology calendar. With more than 40,000 attendees from around the world, ASCO provides a unique opportunity to share the latest advances in cancer prevention, screening, diagnosis, and treatment. The meeting often offers an early glimpse of innovations that may shape future standards of care and influence both clinical practice and research priorities. This year’s meeting highlighted the rapid evolution of precision oncology, the growing role of genomic testing, the continued expansion of targeted therapies, and the promise – and challenges – of earlier cancer detection. Together, these developments reinforce how a deeper understanding of tumor biology is translating into more personalized approaches to cancer care. The following highlights provide a snapshot of some of the most notable studies presented at ASCO 2026. They illustrate both the remarkable progress being made and the significant challenges that remain.

Pancreatic cancer remains one of the most lethal solid organ malignancies, largely because it is often diagnosed at an advanced stage and has limited effective treatment options. Despite advances in oncology, survival outcomes have improved only modestly over recent decades, and metastatic disease continues to carry a particularly poor prognosis. Against this backdrop, the emergence of targeted therapies offers a rare and encouraging development in a disease that has long been associated with significant unmet clinical need. The headline, “Daily pill can double survival time for world’s deadliest cancer, trial shows,” (The Guardian newspaper)1 highlights a potentially significant advance in the treatment of pancreatic cancer. Current first-line chemotherapy typically involves the intensive FOLFIRINOX regimen; however, median overall survival remains limited at approximately 11.1 months.2 Once disease progresses, second-line treatment options provide only modest clinical benefits, highlighting the substantial unmet need in this setting. Approximately 90% of pancreatic cancers harbor a KRAS mutation. KRAS is a molecular on/off switch, critical for regulating cell growth, division, and survival. Daraxonrasib is a novel Ras(On) multi-selective inhibitor designed to switch off KRAS-driven signaling and inhibit tumor growth, regardless of the specific KRAS variant present. In a 500-patient clinical trial, daraxonrasib increased median overall survival in patients with advanced, previously treated pancreatic cancer from 6.7 months to 13.2 months compared with investigator-selected chemotherapy regimens. Importantly, the treatment was also associated with a more favorable safety profile. The most common adverse events with daraxonrasib were rash (13.7%) and stomatitis (12.0%), while chemotherapy was more commonly associated with neutropenia (18.2%) and anemia (16.4%).3 The investigators reasonably concluded that, in this historically challenging disease, “daraxonrasib demonstrated unprecedented improvements in overall survival and progression-free survival versus chemotherapy, in patients with pancreatic ductal adenocarcinoma with or without an identified tumor RAS mutation. Daraxonrasib was generally well tolerated, with a manageable safety profile and with no new safety signals.” These findings support daraxonrasib as a potential new standard of care for previously treated advanced pancreatic cancer. Daraxonrasib (RASONQUE®) received its first regulatory approval from the U.S. FDA in August 2026. Why it matters: Currently, from an insurance medicine perspective, a diagnosis of metastatic pancreatic cancer is typically considered eligible for terminal illness benefits because of its uniformly poor prognosis. If the survival benefits observed with daraxonrasib are confirmed in broader clinical practice and future studies, pancreatic cancer may eventually join the growing list of metastatic cancers for which advanced-stage disease is no longer synonymous with an immediately terminal prognosis.

Highlight 1: Targeted therapy offers new hope in advanced pancreatic cancer

BACK

A prominent theme emerging from this year’s ASCO Annual Meeting was the continued evolution of precision oncology and the move away from traditional chemotherapy toward targeted therapies designed to exploit specific molecular drivers of cancer. One example was the growing body of evidence supporting the use of amivantamab, a bispecific antibody targeting both epidermal growth factor receptor (EGFR)-directed and mesenchymal-epithelial transition (MET) receptor pathways. It is currently the first treatment for adults with non-small cell lung cancer harboring epidermal growth factor receptor (EGFR) exon 20 insertion mutations, a subgroup that historically had limited therapeutic options and poor outcomes.At ASCO, investigators presented results from the OrigAMI-4 study4 evaluating amivantamab in patients with HPV-unrelated recurrent or metastatic head and neck squamous cell carcinoma (HNSCC) who had already progressed despite treatment with both immunotherapy and chemotherapy. This is a particularly challenging clinical setting, as treatment responses are often limited, and survival outcomes remain poor. HPV-positive head and neck cancers generally have a more favorable prognosis than HPV-negative disease. Patients with HPV-unrelated tumors typically experience more aggressive disease and have fewer effective treatment options, once standard therapies have been exhausted. Against this backdrop, the OrigAMI-4 findings3 were particularly encouraging. Among treated patients, 79% experienced shrinkage of their target lesions with a median duration of response of 7.2 months. Such levels of tumor response are uncommon in a population that has already progressed on multiple lines of therapy. While larger studies will be needed to confirm these results, the findings suggest that dual EGFR/MET inhibition may offer a promising new treatment approach for patients with significant unmet clinical needs. The study also highlights a broader trend increasingly evident across ASCO presentations: Therapies originally developed for one cancer type are being successfully applied across other tumor types that share similar biological pathways. This shift reflects the growing emphasis on tumor biology and molecular targets, rather than on tumor location alone, when selecting treatment strategies. The safety profile of amivantamab was consistent with previous experience in lung cancer and other EGFR/MET-directed therapies. The most common adverse events included hypoalbuminemia, rash, acneiform dermatitis, paronychia, stomatitis, and fatigue. Why it matters: The OrigAMI-4 results underscore how targeted therapies are continuing to expand beyond their initial indications, offering new hope for patients with advanced cancers that have historically been difficult to treat. They represent another step toward a future where cancer treatment is increasingly driven by molecular characteristics, rather than the organ in which the cancer originated.

Highlight 2: Expanding the reach of targeted therapies beyond lung cancer

Another important theme at ASCO was the continued move toward intensifying treatment earlier in the disease course to improve long-term outcomes. This was exemplified by the PROTEUS trial, which evaluated the addition of apalutamide, an androgen receptor pathway inhibitor, to standard androgen deprivation therapy (ADT) in men with high-risk, locally advanced prostate cancer undergoing radical prostatectomy. Patients with high-risk disease face a substantial risk of recurrence following surgery, with many ultimately requiring additional treatment and some progressing to metastatic disease, which is associated with significantly poorer outcomes. The PROTEUS study5 enrolled more than 2,100 patients to determine whether adding one year of perioperative apalutamide could improve long-term disease control. The results were compelling. Fewer patients receiving apalutamide plus ADT required subsequent therapy compared with those receiving ADT alone (42.4% vs. 56.7%). The combination also substantially prolonged the time to first subsequent treatment (74.2 vs. 41.5 months) and improved event-free survival (57.1 vs. 38.4 months). Importantly, testosterone recovery following treatment was only modestly delayed, with a median recovery time of 8.1 months compared with 6.6 months in the control arm. Why it matters: The PROTEUS findings demonstrate the value of earlier, more intensive treatment in high-risk prostate cancer and are expected to establish perioperative apalutamide plus ADT as a new standard of care for patients undergoing surgery for high-risk locally advanced disease.

Highlight 3: Perioperative therapy raises the bar in high-risk prostate cancer

A theme at ASCO was the increasing use of genomic tools to personalize treatment and identify patients who can safely avoid more intensive therapies without compromising outcomes. In breast cancer, this approach has the potential to reduce treatment-related toxicity while maintaining excellent disease control. The OPTIMA trial evaluated the use of the Prosigna (PAM50) genomic assay in patients with estrogen receptor-positive (ER+) HER2-negative early breast cancer who would traditionally be considered candidates for chemotherapy.6 The test generates a Risk of Recurrence (ROR) score, helping clinicians identify which patients are most likely to benefit from chemotherapy and which may be adequately treated with endocrine therapy alone. Patients were randomized to receive standard chemotherapy followed by endocrine therapy or treatment guided by the Prosigna test. Those with higher-risk tumors (ROR >60) received chemotherapy and endocrine therapy, while patients with lower-risk tumors (ROR ≤60) received endocrine therapy alone. The results demonstrated no meaningful difference in invasive breast cancer-free survival between the treatment groups, indicating that patients with a low Prosigna ROR score can safely forgo chemotherapy without compromising outcomes. Importantly, this finding extended to groups traditionally viewed as higher risk, including premenopausal women and patients with more extensive nodal involvement. Why it matters: The OPTIMA findings represent another step toward truly personalized cancer care, enabling more patients to avoid the short- and long-term toxicities of chemotherapy while maintaining excellent outcomes. The study also strengthens the case for broader adoption and routine funding of genomic testing to guide treatment decisions in early breast cancer.

Highlight 4: Tailoring treatment – sparing patients from unnecessary chemotherapy in breast cancer

Early cancer detection remains one of the most effective ways to improve cancer outcomes. A key objective of the NHS England Cancer Plan is to increase the proportion of cancers diagnosed at stage I or II from approximately 50% to 75%, and multi-cancer early detection (MCED) blood tests have generated considerable interest as a potential tool to help achieve this goal. One of the most anticipated studies presented at ASCO was the NHS-Galleri trial, a large randomized controlled study involving more than 142,000 asymptomatic individuals aged 50-79 years. The trial evaluated whether annual screening with the Galleri® MCED blood test, in addition to standard cancer screening programs, could reduce the incidence of advanced cancers by detecting malignancies earlier.7 The results were mixed. The study did not meet its primary endpoint of demonstrating a statistically significant reduction in stage III and IV cancers across 12 pre-specified cancer types. However, annual MCED testing increased the number of screen-detected cancers approximately fourfold, reduced stage IV cancer diagnoses by 14%, and decreased emergency cancer presentations by 21%. These findings highlight both the potential and the challenges of MCED screening. Although the primary endpoint was not achieved, the study demonstrated an ability to detect more cancers and shift some diagnoses away from emergency presentations, where outcomes are often poorer. In a separate but related development, on September 23, 2026, the FDA advisory committee voted in favor of approving GRAIL's Galleri® test for premarket approval, an important step in the ongoing evaluation of MCED technology. Why it matters: The NHS-Galleri study suggests that MCED testing is not yet the transformative population screening solution many had hoped for, but neither is it a failure. Despite ongoing questions regarding sensitivity, cost-effectiveness, implementation, and long-term impact on cancer mortality, increased detection of cancer and reductions in late-stage cancer diagnoses, together with the FDA advisory committee's recommendation to approve Galleri®, suggest that MCED technology may still play an important role as a supplement in cancer screening strategies. FDA approval would be a significant milestone that could accelerate adoption and generate further evidence on the technology's clinical utility. For insurers, MCED testing may warrant continued monitoring as evidence emerges regarding its potential impact on cancer detection, outcomes, and risk assessment.

Highlight 5: The promise and challenges of multi-cancer early detection (MCED)

The remarkable expansion of GLP-1 receptor agonists continues to be one of the most intriguing stories in medicine. Initially developed for diabetes management and now widely used for obesity treatment, these agents have already demonstrated substantial benefits in cardiovascular disease. Emerging evidence presented at ASCO suggests their impact may extend even further. In a large matched-cohort study, patients with a history of early-stage cancer who were exposed to GLP-1 receptor agonists were less likely to progress to stage IV metastatic disease across six of seven cancer types evaluated.8 Significant benefits were observed in lung, breast, colorectal, and liver cancers. Investigators also reported that higher tumor expression of GLP-1 receptors was associated with improved survival across multiple tumor types, with particularly notable findings in breast cancer. These findings remain observational and do not establish a causal relationship. However, the consistency of the signal across several cancers has generated considerable interest. As noted by ASCO presenters, the known anti-inflammatory and immune-modulating effects of GLP-1 therapies provide a biologically plausible explanation and warrant further investigation in prospective randomized trials. Why it matters: While far from practice-changing, these findings add to the growing list of potential benefits associated with GLP-1 therapies and raise the intriguing possibility that these agents may one day play a role in cancer prevention or progression, in addition to their established metabolic and cardiovascular effects.

Highlight 6: Could GLP-1 therapies influence cancer outcomes?

ASCO 2026 showcased both the complexity of cancer and the extraordinary progress being made in understanding and treating the disease. From novel, targeted therapies to advances in screening and treatment individualization, the common theme was a move toward more precise, effective, and patient-centered care. However, as the incidence of cancer continues to rise globally, prevention remains the greatest opportunity to reduce the burden of disease. The World Health Organization estimates that up to 40%9 of cancers could be prevented, underscoring the continuing importance of public health initiatives alongside therapeutic innovation.

Conclusion

Stephen Falk qualified from Liverpool University in 1983 and spent a period in Liverpool undertaking a general medical rotation. He then trained in oncology in Cardiff and Cambridge, becoming a Consultant in Bristol in 1994. His major clinical interest is gastro-intestinal cancer (esophagus, stomach, pancreas, and colorectal). A further interest in lung cancer and clinical research was stimulated as an MRC Clinical Scientist in Norman Bleehens' unit in Cambridge where the MRC Cancer Trials Unit was situated at the time. An MD related to lung cancer and scheduling of topoisomerase poisons with and without radiation is the basis of on-going interest in chemo-radiation interactions. Stephen retired from both NHS and private practice on July 31, 2025, after more than 31 years as a consultant. Previously he had a challenging clinical load seeing up to 1,000 new patients with cancer annually. Until 2018, he was Clinical Director of the West of England Clinical Research Network, whose job is to fund the NHS costs of delivering research activity in the NHS, and chaired the NCRI research Pancreatic Sub Group until 2018. Stephen has provided oncology advice to Scottish Widows/HBOS from 2008 and more recently has started assisting RGA.

About the author

Dr. Stephen Falk

Consultant Oncologist

Stephen.Falk@rgare.com

Older-Age Underwriting: Rethinking mortality risk

Aging is defined as “the process of accumulation of consequences of life, such as molecular and cellular damage, that leads to functional decline, chronic diseases, and ultimately mortality.”1 Although individuals age chronologically at the same rate, biological changes, such as molecular and cellular damage accumulation, can occur at different rates. While some people maintain good overall health in old age, others are impacted by negative external causes such as lifestyle and environmental factors. Understanding the difference between biological and chronological age – referred to as age deviation – and the causes behind this can allow for better mortality risk assessment at the underwriting stage. A key consideration is when biological markers are first noted to be associated with accelerated aging and early mortality. Early influences on health, such as physical inactivity, being overweight, lifestyle (e.g., smoking, alcohol consumption), antibiotic use, and environmental factors (e.g., air pollution) have been associated with accelerated biological aging later in life. This article examines important markers of mortality and the challenges faced by insurers in underwriting older lives.

Introduction

References

1. Moqri, M. et al. (2023). Biomarkers of aging for the identification and evaluation of longevity interventions. Cell; 186(18), August 31, 2023: 3758-3775. Available from: Biomarkers of aging for the identification and evaluation of longevity interventions: Cell2. Bakrania, K. et al. (2026). Resting heart rate: the heart of underwriting decisions. RGA Knowledge Center, Research and White Papers; June 2026. Available from: https://docs.google.com/spreadsheets/d/1M1x9O55ik1so204-rjRgxmxNQEMF2phyKlkMO_tnXNA/edit?gid=526623547#gid=5266235473. Crimmins, E. et al. (2008). Biomarkers related to aging in human populations. Adv Clin Chem 2008; 46:161-216. Available from: Biomarkers related to aging in human populations – PubMed4. Markon, K. et al. (2025). Aging dimensions and markers as relative predictors of mortality in a longitudinal epidemiological sample. PLoS One; 20(6): e0324156. Available from: Aging dimensions and markers as relative predictors of mortality in a longitudinal epidemiological sample | PLOS One5. Mitnitski, A. et al. (2015). Age-related frailty and its association with biological markers of ageing. BMC Medicine 13; 161 (2015). Available from: Age-related frailty and its association with biological markers of ageing | BMC Medicine | Springer Nature Link6. Garmany, A. Terzic, A. (2024). Global Healthspan-lifespan gaps among 183 World Health Organization member states. JAMA Network Open 2024; 7 (12): e2450241. Available from: garmany_2024_oi_241395_1733259944.03263.pdf7. GBD 2021 Fertility and Forecasting Collaborators (2024). Global fertility in 204 countries and territories, 1950-2021, with forecasts to 2100: a comprehensive demographic analysis for the Global Burden of Disease Study 2021. The Lancet; 403 (10440): 2057-2099. Available from: Global fertility in 204 countries and territories, 1950-2021, with forecasts to 2100: a comprehensive demographic analysis for the Global Burden of Disease Study 2021 – ScienceDirect8. Utmost Group (2025). Wealth management market study identifies size of the HNW international life insurance market for the first time. 2024 Wealth management market study. Available from: Wealth-Management-Market-Study-Press-Release-June-2025.pdf9. ABI (2026). Larger pension pots drive record-breaking year for individual annuity premiums. The ABI, 12 Feb 2026. Available from: Larger pension pots drive record-breaking year for individual annuity premiums | ABI10. Wang, G. et al. (2026). Artificial intelligence approaches in biological age prediction: current status and challenges. British Journal of Biomedical Science, 10 Jun 2026; 83:16141. Available from: Frontiers Publishing Partnerships | Artificial intelligence approaches in biological age prediction: current status and challenges

About the author

Hilary Henly, FCII, is a Global Medical Researcher with RGA’s Strategic Research team. Based in Ireland, she is a Fellow of the Chartered Insurance Institute (FCII) and has more than 30 years of experience in underwriting, claims, and mortality and morbidity research.

Hilary Henly

Global Medical Researcher, Global R&D

hhenly@rgare.com

People are living longer, but increased longevity does not always translate into improved health.

Chronological age does not fully capture variation in health and mortality risk between individuals. The growing population of older-age applicants creates both underwriting challenges and significant business opportunities for insurers. Future underwriting approaches will need to combine age-specific risk factors with advances in technology and data collection.

Key takeaways

Insurance challenges in underwriting older-age lives Within the insurance industry, the term “older-age lives” typically refers to individuals aged 65 and above, although certain products and market segments target applicants from age 55 onward. It is important for insurers to adopt risk segmentation approaches that more accurately reflect the mortality and morbidity experience of older populations. There is an opportunity for insurers to maximize underwriting potential of life insurance products in this age group by designing products and underwriting journeys that reflect their needs while exploring the use of technology to help gather relevant information to appropriately assist with underwriting risk assessment. In doing so, insurers can not only improve their overall underwriting experience but also position themselves to offer a customer-centric journey and appropriate terms through more precise assessment of mortality risk presented by older lives. Challenge 1: Changing demographicsPeople are living longer, but increased longevity does not always translate into good health. Surveillance of healthspan-lifespan gaps across 183 World Health Organization (WHO) member states, representing the number of years burdened by disease, identified a widening healthspan-lifespan gap worldwide of 9.6 years, with the U.S. showing the largest gap of 12.4 years.6 Furthermore, there has been a significant and sustained downturn in fertility rates, a key driver of global population aging. Between 1950 and 2021, the global total fertility rate across 204 countries and territories more than halved, from 4.84 to 2.23 per female.7 The changing global demographic landscape brings significant challenges for insurers but also a fresh opportunity to revise product, underwriting, and service propositions. Challenge 2: Insurance activity and size of opportunityThere is growing expectation that insurers will need to adapt product and underwriting strategies to reflect longer working lives that result in: Greater accumulation of assets over an extended earning period A longer phase of asset decumulation across an extended retirement and lifespan In addition, economic factors such as interest rates and changes in tax policies impact demand for insurance products. Insurance protection sales for older lives show strong and significant growth in many countries, particularly within the high-net-worth (HNW) underwriting space. A 2024 wealth management market study reported that global HNW annual new business sales reached £41 billion (GBP) in 2024, a 25% growth from 2023, and are projected to grow to £67 billion by 2030.8 Increased demand for insurance products is in part being driven by economic factors, rising wealth among retirees with an increased interest in wealth planning, aging populations, and increasing life expectancy. As retirees seek secure and predictable retirement incomes, annuity markets also continue to expand. Asset decumulation through annuity products continues to gather pace; for example, in the UK, the total value of premiums paid into individual pension annuities grew 4% to £7.4 billion in 2025, alongside an 8% rise in the number of people aged 70 and over purchasing an annuity.9 Tailwinds propelling growth are projected to be long term, which may provide insurers with the reassurance to pursue investments in strategies to revise products and develop tailored underwriting solutions for life insurance. Challenge 3: The application process for life insuranceInsurers should tailor application journeys, underwriting requirements, and medical screenings according to applicant age, reflecting the distinct mortality and morbidity risks associated with increasing age. There are many well-established medical parameters associated with health outcomes and longevity; however, underwriting should also consider age-specific markers of mortality that can be reliably measured, objectively verified, and appropriately documented throughout the application process, including: Enhanced medical screening leveraging digital technologies, which provides significant opportunities to embed new technology within the underwriting process. Assessing frailty, cognitive function, and functional capacity, as these factors are often highly predictive of mortality and morbidity in older adults. Challenge 4: Capturing appropriate dataCurrent insurance customer journeys typically involve ineffective data capture processes, limiting the effective assessment and underwriting of older-age lives: Older applicants present different mortality and morbidity risk profiles than younger applicants, yet often the underwriting journeys do not reflect these differences. Many application forms lack age-specific differentiation, resulting in limited use of targeted questions more appropriate for older applicants. Medical screening requirements and supplementary questionnaires do not always differentiate between age groups, with insufficient tailoring of risk factors most relevant to older lives. Underwriting processes may not place enough emphasis on the specific challenges with assessing older lives. There is often limited consideration given to differences in regional health screening protocols and insufficient use of valuable insights available from age-related screening programs in risk assessment.

Important markers of mortality in older lives At present, there is little consensus on what components should be included as measurements of biological aging or how they perform in predicting mortality. Physiological biomarkers are used regularly in the underwriting process (e.g., blood pressure), but other markers such as resting heart rate (RHR) historically have not formed part of the underwriting process, particularly for older individuals. Research shows that the risk of mortality increases as RHR increases and that it is a powerful predictor of mortality. Importantly, this association remains consistent across different age groups. RHR is an easily measured metric, because it is non-invasive, can be completed as part of a medical examination, and can be easily integrated into current underwriting practices. A recent collaborative study by RGA and the University of Leicester found that RHR significantly outperformed total cholesterol as a mortality predictor and has the potential to replace body mass index (BMI) without meaningful loss of predictive accuracy.2 While RHR measurements should be interpreted in context (e.g., medication use or prevalent disease), it adds incremental value to traditional underwriting metrics.

To share your thoughts about the article, or to suggest topics we could cover in future issues, please click here to send us an email.

Rate this article

BACK

Solutions/the future in risk assessment Digital biomarkers have recently been proposed as novel tools for older age risk assessment. Advances in digital technology and artificial intelligence (AI) are revolutionizing the assessment of biological age through the capture, integration, and interpretation of complex biological data. Wearables, facial imaging, and AI-enabled risk assessment tools are attracting growing interest for their potential to provide additional insights into an individual's health status and longevity risk.10 Using alternative biometrics, many of which are becoming easier to capture with modern technology, may improve the ability to more accurately differentiate morbidity and mortality risks in older lives. They may even be used as substitutes for routine physiological markers of health in the future. Others, such as self-reported dietary insights, walking speed, and step-count are not as easy to apply in the risk assessment process, despite research showing their association with mortality.

Conclusion While many individuals are affected by age-related decline at an early chronological age, others maintain good health and overall function well into older age. Novel biomarkers of health and longevity may help improve mortality risk assessment, particularly in older age groups. However, there is no single study providing information on all biomarkers that indicates the rate of aging in its entirety. Insurers have a significant opportunity to revisit product design and underwriting journeys to better reflect the risk profiles of older-age applicants, including the use of age-specific mortality indicators. Advances in technology can streamline the underwriting process, improve the applicant’s experience, and reduce the costs associated with obtaining traditional medical evidence when medically validated. Importantly, underwriting frameworks should incorporate alternative approaches to risk assessment for older lives, recognizing that the most relevant predictors of risk often differ from those used for younger populations.

Steve Simkin, Chartered Insurer (ACII), is the Technical Underwriting Manager for RGA’s Protection business in the UK. Steve joined RGA’s underwriting team in 2008 and helps manage the underwriting philosophy for RGA in the UK as well as working with clients around risk management improvements; particularly in areas such as underwriting philosophy, quality assurance, and underwriter training and development.

About the author

Steve Simkin

Technical Underwriting Manager, Underwriting

ssimkin@rgare.com

Increasing evidence shows that pulse pressure (the difference between systolic and diastolic blood pressure) is predictive of coronary heart disease in middle and older age, while the use of N-terminal pro B-type natriuretic peptide (NT-proBNP), a protein produced by the heart, has potential use in assessing mortality risk for older-age applicants with cardiovascular disease.3 High levels (3-10 mg/dl) of the inflammatory marker C-reactive protein (CRP) are also predictive of early mortality and physical decline. Age-related increases of interleukin IL-6 have been associated with cardiovascular disease, certain cancers, Alzheimer’s disease, and osteoporosis, while low levels of insulin-like growth factor-1 (IGF-1) have been correlated with increased mortality. Several of these markers, as well as high systolic blood pressure (OR 1.16), fibrinogen (OR 1.29), and creatinine clearance (OR 1.31) were significantly related to mortality in the National Health and Nutrition Examination Survey (NHANES) III sample of middle-aged and older adults.3 Another key marker of mortality risk in older lives is cognitive function, which appears to be one of the strongest and most consistent predictors of mortality. Functional capacity, particularly the ability to perform activities of daily living (ADLs), is also strongly associated with mortality risk, as too are frailty and walking pace – markers of physiological decline that are specific to older age.4 A frailty index (FI), noted to be a strong predictor of mortality, could be used to quantify frailty based on accumulation of various health-related deficits and has been shown to outperform chronological age as a predictor of mortality.5

BACK

Article

Dr. Georgiana Willwerth-Pascutiu Vice President and Medical Director, Global Medical Richard Russell Vice President, Biometric Research, Global Research and Development

What Hantavirus and Ebola Outbreaks Remind Us About Pandemic Risk

Article

Dr. Adela Osman Senior Vice President, Head of Global Medical Dr. Nico van Zyl Senior Vice President, Chief Medical Director

FDA Peptide Debate Points to Broader Trend for Insurers

Article

Kaitlyn Fleigle Actuary, Strategic Research, Innovation & ContentHussain Boricha Actuary, Americas Data SolutionsNetanya Martin Data Scientist, Global Research and Development

Early-Onset Cancer Is Reshaping Risk: A look into the research

video

Kaitlyn Fleigle Actuary, Strategic Research, Innovation & Content

The Early Exit – Insights into Lapse Behavior

VIDEO

Julianne Callaway Vice President and Senior Actuary, Strategic Research

Real Hope — An Update on Dementia

Article

Hilary Henly, Global Medical Researcher, Strategic Research, RGA

The Cost of Cancer: Price Versus Life

For recent ReFlections articles, click here.

You might also be interested in

Editor’s note: As GLP-1 therapies become increasingly important in diabetes and obesity management, this study highlights that genetic factors may influence treatment effectiveness. For insurers, the findings underscore the potential value of precision medicine approaches and the need to recognize variability in therapeutic response when assessing future metabolic risk.

Gloyn AL, Stoffel M, et al.

Why Ozempic Doesn’t Work for Everyone: Scientists Identify a Genetic Cause of ‘GLP-1 Resistance’

Editor's note: Excess weight may contribute to more than 10% of all cancers, substantially higher than previous estimates. The findings also suggest that BMI alone may underestimate obesity-related cancer risk, with measures of central adiposity, such as waist circumference and waist-to-hip ratio, providing a more accurate assessment. For insurers, this highlights the growing impact of obesity on future cancer risk and the importance of monitoring obesity-related morbidity and mortality trends. It also questions the use of standard BMI in underwriting and raises support for other measurements, such as waist-to-hip ration.

Adhikari LK, Safizadeh F, Mandic M, Brenner H.Cancer Communications, 2026; 46:0040 DOI: 10.34133/cancomm.0040

Excess Weight May Account for More Than 10% of All Cancers: The Underestimated Impact of the Obesity Epidemic

Fauchier L, Guglieri M, Lenormand T, Bisson A, Lip GYHEuropean Heart Journal, 2026; ehag543https://doi.org/10.1093/eurheartj/ehag543

Editor’s note: For life and health insurers, these findings reinforce the importance of vaccination status as a potentially relevant marker of both preventive health behavior and future cardiovascular risk. While causality cannot be confirmed from observational data, the results underscore the value of continued monitoring of vaccine-related cardiovascular outcomes, particularly in older adults and individuals with established cardiometabolic disease.

Influenza vaccination linked to lower atrial fibrillation and cardiovascular risk

A new study from Stanford Medicine and international collaborators suggests that approximately 10% of individuals may carry genetic variants associated with a newly identified phenomenon termed “GLP-1 resistance.” These individuals have higher-than-normal levels of the hormone GLP-1, which plays a key role in glucose regulation, yet appears to respond less effectively to its actions. The finding may help explain why some patients experience limited benefit from GLP-1 receptor agonists such as Ozempic (semaglutide) and related therapies. The research combined human and animal studies with analyses of clinical trial data collected over a decade. Investigators found that individuals carrying specific genetic variants achieved less improvement in blood glucose control following treatment with GLP-1 receptor agonists compared with non-carriers. In some clinical trials, these patients showed a reduced ability to lower blood glucose levels after six months of therapy, suggesting a diminished treatment response. While the study focused primarily on diabetes outcomes, the researchers noted that it remains unclear whether these variants also influence the magnitude of weight loss achieved with GLP-1 therapies used for obesity treatment. The findings highlight the potential for more personalized approaches to selecting metabolic therapies and may represent an important step toward precision medicine in diabetes care.

Excess weight is already recognized as an important cancer risk factor, with previous estimates suggesting that 2%-8% of all cancers are attributable to overweight and obesity. However, this analysis of 458,543 participants from the UK Biobank suggests that the true burden may be substantially higher. Researchers evaluated multiple measures of adiposity, including body mass index (BMI), waist circumference, and waist-to-hip ratio, and addressed a key source of bias: Many cancers cause weight loss before diagnosis, which can weaken the apparent association between obesity and cancer risk. Participants were followed for a median of 12 years, during which more than 50,000 incident cancers occurred. Measures of central adiposity, such as waist circumference and waist-to-hip ratio, proved more strongly associated with cancer risk than did BMI alone. When more accurate adiposity measures were combined with methods to account for pre-diagnostic weight loss, the estimated proportion of cancers attributable to excess weight increased from approximately 5.5% to as high as 11.5%, suggesting that more than one in 10 cancers may be linked to overweight or obesity. The association appeared particularly pronounced among women and adults younger than 60 years. Overall, the findings indicate that the cancer burden attributable to excess weight may have been substantially underestimated. The authors suggest that reliance on BMI alone and failure to account for pre-diagnostic weight loss may have obscured the true impact of excess adiposity on cancer incidence.

Influenza infection is a recognized trigger for acute cardiovascular events. Previous studies have demonstrated the cardiovascular benefits of influenza vaccination in individuals at elevated cardiovascular risk. Building on this evidence, this large real-world retrospective cohort study used the TriNetX Global Collaborative Federated Research Data Network to evaluate the association between influenza vaccination, incident atrial fibrillation (AF), and cardiovascular outcomes. Adults with a routine healthcare visit in 2022 were included, with influenza vaccination in the preceding year used to define exposure. After 1:1 propensity score matching across 65 baseline variables, the study compared 138,444 vaccinated individuals with the same number of unvaccinated controls. Over a mean follow-up of 2.7 years, influenza vaccination was associated with a lower risk of incident AF, with an annualized AF incidence of 0.9% in vaccinated individuals compared with 1.1% in unvaccinated individuals. Vaccination was also associated with lower risks of all-cause mortality, myocardial infarction, incident heart failure, and cardiovascular-related hospitalization. These associations were consistent across age groups and among individuals both with and without established cardiovascular disease. Overall, the findings suggest that the cardiovascular benefits of influenza vaccination may extend beyond prevention of infection, potentially reducing the risk of AF and other major cardiovascular events. However, as this was an observational study, randomized trials are needed to confirm causality and further define the magnitude of benefits.

Publications relevant to insurance medicine appearing recently in research literature.

Dr. Adela Osman Senior Vice President, Head of Global Medical

Research Watch

References: https://www.nature.com/articles/s41746-024-01073-0 https://www.theactuary.com/2024/04/04/power-two-digital-twins-insurance

The Food and Drug Administration (FDA) has approved TREGZI™ in June 2026, clinically known as Orca-T®, the first regulatory T-cell (Treg) immunotherapy designed to improve chronic graft-versus-host disease (GVHD)-free survival in adults with high-risk blood cancers undergoing allogeneic stem cell transplantation. This approval represents a significant advance in cellular therapy and transplant medicine. TREGZI is intended for patients with hematologic malignancies – such as acute leukemia and high-risk myelodysplastic syndrome – who require donor stem cell transplantation. While transplantation can be curative, a dreaded complication is chronic GVHD, in which donor immune cells attack the recipient’s tissues, leading to potentially severe and long-lasting organ damage, disability, and even death. The therapy consists of three donor-derived cellular components: hematopoietic stem and progenitor cells (HSPCs), Tregs, and conventional T-cells. Tregs play a key role in controlling immune responses and maintaining immune tolerance. By introducing these cells in a carefully engineered composition, TREGZI aims to substantially reduce the risk of chronic GVHD while the patient’s immune system rebuilds. The FDA approval was supported by results from a randomized clinical trial involving 187 patients. At one year, 78% of patients treated with TREGZI were alive and free from moderate to severe chronic GVHD, compared with 38.4% of patients receiving the standard transplantation protocol. For health insurers, TREGZI may increase upfront treatment costs due to its highly specialized cellular manufacturing process but could generate substantial downstream savings by reducing chronic GVHD, hospitalizations, long-term immunosuppressive therapy, and disability-related expenses. TREGZI is expected to cost $428,000 as a one-off treatment. For life insurers, improved transplant outcomes could potentially translate into better long-term survival and lower post-transplant mortality among eligible blood cancer patients. Nevertheless, insurers will need to monitor real-world outcomes, durability of benefit, evolving treatment costs, and market adoption before adjusting assumptions. Reference: https://www.nature.com/articles/d41573-026-00119-0

World’s First Cell Therapy Against GVHD

Dr. Steve Woh Vice President, Global Medical Director

Health View:

References: https://www.dw.com/en/how-australias-social-media-ban-could-affect-us/video-70972818 https://oxford.shorthandstories.com/social-media-digital-mental-health/index.html https://theconversation.com/australias-social-media-ban-for-kids-under-16-just-became-law-how-it-will-work-remains-a-mystery-244736 https://theconversation.com/the-government-has-introduced-laws-for-its-social-media-ban-but-key-details-are-still-missing-244272

In June 2026, the Food and Drug Administration (FDA) approved the world’s first AI tool to detect multiple structural heart diseases from a standard 12-lead electrocardiogram (ECG). The technology, called EchoNext, received clearance for six cardiovascular indications and is designed to screen and identify patients at risk of serious heart conditions that may not be detectable through routine clinical interpretation of ECGs. Identified patients would then undergo an echocardiogram to confirm diagnosis. EchoNext can flag conditions including left- and right-sided heart failure, valvular heart disease, pulmonary hypertension, and severe hypertrophy associated with infiltrative cardiomyopathies. The AI model was developed using more than 700,000 paired ECG-echocardiogram records and reportedly demonstrated higher diagnostic accuracy than cardiologists, including those using AI assistance. Its performance was shown to be robust through validation studies that involved over 500,000 patients across more than 20 hospitals in the U.S. and Canada. The company cited a case published in Nature Medicine1 where EchoNext identified previously undiagnosed heart failure that ultimately led to the first heart transplant resulting from AI-enabled disease detection. The company also announced a partnership with OpenEvidence,2 a clinical AI platform reportedly used by more than half the physicians in the U.S., enabling wide-scale distribution of the tool at the point of care. For healthcare payers, earlier detection of cardiovascular disease may increase short-term healthcare utilization through additional diagnostic testing, specialist consultations, and treatment, while potentially supporting improved long-term health outcomes and reducing the burden of advanced cardiac disease. Widespread adoption may increase detection rates of previously undiagnosed cardiovascular disease and provide new insights into disease prevalence, with the potential to inform future risk assessment and portfolio management considerations. References: https://www.nature.com/articles/s41591-026-04454-y https://www.businesswire.com/news/home/20260623150207/en/Pathway-Labs-Announces-Worlds-First-FDA-Clearance-for-Multicondition-AI-in-Cardiology-and-Partnership-with-OpenEvidence

World’s First AI Tool for Early Detection of Heart Disease

Dr. Steve Woh Vice President, Global Medical Director

Health View:

What’s New?

Health View

Industry Event

Research Watch

RGA Thought Leadership

Editor’s note: Current liver cancer screening strategies may miss a substantial proportion of at-risk individuals, as many cases occur in people without recognized liver disease. This study demonstrates the potential of AI to improve early-risk identification using routine clinical data. For insurers, more accurate risk stratification could support earlier detection of high-risk individuals, improve understanding of future morbidity trends, and reinforce the growing role of AI-driven predictive analytics in risk assessment and prevention strategies.

Zhang DY, van Haag F, El Nahhas OSM, et al.

AI Model Predicts Liver Cancer Risk Using Routine Clinical Data

Researchers from RWTH Aachen University and the Technical University of Dresden developed a machine learning model capable of predicting the risk of hepatocellular carcinoma (HCC), the most common form of liver cancer, using routinely collected clinical data such as demographics, medical history, and standard laboratory results. The model was developed using data from more than 500,000 participants in the UK Biobank, including 538 HCC cases, and subsequently validated in an independent U.S. cohort. Notably, 69% of HCC cases occurred in individuals without a prior diagnosis of cirrhosis, viral hepatitis, or other chronic liver diseases, highlighting a major limitation of current screening approaches that focus primarily on patients with known liver disease. The machine learning model outperformed existing HCC risk prediction tools and accurately identified individuals at elevated risk using information already available in routine clinical practice. Importantly, the addition of more complex genomic or metabolomic data did not meaningfully improve predictive performance, suggesting that readily available clinical data may be sufficient for effective risk stratification. The investigators suggest that the tool could enable earlier identification of high-risk individuals, including those who would not currently qualify for liver cancer surveillance. Such an approach may facilitate more targeted screening programs and improve opportunities for earlier diagnosis and treatment.

Genome Medicine, 2026Reported by Stanford Medicine; published 29 March 2026

Genetic determinants of GLP-1 resistance and response to GLP-1 receptor agonists

Cancer Discovery, 2026 DOI: 10.1158/2159-8290.CD-25-1323

Machine learning predicts hepatocellular carcinoma risk from routine clinical data: a large population-based multi-centric study

content

Inside the Study: Underwriting’s future is taking shape, but risks remain

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Nam vitae nisi ac dolor consequat ultrices. Cras iaculis est quis sapien lacinia sollicitudin. Nullam eros eros, porta eu risus at, fringilla ornare massa. Vivamus maximus elit eu interdum iaculis. Maecenas porttitor efficitur pretium. Pellentesque dignissim volutpat arcu, eu imperdiet orci eleifend et. Nam molestie aliquet augue, a fringilla ipsum lacinia quis. Curabitur porta, felis quis finibus rhoncus, tellus leo tempor mauris, ut varius nulla ligula vitae libero. Duis nec mauris quis odio vulputate luctus ac id magna. Nunc a libero vitae lectus dictum dignissim vel quis lacus. Vivamus leo turpis, aliquet quis lectus in, imperdiet ultricies metus. Mauris scelerisque ornare auctor. Phasellus maximus nulla mauris, a molestie nisl ullamcorper rutrum. Quisque non enim non velit semper dignissim. Etiam nec dui sed velit efficitur gravida eget vel risus. Nunc ut mattis massa. In sit amet egestas sem, at dictum odio. Vestibulum dignissim eget felis in tempor. Integer suscipit porttitor urna at sollicitudin. Maecenas eu neque metus. Morbi ullamcorper et ipsum id imperdiet. Donec nisl lacus, rutrum rhoncus sapien eget, consequat condimentum magna. Integer pellentesque turpis vitae hendrerit scelerisque. In hac habitasse platea dictumst. Donec luctus nibh sed turpis posuere, quis viverra arcu commodo. Aliquam eget imperdiet magna. Etiam at vestibulum mi. Vestibulum id laoreet lorem, sed auctor nisl. Duis mollis magna quis massa vehicula cursus. Cras nisi sem, tempor quis laoreet vel, ullamcorper at ex. Cras facilisis, dolor quis auctor scelerisque, velit ante placerat risus, et luctus eros dui eget nunc. Sed nec magna urna.

In hac habitasse platea dictumst

Underwriting Updates

In hac habitasse platea dictumst

content

Mortality Impact of Indirect Alcohol Biomarkers

content

Measuring Mortality Slippage in Accelerated Underwriting

Please click the articles below to see some of RGA's recently published thought leadership on underwriting topics.

The latest guideline updates to RGA’s Global Underwriting Manual (GUM)* include proteinuria, glomerulonephritis, and non-Hodgkin lymphoma, along with the launch of a new kidney disorders resource hub. These updates reflect RGA’s commitment to applying expert, data-driven insights to support consistent, informed underwriting decisions. By translating research-based evidence into clear, practical guidance, we continue to help underwriters navigate complex risks with greater confidence and efficiency. This quarter, we are also celebrating a major milestone: more than one million GUM logins over the past year. With more than 3,800 unique users accessing the platform in a single day, these results underscore the growing value and global reach of RGA’s underwriting expertise. To support ongoing professional development, underwriters are encouraged to explore the related Learning Gateway modules. Five new learning resources have been published: Cancers: They’re Complicated! Underwriting Multiple Cancers MASLD – The New Liver Risk The Test of Time: Frailty, Falls, and Older Age Medical Gray Zones in Life Risk Assessment – Practical Case Reviews Relevant to Insurance Decisions AI and Medical Science – Early Applications and Lingering Questions Select the Learning Gateway tile on RGA Central to browse our full library of courses. We hope this edition inspires you to discover new content, deepen your expertise, and engage with the global underwriting community. Together, we will continue advancing underwriting excellence and building on the momentum of this significant milestone.

BACK

We look forward to sharing further Global Underwriting Manual updates in future editions of ReFlections. *Global Underwriting Manual (GUM) guidelines in North America may differ from those applied in other markets. For RGA clients in the U.S. and Canada, please visit RGA Central to access your customized GUM content.

By Brooke Butler, Director, Marketing and Communications – Global Underwriting Philosophy

Case ReView

Dr. Sheetal Salgaonkar, MBBS, MD, DBIM, FALU

Vice President and Global Medical Director

ssalgaonkar@rgare.com

A Matter of Counts: Leukopenia and Neutropenia

Case summary:A 52-year-old woman applied for a life insurance policy with a face amount of $20 million. As part of the underwriting assessment, she submitted medical evidence including a Medical Examination Report (MER), complete blood count (CBC), fasting blood glucose, electrocardiogram (ECG), and liver and kidney function tests. She had a 10-year history of hypothyroidism and was receiving thyroxine replacement therapy. Her thyroid profile – including ultrasensitive TSH, free T3, and free T4 – was within normal limits. Her CBC is shown here; all other investigations were normal.

Q3. What are the causes of leukopenia and neutropenia?Leukopenia and neutropenia may be transient or chronic, congenital or acquired. Mild isolated neutropenia is a relatively common incidental laboratory finding. Common causes include: Infection related: Viral illnesses – including influenza, hepatitis, EBV, HIV, measles, rubella, and varicella – can cause transient leukopenia and neutropenia. Post-infectious neutropenia is often self-limiting. Drug induced: Relevant drugs include chemotherapy, clozapine, antithyroid agents, sulfasalazine, trimethoprim-sulfamethoxazole, beta-lactam antibiotics, NSAIDs, and certain antipsychotics. Autoimmune disease: Neutropenia may occur in systemic lupus erythematosus, rheumatoid arthritis, Sjögren syndrome, autoimmune thyroid disease, and Felty syndrome. Nutritional deficiency: Vitamin B12, folate, copper deficiency, and severe malnutrition can result in cytopenias, usually affecting more than one cell line. Hypersplenism: Splenic sequestration may result in leukopenia, thrombocytopenia, or pancytopenia. Bone marrow or haematological disorders: Leukaemia, myelodysplastic syndromes, lymphoma, myeloma, marrow infiltration, and aplastic processes should be considered when cytopenias are severe, progressive, persistent, or associated with smear abnormalities. Constitutional or familial neutropenia: This is usually mild and chronic, often with an ANC >1,000/µL and no history of recurrent infections. Q4. What are the underwriting red flags? Progressive decline in WBC count or ANC on serial CBCs Associated anaemia, thrombocytopenia, pancytopenia, or an abnormal differential count Peripheral smear abnormalities, including blasts, dysplasia, atypical lymphocytes, or immature myeloid cells Elevated ESR or other inflammatory markers Recurrent or unusual infections, oral ulcers, fever, or poor wound healing Clinical features suggestive of autoimmune disease, splenomegaly, nutritional deficiency, or haematological malignancy Q5. What was the underwriting decision in this case?The applicant was asymptomatic. The initial CBC showed leukopenia and neutropenia, with preservation of the other cell lines, with no anaemia or thrombocytopenia. Therefore, rather than immediately postponing the case, a repeat CBC and peripheral smear examination were requested. The repeat CBC showed worsening leukopenia, with a total WBC count of 2,300/µL and neutrophils of 25%. This corresponded to an ANC of approximately 575/µL, indicating moderate neutropenia. Mild thrombocytopenia was also present, while haemoglobin remained normal. The peripheral smear showed leukopenia with marked neutropenia, relative lymphocytosis, occasional reactive lymphocytes, and reduced platelet numbers. There were no blasts, dysplastic cells, nucleated red blood cells, or abnormal immature myeloid cells. The progression of neutropenia together with new thrombocytopenia represents bicytopenia. This means a reduction in any two of the three major peripheral-blood cell lines: red cells, white cells, and platelets. This may indicate transient viral marrow suppression but also warrants exclusion of systemic disease, immune-mediated cytopenia, hypersplenism, nutritional deficiency, and underlying bone-marrow or haematological disorders. The case requires further evaluation, including bone marrow assessment if clinically indicated. The case was therefore postponed for a complete evaluation. Key takeaways: Mild isolated leukopenia and neutropenia are relatively common incidental findings on CBC testing. A single low WBC count or ANC may be transient, particularly following a viral illness or drug exposure. A falling ANC or worsening leukopenia/neutropenia is more significant than an isolated mildly abnormal result. Associated anaemia, thrombocytopenia, pancytopenia, abnormal peripheral smear findings, recurrent infections, or systemic symptoms warrant structured clinical evaluation.

BACK

Key underwriting questions: Q1. What is the abnormality in the complete blood count report? The CBC demonstrates leukopenia with neutropenia. Leukopenia refers to a reduced total white blood cell (WBC) count, commonly defined as being below the lower limit of the reference range, often less than 4.0 × 109/L or 4,000/µL, depending on the laboratory. Since neutrophils are the predominant circulating white cell subtype, leukopenia is often due to neutropenia. However, lymphopenia (reduced lymphocyte count) or reductions in other leukocyte populations may also contribute. Q2. What is neutropenia and how is Absolute Neutrophil Count (ANC) calculated?Neutropenia is generally defined as an absolute neutrophil count (ANC) below 1,500/µL. The ANC is calculated using the total WBC count and the proportion of neutrophils, including segmented neutrophils and band forms, where reported:

ANC = Total WBC × Neutrophil percentage/100 In this case, the WBC count was 2.9×109/L and neutrophils accounted for 35% of leukocytes:2.9×0.35=1.015×109/L Therefore, the estimated ANC was approximately 1.02×109/L, or 1,015/µL.

BACK

Please click the articles below to see some of RGA's published thought leadership on claims topics.

Claims Updates

Behind every claim is a person seeking clarity, support, and a timely decision. Delivering these outcomes requires not only technical expertise but also access to trusted guidance and ongoing professional development. Through the Global Claims Manual (known as the Global Claims Guide in the U.S.) and RGA’s Pathfinder training program, we continue to support claims professionals with practical resources that build confidence, capability, and consistency across global markets.

Pathfinder Get Started: Building confidence in disability income claims Disability income claims – known as income protection in some markets – are among the most complex claims that assessors manage. Effective management requires consideration of multiple factors, including medical evidence, occupational demands, biopsychosocial influences, and policy interpretation. Building the confidence and capability required to make consistent, high-quality decisions takes time. RGA’s new Get Started training series, available to clients, provides a practical springboard for developing these skills. The 18-module series is the latest addition to RGA’s Pathfinder professional claims training program, offering practical guidance and structured, scenario-based learning specifically designed for emerging claims professionals. Topics covered range from assessment fundamentals and functional capacity evaluation to occupational assessment and effective claimant communication. The self-paced modules can each be completed in 15-45 minutes, making it easy to build knowledge alongside the demands of managing a claims portfolio. → Clients: Discover Disability Income Get Started

Pathfinder Training: Building capability with practical learningIn addition to disability income, Get Started modules are also available to clients for death, critical illness, total and permanent disability, and health claims. Assessors can tailor their learning journey by selecting modules most relevant to their role, level of experience, and the claims they manage. The Pathfinder training library continues to expand, offering assessors even greater opportunities to build technical knowledge and strengthen decision-making skills. Explore the Pathfinder library to discover practical, claims-focused learning designed for real-world application. → View the Pathfinder menu

Tough Customer Conversations: Turning discomfort into constructive dialogue Navigating difficult conversations with claimants remains one of the most challenging aspects of claims management. To support claims professionals in this area, RGA recently delivered to clients a global webinar exploring how behavioral science principles can help turn uncomfortable interactions into more constructive and productive discussions. The session combined behavioral science insights with practical claims scenarios, providing assessors with techniques they can apply to their day-to-day interactions with claimants. → Clients: Watch the webinar

Explore Pathfinder. All Pathfinder modules are available through the Global Claims Manual. Click here for a menu of Pathfinder modules by benefit type and experience level.

By Jennie Calder Brown, Executive Director, Global Claims Philosophy and Education Gayle Kanchanapume, Executive Director, Global Claims Philosophy and Education

By combining technical expertise with communication skills, claims professionals can improve claimant experiences while supporting more effective claims outcomes. We remain committed to developing practical resources that help assessors build capability, navigate complexity, and deliver consistent, customer-focused claims decisions. For more information or to access these resources, please contact your local RGA claims representative or email the Global Claims team.

Article

The Promise and Peril of Seamless Integration in Disability Claims

Article

Why Most Life Insurers Under-Detect Fraud – and How to Fix It

Article

How Fraud Becomes the Carbon Monoxide of the Insurance Industry

Excellence in claims management demands deep technical expertise and a human touch. To support this, RGA delivers high-quality training and continuing education through our Pathfinder claims training program, available exclusively to RGA clients via the Global Claims Manual/Guide.

The people most vulnerable to the ketogenic diet’s hidden cardiac risk are, oddly, the healthiest ones. In a subset of lean adults, the diet sends cholesterol to levels usually seen only in inherited lipid disorders, while sparing everyone else. For Dr. Max Petersen, that paradox has become the focus of his research at Washington University School of Medicine in St. Louis (WashU Medicine), where he studies why cholesterol metabolism responds so differently from one individual to the next. Dr. Petersen is featured in the second installment of the Longer Life Foundation's Voices of Discovery video series, a multipart initiative designed to introduce LLF’s community of researchers to the public. In this installment, Dr. Petersen shares what drew him to metabolism research, why the ketogenic diet’s effects on the heart demand closer study, and what his LLF-supported research may mean for the future of personalized nutrition. The series is available on the LLF LinkedIn page and at longerlife.org, with new installments planned throughout 2026.

Click here to read the newsletter

BACK

What the research reveals WashU Medicine Assistant Professor of Medicine Dr. Max Petersen leads research investigating ketogenic diet-induced hypercholesterolemia, a condition in which blood cholesterol rises sharply in certain individuals following the diet. His LLF-supported project examines the two routes by which cholesterol reaches the bloodstream (production in the liver and absorption from food in the gut) to determine which pathway goes awry. Using tracer techniques that measure cholesterol kinetics directly, his team compares individuals on the diet whose cholesterol climbs against those whose levels remain stable. This is fundamental science, aimed at some of the most basic open questions in cholesterol metabolism. Why do individuals with nearly identical health profiles respond differently to the same dietary intervention? Can susceptibility be identified, or the response interrupted, before cardiovascular damage accumulates? The answers have implications that extend well beyond the laboratory. A clearer understanding of how diet reshapes cholesterol metabolism could eventually support screening for at-risk individuals, inform safer use of ketogenic therapies, and refine how clinicians and underwriters think about dietary risk.

Why study the ketogenic diet? Few eating patterns have attracted as much popular attention as the ketogenic diet. Millions of people worldwide have adopted it for weight management, glycemic control, or therapeutic reasons. Despite the enthusiasm, however, the diet’s potential risks remain far less well understood than its promise. For a subset of individuals, the same diet that helps others can drive LDL cholesterol to strikingly high levels. Dr. Petersen’s research is not an argument for or against the diet. It is an effort to replace hype with evidence, and to clarify who is vulnerable to this side effect and why, so that individuals and their clinicians can weigh the diet’s benefits and risks with a balanced, informed view. Insurers and risk professionals, in particular, are taking note as evidence accumulates that population-level dietary guidance conceals substantial individual variation. Elevated LDL cholesterol is a principal driver of atherosclerotic disease, which remains a leading cause of death worldwide. Research that clarifies who is vulnerable to diet-induced cholesterol elevation and why it could, over time, support earlier detection of individuals at elevated risk, leading to increased underwriting accuracy and improved product design.

The LLF model: Early investment, long-term return Dr. Petersen’s work is supported by the Longer Life Foundation, a collaboration between the RGA Foundation and WashU Medicine to support early-stage research into factors that influence longevity, health, and wellness. This early-stage, hypothesis-driven research rarely makes headlines but consistently seeds the discoveries that do. The foundation has operated since 1998, backing researchers across more than 170 investigations, with results that have appeared in more than 180 peer-reviewed publications. LLF’s Voices of Discovery series tells the stories behind the data, featuring each researcher speaking directly about what motivates their work, what they have found, and what they hope it ultimately will mean. To watch the series and explore the full archive of LLF-supported researcher profiles, visit longerlife.org/meet-the-investigators. Learn more about LLF’s many recent activities in our September 2026 newsletter. The Longer Life Foundation is a collaboration between the RGA Foundation and Washington University School of Medicine in St. Louis, Missouri, that was established in 1998 to fund early-stage research into factors that influence longevity, health, and wellness.

Click here to read the newsletter

An RGA Foundation/WashU Medicine Collaboration

Pioneering Research in Cholesterol Metabolism: Dr. Max Petersen