Ebola Reaches Nigeria, AI Eyes Pancreatic Cancer Early, and a New Hepatitis B Drug Arrives

A week that began with alarm in Abuja ends with cause for guarded optimism on other fronts: a new drug class arrives for hepatitis B, an artificial-intelligence tool may crack one of oncology's hardest early-detection problems, and a long-contested hypothesis about aspirin and the brain gets a precision-medicine twist. Here is what moved in global health this week.

Ebola Passes a New Threshold: Nigeria Reports Its First Case

The Bundibugyo Ebola outbreak, declared a Public Health Emergency of International Concern (PHEIC) by the WHO on 8 October, has now claimed more than 9,700 confirmed cases and 4,687 deaths across the Democratic Republic of the Congo — a crude case fatality ratio of 48.3 percent. As of 20 October, eight provinces and 66 health zones in the DRC had been affected.

The geography deteriorated further last week. Nigeria confirmed its first case on 20 October: a woman in Abuja who had recently arrived from Uganda. Kenya, which had already reported eight imported cases, now joins Nigeria as the outbreak's newest export destination. Cameroon, Benin and Madagascar have all tightened border health screening, and none has reported confirmed transmission.

The Bundibugyo strain carries a particular sting for responders: unlike the Zaire strain that anchored the 2014–16 West Africa epidemic, it has no approved vaccine and no specific antiviral treatment. With the virus now circulating in a densely populated federal capital and in two countries outside the original DRC epicentre, global health authorities face a harder containment calculation than they did even a fortnight ago. The WHO has urged affected and neighbouring nations to accelerate contact-tracing and border-surveillance capacity.

AI System REDMOD Flags Pancreatic Cancer Years Before Scans Can

Pancreatic ductal adenocarcinoma kills roughly 90 percent of patients within five years of diagnosis, largely because by the time symptoms appear the disease has already metastasised. A new artificial-intelligence system called REDMOD may change that calculus.

Researchers reporting this week found that REDMOD identified hidden molecular warning signs in patients who went on to develop pancreatic cancer, flagging them an average of several years before the tumour became visible on conventional CT scans or MRI. The system analyses patterns in routine blood tests and imaging metadata — data already collected in standard clinical care — meaning deployment would not require expensive new diagnostic infrastructure.

Pancreatic cancer's stealth is partly explained by the tumour microenvironment's ability to suppress immune recognition; separate research reported concurrently showed how cancer cells reorganise chromatin structure to resist chemotherapy. REDMOD is not yet in clinical use and must validate in large, independent prospective cohorts. But if it performs outside its development dataset as well as it did within it, earlier interception of the disease could meaningfully shift five-year survival rates for one of oncology's most refractory cancers.

FDA Clears Bepirovirsen as First Antisense Therapy for Chronic Hepatitis B

An estimated 296 million people worldwide live with chronic hepatitis B infection, and existing antivirals — though effective at suppressing viral replication — rarely clear the infection entirely. The US Food and Drug Administration changed the treatment landscape on 21 October by approving bepirovirsen (brand name Hibsago), developed by GSK.

Bepirovirsen is an antisense oligonucleotide: rather than blocking viral enzymes, it targets the messenger RNA produced by the hepatitis B virus and silences it at the genetic level. The FDA approved it as an add-on to existing nucleoside or nucleotide analogue therapy for non-cirrhotic adults with chronic HBV infection.

The approval signals a shift in what hepatitis B treatment can aspire to. Current standard-of-care suppresses the virus for life but rarely eliminates it. A drug that acts on the virus's genetic machinery opens a path toward what hepatitis specialists call a "functional cure" — sustained suppression that persists after treatment ends. Whether that goal is achievable at a population scale with bepirovirsen alone, or whether combination regimens will be needed, will be the focus of ongoing clinical development.

Low-Dose Aspirin Cuts Dementia Risk 70% — But Only in One Genetic Subgroup

For most of the past decade, aspirin for primary prevention has been in retreat. Large randomised trials in otherwise healthy older adults found that bleeding risk outweighed cardiovascular benefit, and clinical guidelines across the US, UK and Europe walked back longstanding recommendations. A study published this week complicates that consensus in an important way.

Researchers found that older adults carrying a specific genetic variant had a 70 percent lower relative risk of dementia when taking low-dose aspirin compared with non-users who shared the same variant. No significant protective effect was found in people without that variant, and adverse effects remained a concern in that group.

The result is an early but striking example of pharmacogenomics applied to dementia prevention: the same drug that looks neutral or harmful at the population level may be highly beneficial for a biologically distinct subgroup. Clinical translation is still years away — the variant must be replicated and the mechanism explained, and a confirmatory trial would need to enrich enrolment using genetic screening. But the finding reopens a door that many had considered firmly shut.

Lorundrostat Cuts Blood Pressure 9.6 mmHg in Resistant Hypertension With CKD

Resistant hypertension — blood pressure that remains above target despite three or more medications — affects millions of patients with chronic kidney disease and dramatically raises their risk of heart attack, stroke, and accelerated kidney failure. Phase 3 Launch-HTN trial data presented this week showed that lorundrostat, an aldosterone synthase inhibitor, reduced systolic blood pressure by 9.6 mmHg versus placebo in this population.

Aldosterone, the hormone that lorundrostat inhibits at its source, drives salt retention and vascular stiffness. Existing mineralocorticoid receptor antagonists like spironolactone block aldosterone's action downstream, but they carry a meaningful risk of hyperkalaemia — dangerously elevated potassium — in CKD patients, which has historically limited their use in this specific population. Aldosterone synthase inhibition, which prevents aldosterone production rather than blocking its receptor, may offer comparable blood-pressure benefit with a more manageable electrolyte profile.

If lorundrostat's regulatory submission proceeds and is approved, it would be the first new mechanistic approach to resistant hypertension in CKD in over a decade — and a benchmark against which future aldosterone-targeting drugs will be measured.