Alumis Inc. (NASDAQ:ALMS) reported that the 408-patient Phase 2b LUMUS trial of envudeucitinib in moderate-to-severe systemic lupus erythematosus did not meet its primary or secondary endpoints in the overall population. Envudeucitinib is an investigational oral inhibitor of tyrosine kinase 2, or TYK2.
The primary endpoint was the British Isles Lupus Assessment Group-based Composite Lupus Assessment, or BICLA, response at Week 48. The 20-milligram once-daily, 20-milligram twice-daily and 40-milligram twice-daily regimens produced response rates of 40.0%, 42.2% and 41.0%, respectively, compared with 35.7% for placebo. The corresponding p-values were 0.5018, 0.3455 and 0.3740.
Alumis Inc. nevertheless plans end-of-Phase 2 meetings with the FDA and European Medicines Agency after a prespecified subgroup with a high interferon gene signature, or IFNGS-high, showed numerically higher responses. Those discussions will address the Phase 3 dose, design, size, and biomarker-based enrollment criteria.
Bull Case
The subgroup was defined before the analysis and identified at baseline using a commercially available four-gene mRNA assay. IFNGS-high patients represented approximately 59% to 63% of each LUMUS treatment group. Alumis Inc. said this was below the approximately 70% generally expected among moderate-to-severe lupus patients.
Within the IFNGS-high subgroup, the 20-milligram once-daily, 20-milligram twice-daily and 40-milligram twice-daily regimens produced numerical BICLA response rates of 40.7%, 49.5% and 52.6%, respectively, compared with 28.6% for placebo. The covariate-adjusted numerical differences were 11.6, 21.3 and 24.0 percentage points.
All three active groups also numerically exceeded placebo on SRI-4, CLASI-50, low disease activity, joint-count improvement, and corticosteroid-reduction measures in IFNGS-high patients. The direction across different manifestations supports the biomarker hypothesis rather than relying on one favorable endpoint.
Pharmacodynamic data showed dose-dependent interferon-pathway target engagement, with maximal suppression at 40 milligrams twice daily. Envudeucitinib was generally well tolerated, and Alumis Inc. reported no new safety signals.
Bear Case
The randomized trial still failed in the population it was designed to test. Subgroup selection can support another study, but it does not convert the overall LUMUS result into a successful Phase 2b trial.
The subgroup evidence remains incomplete. Alumis Inc. disclosed numerical response rates and adjusted BICLA differences but did not provide subgroup-specific confidence intervals or p-values. Investors therefore cannot fully assess statistical precision, multiplicity, or whether the apparent treatment-by-biomarker interaction is reproducible.
The dose pattern was not uniform across endpoints. The lowest daily dose generated the strongest SRI-4 and joint-count responses, while the highest dose produced the strongest BICLA result. That complicates dose selection despite the clear pharmacodynamic dose-response.
Placebo behavior is another issue. Among IFNGS-low patients, the BICLA placebo response was 47.5%, compared with 39.0%, 30.0% and 24.4% for the active groups. Alumis Inc. said the higher-than-expected IFNGS-low share drove a high placebo response and reduced overall efficacy. The subgroup pattern is consistent with that explanation but does not establish that patient mix caused the overall miss.
A biomarker-enriched trial would also narrow enrollment, require regulatory agreement on the assay, and add development time and cost.
Hedge Fund Sentiment
The filings available so far reflect positions held before Alumis Inc. reported the LUMUS Phase 2b results. Insider Monkey’s database showed 41 hedge funds holding Alumis Inc. at the end of 2Q2026, down from 46 funds three months earlier.
Conclusion
The IFNGS-high results provide a credible basis for regulatory discussions because the subgroup was prespecified and the numerical BICLA difference increased with dose. However, the proposed path is a biomarker-enriched development strategy after an overall trial failure. A Phase 3 program becomes compelling only if regulators accept the enrichment approach and complete subgroup statistics support a reproducible, properly powered design.
READ NEXT: Main Street Capital’s (MAIN) Blowout Exit Fuels A Bigger Dividend and Here is Why Chevron (CVX) is a Favorite Among Hedge Funds
This article is originally published at Insider Monkey.
Hedge Fund Resource Center