The effectiveness of this season’s flu vaccine in Canada is 40% against medically attended infection with influenza A(H3N2) viruses, 37% against newly emerged and predominant subclade K of the H3N2 strain, and 31% against the H1N1 influenza A strain, an interim analysis estimates.

Researchers from the Canadian Sentinel Practitioner Surveillance Network (SPSN) conducted the test-negative study, which evaluated samples from patients aged one year or older who had acute respiratory illness. Community-based sentinel health care providers in Alberta, British Columbia, Ontario, and Quebec collected the specimens from October 26, 2025, to January 10, 2026, and the findings were published yesterday in Eurosurveillance.

Important mutations in H3N2 viruses 

In Canada, only trivalent (three-strain) flu vaccines were available, and nearly all publicly funded vaccines in SPSN provinces were inactivated and egg-based. Adjuvanted vaccine (those with added substances to enhance immune response) was also available for adults aged 65 years and older, and this age-group received high-dose vaccines in Ontario. 

This season’s H3N2 vaccine strain was updated to subclade J.2 but stayed the same since 2023-24 for H1N1 (clade 5a.2a.1, subclades C.1.1 and D) and since 2022-23 for influenza B (Victoria lineage; clade V1A.3a.2, subclade C).

This viral evolution culminated in sweeping predominance of subclade K, an antigenically distinct variant showing early surge in Europe and elsewhere, and leading to further threat assessments and vaccine mismatch concerns expressed by others.

“Canadian investigators used global genomic and antigenic surveillance data to highlight important mutations in circulating influenza A(H3N2) viruses,” the authors wrote. Antigens are molecules that trigger an immune response, in this case against viruses.

“This viral evolution culminated in sweeping predominance of subclade K, an antigenically distinct variant showing early surge in Europe and elsewhere, and leading to further threat assessments and vaccine mismatch concerns expressed by others,” the researchers added. 

Higher efficacy in ages 18 to 29 years than 30 to 49 

Of 4,873 samples, 44% tested positive for influenza. Overall vaccine effectiveness (VE) against medically attended H3N2 was 40% at a time when infections were primarily driven by subclade K. For sequenced viruses, VE was comparable for subclade K (37%) and K-like J.2.4 viruses, at 32%, although the confidence interval (CI) for the latter finding ranged below zero (95% CI, –25% to 63%), which typically indicates the results are not statistically significant. 

Estimated VE against H3N2 by age-group was 36% for ages 1 to 17 years (95% CI, 10% to 55%) and 48% (95% CI, 33% to 60%) for 18 to 64 years, similar to subclade K viruses. Estimates suggested higher VE for adults aged 18 to 29 years (63%; 95% CI, 32% to 80%) than for those aged 30 to 49 (30%; 95% CI, 0% to 51%) but similar efficacy among those 50 to 64 (56%; 95% CI, 29% to 73%). 

VE was lowest in adults aged 65 years and older (25%; 95% CI, –7% to 48%). Estimated VE against H1N1 was 31% (95% CI, 3% to 50%), but the authors noted a small sample size and wide CIs with age grouping.

91% of H3N2 viruses vaccine-mismatched 

Nearly all (99%) of viruses were influenza A. A total of 86% of the 1,959 subtyped influenza A specimens were H3N2 only, while 13% were H1N1 only, and two were coinfections (one H3N2 plus B and one H3N2 plus H1N1). All 835 sequenced H3N2 viruses were clade 2a.3a.1, with 87% being subclade K.

The vast majority of H3N2 viruses (91%) were vaccine-mismatched. Of the 12 antigenically characterized viruses that were also genetically sequenced, 10 non–subclade K variants were vaccine-matched, except for two of 12 J.2.3 viruses, which have two more mutations than the J.2 vaccine strain. All 79 subclade K and K-like viruses were vaccine-mismatched.

Of the 264 sequenced H1N1 viruses, 99% were vaccine clade 5a.2a.1 but belonged to descendant subclades D.3.1 (28%) and D.3.1.1 (72%). All but one of the 53 H1N1 viruses that underwent antigenic characterization were vaccine-matched.

Potential role for pre-existing immunity

Since 2012-13, the SPSN has produced eight mid-season estimates of H3N2 VE: two were less than 20%, three were 40% to 45%, and three were 54% to 62%, placing this season’s VE in the low to mid range, the authors said.

Interim findings reinforce the importance of incorporating immuno-epidemiological data into influenza risk assessment, vaccine strain selection, and VE interpretation.

“Our 2025/26 interim A(H3N2) estimates are lower than early reports from England (weeks 40–44) and Europe (weeks 41–49),” they wrote. “Nonetheless, all VE estimates to date may be higher than anticipated from virological assessments.” 

The researchers noted the substantial mismatch between subclade K and cell-based flu vaccines and the potentially even greater mismatch with the egg-based J.2 vaccine due to an additional mutation arising from egg-adaptation of the high-growth reassortant (hybrid virus) used in inactivated vaccines.

The vaccine protection “may be inconsistent with genetic and antigenic indicators of substantial vaccine mismatch, but age-related variation in participant profiles signals a potential role for pre-existing immunity as an effect modifier,” they wrote. “Interim findings reinforce the importance of incorporating immuno-epidemiological data into influenza risk assessment, vaccine strain selection, and VE interpretation.”