CYP2C19 Genotype Did Not Predict Escitalopram Outcomes in a Post-Stroke Depression Study

A 2026 study from China found no significant difference in eight-week escitalopram outcomes between patients with post-stroke depression classified as slower CYP2C19 metabolizers and those classified as normal or rapid metabolizers.

The study included 148 people treated with escitalopram. Median depression-score reduction was 10 points in the slower-metabolizer group and 11 points in the normal-or-rapid group. The difference was not statistically significant.

What the study found

Researchers analyzed adults admitted to the neurology department of Tongji Hospital in Wuhan after a stroke. Participants had a HAMD-17 depression score of at least 14 and had not previously received antidepressant treatment.

Among the 82 patients in the slower CYP2C19 group, 50—61.0%—met the study’s definition of response after eight weeks. Response occurred in 46 of 66 patients—69.7%—in the normal-or-rapid group. This difference was not significant.

Remission rates were also similar: 48.8% in the slower group and 51.5% in the normal-or-rapid group. An adjusted analysis likewise found no association between CYP2C19 group and symptom reduction.

At week four, response occurred in 41.5% of slower metabolizers and 33.3% of normal-or-rapid metabolizers. That difference was not significant either.

How the groups were defined

Of the escitalopram-treated patients, 18 were classified as poor, 64 as intermediate, 63 as normal and three as rapid CYP2C19 metabolizers. The researchers combined poor and intermediate metabolizers into a slower group, and normal and rapid metabolizers into a second group.

Escitalopram is described in the study as being broken down mainly by CYP2C19, a gene-related enzyme pathway. The researchers examined whether differences in that pathway were associated with symptom improvement, response or remission while patients received a fixed 10-milligram daily dose.

The study did not compare blood levels or side effects by genotype, so it does not show whether CYP2C19 status affects tolerability or the most suitable dose.

Other factors associated with outcomes

The analysis identified some associations with treatment outcomes that were separate from CYP2C19 grouping. Each one-point increase on the modified Rankin Scale was associated with lower odds of response and remission. Each additional baseline HAMD-17 point was associated with higher odds of response in the escitalopram group.

These findings describe statistical associations within the study. They do not establish that stroke-related disability or initial depression severity caused a particular treatment outcome.

Why the findings need careful interpretation

The study enrolled 313 patients and analyzed 261 adults aged 18 to 75. It was conducted at one hospital, was observational rather than randomized, and had no placebo group. Treating doctors selected whether patients received escitalopram or venlafaxine, so the study cannot provide a randomized comparison of the two medicines.

The CYP2C19 comparison also included relatively few people at the extremes. Only 18 patients were poor metabolizers and three were rapid metabolizers, while no patients were ultrarapid metabolizers. Combining metabolizer categories may have obscured an effect limited to a smaller subgroup.

The results came from hospitalized Han Chinese adults with moderate-to-severe post-stroke depression. People with aphasia or major cognitive problems were excluded, and most follow-up ratings were conducted by phone or video call. The findings may therefore not apply to all people who develop depression after a stroke.

The fact sheet also reports an inconsistency in the number of normal-or-rapid metabolizers: the abstract lists 64, while the results text and main table list 66. The reported analyses used 66.

What this means for pharmacogenomic testing

The researchers concluded that gene testing may not yet be sufficiently useful on its own for selecting antidepressants in this setting, favoring practical and lower-cost approaches over costly testing alone.

That conclusion is consistent with the study’s narrow design, but it does not prove that CYP2C19 testing has no value. The observational data cannot rule out a modest effect, an effect on side effects or dosing, or differences that might emerge in other populations.

Other research cited by the source has also been mixed. In the GENDEP study, CYP2C19 type was associated with escitalopram blood levels but did not predict who improved. A 2024 meta-analysis found somewhat higher remission rates among CYP2C19 poor metabolizers than normal metabolizers, but the result did not remain significant after correction for multiple testing. In the PRIME Care trial, pharmacogenomic results changed prescribing away from medicines with predicted gene interactions, while the improvement in remission was modest and was not significant at 24 weeks.

For readers, the central finding is narrower than a general verdict on genetic testing: in this particular hospitalized post-stroke population, the CYP2C19 categories studied were not significantly associated with escitalopram response or remission over eight weeks.

AI tools were used to assist with the preparation of this article under our editorial standards.