In advanced thyroid cancer, lenvatinib offers similar survival but more hypertension and proteinuria vs other systemic therapies, real-world analysis suggests
Presenter: Eleana Estrella, DO, University of Central Florida, Orlando, FL
Five-year survival and toxicity of lenvatinib versus other systemic therapies in thyroid cancer: a propensity-matched real-world TriNetX study. Presented April 23, 2026.
Among patients with advanced thyroid cancer treated in real-world settings, lenvatinib conferred a survival benefit similar to other systemic therapies, but with the toxicity profile of a vascular endothelial growth factor (VEGF) inhibitor observed in clinical trials, according to a large propensity-matched analysis presented at AACE 2026.1
At 5 years, the all-cause mortality rate was roughly 35%, both for patients treated with lenvatinib and for those treated with sorafenib, pembrolizumab, or cabozantinib, according to the analysis across more than 1,000 matched pairs drawn from the TriNetX federated electronic health record network.
While safety profiles were largely similar, lenvatinib was associated with significantly more new-onset hypertension and proteinuria, two adverse events that were characteristic of the drug in clinical trials.
Pivotal evidence for lenvatinib comes largely from the Study of Lenvatinib in Differentiated Cancer of the Thyroid (SELECT) trial,2 which was designed to test efficacy against placebo in a tightly defined population of patients with progressive, radioiodine-refractory differentiated thyroid cancer.
“Clinical trials like SELECT tell us what lenvatinib can do under tightly controlled conditions, but real patients are often older, have more comorbidity, receive dose reductions, and often have more heterogeneous disease biology and prior therapies,” Dr. Estrella said. “Real-world data help answer whether the trial signal remains recognizable once lenvatinib is used outside ideal trial conditions.”
Study design
Dr. Estrella and co-investigators identified adults age 18 to 80 with ICD-10 code C73 (thyroid cancer) in the TriNetX federated electronic health record network. The lenvatinib cohort received the drug within 1 year of diagnosis; the comparator cohort received sorafenib, pembrolizumab, or cabozantinib instead.
After propensity score matching on demographics and comorbidities, 1,133 patients remained in each group, with similar median follow-up (523 vs 505 days). Outcomes were assessed over 1,825 days and included all-cause mortality and prespecified toxicities: hypertension, renal impairment, hepatotoxicity, cardiac dysfunction, hemorrhage, diarrhea, and proteinuria.
Real-world efficacy
The 5-year all-cause mortality rate was 34.4% with lenvatinib and 34.8% with other therapies. Median survival was 1,672 days with lenvatinib versus 1,617 days in the comparator group.
The efficacy findings are notable because the real-world cohort was substantially broader than the SELECT trial population, Dr. Estrella said. SELECT enrolled 392 patients with a tightly defined clinical profile. By contrast, this analysis drew on matched patients treated across a federated electronic health record network, with wider variation in age, comorbidity, prior therapy, and follow-up intensity, according to the researcher.
Characteristic toxicities
Two adverse events separated lenvatinib from the comparator group: new-onset hypertension occurred in 35.0% of lenvatinib-treated patients versus 24.5% of comparator patients (risk ratio 1.43; hazard ratio 1.53; log-rank P < .001). Proteinuria occurred in 7.6% versus 3.8% (risk ratio 1.97; hazard ratio 1.99; log-rank P < .001).
Risks of renal impairment, hepatotoxicity, cardiac dysfunction, hemorrhage, and diarrhea were broadly similar between groups, with small differences in risk, confidence intervals that crossed 1, and log-rank tests that did not return statistical significance.
This adverse event pattern matched the known VEGF-inhibitor profile seen in SELECT, though absolute event rates were lower in this real-world cohort, according to Dr. Estrella.
In SELECT, hypertension occurred in 69% of lenvatinib-treated patients, with grade 3 or 4 hypertension in 43%, and proteinuria occurred in 32%; the lower rates in the TriNetX analysis likely reflect how toxicity is captured rather than how often it occurs, she explained.
“EMR-coded clinical events tend to under-capture adverse events compared with prospective trials, which helps explain why diarrhea, hypertension, and proteinuria rates here are much lower than in SELECT,” Dr. Estrella said. “It also means these toxicity endpoints depend on diagnosis coding behavior, not standardized CTCAE reporting.”
Clinical implications
Taken together, these findings suggest that in clinical practice, lenvatinib retains its recognizable VEGF-inhibitor toxicity signature, especially hypertension and proteinuria, Dr. Estella said.
However, in this matched observational analysis, it did not show a survival advantage over a mixed group of other systemic therapies, she added.
“The drug remains a valid and active option for advanced thyroid cancer, but its value in practice depends heavily on careful patient selection, close toxicity monitoring, and comparison against the right alternative for the right disease setting,” Dr. Estrella said.
Crucially, hypertension and proteinuria are well-characterized adverse effects of VEGF inhibitor therapy that can be routinely monitored and managed with standard interventions and dose modifications, Dr. Estrella and coauthors noted.
Disclosures
Dr. Estrella reported no disclosures related to the presentation.
References
- Estrella E, Papp SR, Gill P, Cream L, Al-Moussally F. Five-year survival and toxicity of lenvatinib versus other systemic therapies in thyroid cancer: a propensity-matched real-world TriNetX study. Endocrine Pract 2026; 32(suppl). https://doi.org/10.1016/j.eprac.2026.03.080
- Schlumberger M, Tahara M, Wirth LJ, et al. Lenvatinib versus placebo in radioiodine-refractory thyroid cancer. N Engl J Med 2015; 372(7):621–630. doi:10.1056/NEJMoa1406470