Identification of a Novel Subtype of Ocular Hypertensive Eyes with the Lowest Risk of Progression Under Treatment

Asma Poursoroush; Vida Abedi; Xiaoqin Huang; Jiang Li; Alireza Vafaei Sadr; Amy L de Jongh Curry; Louis R Pasquale; Michael V Boland; Chris A Johnson; Siamak Yousefi

doi:10.1007/s40123-025-01174-3

Identification of a Novel Subtype of Ocular Hypertensive Eyes with the Lowest Risk of Progression Under Treatment

Ophthalmol Ther. 2025 Aug;14(8):1893-1903. doi: 10.1007/s40123-025-01174-3. Epub 2025 Jun 28.

Authors

Affiliations

¹ Department of Biomedical Engineering, University of Memphis and University of Tennessee Health Science Center, Memphis, USA.
² Department of Public Health Sciences, College of Medicine, The Pennsylvania State University, Hershey, PA, 17033, USA.
³ Department of Ophthalmology, University of Tennessee Health Science Center, 930 Madison Ave, Suite 471, Memphis, TN, 38163, USA.
⁴ Department of Molecular and Functional Genomics, Geisinger Health System, Danville, PA, 17822, USA.
⁵ Uniklinik RWTH Aachen, Aachen, North Rhine-Westphalia, Germany.
⁶ Department of Ophthalmology, Icahn School of Medicine at Mount Sinai, New York, USA.
⁷ Department of Ophthalmology, Massachusetts Eye and Ear, Harvard Medical School, Boston, USA.
⁸ Department of Ophthalmology and Visual Sciences, University of Iowa Hospitals and Clinics, Iowa City, IA, USA.
⁹ Department of Ophthalmology, University of Tennessee Health Science Center, 930 Madison Ave, Suite 471, Memphis, TN, 38163, USA. siamak.yousefi@uthsc.edu.
¹⁰ Department of Genetics, Genomics, and Informatics, University of Tennessee Health Science Center, Memphis, TN, USA. siamak.yousefi@uthsc.edu.

Abstract

Introduction: Our aim in this work is to identify a subtype of ocular hypertensive eyes that benefited the most from topical hypertensive medication.

Methods: We included 1636 patients (817 patients in the active group and 819 in the observation group) who participated in the Ocular Hypertension Treatment Study (OHTS). We developed principal component analysis (PCA), t-distributed stochastic neighbor embedding (t-SNE), and unsupervised density-based clustering, then applied to the baseline visual fields (VFs) to identify different clusters of eyes. For each cluster, the rate of mean deviation (MD) worsening was calculated based on 7 years of follow-up. To determine which cluster (subtype) of eyes responded better to the ocular hypertensive medications, MD worsening was compared between the active and observation groups within each cluster of eyes separately.

Results: We identified 13 clusters based on visual field (VF) tests at baseline visits. Among these, one cluster demonstrated a significantly better response to ocular hypotensive medication compared to the others. Key distinguishing characteristics of this cluster included a lower frequency of family history of glaucoma (p = 0.006) but a higher prevalence of heart disease (p < 0.001) and high blood pressure (p < 0.001). Interestingly, although eyes in this cluster had a worse baseline MD than those in all other clusters (- 2.2 ± 0.1 dB vs. + 0.47 ± 1.2 dB, p < 0.001) and patients were older (64.3 ± 9.5 vs. 54.7 ± 8.94 years, p < 0.001), this was the only cluster where the MD progression rate in the treatment (active) group was significantly lower than in the observation group (+ 0.04 dB/year vs. - 0.09 dB/year, p = 0.03). Furthermore, treated eyes in this cluster had a significantly better MD progression rate than treated eyes in all other clusters (+ 0.04 dB/year vs. - 0.07 dB/year, p < 0.001).

Conclusions: We identified a subtype of ocular hypertensive eyes in the OHTS that showed a better response to topical hypotensive medications, as indicated by the least risk of progression.

Keywords: Clustering; Glaucoma; Ocular hypertension; Personalized medicine; Unsupervised machine learning.

Grants and funding

R01 EY033005/EY/NEI NIH HHS/United States