Journal Publication

Pooled data show revakinagene taroretcel slows MacTel type 2 progression over 24 months

A 295-patient pooled analysis of three sham-controlled trials found the gene therapy cut ellipsoid zone loss, reading speed decline, and retinal sensitivity loss versus sham.

A pooled analysis of one Phase 2 and two Phase 3 sham-controlled trials of revakinagene taroretcel-lwey in macular telangiectasia type 2 (MacTel) reported reductions in anatomic and functional disease progression over 24 months.

By AppliedXLJuly 20, 2026

Executive Summary

A pooled analysis combined three sham-controlled trials of a cell-based gene therapy in macular telangiectasia type 2, a retinal disease with no approved treatment that slows anatomic degeneration.
Treated eyes lost less photoreceptor-supporting tissue and retinal sensitivity over two years than sham-treated eyes, and lost less reading speed, though visual acuity barely moved in either group.
The three trials it draws on are the drug's own registration program rather than independent confirmation, and the therapy's benefit shows up on structural and functional surrogates rather than on the acuity measure patients and clinicians weight most.

The question

Macular telangiectasia type 2 (MacTel) is a retinal disease that progressively degrades the photoreceptor-supporting ellipsoid zone, eroding reading speed and retinal sensitivity even when visual acuity itself stays relatively stable. The pooled analysis set out to combine anatomic and visual-function data across a single drug's full sham-controlled trial program to test whether revakinagene taroretcel-lwey, a cell-based therapy, changes that trajectory. Pooled

How it was done

The analysis pooled one Phase 2 trial (NTMT-02, 67 participants, 99 eyes) and two identically designed Phase 3 trials (NTMT-03-A, 115 participants and eyes; NTMT-03-B, 113 participants and eyes), all randomized and sham-surgery-controlled. The combined population totaled 327 eyes from 295 participants, 165 who received revakinagene taroretcel-lwey and 162 who underwent the sham procedure. Investigators assessed both the individual trials and the single pooled population for the rate of ellipsoid zone (EZ) area loss over 24 months, alongside changes from baseline in monocular reading speed, aggregate retinal sensitivity measured by microperimetry, and best-corrected visual acuity (BCVA). Pooled

The anatomic result

Treated eyes lost ellipsoid zone area at 0.118 mm2 over 24 months (95% CI, 0.087 to 0.149) versus 0.185 mm2 with sham (95% CI, 0.154 to 0.217), a 36.2% reduction in the rate of loss with a nominal P value of 0.003. Aggregate retinal sensitivity loss, measured by microperimetry across the three trials, was 28.47 decibels with revakinagene taroretcel-lwey versus 44.85 decibels with sham, a 36.6% reduction with a nominal P value below 0.001. Pooled

The functional result

Monocular reading speed fell by 4.65 words per minute from baseline in the treated group versus 14.60 words per minute with sham at 24 months, a 68.2% reduction in reading-speed loss. The treated-arm confidence interval spanned -9.58 to 0.27 words per minute, crossing zero, while the sham interval spanned -20.41 to -8.79 words per minute and stayed negative throughout, indicating a more consistent decline in the control arm. Best-corrected visual acuity changed minimally in both groups over the 24-month period. Pooled

What it changes

The pooled dataset extends revakinagene taroretcel-lwey's evidence base beyond any single trial by combining the drug's Phase 2 and both Phase 3 programs into one 327-eye analysis, giving the anatomic and sensitivity findings more statistical support than any individual trial could carry alone. The gap between the anatomic and sensitivity results, both nominally with narrow confidence intervals, and the acuity result, essentially flat in both arms, indicates the therapy's measurable effect concentrates in structural preservation and functional tests more sensitive than acuity, rather than in acuity itself. Pooled

This analysis was produced using AI-assisted reporting systems, AppliedXL data, and official public records. These systems undergo editorial review, quality checks, and regular audits by human experts. Errors may still occur, as with any automated system. Always consult the linked primary sources. Read our AI Editorial Policy.