Retinal Vascular Abnormalities Related to Neurofibromatosis Type 1

Abstract

Purpose: To analyze and classify neurofibromatosis Type 1 (NF1)-related retinal vascular abnormalities (RVAs), their natural history and correlation with disease severity, in a large cohort of patients. Methods: This was an observational longitudinal study with prospective enrollment. Four hundred and seventy-three patients affected by NF1 and 150 age-matched healthy subjects were consecutively enrolled. Retinal vascular abnormalities were detected by means of near-infrared reflectance and studied by optical coherence tomography angiography. The superficial vascular plexus and the deep vascular complex (DVC) were quantitatively and qualitatively analyzed. Results: We identified RVAs in 82 of 473 (17%) NF1 patients, but in none of the 150 healthy subjects. A comparison revealed that NF1 patients with RVAs showed a higher number of NF1 diagnostic criteria (4.3 ± 1.5 vs. 3.9 ±1.5, respectively; P = 0.02) than patients without RVAs. Three different RVA types were identified on optical coherence tomography angiography: macrovascular angiomatosis of the sole superficial vascular plexus; macrovascular angiomatosis of the superficial vascular plexus combined with microvascular angiomatosis of the deep vascular complex; and combined macrovascular angiomatosis of both superficial vascular plexus and deep vascular complex. The prospective analysis of optical coherence tomography angiography images showed no significant longitudinal evolution of RVAs (mean follow-up: 3.7 ± 2.8 years). A single patient developed a de novo single RVA, and two RVAs showed detectable changes during follow-up. Conclusion: In NF1 patients, RVAs are a characteristic sign that correlates with a more severe systemic disease expression, usually remaining stable during time. Optical coherence tomography angiography allows for the identification of different RVAs subtypes.

Retinal Vascular Abnormalities in Children With Neurofibromatosis Type 1

“…2,6,10 Parrozzani et al found a prevalence of 6.1% in 2018 and 17% in 2020, but this can probably be explained by the fact that they only reported corkscrew pattern abnormalities. 13,14 This prevalence is actually similar to the prevalence of corkscrew pattern abnormalities in the present series26.4%. Moramarco et al 6 recently assessed RVAs in 334 patients, including 57 children younger than 16 years.…”

“…No correlation between choroidal abnormalities and RVAs was found, which is similar to both Parrozzani et al's series. 13,14 This does not support the hypothesis suggested by Abdolrahimzadeh et al, 10 according to which RVAs are overly choroidal abnormalities. We found a positive association between LNs and RVAs, like Parrozzanni et al 13,14 The age was used as a covariate to take into account its potential confounding effect.…”

“…13,14 This does not support the hypothesis suggested by Abdolrahimzadeh et al, 10 according to which RVAs are overly choroidal abnormalities. We found a positive association between LNs and RVAs, like Parrozzanni et al 13,14 The age was used as a covariate to take into account its potential confounding effect. Lisch nodules are found in 90% to 100% of adults with NF1 and are among the most consistent feature of NF1 15 ; their diagnostic value has been established long ago.…”

“…This study confirms the recently reported fact that RVAs appear and evolve over time. 11,14 We hypothesize that NF1 leads to an overall trend to tortuous vessel (arteries/veins), but that only small venules can evolve into a corkscrew pattern, because of their more deformable properties. Retinal vascular abnormalities in NF1 therefore seem to form a spectrum of vessel tortuosity, from a simple vascular tortuosity to a highly tortuous “venous mesh.” Their natural history can probably explain why the most severe forms were only found in series of patients older than those included in the present one.…”

Correspondence

“…Thus, the natural history of RVAs may explain why the most complex forms were only found in series of patients older than those included in Touzé et al 1,2 This questions the reliability of previous descriptions of congenital and stable nature of RVAs. 3 Interestingly, the prevalence of RVAs evolving over time presented by Touzé et al 1 differs from that of Parrozzani et al 4 reporting only one patient of 473 to develop de novo RVAs and two patients showing progressive changes during the follow-up. This may be related to the higher mean age of the study participants in Parrozzani et al 4 compared with Touzé et al 1 (mean age: 17.8 years vs. 8.8 years), suggesting that variations in the RVAs may have developmental features in children with a tendency to stabilize in the more advanced age.…”