Frequency of colour vision deficiencies in melanoma patients: results of a prospective comparative screening study with the Farnsworth panel D 15 test including 300 melanoma patients and 100 healthy controls.

Patients with melanoma may experience a variety of different vision symptoms, in part associated with melanoma-associated retinopathy. For several melanoma patients with or without melanoma-associated retinopathy, colour vision deficiencies, especially involving the tritan system, have been reported. The frequency of colour vision deficiencies in a larger cohort of melanoma patients has not yet been investigated. The aim of this study was to investigate the frequency of colour vision deficiencies in melanoma patients subject to stage of disease, prognostic factors such as tumour thickness or Clark level, S100-beta and predisposing diseases that may have an impact on colour vision (hypertension, diabetes mellitus, glaucoma or cataract). Three hundred melanoma patients in different tumour stages and 100 healthy age-matched and sex-matched controls were examined with the saturated Farnsworth panel D 15 test. Seventy out of 300 (23.3%) melanoma patients and 12/100 (12%) controls showed pathologic results in colour testing. This discrepancy was significant (P < 0.016; odds ratio = 2.23, 95% confidence interval 1.15-4.32). Increasing age was identified as a highly significant (P = 0.0005) risk factor for blue vision deficiency. Adjusting for the age and predisposing diseases, we could show that melanoma was associated with the risk of blue vision deficiency. The frequency of blue vision deficiency in 52/260 melanoma patients without predisposing diseases (20%) compared with 4/78 controls without predisposing diseases (5.1%) differed significantly (odds ratio 4.441; confidence interval 1.54-12.62; P < 0.004). In 260 melanoma patients without predisposing diseases, blue vision deficiency, as graded on a 6-point scale, showed a weak positive correlation (Spearman) with tumour stage (r = 0.147; P < 0.01), tumour thickness (r = 0.10; P = 0.0035), Clark level (r = 0.12; P = 0.04) and a weak negative correlation with time since initial diagnosis (r = -0.11; P = 0.0455). Blue vision deficiency is associated with melanoma, but is only weakly related to stage of disease. Although we saw a positive correlation with well-known prognostic markers, such as tumour thickness and Clark level, blue vision deficiency as assessed by the Farnsworth panel D 15 test in general is inappropriate as a marker of tumour progression. For the use of blue vision deficiency in melanoma patients without predisposing diseases, a diligent test performance and interpretation is very important.