Subclinical visual impairment in phenylketonuria. A neurophysiological study (VEP-P) with clinical, biochemical, and neuroradiological (MRI) correlations.

Abstract

During detailed visual function testing, pattern-reversal visual evoked potentials (VEP), generated by different spatial frequencies (3 c/d, 1 c/d and 0.6 c/d) and visual contrasts (100% and 10%) were recorded in 21 adolescent and young adult phenylketonuric (PKU) patients (11 females and 10 males; mean age 14.8 years, range 9-22.8) on and off diet. In 14 of the 21 patients, disease had been detected at neonatal screening and in 7 later. Ten age-matched healthy subjects acted as controls. Recordings in more than 40% of eyes in the whole group and 30% of eyes in the screening subgroup showed a prolonged P100 latency. All visual pattern stimuli elicited a significantly longer P100 latency in PKU patients than in controls. VEP latencies to 3 c/d, 1 c/d and 1 c/d with 10% contrast--but not to 0.6 c/d--were longer in patients off diet than in patients on diet. No differences were found between VEP latencies in early- and later-detected subjects. To study the link between biochemical variables and VEP latencies, we envisaged either a linear relationship between recent exposure to phenylalanine (Phe) and VEP abnormalities or a threshold model considering phenylalanine (Phe) concentrations among the factors influencing VEP latencies. The correlation analysis detected an association between plasma Phe concentrations and abnormal VEP latencies, predicting that plasma Phe concentrations > 901 mumol/L would prolong VEP latencies to 1 c/d; concentrations > 879 mumol/L would prolong latencies to 3 c/d; and concentrations > 898 mumol/L would prolong latencies to 1 c/d with 10% contrast.