Phenotypic features, androgen receptor binding, and mutational analysis in 278 clinical cases reported as androgen insensitivity syndrome.

Androgen insensitivity syndrome (AIS) is the most common single entity that results in male under-masculinization, but large cohort studies of AIS have rarely been performed. Over the last decade, nationwide cooperation between pediatric endocrinologists in the United Kingdom has allowed the creation of a database of cases of intersex and ambiguous genitalia where detailed clinical information on every notified case has been collected via a questionnaire. Among the 816 entries recorded by January 1999, there were 105 clinically diagnosed cases of complete AIS (CAIS) and 173 cases of partial AIS (PAIS). A masculinization score was devised by scoring the external phenotype, and a score of 12 represented normal masculinization. Androgen receptor (AR) binding was determined by studying binding capacity (Bmax) and receptor affinity (K(d)), and cases were classified as either zero, abnormal, or normal binding. Mutation screening of all eight exons of the AR gene was performed by single-strand conformational polymorphism analysis, followed by direct DNA sequencing. All cases of PAIS presented within the first month of birth. The median age at presentation of children with CAIS was 1 yr (P10,P90: 0.1,10.4). The testes were palpable in the labioscrotal folds or the inguinal region in 77% and 41% of cases of CAIS and PAIS, respectively. There was marked overlap between the masculinization score of those children with PAIS reared as girls [2.5(P10,P90:1, 6)] and those reared as boys [3(P10,P90:2, 7.5)]. Gonadectomy was performed prepubertally in 66% and postpubertally in 29% of the cases of CAIS. The median age of the latter group was older at 14 yr (P10,P90:0.1,18). No cases of malignancy or carcinoma in situ were reported in the 121 cases of AIS where histology results were available. Biochemical endocrine investigations were reported to have been performed in a greater number of cases of PAIS than CAIS (98% vs. 48%). AR binding was abnormal in 44 of 51 (86%) and 40 of 113 (35%) cases of CAIS and PAIS, respectively. Zero binding was encountered in 29 of 43 (67%) and 1 of 55 (2%) cases of CAIS and PAIS, respectively. Mutational analysis of the AR gene, performed in 102 index cases was positive in 57 of 69 (83%) cases of CAIS and 12 of 43 (28%) cases of PAIS. In 24 of these cases, the mutation identified was novel. The mutations in PAIS cases were all missense, whereas in CAIS the mutations were more diverse. AR binding was only normal in 3 of 69 mutation-positive cases. In the PAIS group, mutation-positive cases had a significantly higher Kd and Bmax compared to the mutation negative cases. The clinical diagnosis of AIS can be confirmed in a significant number of cases by a combination of androgen-binding studies and mutational analysis. There is some correlation between the phenotypic features and the abnormalities discovered on mutational analysis of the AR gene, but there is a need to improve this further by developing optimal bioassays of AR function. The phenotypic heterogeneity among clinically diagnosed cases of AIS emphasizes the need for appropriate comprehensive evaluation of male under-masculinization.