P-861: On-line program for a novel individual classification of female functional androgenization (FA) using algorithmic group-specific parameter clusters

Abstract

FA consists of various heterogenous entities. Therefore, a novel FA classification which defines underlying dysfunctions by a new sonographic functional ovary score (OS) (a) and a compact biochemical testing procedure has been developed (b). In order to strengthen the reproducibility, an on-line classification program, FuncAndroAid, was created using a strict algorithm. Software for a revised classification of FA patients analysed in a retrospectively controlled study. 126 patients (25.8 ± 5.3 yrs) with cutaneous androgenizing symptoms (CAS), hyperandrogenemia (HA), hyperinsulinemia (HI) and/or obesity (Ob) (b) were re-classified using an algorithm of group-specific parameter clusters. BMI, CAS, OS, LH/FSH, testosterone (T), free T/SHBG, prolactin (PRL) and lipoproteins (LPs), as well as, glucose (Gluc) and insulin (Ins) prior to (0), and 60 min after (1) an oral glucose loading were determined inter alia. FA was subdivided into 5 groups termed as: Functional Cutaneous Androgenization (FCA): cutis; Functional Androgenizing Syndrome (FAS) I to IV: ovary; adrenal gland; HA, HI and Ob; miscellaneous. Grouping was performed through a group-specific obligate 1stcategory parameter core cluster (1stCP-CC) (subgroup B) and a 1stCP-full blown cluster (1stCP-FBC) (subgroup A); e. g.: CAS and otherwise normally ranging 1stCPs conditioned subgroup A of FCA. Group′s further characterization was carried out by secondary 2ndCPs (e.g. PRL, LPs). The data of 8 volunteers (29.7 ± 4.7 yrs) served as a control (C). The web application is based on PHP language and a MySQL database. After anonymously submitting patients‘ data on the website, they were checked to see if they match a specific diagnostic group and then stored into the database for further use. Grouping is shown in Table 1. For example, FAS III revealed per defintionem significantly elevated T and OS values vs. C (P<.0001; Wilcoxon Two-Sample test) and the highest levels of BMI, FT/SHBG, Gluc 1 and Ins 0/1 vs. FCA, FAS I, FAS II and C (P<.0001-.03). Matched data showed full agreement indicating the software′s sufficient reliability. Table 1: Results of patients’ revised classificationTabled 1*Values are mean ± SD; data of FAS IV (n=18), subgroups A and B, and further 1st CPs are not shown.**Statistics of FAS III as an example: see under RESULTS. *Values are mean ± SD; data of FAS IV (n=18), subgroups A and B, and further 1st CPs are not shown. **Statistics of FAS III as an example: see under RESULTS. The revised FA classification using an exactly defined algorithm of parameter clusters served as data base for establishing the program (www.funcandroaid.com). This novel on-line program provides an indivudual, objective, standardized, reproducable and fast diagnostic approach for female FA. The software is developed modularly, including many possibilities for later extensions (e.g. therapeutic algorithm), and can be used for multi-purpose tasks. Furthermore, it could be useful for clinic and basic research. a.) Wetzka et al (2005) Hum Reprod 20, Suppl.1, i138. b.) Geisthövel et al (2005) Hum Reprod 20, Suppl.1, i174.