ESPE Abstracts (2014) 82 P-D-2-2-576

The Utility of AMH for Predicting Testosterone Response to hCG Stimulation in Children with Suspected DSD

Andreas Kyriakoua, Jane D McNeillyb, Guftar M Shaikha, Avril Masona, David Shapiroc & Syed Faisal Ahmeda


aDevelopmental Endocrine Research Group, University of Glasgow, Glasgow, UK; bDepartment of Biochemistry, RHSC, Glasgow, UK; cDepartment of Biochemistry, Glasgow Royal Infirmary, Glasgow, UK


Background: In children undergoing investigation of testicular function the relationship between serum anti-Mullerian hormone (AMH) and the testosterone response to hCG stimulation test (HST) is unclear.

Method: 75 children (3F, 72M) with a median age of 1.08 years (range, 0.003, 14.3) were investigated for suspected DSD by AMH on D1 and testosterone on D1 and D4, before and after 3-day HST. Of these children, 27 had an additional prolonged HST. Normal testosterone response to HST was defined as a T>upper prepubertal limit or a testosterone increment (ΔT)> twice the baseline value. A low AMH was defined as below the 5th centile for age.

Results: The D4 testosterone response was normal in 63 with a median testosterone of 9.4 nmol/l (1.0, 40.7) and a median ΔT of 10.0 (0.8, 59.2) and abnormal in 12 with a median testosterone of 0.55 nmol/l (0.5, 3.0) and a median ΔT of 1.0 (0.7, 1.8). AMH was low in 16/75 children and in 8 (50%) of these cases a low D4T was observed. An AMH >5th centile was associated with a low D4T in only 4/59 cases (7%; P<0.0001, ppv 93%)). Median AMH in the two groups of patients who responded and did not respond by D4 was 734 pmol/l (97, 1926) and 93.6 pmol/l (0.4, 256; P<0.0001). The testosterone response after prolonged HST was normal in 23/27 children with a median testosterone of 1.3.5 nmol/l (0.8, 43.4) and a median ΔT of 18.2 (0.57, 62.0) and abnormal in 4 with a median testosterone of 0.55 nmol/l (0.5, 1.0) and a median ΔT of 1.0 (0.6, 1.2). AMH was low in seven children and in 3 (37.5%) of these cases a low D22 testosterone was observed. An AMH>5th centile was associated with a normal D22 testosterone in 19/20 cases (95%) (P<0.0001, ppv 95%). Median AMH in children who responded and did not respond at D4 and D22 was 420 pmol/l (100, 1664) and 2.8 pmol/l (1.5, 214; P<0.0001).

Conclusion: A normal AMH may provide valuable information on overall testicular function. However, a low AMH does not necessarily predict a sub-optimal testosterone response to hCG stimulation.

Article tools

My recent searches

No recent searches.