Can the TW3 bone age determination method provide additional criteria for growth hormone treatment in adopted girls with early puberty? A comparison of the Tanner-Whitehouse 3 method with the Greulich-Pyle and the Tanner-Whitehouse 2 methods

Abstract

Background/aims: Gonadotropin-releasing hormone analogues (GnRHa) are the accepted treatment of idiopathic central precocious puberty. As it has been found that growth velocity may be decreased with GnRHa treatment, clinical trials with GnRHa combined with growth hormone (GH) have been carried out. In a recent study 46 adopted girls with early or precocious puberty were randomly assigned to treatment with either GnRHa or GnRHa combined with GH, and followed to final height (FH). It was found that FH was significantly higher in the combined treatment group, 158.9 compared with 155.8 cm in the GnRHa treated group. In order to select the patients who could benefit from added GH, predictions of FH at the start of treatment according to the methods of bone age determination of Greulich-Pyle (GP) and Tanner-Whitehouse 2 (TW2) were compared. It was found that the GP method was the most useful method for patient selection. Recently, a revision of the Tanner-Whitehouse method, named Tanner-Whitehouse 3 (TW3), has been developed. The present study examined the usefulness of the TW3 method in selecting suitable patients for combined treatment.

Method: The TW3 method bone age determinations of the 46 girls were compared to the GP and TW2 method determinations, using the differences between actual FH and predicted adult height (PAH). Beside accuracy of prediction of FH, the criteria of efficiency of selection and replicability were applied in the comparison.

Results: We found that the GP method, also when compared to the TW3 method, gave the most accurate prediction of the FH on only GnRHa treatment. This gives the best ground for selection of patients who can benefit from combined treatment. The GP method was also the most efficient in selecting patients, i.e., it could select the least number of patients that needed the combined treatment. The only drawback of the GP method was that it requires an experienced pediatric radiologist. Automated methods are being developed and may soon facilitate the use of the GP method for those less experienced. The FH after combined treatment could be predicted with an equation including PAH GP as well as PAH TW3 as variables.

Conclusion: The GP method remains the most useful method for selection of those patients who will benefit most from the addition of GH to GnRHa in the treatment of idiopathic central precocious or early puberty. FH prediction after combined treatment requires PAH GP as well as PAH TW3.