Endocrine Abstracts (2004) 7 P168

Variability in GH assays undermines the value of consensus criteria for the diagnosis of adult GH deficiency and acromegaly

A Pokrajac-Simeunovic1, GE Wieringa2, AK Ellis3 & PJ Trainer1


1Department of Endocrinology, Christie Hospital NHS, Manchester, UK; 2Department of Biochemistry, Christie Hospital NHS, Manchester, UK; 3Department of Clinical Biochemistry, Royal Infirmary, Edinburgh, UK.


There is increasing reliance on published consensus criteria for clinical decision-making in states of GH excess and deficiency. NICE eligibility criteria for GH treatment include a peak GH response <9 mU/L during an ITT. To determine the adequacy of GH assay performance for diagnosing GH deficiency (GHD), we have assessed the variability in 101 UKNEQAS reported results from a single sample with a value close to 9.0 mU/L. For all laboratories (n=101) median GH was 11.1 mU/L (range: 6.6-13.9). 15% of laboratories reported GH <9 mU/L, so fulfilling NICE criteria for GH replacement. The most negatively biased method (n=11 labs, median 7.1 mU/L, range 6.6-7.7) classified all patients as GHD; the most positively biased (n=44, median 11.95 mU/L, 9.8-13.9) classified all patients as normal.

The current consensus for acromegaly states that the diagnosis is excluded by a nadir GH during an OGTT of <1 ug/L but with no stated mU/L equivalent. In the absence of an evidence base for an appropriate conversion factor, we have assessed diagnostic accuracy in 104 UKNEQAS reported results from a single sample using conversion factors of 2.6 and 3.0 (ug/l to mU/l). The 'all-laboratory' (n=104) median GH was 2.6 mU/L (range 1.04-3.5). Applying a conversion factor of 2.6, the most negatively biased method (n =10, median 0.69 ug/L, range 0.61-0.81) classified all the values as normal, while the most positively biased (n= 42, median 1.08 ug/L, 0.9-1.23) classified 93% of values as consistent with acromegaly. Applying a conversion factor of 3.0, 11% labs of reported values were consistent with acromegaly, while 89% were normal.

In summary, variability in GH assays undermines the applicability of international consensus criteria to local practice. A consensus on GH assay performance, taking into account the availability of an international GH calibrant and recently introduced EU legislation, is urgently required.

Article tools

My recent searches

No recent searches.