Treatment of acromegaly with SS analogues: should GH and IGF-I target levels be lowered to assert a tight control of the disease?

Abstract

Background: in acromegaly, the criteria for the cure of the disease after neurosurgery have become tighter and tighter. In contrast, the evaluation of control of disease activity during medical treatment is based upon the normalisation of IGF-I levels and epidemiological criteria, i.e. lessening GH (assessed by RIA) to levels reported to normalise increased mortality. The aim of this study was to evaluate GH and IGF-I suppression during prolonged SS analogues (SA) treatment. The concordance between "safe" GH and normalised IGF-I levels during SA was also assessed, according to gender and gonadal status.

Design: multicentre, retrospective. Patients. GH/IGF-I levels were evaluated in 207 acromegalic patients (132 females, aged 20-85 yr) during a prolonged treatment (longer than 12 months) with individually tailored doses of depot SA( lanreotide or octreotide-LAR in 97 and 110 patients, respectively). Final IGF-I levels were transformed in z-scores using data collected in a large cohort of normal subjects of 3 different age groups (20-40 yr old, 41-60 yr old, 61-80 yr old, n=160, 148, 115, respectively), that allowed to set up quartiles of normality (I = 3rd-25th percentile, II = 26th-50th, III = 51th-75th, IV = 76th-97th).

Results: fifty-nine and 19.3% of patients achieved GH levels <2.5 and <1 microg/l, respectively. IGF-I were normalised (z-score between 2 and -2) in 58.4% of patients. The distribution of normal IGF-I values among quartiles was uneven: 7%, 19%, 25%, and 49% of values were distributed in the I, II, III, and IV quartile, respectively. The concordance between GH and IGF-I values was poor: 28.4% of patients attaining GH values <2.5 microg/l had still pathological IGF-I (even 12.5% of those with GH <1 microg/l), and 39.3% of those with GH levels still above the "safe" limit had "nor IGF-I. Although proportions of IGF-I normalisation were not different between males and females, the regression line obtained between GH and IGF-I z-score showed the same slope but with a significantly lower intercept in regularly cycling women than in males and in postmenopausal females. Thus for any GH value, cycling females had lower IGF-I than menopausal women and males, and their IGF-I normalisation could be achieved by higher GH values. By ROC analysis, the achievement of normal IGF-I was predicted by the GH value of 1.8 microg/l in males and 2.4 microg/l in females.

Conclusions: in acromegalic patients on SA treatment, GH and IGF-I levels are often not concordant. In addition to age, sex is to be taken into account in the evaluation of hormonal targets. A better refinement of GH and IGF-I targets to be reached while on treatment with SA is warranted.