Acromegalia: comparación de los niveles séricos de IGF-I por dos inmunoensayos y su correlación con la prueba de tolerancia oral a la glucosa / Acromegaly: Comparison of serum IGF-1 levels measured by two immunoassays and their correlation with the Oral Glucose Tolerance Test

RESUMEN

Introducción: La determinación de IGF-I en suero o plasma es una herramienta esencial en el diagnóstico y seguimiento de la acromegalia. Sin embargo, se deben tener presentes algunos inconvenientes en su medición por diferentes inmunoensayos. Objetivos: Evaluar dos inmunoensayos para la determinación de IGF-I y su correlación con el nadir de GH en el TTOG en pacientes acromegalicos. Materiales y métodos: Se analizaron 37 pacientes acromegálicos, 20 mujeres y 17 hombres. IGF-I fue determinada por Immulite 1000, (IMM) y por IRMA (DSL). Se realizó el TTOG y se determinó glucosa y GH en todos los tiempos (basal, 30, 60, 90 y 120min). Se consideró respuesta normal un nadir de GH <1ng/ml. Nueve pacientes se encontraban bajo tratamiento y 28 sin tratamiento. Análisis estadístico se utilizaron el test de Wilcoxon, de Bland y Altman y curvas ROC. Se consideró significativa una p<0,05. Resultados: Las concentraciones basales de glucosa fueron 97,86±10,91 mg/dl, de GH 2,8 (1,59-14,4) ng/ml, de IGF-I por IMM 602±318 ng/ml y por DSL 1006±596 ng/ml. IGF-I por IMM y DSL mostró una diferencia significativa con p <0,01 y un bias de - 403,2 ng/ml con valores menores por IMM. IGF-I elevada por IMM y DSL, se encontró en el 84% y en el 97% respectivamente. IGF-I elevada con nadir de GH >1ng/ml se encontró en el 70%, con nadir de GH normal en el 13,5%. IGF-I normal con nadir >1ng/ml en el 2,7% y con nadir de GH normal en el 13,5%. El área bajo las curvas ROC no mostró diferencias significativas. Conclusiones: Los niveles de IGF-I determinados por IMM y DSL fueron significativamente diferentes mostrando un bias negativo para IMM. La mayoría de los valores del nadir de GH fueron consistentes con los niveles de IGF-I observándose una discrepancia en el 30% de los pacientes, estuvieran o no bajo tratamiento.

ABSTRACT

Introduction: IGF-I determination in serum or plasma is an essential tool in the diagnosis and follow-up of acromegaly. Hepatic production of IGF-I is regulated by GH and circulates bound to several IGF-I binding proteins which extends its half life. IGF-I is not released in a pulsatile pattern and has no significant variability in 24 h. Objective: To evaluate two different methodologies in IGF-I levels determination and their correlation with GH nadir in OGTT in acromegalic patients. Material and methods: We analyzed 37 acromegalic patients, 20 women and 17 men, mean age was 45±12 years. IGF-I levels were assayed by Immulite 1000, DPC (IMM) and DSL-5600 ACTIVE® IGF-I Coated-Tube IRMA (DSL) and OGTTs (at baseline and at 30, 60, 90 and 120 minutes) were performed by measuring plasma glucose and GH assay by immunochemiluminometric assay (Access); we considered a nadir <1ng/ml as normal response. Nine patients were under medical treatment (cabergoline 4, octeotride 4, and cabergoline plus octeotrite 1) and 28 without treatment. Statistical analysis: Wilcoxon and, Bland and Altman tests and ROC curves. Differences were considered significant at p< 0.05. Results: Basal glucose levels were 97.86±10.91 mg/dl and mean GH was 2.8 (1.59-14.4) ng/ml. Mean IGF-I levels performed by IMM were 602±318 ng/ml and 1006±596 ng/ml by DSL. There was a statistically significant difference between both methodologies (p<0.01). Bland and Altman test showed a bias of - 403.2 ng/ml with lower values by IMM. We observed elevated IGF-I levels in 84% by IMM and in 97% by DSL, and only one patient had normal levels with both methodologies. Elevated IGF-I levels and GH nadir >1ng/ml were observed in 70% of the patients, increased IGF-I with normal GH nadir in 13.5%, normal IGF-I with GH nadir >1ng/ml in 2.7% and normal IGF-I with normal GH nadir in 13.5%. Patients under treatment 3 showed normal GH nadir with elevated IGF-I levels, in 2 of them by both methodologies, and in the other one it was normal by IMM and elevated by DSL; the other 6 showed GH nadir > 1ng/ml, 5 of them presented elevated IGF-I by both methodologies and the other one showed discrepancy in IGF-I levels. The under ROC curve area and confidence interval (CI) of 95% for IGF-I IMM and DSL were 0.96 (0.90-1.00) and 0.91 (0.82-1.00) respectively. Differences between the ROC curves areas were not significant Conclusions: IGF-I levels determined by IMM and DSL were statistically significantly different. IGF-I levels showed a negative bias by IMM. Most of the results of GH nadir were consistent with IGF-I levels but we observed discrepancy in 30% of the patients, regardless of whether they were under treatment or not.