Acromegaly is a rare and heterogeneous disorder caused by chronic growth hormone (GH) excess, most often due to a pituitary tumor1. The clinical presentation and biochemical profile can vary substantially depending on age, tumor imaging characteristics, granulation pattern, and diagnostic timing2,3. Patients with acromegaly often present with a wide range of phenotypes: from young individuals with large, invasive macroadenomas and highly elevated GH and insulin-like growth factor 1 (IGF-1) levels, to older patients with subtle symptoms, smaller tumors, and only mild biochemical abnormalities. While macroadenomas represent the majority of GH-secreting pituitary tumors at diagnosis, increased awareness and the widespread use of brain imaging have led to a growing recognition of microadenomas and incidental diagnosis4,5. Up to 50% show discordant profiles, with the predominant discordant form being elevated IGF-1 and normal/low GH levels6, 7, 8, 9, 10. In addition, atypical biochemical phenotypes, including patients with extremely high GH ≥100 ng/mL11 to normal or mildly elevated basal GH levels but elevated IGF-1 concentrations are increasingly being reported12. This latter group is sometimes referred to as “micromegaly”, which is an entity encompassing patients with obvious clinical features of acromegaly who have elevated age-adjusted IGF-1 concentrations alongside seemingly normal baseline GH levels and frequently exhibit glucose-suppressed GH values below 1 ng/mL 13, 14, 15. Over the past two decades, international guidelines have progressively shifted from relying primarily on GH values to favouring IGF-1 as the key biochemical marker for the diagnosis and follow-up of acromegaly, given the various limitations of GH assays and the greater reliability of serum IGF-116. In addition, previous studies assessing disease severity or classifying biochemical profiles in acromegaly have relied primarily on absolute GH values11, 12, 13,17. However, there is considerable heterogeneity among the thresholds used. While some series have defined normal GH levels as those below 4.7 ng/mL 12,13,17, others adopted more stringent cut-offs, such as 2.5 ng/mL18 or 2.0 ng/dL11. This lack of consensus is compounded by the fact that different immunoassay platforms vary widely in their sensitivity and reference ranges, making the interpretation of absolute serum GH concentrations challenging and potentially misleading in multicenter cohorts19 .

Additionally, there is a lack of real-world data from large unselected national cohorts addressing whether GH stratification at diagnosis correlates with clinical manifestations, comorbidities, or tumor aggressiveness. A GH standardization by upper limit of normal (ULN) and data-driven approach such as cluster analysis may help uncover naturally occurring patient subgroups that share similar tumor behavior, biochemical profiles, or therapeutic responses. This approach has been previously applied in one acromegaly cohort2, but has not been widely reproduced in other populations.

To address this gap, we conducted a comprehensive analysis of a large cohort of patients with acromegaly, stratified by tumor size (microadenomas vs. macroadenomas) and by basal GH standardizing values to the assay-specific ULN levels grouped into clinically relevant categories. Additionally, we applied unsupervised cluster analysis to identify distinct clinical-radiological phenotypes within the cohort. Our study aimed to overcome the limitations of absolute GH values by applying a standardization strategy and unsupervised clustering to identify clinically meaningful acromegaly phenotypes and to evaluate their clinical relevance.