Radiotherapy is a component of treatment for many intracranial tumours, head and neck cancers and haematological malignancies. In these clinical scenarios the hypothalamic‑pituitary (HP) axis is frequently incorporated into the radiation field and therefore functions as an organ at risk. As survival after these conditions improves, more patients are living long enough to develop late endocrine complications, with growth hormone deficiency (GHD) identified as a particularly important consequence.
Diagnosing GHD in patients who have received radiotherapy involving the HP axis is clinically challenging. Much of the diagnostic evidence supporting commonly used GH stimulation tests originates from populations with different underlying disease mechanisms than radiotherapy survivors. This difference in disease context may affect test performance and limits the direct applicability of existing accuracy estimates to the post‑radiotherapy population. The protocol described here defines a systematic approach to summarise and, where appropriate, pool the available diagnostic accuracy data for GH stimulation tests in this specific population.
The primary objective is to assess the diagnostic accuracy of GH stimulation tests for diagnosing GHD in children, teenagers, young adults and adults who have received radiotherapy involving the HP axis. Where studies are sufficiently comparable, a meta‑analysis of sensitivity and specificity will be performed using recommended hierarchical models.
Eligible studies will include patients who were exposed to radiotherapy that involved the HP axis and who subsequently underwent GH stimulation testing for suspected growth hormone deficiency. The review will include studies across pediatric and adult age ranges as specified in the objective, encompassing children, teenagers, young adults and adults treated with radiotherapy to the HP region.
Index tests considered in the review include commonly used GH stimulation procedures: the insulin tolerance test, glucagon stimulation test, arginine stimulation test, clonidine stimulation test, GH‑releasing hormone (GHRH)‑arginine stimulation test and macimorelin stimulation test.
Reference standards for determining GHD will include clinical criteria such as the presence of two or more additional pituitary hormone deficiencies and, where used, alternative GH stimulation tests serving as comparator reference standards. The protocol explicitly accepts either clinical criteria or alternative stimulation tests as reference standards, reflecting the varied approaches used in the literature.
Searches will be conducted in electronic databases including MEDLINE, Embase, the Cochrane Library and PubMed. The protocol specifies comprehensive database searching to identify studies that report diagnostic accuracy of GH stimulation tests in patients previously treated with radiotherapy involving the HP axis.
Two independent reviewers will screen identified records for eligibility and will independently extract data from included studies. Extracted data will include measures necessary to calculate diagnostic accuracy (sensitivity and specificity) and information to characterise study populations and testing procedures.
Risk of bias in included diagnostic accuracy studies will be evaluated using the Quality Assessment of Diagnostic Accuracy Studies‑2 (QUADAS‑2) tool. Reporting quality will be assessed using the Standards for Reporting Diagnostic Accuracy Studies (STARD) criteria. Two independent reviewers will perform these assessments.
Sensitivity and specificity will be extracted from each study. Where studies are sufficiently comparable with respect to index test, reference standard and population, pooled estimates will be generated using hierarchical bivariate and summary receiver operating characteristic (SROC) models. The protocol describes the use of recommended hierarchical methods for pooling diagnostic accuracy metrics.
Prespecified subgroup analyses will explore potential sources of heterogeneity in test performance, including patient age group, body mass index, ethnicity, radiotherapy dose and time elapsed since treatment. These analyses aim to identify factors that may influence diagnostic accuracy in the post‑radiotherapy population.
Ethical approval is not required for this systematic review because it will summarise published data only. The authors plan to disseminate findings through peer‑reviewed publications, conference presentations and engagement with relevant clinical teams to inform and improve diagnostic decision‑making for patients who have received radiotherapy involving the HP axis.
The review protocol is registered on PROSPERO with registration number CRD420261415689.