A Permissive Role of Growth Hormone on the Development of Second Brain Tumors after Radiotherapy? a Systematic Review

Presentation Number: SUN 441
Date of Presentation: April 2nd, 2017

Pia Burman*1 and Olaf M Dekkers2
1Skane University Hospital, University of Lund, Malmö, Sweden, 2Leiden University Medical Center, Netherlands


Context: Recent studies have shown that patients treated with radiotherapy (RT) for pituitary adenomas (secreting/non-secreting) have a 4-5 fold higher risk of second brain tumors (SBT) compared with patients treated with surgery/medications only. Besides age at RT, other risk factors have not been identified. Acromegaly has been associated with an increased risk of colon neoplasia. There are no reports of an increased incidence of brain tumors. In vitro, growth hormone (GH) has been shown to stimulate growth of tumor cells. Whether acromegaly is a risk factor for RT induced brain tumors is unknown.

Objective: To investigate if GH overproduction could have a permissive role on the development of RT-induced secondary brain tumors.

Method: A systematic review for studies reporting an incidence of SBT (malignancies and meningiomas) after RT treatment of pituitary adenomas stratified by hormonal secretory profiles. Subsequently, a meta-analysis was performed to compare risk of secondary brain tumor in GH secreting adenomas versus non-GH secreting adenomas.

Results: Four eligible studies were identified; the studies comprised in total 298 GH secreting adenomas, 2888 non-GH secreting adenomas; median follow-up time since RT 11.5 years.
Relative risks for second brain tumor development in patients with acromegaly were consistently above 1.0 and ranged from 1.46 to 2.40. The weighted average compare risk of secondary brain tumor in GH secreting adenomas versus non-GH secreting adenomas was 1.83 (95% CI 0.73-4.59).

Conclusion: The results suggest that GH excess at time of RT confers an additional risk for a subsequent second brain tumor. Given the low incidence of second brain tumors and therefore the low power to detect a difference according to biochemical profile, it should be emphasized that larger cohorts are needed to confirm this finding.


Nothing to Disclose: PB, OMD