Surgical interventions and medical treatments in treatment-naïve patients with acromegaly: Systematic review and meta-analysis

Context: Acromegaly is usually treated with surgery as a first-line treatment, although medical therapy has also been used as an alternative primary treatment.

Objective: Weconducted a systematic review and meta-analysis to synthesize the existing evidence comparing these two approaches in treatment-naïve patients with acromegaly.

Data Sources: This study performed a comprehensive search in multiple databases, including Medline, EMBASE, and Scopus from early inception through April 2014.

Study Selection: The study used original controlled and uncontrolled studies that enrolled patients with acromegaly to receive either surgical treatment or medical treatment as their first-line treatment.

Data Extraction: Reviewers extracted data independently and in duplicates. Because of the noncomparative nature of the available studies, we modified the Newcastle-Ottawa Scale to assess the quality of included studies. Outcomes evaluated were biochemical remission and change in IGF-1 or GH levels. We pooled outcomes using the random-effects model.

Data Synthesis: The final search yielded 35 studies enrolling 2629 patients. Studies were noncomparative series with a follow-up range of 6-360 months. Compared with medical therapy, surgery was associated with a higher remission rate (67% vs 45%; P = .02). Surgery had higher remission rates at longer follow-up periods (≥24 mo) (66% vs 44%; P = .04) but not the shorter follow-up periods (≥6 mo) (53% vs 26%; P = .02). Surgery had higher remission rates in the follow-up levels ofGH(65% vs 46%; P=.05). In one study, the IGF-1 level was reduced more with surgery compared with medical treatment (-731μg/L vs-251μg/L; P=.04). Studies in which surgery was performed by a single operator reported a higher remission rate than those with multiple operators (71% vs 47%; P = .002).

Conclusions: Surgery may be associated with higher remission rate; however, the confidence in such evidence is very low due to the noncomparative nature of the studies, high heterogeneity, and imprecision.