Mammosomatotroph and mixed somatotroph-lactotroph adenoma in acromegaly: a retrospective study with long-term follow-up

Liang Lv,Yong Jiang,Senlin Yin,Yu Hu,Cheng Chen,Weichao Ma,Shu Jiang,Peizhi Zhou
DOI: https://doi.org/10.1007/s12020-019-02029-1
2019-01-01
Endocrine
Abstract:Purpose Although well-documented from pathological aspect, the clinical features and outcomes of acromegaly with mammosomatotroph (MSA) and mixed somatotroph-lactotroph adenoma (MSLA) are seldom reported. Thus, in this study, we analyzed and reported the clinical data about MSAs and MSLAs. Methods We retrospectively reviewed medical records of patients with acromegaly in our institution during 2008–2017. Growth hormone (GH)-secreting adenomas were categorized into pure somatotroph adenoma (PSA), MSA and MSLA based on inclusion and exclusion criteria. Clinical information and treatment outcomes during follow-up were analyzed by univariate and multivariate methods. Results Among 94 patients within this cohort, PSAs, MSAs, and MSLAs accounted for 53, 28 and 13 cases, respectively. MSAs often had smaller size, lower frequency of cavernous sinus invasion and higher gross total resection (GTR) rate. MSLAs were characterized by bigger tumor size, higher frequency of preoperative hyperprolactinemia, and lower GTR rate. Thus, MSLAs had worse long-term biological remission rate than MSAs and PSAs (15.4% vs. 50.0% and 26.4%, p = 0.0371). Gender (male, OR = 0.784, p = 0.011) and tumor volume (OR = 0.784, p = 0.020) were independent predictors for long-term biological remission in binary logistic regression. Subgroup analyses indicated that postoperative nadir GH level (GH-7, HR = 1.242, p = 0.001) was the only risk factor for tumor recurrence for patients with GTR. Conclusions Our results provide valuable insights into clinicopathological features of acromegaly. MSAs were relatively smaller lesions with better prognosis. MSLAs were more aggressive with massive size, invasiveness and preoperative hyperprolactinemia. Tumor size and GH-7 were significantly associated with biological remission and tumor relapse after GTR, respectively.
What problem does this paper attempt to address?