Thapar K, Kovacs K, Stefaneanu L, Scheithauer B, Killinger D W, Lioyd R V, Smyth H S, Barr A, Thorner M O, Gaylinn B, Laws E R
Department of Neurosurgery, University of Toronto, Ontario, Canada.
Am J Pathol. 1997 Sep;151(3):769-84.
The clinical behavior of growth hormone (GH)-producing pituitary tumors is known to vary greatly; however, the events underlying this variability remain poorly understood. Herein we demonstrate that tumor overexpression of the GH-releasing hormone (GHRH) gene is one prognostically informative event associated with the clinical aggressiveness of somatotroph pituitary tumors. Accumulation of GHRH mRNA transcripts was demonstrated in 91 of a consecutive series of 100 somatotroph tumors by in situ hybridization; these findings were corroborated by Northern analysis and reverse transcriptase polymerase chain reaction, and protein translation was confirmed by Western blotting. By comparison, transcript accumulation was absent or negligibly low in 30 normal pituitary glands. GHRH transcripts were found to preferentially accumulate among clinically aggressive tumors. Specifically, GHRH mRNA signal intensity was 1) linearly correlated with Ki-67 tumor growth fractions (r = 0.71; P < 0.001), 2) linearly correlated with preoperative serum GH levels (r = 0.56; p = 0.01), 3) higher among invasive tumors (P < 0.001), and 4) highest in those tumors in which post-operative remission was not achieved (P < 0.001). Using multivariate logistic regression, a model of postoperative remission likelihood was derived wherein remission was defined by the single criterion of suppressibility of GH levels to less than 2 ng/ml during an oral glucose tolerance test. In this outcome model, GHRH mRNA signal intensity proved to be the most important explanatory variable overall, eclipsing any and all conventional clinicopathological predictors as the single most significant predictor of postoperative remission; increases in GHRH mRNA signal were associated with marked declines in remission likelihood. The generalizability of this outcome model was further validated by the model's significant performance in predicting postoperative remission in a random sample of 30 somatotroph tumors treated at another institution. These data indicate that overexpression of GHRH gene is an event associated with the neoplastic progression and clinical aggressiveness of somatotroph adenomas. More generally, these data merge essential elements of the hypothalamic and pituitary hypotheses of pituitary tumorigenesis, providing for a more unified concept of neoplastic progression in the pituitary.
已知分泌生长激素(GH)的垂体瘤临床行为差异很大;然而,这种变异性背后的机制仍知之甚少。在此我们证明,生长激素释放激素(GHRH)基因在肿瘤中的过表达是与生长激素细胞垂体瘤临床侵袭性相关的一个具有预后信息价值的事件。通过原位杂交在连续的100例生长激素细胞肿瘤系列中的91例中证实了GHRH mRNA转录本的积累;Northern印迹分析和逆转录聚合酶链反应证实了这些发现,蛋白质翻译通过Western印迹得到确认。相比之下,在30个正常垂体中未发现转录本积累或其水平极低可忽略不计。发现GHRH转录本在临床侵袭性肿瘤中优先积累。具体而言,GHRH mRNA信号强度:1)与Ki-67肿瘤生长分数呈线性相关(r = 0.71;P < 0.001),2)与术前血清GH水平呈线性相关(r = 0.56;p = 0.01),3)在侵袭性肿瘤中更高(P < 0.001),4)在术后未实现缓解的肿瘤中最高(P < 0.001)。使用多变量逻辑回归,得出了一个术后缓解可能性模型,其中缓解定义为口服葡萄糖耐量试验期间GH水平可被抑制至低于2 ng/ml这一单一标准。在这个结果模型中,GHRH mRNA信号强度被证明是总体上最重要的解释变量,作为术后缓解的单一最显著预测因子,超过了任何和所有传统的临床病理预测指标;GHRH mRNA信号的增加与缓解可能性的显著下降相关。通过该模型在预测另一机构治疗的30例生长激素细胞肿瘤随机样本的术后缓解方面的显著表现,进一步验证了该结果模型的可推广性。这些数据表明,GHRH基因的过表达是与生长激素细胞腺瘤的肿瘤进展和临床侵袭性相关的一个事件。更一般地说,这些数据融合了垂体瘤发生的下丘脑和垂体假说的基本要素,为垂体肿瘤进展提供了一个更统一的概念。