- 學位論文
腦下垂體腫瘤體積精密計算法在臨床上的重要性及其在評估病人預後與治療成效上之應用
Clinical Importance of Precise Volumetric Measurement of Pituitary Adenoma and Its Application in Patient Prognosis and Treatment Outcome Evaluation
摘要
腦下垂體是人體荷爾蒙的控制中心，位於顱底正中央的蝶鞍（sella turcica）凹陷裡。這個位置正處於兩條視覺神經交叉點之下。腦下垂體產生多種激素來維持身體機能恆定：腎上腺皮質刺激素刺激腎上腺分泌腎上腺皮質素；甲狀腺刺激素刺激甲狀腺分泌甲狀腺激素；黃體刺激素和濾泡刺激素與性器官產生作用；生長激素會幫助糖類新陳代謝和細胞成長；泌乳激素會影響泌乳。腦下垂體腺瘤約佔所有腦瘤10～15%，在眾多腦腫瘤中是屬於生長緩慢的。若以大小來分類，較小的腫瘤（少於1公分直徑）稱為微腺瘤（microadenoma），較大的腫瘤（大於1 公分直徑）稱為巨腺瘤（macroadenoma）。腦下垂體腫瘤若依功能來分類可分為功能性和非功能性腫瘤。功能性腫瘤會產生激素，影響生理機能，較容易被發現，此時腫瘤一般不會太大。相對地，非功能性腫瘤因不產生激素，不容易被診斷出來，直到大到一定程度後有視力障礙或內分泌功能低下才會檢查出來，此時往往腫瘤已經長到2～3 公分了。非功能性的巨腺瘤因為壓迫腦下垂體，導致性腺素、腎上腺皮質素、甲狀腺素及抗利尿激素低下，程度較嚴重的病人即使接受手術，術後仍需長期或終生補充口服或針劑荷爾蒙。這些荷爾蒙不足影響比想象中深遠，病人日常生活表現差、肌肉萎縮、提不起精神整天病懨懨。因此，如何在術前預估腦下垂體功能於術後是否能復原，就可以於手術前強調及早手術的必要性以及術後需補充荷爾蒙的可能性。以往文獻也提到腫瘤的直徑大小、腫瘤切除的多寡等與荷爾蒙低下改善的相關性，但是多以腫瘤直徑，也就是一維空間的長度來做判斷預估的標準，不僅不夠嚴謹，當遇到不規則型的腫瘤時更難界定。所以我們利用計算核磁共振影像上腫瘤體積的方法，以三維空間更精密的數據來預估是否須於術後長期補充荷爾蒙。再者，藉由體積可以得到腫瘤切除率（殘存率），亦可以探討全切除（grossly total resection）、次全切除（subtotal resection）、腫瘤殘存率（residual ratio）是否與補充荷爾蒙有關。 另外，用來計算腫瘤體積的方法長久以來一直是以1/2（長x寬x高）來估算，甚至只以一維長度來測量並依此來判斷腫瘤為穩定（stationary）、生長（growing）或是萎縮（regressive），這樣的方法在形狀規則或是（橢）圓形的病灶其估算出來的體積並不會有太大誤差，但是一但遇到侵犯性高（highly invasive）生長快速或是手術後導致形狀不規則的腫瘤，以傳統估算方法不但無法精確計算，更會造成後續對腫瘤現狀判斷的誤差，甚至導致病患接受不必要的治療，如再次手術、放射治療等等。這些不必要的治療或多或少會有併發症，包括腦下垂體功能失調、續發性腦部癌變、腦部認知功能缺損等等沒有必要且可以事先避免的副作用。因此，我們在研究中所採取的方法為OsiriX，此法利用segmentation process，且已經多次被證實可以用來精確計算腫瘤體積，其結果也被刊登在不同的期刊上。經由與傳統算法比較後可以發現：OsiriX方法優於傳統算法，尤其是對於形狀不規則或是術後的腫瘤 (p<0.0001)。除此之外，我們更採用了另一種3D slicer segmentation process的方法來計算，並比較與OsiriX法兩者之間的差距，統計結果顯示這兩種電腦軟體不論在術前(p=0.4964)或術後(p=0.4062)均沒有估算體積上的差別。 總之，以精密方法計算腫瘤體積，不但可以先行預測腫瘤可能會引起的臨床效應，更是腫瘤後續追蹤以及決定下一步治療方式的關鍵因素。尤其對於形狀不規則的腫瘤，更是應該以精密的方式來追蹤與計算，才可以避免因誤判腫瘤現況與後續不必要的治療所帶來的併發症。
並列摘要
The pituitary gland regulates the secretion of hormones in the human body. It is located on the sella turcica, a bony cavity in the center of the cranial fossa and below the optic chiasm. Normally, the pituitary gland produces several hormones to maintain constant body functions and homeostasis. Pituitary adenomas accounting for approximately 10%–15% of all brain tumors, are relatively slow-growing tumors. When classified by size, tumors that are < 1 cm in diameter are defined as microadenomas, whereas those ≥ 1 cm as macroadenomas. Pituitary adenomas are also classified as functioning and nonfunctioning adenomas. Functioning adenomas produce excess hormones and are discovered by hypersecretion functions. By contrast, nonfunctioning adenomas do not produce hormones and are thus not detected until visual impairment or hypopituitarism occurs. These tumors are generally 2–3 cm in diameter when discovered. The external mass compression of a macroadenoma damages gland functions, resulting in different kinds of hormone deficiency. Patients with severe hormone-related problems require long-term hormone supplementation postoperatively. Hormone deficiency has a greater impact on patients than was previously thought, and these patients may experience fatigue and decreased muscle strength that interfere with their daily activities. Therefore, predicting the likelihood of postoperative pituitary function recovery can enable us to emphasize the necessity of early surgery to avoid the requirement of lifelong hormone supplementation. Previous studies have investigated whether tumor size and the amount of tumor resection are correlated with hormone deficiency improvement. However, the majority of these studies predicted using tumor diameter (i.e., one-dimensional length). This approach not only lacks rigorousness, but also presents difficulties when defining irregularly shaped tumors. Therefore, we calculated tumor volume using this precise three-dimensional data and then predicted the need of long-term postoperative hormone supplementation. In addition, volumetric measurement of tumors has long been calculated by the geometric formula (length × width × height)/2. Some clinicians have even determined whether a tumor is stationary, growing, or regressive using only its diameter. This method does not result in many errors when used for regularly shaped tumors. However, it cannot make accurate calculations for highly invasive tumors with rapid growth or irregularly shaped tumors that have formed postoperatively, and it may lead to errors in determining tumor status and even unnecessary treatment such as surgery and radiotherapy, which may lead to complications that could have been avoided, including pituitary insufficiency, secondary malignancy, and cognitive dysfunction. To prevent such problems, our second study used OsiriX to perform the precise volumetric measurement of tumors. The segmentation process has been long employed for its accuracy, and its outcomes have been published in various journals. A comparison revealed that OsiriX yielded more favorable outcomes than the traditional method, particularly for post-operative irregularly shaped tumors (p<0.0001). We also compared the OsiriX method with another 3D slicer segmentation process, and the statistical results revealed that no differences between these two softwares in estimating pre-operative (p=0.4964) or post-operative tumor volume (p=0.4062). In summary, precise volumetric measurement not only predicts the potential clinical effects of a tumor, but is also critical for residual tumor follow-ups and determining whether subsequent treatment is necessary. Irregularly shaped tumors should be identified and tracked using this method to avoid misjudgment of the current status and complications caused by unnecessary treatment.