目的：評估在三度空間立體順形放射治療技術( 3-D conformal radiotherapy' 3-D C前)治療攝護腺癌的病患，在病患位置固定上不使用骨盆腔固定器或是使用不同骨盆腔固定器間的差異性，並比較在使用各種固定技術上因病患體重及病患臀圍大小所造成病患的擺設位置誤差(setup error)上所產生的效果作比較。材料與方法：全部病患共48個人使用不同3-D CRT固定技術，病患以仰臥或俯臥姿勢治療，選擇不用骨盆腔固定器或使用四種骨盆腔固定器中的一種的治療方式。病患分成五組固定情形作比較： (A)病患仰臥不使用骨盆腔固定器治療； (8)病患俯臥及使用腹板( belly-board)但不使用骨盆腔固定器； (C)病患俯臥及使用腹板加熱塑塑膠(cast)骨盆腔固定器； (D)病患仰臥使用發泡型固定器 (alpha cradle)從腰部固定到膝蓋以下； (E)病患仰臥並使用立體定位緊焦軀幹固定器( stereotactic body frame )。於治療前一天內需先完成驗證片的照射，並在治療期間每二週照射一次驗證片，再將 驗證片與模擬定位片作比較分析，記錄下來。我們研究在使用各種骨盆腔固定器及不使用骨盆腔固定器下各個軸位及接轉角度之間的擺設位置誤差，加上分析病患體重及骨盆周界大小對擺設位置誤差的影響。結果：以Kruskal-Wallis方法試算分析：A組在AP-X (p=0.044)有最小的擺設位畫誤差；B組在AP是 ( p=0.065)有最小的擺設位置誤差；D組在AP-Ang. (p=0.067)有最小的擺設位置誤差；E組在Lat-Ang. (P=0.078)有最小的擺設位置誤差。以Mann-Whitney方法試算分析：無論病患是否胖瘦，在各個方向角度的擺設位置誤差上並沒有出現特別的差異性。以Pearson Correlation方法試算分析，體重較重的病患在AP去( p=0.011)產生較小的擺設位置誤差。Lat-Z和AP-Z (p<0.001 )的擺設位置誤差有正向的相互關係， AP-X和Lat-Ang. (p=0.037和p=0.05 )的擺設位置誤差有正向相互關係，但在 PC-AP . PC-Lat上沒有任何關連。結論：A組病患在AP-X和Lat-Z上有較小的擺設位置誤差. B組病患在AP-Z上有較小的擺設位置誤差， D組病患和E組病患在旋轉的角度上有較小的擺設位置誤差， C組病患在Lat-Y上有較小的擺設位置誤差，但對以上A組的Lat-Z和c組的Lat-Y而言並不具統計學上的差異。一般而言，病患仰臥不使用骨盆腔固定技術在治療攝護腺癌上是適當且方便的方法。

Purpose: To determine the variability of patient setup errors during three-dimensional conformal radiotherapy (3D-CRT) for prostate cancer treatment without immobilization or with different immobilization devices. Materials and Methods : Forty-eight prostate cancer patients treated with 3D-CRT were prospective evaluated. They were treated by the same physician, technician and machine. Five groups of patients and technique were compared: (A) supine position without immobilization; (B) prone position without immobilization, but with belly-board; (C) prone position with belly-board and cast immobilization; (D) supine position with alpha cradle from waist to below knee immobilization; (E) supine position with stereotactic body frame immobilization. Serial portal films were recorded and compared with simulation films. The deviation in each axis and rotation angle were compared. Results: Group A has the least deviation in X-axis of anterior-posterior view CAP-X; p = 0.044); group B has the least deviation in Z-axis of anterior-posterior view (AP-Z; p = 0.065); group 0 has the least deviation in rotation angle of anterior-posterior view (AP-Ang.; p = 0.067); group E has the least deviation in rotation angle of lateral view (Lat-Ang. ; p = 0.078). There is no significant difference in deviation both in each directions and angles according to body habitus as normal or obese. Heavier weight has less AP-Z and Lat-Z deviation (p = 0.011 and p = 0.108); the deviation of Lat-Z has positive correlation with AP-Z (p < 0.001), deviation AP-X has positive correlation with deviation Lat-Z and Lat-Ang. (p = 0.037 and p = 0.05 ). For patients with thickness of pelvic circumference (PC-AP) < 19.0 cm have the least deviation in AP-Z (p = 0.034), Lat-Ang. (p = 0.034) in group B and the least AP-Ang. (p = 0.031) in group D; for patients with width of pelvic circumference (PC-Lat) < 35.0 cm have the least deviation in AP-X (p = 0.013) in group A, Lat-Z in group C; for patients with body weight < 67.5 kg have the least deviation in AP-X (p = 0.039) in group A. Conclusion: Group A has better setup error in AP-X and Lat-Z with statistics significant (p < 0.05), group B has better setup error in AP-Z; group D and E have better setup error in rotation angle; group C has less setup error in Lat-Y, but without statistical significance. We conclude supine position without immobilization fixation technique seems to be a suitable method for treating patients with localized prostate cancer either in accuracy or conveniency.