doi:10.1016/j.ijrobp.2004.05.028 PHYSICS CONTRIBUTION COMPARATIVE STUDY OF REFERENCE POINTS BY DOSIMETRIC ANALYSES FOR LATE COMPLICATIONS AFTER UNIFORM EXTERNAL RADIOTHERAPY AND HIGH-DOSE-RATE BRACHYTHERAPY FOR CERVICAL CANCER SHANG-WEN CHEN, M.D.,* ‡§ JI-AN LIANG, M.D.,* ‡ LIAN-SHUNG YEH, M.D., †‡ SHIH-NENG YANG, M.D.,* ‡ AN-CHENG SHIAU, M.S.,* ‡ AND FANG-JEN LIN, M.D., PH.D.* §‡ Departments of *“Radiation Therapy and Oncology”, and † “Obstetrics and Gynecology”, and ‡ School of Medicine, China Medical University Taichung, Taiwan, § School of Medicine, National Taiwan University, Taipei, Taiwan,  Department of Radiation Therapy and Oncology, Shin Kong Memorial Hospital, Taipei, Taiwan Purpose: This study aimed to correlate and compare the predictive values of rectal and bladder reference doses of uniform external beam radiotherapy without shielding and high-dose-rate intracavitary brachytherapy (HDRICB) with late sequelae in patients with uterine cervical cancer. Methods and Materials: Between September 1992 and December 1998, 154 patients who survived more than 12 months after treatment were studied. Initially, they were treated with 10-MV X-rays (44 to 45 Gy/22 to 25 fractions over 4 to 5 weeks) to the whole pelvis, after which HDRICB was performed using 192 Ir remote afterloading at 1-week intervals for 4 weeks. The standard prescribed dose for each HDRICB was 6.0 Gy to point A. Patient- and treatment-related-factors were evaluated for late rectal complications using logistic regression modeling. Results: The probability of rectal complications showed better correlation of dose–response with increasing total ICRU (International Committee on Radiotherapy Units and Measurements) rectal dose. Multivariate logistic regression demonstrated a high risk of late rectal sequelae in patients who developed rectal complications (p  0.0001;relative risk, 15.06;95% CI, 2.8943.7) and total ICRU rectal dose greater than 16 Gy (p  0.02;relative risk, 2.07;95% CI, 1.134.55). The high risk factors for bladder complications were seen in patients who developed rectal complications (p  0.0001;relative risk, 15.2;95% CI, 2.8144.9) and total ICRU bladder dose greater than 24 Gy (p  0.02;relative risk, 8.93;95% CI, 1.7933.1). Conclusion: This study demonstrated the predictive value of ICRU rectal and bladder reference dosing in HDRICB for patients receiving uniform external beam radiation therapy without central shielding. Patients who had a total ICRU rectal dose greater than 16 Gy, or a total ICRU bladder dose over 24 Gy, were at risk of late sequelae. © 2004 Elsevier Inc. : Cervical cancer, Brachytherapy, Late sequelae, Radiation dose. INTRODUCTION High-dose-rate intracavitary brachytherapy (HDRICB) has been widely used in the treatment of cervical cancer for more than 20 years. Despite its questionable radiobiological advantage (1), its application has been increasing (2, 3). In the analysis of the radiotherapy (RT) outcome, it is impor- tant to assess not only local tumor control, but also the incidence of late sequelae caused by the treatment. Among such sequelae are rectal and bladder complications that include proctitis, cystitis, and fistulae (4-19). The risk fac- tors of late sequelae in patients undergoing definitive RT have been evaluated and include age, stage, external beam radiotherapy (EBRT) dose, number of HDRICB treatments, point A dose, and rectal and bladder reference doses. With HDRICB, the complication rate increases with in- crease of dose per fraction. Currently, the most popular dose fractionation schedules are 3–10 fractions with the point A dose per fraction ranging from 4 to 8 Gy. Because HDRICB reference points usually have a larger dose per fraction than does an external radiation dose, simple addition of the doses may not represent the cumulative effect (6). Nonetheless, there are many reports in the literature that use summation of biologically effective dose (BED) for both EBRT and intracavitary brachytherapy (ICB) for analysis of the asso- ciation between late complications and total cumulative irradiation dosage (6, 15-18, 20, 21). However, summing EBRT-BEDs and ICB-BEDs can indeed be problematic for two main reasons. Firstly, the irradiated volume and isodose Reprint requests to: Shang-Wen Chen, Department of Radiation Therapy and Oncology, China Medical University Hospital, No. 2, Yuh-Der Road, Taichung, Taiwan 404. Tel: (+886) 4-2205-2121, ext 2411; E-mail: vincent1680616@yahoo.com.tw Received Nov 19, 2003 and in revised form May 3, 2004. Accepted for publication May 10, 2004. Int. J. Radiation Oncology Biol. Phys., Vol. 60, No. 2, pp. 663– 671, 2004 Copyright © 2004 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/04/$–see front matter 663