- Email: [email protected]
Combination external beam radiotherapy and high-dose-rate intracavitary brachytherapy for uterine cervical cancer: Analysis of dose and fractionation schedule
Int. J. Radiation Oncology Biol. Phys., Vol. 56, No. 5, pp. 1344 –1353, 2003 Copyright © 2003 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/03/$–see front matter
doi:10.1016/S0360-3016(03)00288-8
CLINICAL INVESTIGATION
Cervix
COMBINATION EXTERNAL BEAM RADIOTHERAPY AND HIGH-DOSE-RATE INTRACAVITARY BRACHYTHERAPY FOR UTERINE CERVICAL CANCER: ANALYSIS OF DOSE AND FRACTIONATION SCHEDULE TAKAFUMI TOITA, M.D.,* YASUMASA KAKINOHANA, PH.D.,* KAZUHIKO OGAWA, M.D.,* GENKI ADACHI, M.D.,* HIDEHIKO MOROMIZATO, M.D.,† YUTAKA NAGAI, M.D.,† TOSHIYUKI MAEHAMA, M.D.,† KAORU SAKUMOTO, M.D.,† KOJI KANAZAWA, M.D.,† AND SADAYUKI MURAYAMA, M.D.* Departments of *Radiology and †Obstetrics and Gynecology, University of the Ryukyus School of Medicine, Okinawa, Japan Purpose: To determine an appropriate dose and fractionation schedule for a combination of external beam radiotherapy (EBRT) and high-dose-rate intracavitary brachytherapy (HDR-ICBT) for uterine cervical cancer. Methods: Eighty-eight patients with uterine cervical squamous cell carcinoma treated with EBRT and HDRICBT were analyzed. Twenty-five patients were classified as early disease (nonbulky Stage I/II, less than 4-cm diameter) and 63 patients as advanced disease (greater than 4 cm diameter or Stage IIIB) according to the American Brachytherapy Society definition. Tumor diameter was measured by MRI. Pelvic EBRT was delivered before applications of ICBT. HDR-ICBT was performed once a week, with a fraction point A dose of 6 Gy. Source loadings corresponded to the Manchester System for uterine cervical cancer. No planned optimization was done. A Henschke-type applicator was mostly used (86%). Median cumulative biologic effective dose (BED) at point A (EBRT ⴙ ICBT) was 64.8 Gy10 (range: 48 –76.8 Gy10) for early disease, and 76.8 Gy10 (range: 38.4 – 86.4 Gy10) for advanced disease. Median cumulative BED at ICRU 38 reference points (EBRT ⴙ ICBT) was 97.7 Gy3 (range: 59.1–134.4 Gy3) at the rectum, 97.8 Gy3 (range: 54.6 –130.4 Gy3) at the bladder, and 324 Gy3 (range: 185.5– 618 Gy3) at the vagina. Actuarial pelvic control rate and late complication rate were analyzed according to cumulative dose and calculated BED. Results: The 3-year actuarial pelvic control rate was 82% for all 88 patients: 96% for those with early disease, and 76% for advanced disease. For pelvic control, no significant dose–response relationship was observed by treatment schedules and cumulative BED at point A for both early and advanced disease. The 3-year actuarial late complication rates (Grade >1) were 12% for proctitis, 11% for cystitis, and 14% for enterocolitis. There were significant differences on the incidence of proctitis (p < 0.0001) and enterocolitis (p < 0.0001), but not for cystitis by the treatment schedules and cumulative point A BED. All 4 patients treated with 86.4 Gy10 at point A suffered both proctitis and enterocolitis. Patients with cumulative BED at rectal point of >100 Gy3 had significantly higher incidence of proctitis (31% vs. 4%, p ⴝ 0.013). Conclusions: In view of the therapeutic ratio, cumulative BED 70 – 80 Gy10 at point A is appropriate for uterine cervical cancer patients treated with a combination of EBRT and HDR-ICBT. Present results and data from other literatures suggested that cumulative BED at the rectal point should be kept below 100 –120 Gy3 to prevent late rectal complication. © 2003 Elsevier Inc. Cervix neoplasms, Radiotherapy, High-dose-rate brachytherapy, Biologic effective dose.
INTRODUCTION High-dose-rate intracavitary brachytherapy (HDR-ICBT) has been widely used in treatment of uterine cervical cancer in Asia and Europe (1–12). Despite some criticism of HDRICBT (13), its application has been increasing in the United States (14, 15). However, a wide variation of dose and fractionation schedules exists for the combination of external beam radiotherapy (EBRT) and HDR-ICBT (1–12, 14 –
18). An optimum treatment schedule has not yet been clearly determined. In this paper, we retrospectively analyzed the dose–response relationship for local control and late complication for uterine cervical cancer patients treated with a combination of EBRT and HDR-ICBT. The aim of this study was to determine an appropriate dose and fractionation schedule for EBRT and HDR-ICBT in the treatment of uterine cervical cancer.
Address for reprints: Takafumi Toita, M.D., Department of Radiology, University of the Ryukyus School of Medicine, 207 Uehara, Nishihara-cho, Okinawa, 903-0215 Japan. Fax: ⫹81-98895-1420; E-mail: [email protected]
This study was supported by Grants-in-Aid for Cancer Research No. 14 – 6 from the Ministry of Health and Welfare, Japan. Received Oct 3, 2002, and in revised form Mar 3, 2003. Accepted for publication Mar 4, 2003. 1344
High-dose-rate brachytherapy for cervical cancer
Table 1. Patient characteristics Number of patients Median age (years): 68 (30–88) FIGO stage IB IIA IIB IIIA IIIB IVA Hydronephrosis Tumor diameter (maximum)* ⬍20 20–40 40–60 ⱖ60 Pelvic nodal status (ⱖ10 mm)* Negative Positive American Brachytherapy Society stage Early (nonbulky Stage I/II, ⬍4 cm) Advanced (ⱖ4 cm or Stage IIIB)
13 3 31 1 38 2 4 (4.5%) 8 25 41 14 69 19 25 63
* Assessed by MRI.
METHODS AND MATERIALS Patients Three hundred and twelve patients with uterine cervical cancer had been treated with radiotherapy at the University of the Ryukyus Hospital between August 1994 and December 1999. From these, 71 patients treated postoperatively, 65 patients treated with chemotherapy, 53 patients treated with palliative intent (i.e., Stage IVB, recurrence), 15 patients without pretreatment MRI, 10 patients with adenocarcinoma, 6 patients who underwent external beam radiotherapy at other institutions, and 4 patients with stump cancer were excluded. Therefore, 88 patients with previously untreated uterine cervical squamous cell carcinoma were the subjects of this study. Table 1 shows the patients’ characteristics. Patients were jointly staged by gynecologic oncologists and radiation oncologists according to the FIGO staging system. Pelvic examination was performed without general anesthesia. Tumor diameter and pelvic nodal status were assessed by MRI (12). Maximum tumor diameter ranged from 19 to 86 mm with a median diameter of 45 mm. Lymph nodes greater than 10 mm in minimum diameter were assessed as enlarged. No patients had para-aortic lymph node enlargement greater than 10 mm in minimum diameter as assessed by CT. Patients were divided into two categories according to the American Brachytherapy Society (ABS) definition: Early disease is defined as nonbulky Stage I/II, less than 4 cm diameter, and advanced disease is defined as greater than 4 cm diameter or Stage IIIB (19). Median pretreatment hemoglobin level was 12.7 g/dL (range: 8 –15 g/dL). Median height and weight of the patients was 147 cm (range: 138 –160 cm) and 52 kg (range: 38 –78 kg), respectively.
● T. TOITA et al.
1345
External beam radiotherapy (EBRT) All patients were treated in a supine position with an 18-MV photon beam. Treatment was delivered 5 days per week. Seventy patients (80%) were treated over the whole pelvic field. Seventeen patients over 75 years of age were treated over a small pelvic field that excluded the common iliac region, and 1 patient was treated over an extended field. Dose per fraction of EBRT was 2.0 Gy, except for the extended-field radiotherapy, in which 1.8 Gy was used. These were treated through opposing anteroposterior fields. The dose of EBRT was calculated at the midpoint on beam axes using central axis depth– dose tables without inhomogeneity correction. Depth from skin surface to the midpoint on beam axes ranged from 7 to 12.5 cm, with a median of 9 cm. A simple rectangular block, 4 cm in width at midplane, was used as a midline block. It was inserted after the first application of HDR-ICBT. This did not extend to the top of the pelvic field. Boost irradiation was delivered in 19 patients: Eight received a boost to enlarged nodes, 9 to the parametrium, and the remaining 2 to both. A total dose of 6 –10 Gy was delivered using posterior-lateral, anteroposterior-lateral, or 4-field orthogonal beams. Intracavitary brachytherapy (ICBT) The HDR facility was a facts system (Buchler/STS, Germany). The system contained a high-activity 192Ir source (296 GBq at time of installation). HDR-ICBT was administered 289 times. Before December 1994, Fletcher-type ovoids (without lead shielding) were used mainly (21 treatments); these were replaced by a Henschketype applicator (249 treatments, 86%). Three sizes of ovoids were available according to expansion of the vaginal wall. Figure 1 shows the typical application using the Henschke-type (medium-sized) applicator. For patients with massive vaginal infiltration (tumor involved lower 1/2 of vaginal wall), a tandem and vaginal cylinder were used (19 treatments). Applications were usually done under i.v. anesthesia with ECG and O2 saturation monitoring. For some patients with contraindications to i.v. anesthesia, epidural anesthesia was performed. Pentazocin (15– 60 mg) and midazolam (5–10 mg) (or diazepam, 5–10 mg) were used for i.v. anesthesia. In an attempt to displace the bladder and rectum from the applicators, anterior and posterior vaginal packing was done. All the applicator insertions, radiograph generation, and treatment were performed in a dedicated brachytherapy suite equipped with imaging equipment. Therefore, there was no need to move patients during the ICBT procedure. Patients were treated in the dorsal lithotomy position. Applicators were inserted with fluoroscopic guidance for proper placement. An original external fixation device was used to prevent applicator movement. Point A was defined on X-ray films as being 2 cm superior (along the tandem) to the flange abutting external cervical os and 2 cm lateral from the axis of the tandem. For patients with a deep vaginal fornix, the original point A definition of the Manchester System, found by drawing a line connecting the superior aspects of the vaginal
1346
I. J. Radiation Oncology
● Biology ● Physics
Volume 56, Number 5, 2003
Fig. 1. Typical radiograph of high-dose-rate intracavitary therapy. (a) Anterior-posterior view, (b) lateral view.
High-dose-rate brachytherapy for cervical cancer
● T. TOITA et al.
1347
Table 2. Patterns of source loading (University of the Ryukyus Hospital) Dwell time ratio Tandem length 4 cm 5 cm 6 cm 7 cm
Ovoid type Henschke Fletcher Henschke Henschke Fletcher Henschke Henschke Fletcher Henschke Henschke Fletcher Henschke
Tandem*
Ovoid (right/left)
small medium small medium small medium small medium
1.5/1.5/1.5/1.5/1/1/1/1 1.5/1.5/1.5/1.5/1/1/1/1
7/7 8/8
1.5/1.5/1.5/1.5/1/1/1/1/1/1 1.5/1.5/1.5/1.5/1/1/1/1/1/1
7/7 8/8
1.5/1.5/1.5/1.5/1/1/1/1/1/1/1/1 1.5/1.5/1.5/1.5/1/1/1/1/1/1/1/1
7/7 8/8
1.5/1.5/1.5/1.5/1/1/1/1/1/1/1/1/1/1 1.5/1.5/1.5/1.5/1/1/1/1/1/1/1/1/1/1
7/7 8/8
* Stepwise distance in tandem is equal: 0.5 cm.
ovoids (mucous membrane of the lateral fornix) and measuring 2 cm superior along the tandem from the interception with this line and 2 cm perpendicular to this in the lateral direction, was used (20). Source loading corresponded to the Manchester System for uterine cervical cancer (20). Patterns of dwell positions and weightings of the source are shown in Table 2. No planned optimization was done. However, for patients whose rectal dose exceeded point A dose at the first calculation, standard dwell weighting was arranged to decrease the rectal dose. HDR-ICBT was performed once a week with a daily dose of 6 Gy at point A. Figure 2 shows the calculated dose distribution curves.
The rectal, bladder, and vaginal reference points were determined according to the guidelines in ICRU Report 38 (21). We commenced calculation of the ICRU bladder reference dose routinely from May 1995. The ICRU rectal reference and vaginal reference dose were also calculated routinely from November 1995. When the vaginal cylinder was used, the rectal reference dose was not calculated. As a result, rectal reference dose was calculated in 189 treatments (70%), the bladder reference dose in 226 (78%), and the vagina in 193 (72%). Point dose calculations were done for all ICBT sessions in 49 cases (64%) for the rectal dose, 66 cases (75%) for the bladder dose, and 48 cases (62%) for the vaginal dose.
Fig. 2. Isodose curves generated with fixed relative dwell weighting based on the Manchester System.
1348
I. J. Radiation Oncology
● Biology ● Physics
Table 3. Treatment schedules EBRT (Gy) WP
WP with MB
HDR-ICBT (Gy/fr.)
No. of pts. treated
0 0 20 30 30 40 40 40 50
50 50 30 20 20 10 10 10 0
24/4 30/5 24/4 18/3 24/4 12/2 18/3 24/4 6/1
1 5 7 2 9 4 55 4 1
Abbreviations: EBRT ⫽ external beam radiation therapy; WP ⫽ whole pelvic irradiation; MB ⫽ midline block; HDR-ICBT ⫽ high-dose-rate intracavitary therapy; fr. ⫽ fraction; pts. ⫽ patients.
Dose and BED The median cumulative point A dose (sum of the EBRT midline dose and point A dose of ICBT) was 58 Gy (range: 24 – 64 Gy). The biologic effective dose (BED) was calculated to the tumor (␣/ ⫽ 10) and late-responding tissues (␣/ ⫽ 3) for both EBRT and HDR-ICBT (22). The total BED values to point A and ICRU reference points (rectum, bladder, vagina) were the summation of those of EBRT and HDR-ICBT. Median cumulative values of point A BED were 76.8 Gy10 (range: 38.4 – 86.4 Gy10) for early-responding tissue, and 120.8 Gy3 (range: 72–138.8 Gy3) for lateresponding tissue. Median calculated doses at ICRU reference points from single ICBT were 4.5 Gy (range: 2.6 – 8.3 Gy) for the rectum, 4.3 Gy (range: 2.0 – 8.0 Gy) for the bladder, and 13.2 Gy (range: 8.1–25 Gy) for the vagina. Dose at reference points exceeded point A doses in 17 treatments (7.5%) for the bladder dose and in 15 treatments (7.9%) for the rectal dose. The sum of the EBRT midline dose and the reference point dose from ICBT were 23.1– 61.8 Gy (median: 50.1 Gy) at the rectum, 25– 61.4 Gy (median: 52.3 Gy) at the bladder, and 36.6 –105.9 Gy (median: 81.2 Gy) at the vagina. Total BED at reference points from EBRT and ICBT was 59.1–134.4 Gy3 (median: 97.7 Gy3) at the rectum, 54.6 –130.4 Gy3 (median: 97.8 Gy3) at the bladder, and 185.5– 618 Gy3 (median: 324 Gy3) at the vagina.
Treatment schedules Table 3 shows treatment schedules for patients with uterine cervical cancer in our department. EBRT preceded ICBT, except for patients with small Stage IB disease. A midline block was inserted after the first application of ICBT. The most frequent schedule in our department was a whole-pelvic EBRT of 40 Gy followed by HDR-ICBT of 18 Gy and whole-pelvic EBRT of 10 Gy with a midline block (55 patients, 63%). No EBRT was given on the same day as ICBT. Median overall treatment time was 48 days with a range of 36 to 84 days.
Volume 56, Number 5, 2003
Statistics and follow-up Statistical analyses were performed using SPSS version 6.1J for Macintosh computers. Actuarial pelvic control rate, survival rate, and late complication rates were calculated using the Kaplan-Meier method. Statistical significance was determined using the log–rank test. Both radiation and gynecologic oncologists followed the patients treated. The patients were followed every month for the first year, every 2–3 months for the next 2 years, and 3–12 months afterward. Patients without regular visits were followed by telephone. No patients were lost to follow-up. Follow-up procedures included pelvic examination, palpation of supraclavicular nodes, cervical Papanicolaou smear, and review of serum squamous cell carcinoma antigen value. When central and/or parametrial recurrence was suspected by pelvic examination and/or Papanicolaou smears, a biopsy was taken for confirmation. A chest radiograph was taken annually. Other imaging studies, such as CT, MRI, ultrasonography, and bone scintigraphy, were not routinely performed. Late complications were graded according to the RTOG/EORTC late radiation morbidity scoring scheme (23). Patients with symptoms defined in the scheme were counted as having radiation-related complications. Confirmation by endoscopic examinations (cystoscopy/proctoscopy) was encouraged but optional. RESULTS Median follow-up time for all 88 patients was 48 months (range: 8 – 88 months), and 58 months (range: 27– 88 months) for the surviving 65 patients. No patients were lost to follow-up. Treatment results The actuarial 3-year pelvic control rate (PC), disease-free survival rate, and overall survival rate were 82%, 73%, and 77%, respectively. Table 4 shows the actuarial 3-year outcomes according to several factors. FIGO stage, ABS stage, and pelvic nodal status significantly affected both PC and disease-free survival. Overall treatment time was not the significant predictor on both PC and disease-free survival. Table 5 shows PC according to treatment schedules. No significant difference in PC was observed by treatment schedules or cumulative point A BED for patients with both ABS early and advanced disease. Late complications Twenty-four late complications were observed in 19 patients (22%). Proctitis occurred in 10 patients (Grade 2: 9, Grade 3: 1), enterocolitis in 5 (Grade 1: 1, Grade 2: 3, Grade 3: 1), and cystitis in 9 (all Grade 2). Two patients who suffered proctitis also had central recurrence. Median intervals from start of treatment to onset of complications were 20 months for proctitis, 21 months for cystitis, and 10 months for enterocolitis. One patient suffered both Grade 3 proctitis and enterocolitis and required surgery. No vaginal ulceration was observed. No patient suffered rectovaginal
High-dose-rate brachytherapy for cervical cancer
Table 4. The 3-year actuarial outcomes Parameters
No.
DFS (%)
PC (%)
88
73
82
13 34 39 2
100 79 63 0 p ⫽ 0.0012
100 85 74 – p ⫽ 0.031
ABS stage Early Advanced
25 63
91 66 p ⫽ 0.0087
96 76 p ⫽ 0.022
Maximum tumor diameter* ⬍ 40 ⬎ ⱖ 40 ⬍
33 55
84 67 p ⫽ 0.087
88 78 p ⫽ 0.34
Enlarged pelvic nodes* Negative Positive
69 19
85 28 p ⬍ 0.0001
90 53 p ⬍ 0.0001
Overall treatment time (days) ⬍ 49 ⬎ ⱖ 49 ⱕ
35 52
68 77 p ⫽ 0.63
83 81 p ⫽ 0.89
All FIGO stage I II III IVA
Abbreviations: DFS ⫽ disease-free survival; ABS ⫽ American Brachytherapy Society; PC ⫽ pelvic control rate. * Assessed by MRI.
fistula or vesicovaginal fistula. No Grade 4 complication was observed. The actuarial 3-year complication rates (ⱖGrade 1) were 12% for proctitis, 11% for cystitis, and 14% for enterocolitis. Table 6 shows the complication rates according to the treatment schedules. The treatment schedules showed significant differences in the incidence of proctitis and enterocolitis, but not cystitis. No patient treated with total point A BED less than 70 Gy10 experienced proctitis. On the other hand, all 4 patients treated with total point A BED of 86.4 Gy10 (EBRT 40 Gy ⫹ ICBT 24 Gy/4 fractions) suffered Table 5. The 3-year actuarial pelvic control rate by treatment schedule Treatment schedule
Pelvic control rate (%)
WP (Gy)
HDR-ICBT (Gy/fr.)
Point A BED (Gy10)
0 20 30 40 30 40 40
30/5 24/4 18/3 12/2 24/4 18/3 24/4
48 62.4 64.8 67.2 74.4 76.8 86.4
ABS early
ABS advanced
100 (n ⫽ 5) 100 (n ⫽ 7) 100 (n ⫽ 2) – 100 (n ⫽ 2) 89 (n ⫽ 9) – NS
– – – 75 (n ⫽ 4) 100 (n ⫽ 7) 74 (n ⫽ 46) 75 (n ⫽ 4) NS
Abbreviations: WP ⫽ whole pelvic irradiation; HDR-ICBT ⫽ high-dose-rate intracavitary therapy; fr. ⫽ fraction; BED ⫽ biological effective dose; ABS ⫽ American Brachytherapy Society.
● T. TOITA et al.
1349
both proctitis and enterocolitis. Whereas cumulative dose at rectal point did not significantly affect the incidence of proctitis (⬍50 Gy: 6% vs. ⱖ50 Gy: 21%, p ⫽ 0.18), total BED had a significant effect (⬍100 Gy3: 4% vs. ⱖ100 Gy3: 31%, p ⫽ 0.013). Total BED at rectal reference point for 10 patients who suffered proctitis ranged from 97.5 to 134 Gy3 (mean: 115 Gy3). The incidence of cystitis was not significantly affected by cumulative dose (⬍50 Gy: 10% vs. ⱖ50 Gy: 13%, p ⫽ 0.68) or total BED at the bladder reference point (⬍100 Gy3: 11% vs. ⱖ100 Gy3: 12%, p ⫽ 0.94). Other factors, such as previous abdominal/pelvic surgery, cardiovascular disease, history of inflammatory bowel disease, renal failure, overall treatment time, use of vaginal cylinder, use of boost EBRT to lymph nodes/parametrium, did not significantly affect the incidence of late complications. DISCUSSION An appropriate dose and schedule of radiotherapy have been determined mainly on the bases of retrospective dose– response analyses of local control and late complications for several malignant tumors. Dose–response analysis on local control has usually been performed as a function of tumor size and total dose delivered (23). However, there are some difficulties in retrospective evaluation of these two factors for uterine cervical cancer. Table 7 shows the data from previously published literatures on HDR-ICBT for uterine cervical cancer. Although FIGO stage was stated, tumor size in diameter was not described in most series. Without tumor size data, meaningful comparison of local control is difficult. In the present analyses, we adapted the patient categorization scheme defined by the ABS (19). This is determined by the combination of FIGO stage and tumor size (19). Whereas an assessment method of tumor size was not defined, we used MRI for this purpose. MRI is considered to be an excellent modality to objectively assess tumor size in uterine cervical cancer (24). Several investigators have shown that tumor size measured by MRI is a significant prognostic indicator in radiotherapy for uterine cervical cancer (12, 25–27). Therefore, we consider our present results to be reference data for future comparisons. A wide variation of treatment schedules, both for EBRT and ICBT, also makes it difficult to analyze dose–response in uterine cervical cancer. The linear-quadratic model has been considered to be one of the most reliable methods for converting each dose into a comparable value (22). Several investigators have tried dose–response analysis using BED based on a linear-quadratic model of HDR-ICRT of uterine cervical cancer (6, 9, 16 –18, 28, 29). In this study, we adapted BED in addition to cumulative total dose. Pelvic control rate for ABS early disease was excellent in this study. No dose–response relationship was observed within total point A BED range 48 –76.8 Gy10 for early disease. Some investigators have also shown excellent PC for early-stage uterine cervical cancer treated with similar point A BED 46 –79 Gy10 (2, 4, 6, 30). Petereit and Pearcey
1350
I. J. Radiation Oncology
● Biology ● Physics
Volume 56, Number 5, 2003
Table 6. The 3-year actuarial incidence of late complications (Grade ⬎1) by treatment schedules Treatment schedule
Complication rate (%)
WP (Gy)
HDR-ICBT (Gy/fr.)
Point A BED (Gy10)
n
Proctitis
Enterocolitis
Cystitis
0 20 30 40 30 40 40
30/5 24/4 18/3 12/2 24/4 18/3 24/4
48 62.4 64.8 67.2 74.4 76.8 86.4
5 7 2 4 9 55 4
0 0 0 0 11 10 100 p ⬍ 0.0001
0 0 50 0 22 10 100 p ⫽ 0.0001
20 0 0 0 22 13 0 NS
Abbreviations: WP ⫽ whole pelvic irradiation; HDR-ICBT ⫽ high-dose-rate intracavitary therapy; fr. ⫽ fraction; BED ⫽ biological effective dose.
claimed that a BED 46 –79 Gy10 was less than tumoricidal (18). This is equivalent to 38 – 66 Gy for conventional fractionation. We consider also that these doses are inadequate to eradicate malignant epithelial tumors of 40 mm in diameter. However, a high-dose gradient of ICBT should be
taken into account to evaluate dose–response in cervical cancer. After administration of some amount of EBRT, the tumor would shrink in size to be delivered a larger dose than that of point A. As a result, the tumor could be controlled even with a low prescribed total dose at point A. This
Table 7. Treatment schedules and results of HDR intracavitary therapy for uterine cervical cancer Pelvic control Point A dose Total BED of HDR-ICBT at point A (Gy/fr.) (Gy10)
Authors
No. of cases
Central dose of EBRT (Gy)
North America Clark et al. (16)
43
46
30/3
115
Petereit et al. (17)
173
Early 20–30 Advanced 51–60 (1.7 Gy/fr.)
36–41/5 18.5–24.5/5 (point M)
85.3–109.7 85–100.9 (point M)
Japan/Asia Shigematsu et al. (1) Arai et al. (2)
143
20
30/3
1022
Early 0–20 Advance 20–40 Early 30–40 Advance 40 18–22 20 (median) 30 0–51.4 26.6 (mean) 40–44
Kataoka et al. (3) Teshima et al. (4) Ito et al. (5)
220
Ogino et al. (6)
253
Wang et al. (8)
173
Kodaira et al. (11) Lee et al. (29)
265
Hareyama et al. (30) Other Kapp et al. (9)
61
Ferrigno et al. (28) Present study
430 659
162
25.2–53 44.2 (mean) 20–45 40 (median) IIA 0 IIB, III 30–40
Total BED at rectum ICRU 38 (Gy3)
Early Advanced (I–IIB) (IIIⱕ)
Late complication All
Rectum Bladder Grade
26% (crude) NA
118–366 (median 169) NA
NA
NA
NA
NA
NA
71% (3 yr)
84
NA
NA
NA
23–29/4–5 15–24/3–4 24–30/4–5 24–30/4–5 24–35/3–5 30/4 (median) 20–34/4.5–6
45.8–59.9 58.5–86.4 74.4–96 86.4–96 – 76.5 65–90.4
NA NA
86–95% (crude) NA
67% (crude) NA
90% (1 yr) NA
NA
NA
NA
NA
15–35/5–8 29/5 (mean) 21.6/3
– 77.7 85.2–90
NA 147 (mean) NA NA
88–94% (crude) 88–100% (5-yr) 94% (5 yr) –
68% (crude) 74% (5-yr) 72% (5 yr) –
NA
–
–
NA
89% (5 yr)
69% (5 yr)
10–32 – 19.8 (mean) 82 24–51/8–17 – 39/13 (median) 98.7 29/5 45.8 17.3–23/2–4 72.2–80.3
161
24–54 46 (median)
8.5–38.7/2–6 18.5/3 (mean)
138
45
24/4
88
Early 0–40
18–30/3–5
Advance 30–40
12–24/2–4
– 85
93.9–175.8 60–93% (5 yr) (median 137) not ICRU 91.5 83.7–138.3 NA (median 108.5) 48–76.8 59.1–134.4 96% (med 64.8) (3 yr) 38.4–86.4 (med 97.7) (med 76.8)
36% (crude) 18% (crude) NA 18% (crude) NA 4% (crude) 76% 28% (crude) (crude) NA 52% (crude) 83% 38% (5 yr) (5 yr) 71% (5 26% yr) (crude) 95% 19% (crude) (crude) NA 3.5% (5 yr)
Comment
NA
3–4
NA
–
Chemoradiotherapy –
2% (crude) 15% (crude) NA
All?
–
All
–
2–3
–
3% (crude) 7% (crude) NA
2–3
–
All
–
All
–
9% (5 yr) 9% (crude) 5% (crude) 4% (5 yr)
All
–
All
Stage III
All
Stage IB
ⱖ3
Stage II, III
47% (5 yr)
NA
16% (5 yr)
13% (5 yr)
All
–
NA
62% (5 yr) 82% (3 yr)
16% (5 yr) 11% (3 yr)
11% (5 yr) 12% (3 yr)
All
–
All
–
76% (3 yr)
Abbreviations: EBRT ⫽ external beam radiation therapy; HDR-ICBT ⫽ high-dose-rate intracavitary therapy; fr. ⫽ fraction; BED ⫽ biologically effective dose; NA ⫽ not available; med ⫽ median.
High-dose-rate brachytherapy for cervical cancer
suggested that dose schedules with a total point A BED of about 50 – 80 Gy10 are adequate to achieve local control for patients with a tumor less than 40 mm in diameter without massive parametrial disease. Favorable outcome was presented also for patients with ABS advanced disease (IIIB/IVA or ⱖ40 mm) in this series. As for patients with ABS early disease, no dose–response relationship was observed within total point A BED range 67.2– 86.4 Gy10 (One patient received only 38.4 Gy10). This result suggested that these BED values might be adequate, and further dose escalation would not be necessary to control advanced cervical cancer. However, this should be evaluated with caution. First, patient selection bias might exist in our present analysis. Most patients in this category were candidates for chemotherapy-combined treatment and were not included in this study. Second, uncertainty in diagnosis of Stage IIIB should be taken into account. Only 10% of Stage IIIB patients had hydronephrosis in our series. This incidence was lower than that found in other series (31), suggesting that some patients with Stage IIB might migrate into Stage IIIB in our series (32). Therefore, we cannot make a firm conclusion regarding the adequate dose for advanced disease. We consider the actuarial 3-year late complication rates (Grade 1/2) of about 10% in our series to be acceptable. Although bladder complication was not dose dependent, rectal and small intestinal complications significantly correlated to total BED at point A. Only 1 patient treated with less than 70 Gy10 at point A developed enterocolitis. On the other hand, all 4 patients treated with a total point A BED of 86.4 Gy10 suffered both proctitis and enterocolitis. Several investigators have reported the comparable incidence of late complication for patients treated with almost the same BED schedule as ours (2, 4, 30). Other investigators (3, 5, 8, 11) have reported a slightly higher incidence of rectal complications for patients treated with higher point A BED (about 80 –100 Gy10). This suggested that total point A BED less than 80 Gy10 might be appropriate in view of rectal complication. However, Ferrigno et al. have reported a late complication rate similar to that of ours in patients treated with a higher BED of 91.5 Gy10 at point A (28). This discrepancy suggested that an analysis based on only BED at point A might be insufficient. Several investigators have analyzed probability of late complications as a function of total BED at ICRU 38 reference points (6, 9, 16, 28, 29). We also tried to analyze the relationship between the reference point BED and the incidence of late complications. Because grading severity of complications was considered to be highly subjective, we analyzed the incidence of all grade (Grade ⱖ1) complications. Calculated total BED at the bladder reference point had no significant correlation with the incidence of bladder complications in our series. Cystitis occurred even in patients with a lower BED at the bladder reference point. Kapp et al. (9) and Ferrigno et al. (28) also have reported the negative predictive value of BED at the bladder reference point. This indicated the unreliability of the bladder refer-
● T. TOITA et al.
1351
ence point. In contrast to the bladder complication rate, in the rectal complication rate a significant difference was observed between patients who had been delivered ⬍100 Gy3 and ⱖ100 Gy3 at the rectal reference point. Although Ferrigno et al. showed a negative predictive value (28), Ogino et al. demonstrated that total BED values at the rectal point were significantly correlated with the incidence of late rectal complications (6). They have reported the incidence of rectal complication (all grades): 0% for ⬍100 Gy3, 14% for 100 –120 Gy3, 44% for 120 –140 Gy3, and 72% for ⬎140 Gy3 (6). Clark et al. have reported the relationship between BED at the rectal reference point and the incidence of severe late complications (16). They demonstrated a threshold for severe rectal complication of 125 Gy3 at the rectal reference point (16). As mentioned above, Ferrigno et al. reported an incidence of rectal complication similar to ours, even though they used higher dose schedules (28). This could be explained by their acceptable BED (median: 108.5 Gy3) at the rectal reference point. In our series, only one suffered severe (Grade 3) rectal complication. Unfortunately, total BED at the reference point could not be calculated for this case. We consider that the low incidence of severe rectal complication was because of relatively low BED at the rectal reference point in our series. Therefore, total BED at the rectal reference point should be kept less than 100 –120 Gy3. The ABS recommended that attempts should be made to keep the dose to the rectal and bladder points below 80% of the prescribed (point A) ICRT dose (19). In this series, the median dose ratio was 0.75 for the rectal and 0.71 for the bladder points. The dose ratio for the rectal point exceeded 0.8 in 38%, 0.9 in 17%, and 1.0 in 8% of the treatments. This suggested that it is difficult to keep the rectal point dose below 80% of point A dose for many patients using our methods of ICBT. To reduce the dose to rectal and bladder mucosa, two ways can be considered for HDR-ICBT. One is proper displacement from the applicator surface to the rectal/bladder mucosa, and the other is dose optimization. The ABS has recommended some methods to displace the bladder and rectum away from the applicator (19). These included an in-built rectal retractor, radiopaque gauze, a posterior vaginal speculum blade, and an inflatable catheter bulb (19). We applied cotton/gauze packing for this purpose. To do this effectively, we routinely performed i.v. anesthesia. Furthermore, to soften and widen the vaginal cavity, especially for elderly patients, we routinely administered estrogen hormone during the treatment course. Despite these precautions, the dose at the rectal reference point could not be reduced enough to be within 80% of the point A dose. Other than the prescribed value of the point A dose, dose distribution is an important factor that affects doses to both tumor and surrounding normal tissues. For this reason, analysis of the calculation method is important. We have adapted a nonoptimized plan using fixed relative dwell weighting, manually input to the planning system based on the classic Manchester System. Uno et al. have indicated the importance of shape of the dose distribution on the inci-
1352
I. J. Radiation Oncology
● Biology ● Physics
dence of late rectal complication (10). Investigators at the Medical College of Wisconsin have introduced an original dose optimization program (33). They have demonstrated that their optimized calculation offered decreased vaginal wall and rectal doses (33). However, we consider that this optimization should be carefully evaluated in clinical practice. Close observation is necessary to evaluate whether this optimization compromises local control. The ABS has presented recommended dose and fractionation schemes for the combination of EBRT and HDRICBT (19). Calculated total BED at point A was 100 –108 Gy10 for these recommended schedules. A standard treatment schedule for HDR-ICBT has also been stated by the Gynecologic Oncology Group. The standard schedule is EBRT of 45 Gy and HDR-ICBT of 6 Gy ⫻ 5 (unpublished). Its calculated BED value at point A of 101 Gy10 is similar to that of the ABS value. These BED values are quite higher than those we assessed as adequate. Ferrigno et al. tried to calculate rectal BED (Gy3) according to percentage of dose to point A when the Gynecologic Oncology Group HDRICBT fractionation schedule was applied (28). They have demonstrated that the total rectal dose exceeded the threshold of 125 Gy3 for severe rectal complication if rectal dose was over 75% of point A (70%: 122.4 Gy3, 75%: 128.3 Gy3, 80%: 134.4 Gy3) (28). As described in the ABS paper, these
Volume 56, Number 5, 2003
suggested fractionation schemes have not been thoroughly tested (19). To our knowledge, no published clinical data are available concerning both local control and late complication in the sequence of these schedules. In conclusion, we consider that cumulative BED 70 – 80 Gy10 at point A might be appropriate for uterine cervical cancer patients with early disease treated with a combination of EBRT and HDR-ICBT in view of therapeutic ratio. Our two schedules, EBRT 30 Gy followed by HDR-ICBT of 24 Gy and EBRT of 20 Gy with a midline block (point A BED: 74.4 Gy10), and EBRT 40 Gy followed by HDRICBT of 18 Gy and EBRT of 10 Gy with a midline block (point A BED: 76.8 Gy10), correspond to the value. This study also showed that the cumulative BED at the rectal reference point was a significant predictor for late rectal complication. The present data and literature review suggested that this should be kept below 100 –120 Gy3 to prevent rectal complication. To further improve local control for patients with advanced disease, dose escalation is not suitable, because of the risk of increasing the incidence of late complications. However, this was only a retrospective analysis. A prospective study is necessary to clearly determine an optimum dose and fractionation schedules for EBRT and HDR-ICBT.
REFERENCES 1. Shigematsu Y, Nishiyama K, Masaki N, et al. Treatment of carcinoma of the uterine cervix by remotely controlled afterloading intracavitary radiotherapy with high-dose rate: A comparative study with a low-dose rate system. Int J Radiat Oncol Biol Phys 1983;9:351–356. 2. Arai T, Nakano T, Morita S, et al. High-dose-rate remote afterloading intracavitary radiation therapy for cancer of the uterine cervix. A 20-year experience. Cancer 1992;69:175– 180. 3. Kataoka M, Kawamura M, Nishiyama Y, et al. Results of the combination of external-beam and high-dose-rate intracavitary irradiation for patients with cervical carcinoma. Gynecol Oncol 1992;44:48–52. 4. Teshima T, Inoue To, Ikeda H, et al. High-dose rate and low-dose rate intracavitary therapy for carcinoma of the uterine cervix. Final results of Osaka University Hospital. Cancer 1993;72:2409–2414. 5. Ito H, Kutuki S, Nishiguchi I, et al. Radiotherapy for cervical cancer with high-dose rate brachytherapy— correlation between tumor size, dose and failure. Radiother Oncol 1994;31: 240–247. 6. Ogino I, Kitamura T, Okamoto N, et al. Late rectal complication following high dose rate intracavitary brachytherapy in cancer of the cervix. Int J Radiat Oncol Biol Phys 1995;31: 725–734. 7. Pechoux CL, Akine Y, Sumi M, et al. High dose rate brachytherapy for carcinoma of the uterine cervix: Comparison of two different fraction regimens. Int J Radiat Oncol Biol Phys 1995;31:735–741. 8. Wang C-J, Leung SW, Chen H-C, et al. High-dose-rate intracavitary brachytherapy (HDR-IC) in treatment of cervical carcinoma: 5-year results and implication of increased lowgrade rectal complication on initiation of an HDR-IC fractionation scheme. Int J Radiat Oncol Biol Phys 1997;38:391–398.
9. Kapp KS, Stuecklschweiger GF, Kapp DS, et al. Carcinoma of the cervix: Analysis of complications after primary external beam radiation and 192Ir HDR brachytherapy. Radiother Oncol 1997;42:143–153. 10. Uno T, Itami J, Aruga M, et al. High dose rate brachytherapy for carcinoma of the cervix: Risk factors for late rectal complications. Int J Radiat Oncol Biol Phys 1998;40:615–621. 11. Kodaira T, Karasawa K, Tanaka Y, et al. Definitive radiotherapy combined with high-dose-rate brachytherapy for stage III carcinoma of the uterine cervix: Retrospective analysis of prognostic factors concerning patient characteristics and treatment parameters. Int J Radiat Oncol Biol Phys 1998;41:319– 327. 12. Toita T, Kakinohana Y, Shinzato S, et al. Tumor diameter/ volume and pelvic node status assessed by magnetic resonance imaging (MRI) for uterine cervical cancer treated with irradiation. Int J Radiat Oncol Biol Phys 1999;43:777–782. 13. Eifel PJ. High-dose-rate brachytherapy for carcinoma of the cervix: High tech or high risk? Int J Radiat Oncol Biol Phys 1992;24:383–386. 14. Eifel PJ, Moughan J, Owen J, et al. Patterns of radiotherapy practice for patients with squamous carcinoma of the uterine cervix: Patterns of Care Study. Int J Radiat Oncol Biol Phys 1999;43:351–358. 15. Nag S, Orton C, Young D, et al. The American Brachytherapy Society survey of brachytherapy practice for carcinoma of the cervix in the United States. Gynecol Oncol 1999;73:111–118. 16. Clark BG, Souhami L, Roman TN, et al. The prediction of late rectal complications in patients treated with high dose-rate brachytherapy for carcinoma of the cervix. Int J Radiat Oncol Biol Phys 1997;38:989–993. 17. Petereit DG, Sarkaria JN, Potter DM, et al. High-dose-rate versus low-dose-rate brachytherapy in the treatment of cervical cancer: Analysis of tumor recurrence—the University of
High-dose-rate brachytherapy for cervical cancer
18.
19.
20. 21. 22. 23.
24.
25.
26.
Wisconsin experience. Int J Radiat Oncol Biol Phys 1999;45: 1267–1274. Petereit DG, Pearcey R. Literature analysis of high dose rate brachytherapy fractionation schedules in the treatment of cervical cancer: Is there an optimal fractionation schedule? Int J Radiat Oncol Biol Phys 1999;43:359–366. Nag S, Erickson B, Thomadsen B, et al., for the American Brachytherapy Society. The American Brachytherapy Society recommendations for high-dose-rate brachytherapy for carcinoma of the cervix. Int J Radiat Oncol Biol Phys 2000;48: 201–211. Tod M, Meredith W. A dosage system for use in the treatment of cancer of the uterine cervix. Br J Radiol 1938;11:809–824. Dose and volume specification for reporting intracavitary therapy in gynecology. ICRU report 38. Washington: ICRU,1985. Fowler JF. The linear-quadratic formula and progress in fractionated radiotherapy. Br J Radiol 1989;62:679–694. Withers HR, McBride WH. Biologic basis of radiation therapy. In: Perez CA, Brady LW. Principles and practice of radiation oncology. 3rd ed. Philadelphia: Lippincott-Raven; 1998. p. 79–118. Hawnaur JM, Johnson RJ, Buckley CH, et al. Staging, volume estimation, and assessment of nodal status in carcinoma of the cervix: Comparison of magnetic resonance imaging with surgical findings. Clin Radiol 1994;49:443–452. Hricak H, Quivey JM, Campos Z, et al. Carcinoma of the cervix: Predictive value of clinical and magnetic resonance (MR) imaging assessment of prognostic factors. Int J Radiat Oncol Biol Phys 1993;27:791–801. Mayr NA, Yuh WT, Zheng J, et al. Tumor size evaluated by
27.
28.
29.
30.
31.
32. 33.
● T. TOITA et al.
1353
pelvic examination compared with 3-D quantitative analysis in the prediction of outcome for cervical cancer. Int J Radiat Oncol Biol Phys 1997;39:395–404. Kodaira T, Fuwa N, Kamata M, et al. Clinical assessment by MRI for patients with stage II cervical carcinoma treated by radiation alone in multicenter analysis: Are all patients with stage II disease suitable candidates for chemoradiotherapy? Int J Radiat Oncol Biol Phys 2002;52:627–636. Ferrigno R, Novaes PERDS, Pellizzon ACA, et al. High-doserate brachytherapy in the treatment of uterine cervix cancer. Analysis of dose effectiveness and late complications. Int J Radiat Oncol Biol Phys 2001;50:1123–1135. Lee S-W, Suh CO, Chung EJ, et al. Dose optimization of fractionated external radiation and high-dose-rate intracavitary brachytherapy for FIGO stage IB uterine cervical carcinoma. Int J Radiat Oncol Biol Phys 2002;52:1338–1344. Hareyama M, Sakata K, Oouchi A, et al. High-dose-rate versus low-dose-rate intracavitary therapy for carcinoma of the uterine cervix. A randomized trial. Cancer 2002;94:117– 124. Logsdon MD, Eifel PJ. FIGO IIIB squamous cell carcinoma of the cervix: An analysis of prognostic factors emphasizing the balance between external beam and intracavitary radiation therapy. Int J Radiat Oncol Biol Phys 1999;43:763–775. Eifel PJ. Problems with the clinical staging of carcinoma of the cervix. Semin Radiat Oncol 1994;4:1–8. Mai J, Erickson B, Rownd J, et al. Comparison of four different dose specification methods for high-dose-rate intracavitary radiation for treatment of cervical cancer. Int J Radiat Oncol Biol Phys 2001;51:1131–1141.
doi:10.1016/S0360-3016(03)00288-8
CLINICAL INVESTIGATION
Cervix
COMBINATION EXTERNAL BEAM RADIOTHERAPY AND HIGH-DOSE-RATE INTRACAVITARY BRACHYTHERAPY FOR UTERINE CERVICAL CANCER: ANALYSIS OF DOSE AND FRACTIONATION SCHEDULE TAKAFUMI TOITA, M.D.,* YASUMASA KAKINOHANA, PH.D.,* KAZUHIKO OGAWA, M.D.,* GENKI ADACHI, M.D.,* HIDEHIKO MOROMIZATO, M.D.,† YUTAKA NAGAI, M.D.,† TOSHIYUKI MAEHAMA, M.D.,† KAORU SAKUMOTO, M.D.,† KOJI KANAZAWA, M.D.,† AND SADAYUKI MURAYAMA, M.D.* Departments of *Radiology and †Obstetrics and Gynecology, University of the Ryukyus School of Medicine, Okinawa, Japan Purpose: To determine an appropriate dose and fractionation schedule for a combination of external beam radiotherapy (EBRT) and high-dose-rate intracavitary brachytherapy (HDR-ICBT) for uterine cervical cancer. Methods: Eighty-eight patients with uterine cervical squamous cell carcinoma treated with EBRT and HDRICBT were analyzed. Twenty-five patients were classified as early disease (nonbulky Stage I/II, less than 4-cm diameter) and 63 patients as advanced disease (greater than 4 cm diameter or Stage IIIB) according to the American Brachytherapy Society definition. Tumor diameter was measured by MRI. Pelvic EBRT was delivered before applications of ICBT. HDR-ICBT was performed once a week, with a fraction point A dose of 6 Gy. Source loadings corresponded to the Manchester System for uterine cervical cancer. No planned optimization was done. A Henschke-type applicator was mostly used (86%). Median cumulative biologic effective dose (BED) at point A (EBRT ⴙ ICBT) was 64.8 Gy10 (range: 48 –76.8 Gy10) for early disease, and 76.8 Gy10 (range: 38.4 – 86.4 Gy10) for advanced disease. Median cumulative BED at ICRU 38 reference points (EBRT ⴙ ICBT) was 97.7 Gy3 (range: 59.1–134.4 Gy3) at the rectum, 97.8 Gy3 (range: 54.6 –130.4 Gy3) at the bladder, and 324 Gy3 (range: 185.5– 618 Gy3) at the vagina. Actuarial pelvic control rate and late complication rate were analyzed according to cumulative dose and calculated BED. Results: The 3-year actuarial pelvic control rate was 82% for all 88 patients: 96% for those with early disease, and 76% for advanced disease. For pelvic control, no significant dose–response relationship was observed by treatment schedules and cumulative BED at point A for both early and advanced disease. The 3-year actuarial late complication rates (Grade >1) were 12% for proctitis, 11% for cystitis, and 14% for enterocolitis. There were significant differences on the incidence of proctitis (p < 0.0001) and enterocolitis (p < 0.0001), but not for cystitis by the treatment schedules and cumulative point A BED. All 4 patients treated with 86.4 Gy10 at point A suffered both proctitis and enterocolitis. Patients with cumulative BED at rectal point of >100 Gy3 had significantly higher incidence of proctitis (31% vs. 4%, p ⴝ 0.013). Conclusions: In view of the therapeutic ratio, cumulative BED 70 – 80 Gy10 at point A is appropriate for uterine cervical cancer patients treated with a combination of EBRT and HDR-ICBT. Present results and data from other literatures suggested that cumulative BED at the rectal point should be kept below 100 –120 Gy3 to prevent late rectal complication. © 2003 Elsevier Inc. Cervix neoplasms, Radiotherapy, High-dose-rate brachytherapy, Biologic effective dose.
INTRODUCTION High-dose-rate intracavitary brachytherapy (HDR-ICBT) has been widely used in treatment of uterine cervical cancer in Asia and Europe (1–12). Despite some criticism of HDRICBT (13), its application has been increasing in the United States (14, 15). However, a wide variation of dose and fractionation schedules exists for the combination of external beam radiotherapy (EBRT) and HDR-ICBT (1–12, 14 –
18). An optimum treatment schedule has not yet been clearly determined. In this paper, we retrospectively analyzed the dose–response relationship for local control and late complication for uterine cervical cancer patients treated with a combination of EBRT and HDR-ICBT. The aim of this study was to determine an appropriate dose and fractionation schedule for EBRT and HDR-ICBT in the treatment of uterine cervical cancer.
Address for reprints: Takafumi Toita, M.D., Department of Radiology, University of the Ryukyus School of Medicine, 207 Uehara, Nishihara-cho, Okinawa, 903-0215 Japan. Fax: ⫹81-98895-1420; E-mail: [email protected]
This study was supported by Grants-in-Aid for Cancer Research No. 14 – 6 from the Ministry of Health and Welfare, Japan. Received Oct 3, 2002, and in revised form Mar 3, 2003. Accepted for publication Mar 4, 2003. 1344
High-dose-rate brachytherapy for cervical cancer
Table 1. Patient characteristics Number of patients Median age (years): 68 (30–88) FIGO stage IB IIA IIB IIIA IIIB IVA Hydronephrosis Tumor diameter (maximum)* ⬍20 20–40 40–60 ⱖ60 Pelvic nodal status (ⱖ10 mm)* Negative Positive American Brachytherapy Society stage Early (nonbulky Stage I/II, ⬍4 cm) Advanced (ⱖ4 cm or Stage IIIB)
13 3 31 1 38 2 4 (4.5%) 8 25 41 14 69 19 25 63
* Assessed by MRI.
METHODS AND MATERIALS Patients Three hundred and twelve patients with uterine cervical cancer had been treated with radiotherapy at the University of the Ryukyus Hospital between August 1994 and December 1999. From these, 71 patients treated postoperatively, 65 patients treated with chemotherapy, 53 patients treated with palliative intent (i.e., Stage IVB, recurrence), 15 patients without pretreatment MRI, 10 patients with adenocarcinoma, 6 patients who underwent external beam radiotherapy at other institutions, and 4 patients with stump cancer were excluded. Therefore, 88 patients with previously untreated uterine cervical squamous cell carcinoma were the subjects of this study. Table 1 shows the patients’ characteristics. Patients were jointly staged by gynecologic oncologists and radiation oncologists according to the FIGO staging system. Pelvic examination was performed without general anesthesia. Tumor diameter and pelvic nodal status were assessed by MRI (12). Maximum tumor diameter ranged from 19 to 86 mm with a median diameter of 45 mm. Lymph nodes greater than 10 mm in minimum diameter were assessed as enlarged. No patients had para-aortic lymph node enlargement greater than 10 mm in minimum diameter as assessed by CT. Patients were divided into two categories according to the American Brachytherapy Society (ABS) definition: Early disease is defined as nonbulky Stage I/II, less than 4 cm diameter, and advanced disease is defined as greater than 4 cm diameter or Stage IIIB (19). Median pretreatment hemoglobin level was 12.7 g/dL (range: 8 –15 g/dL). Median height and weight of the patients was 147 cm (range: 138 –160 cm) and 52 kg (range: 38 –78 kg), respectively.
● T. TOITA et al.
1345
External beam radiotherapy (EBRT) All patients were treated in a supine position with an 18-MV photon beam. Treatment was delivered 5 days per week. Seventy patients (80%) were treated over the whole pelvic field. Seventeen patients over 75 years of age were treated over a small pelvic field that excluded the common iliac region, and 1 patient was treated over an extended field. Dose per fraction of EBRT was 2.0 Gy, except for the extended-field radiotherapy, in which 1.8 Gy was used. These were treated through opposing anteroposterior fields. The dose of EBRT was calculated at the midpoint on beam axes using central axis depth– dose tables without inhomogeneity correction. Depth from skin surface to the midpoint on beam axes ranged from 7 to 12.5 cm, with a median of 9 cm. A simple rectangular block, 4 cm in width at midplane, was used as a midline block. It was inserted after the first application of HDR-ICBT. This did not extend to the top of the pelvic field. Boost irradiation was delivered in 19 patients: Eight received a boost to enlarged nodes, 9 to the parametrium, and the remaining 2 to both. A total dose of 6 –10 Gy was delivered using posterior-lateral, anteroposterior-lateral, or 4-field orthogonal beams. Intracavitary brachytherapy (ICBT) The HDR facility was a facts system (Buchler/STS, Germany). The system contained a high-activity 192Ir source (296 GBq at time of installation). HDR-ICBT was administered 289 times. Before December 1994, Fletcher-type ovoids (without lead shielding) were used mainly (21 treatments); these were replaced by a Henschketype applicator (249 treatments, 86%). Three sizes of ovoids were available according to expansion of the vaginal wall. Figure 1 shows the typical application using the Henschke-type (medium-sized) applicator. For patients with massive vaginal infiltration (tumor involved lower 1/2 of vaginal wall), a tandem and vaginal cylinder were used (19 treatments). Applications were usually done under i.v. anesthesia with ECG and O2 saturation monitoring. For some patients with contraindications to i.v. anesthesia, epidural anesthesia was performed. Pentazocin (15– 60 mg) and midazolam (5–10 mg) (or diazepam, 5–10 mg) were used for i.v. anesthesia. In an attempt to displace the bladder and rectum from the applicators, anterior and posterior vaginal packing was done. All the applicator insertions, radiograph generation, and treatment were performed in a dedicated brachytherapy suite equipped with imaging equipment. Therefore, there was no need to move patients during the ICBT procedure. Patients were treated in the dorsal lithotomy position. Applicators were inserted with fluoroscopic guidance for proper placement. An original external fixation device was used to prevent applicator movement. Point A was defined on X-ray films as being 2 cm superior (along the tandem) to the flange abutting external cervical os and 2 cm lateral from the axis of the tandem. For patients with a deep vaginal fornix, the original point A definition of the Manchester System, found by drawing a line connecting the superior aspects of the vaginal
1346
I. J. Radiation Oncology
● Biology ● Physics
Volume 56, Number 5, 2003
Fig. 1. Typical radiograph of high-dose-rate intracavitary therapy. (a) Anterior-posterior view, (b) lateral view.
High-dose-rate brachytherapy for cervical cancer
● T. TOITA et al.
1347
Table 2. Patterns of source loading (University of the Ryukyus Hospital) Dwell time ratio Tandem length 4 cm 5 cm 6 cm 7 cm
Ovoid type Henschke Fletcher Henschke Henschke Fletcher Henschke Henschke Fletcher Henschke Henschke Fletcher Henschke
Tandem*
Ovoid (right/left)
small medium small medium small medium small medium
1.5/1.5/1.5/1.5/1/1/1/1 1.5/1.5/1.5/1.5/1/1/1/1
7/7 8/8
1.5/1.5/1.5/1.5/1/1/1/1/1/1 1.5/1.5/1.5/1.5/1/1/1/1/1/1
7/7 8/8
1.5/1.5/1.5/1.5/1/1/1/1/1/1/1/1 1.5/1.5/1.5/1.5/1/1/1/1/1/1/1/1
7/7 8/8
1.5/1.5/1.5/1.5/1/1/1/1/1/1/1/1/1/1 1.5/1.5/1.5/1.5/1/1/1/1/1/1/1/1/1/1
7/7 8/8
* Stepwise distance in tandem is equal: 0.5 cm.
ovoids (mucous membrane of the lateral fornix) and measuring 2 cm superior along the tandem from the interception with this line and 2 cm perpendicular to this in the lateral direction, was used (20). Source loading corresponded to the Manchester System for uterine cervical cancer (20). Patterns of dwell positions and weightings of the source are shown in Table 2. No planned optimization was done. However, for patients whose rectal dose exceeded point A dose at the first calculation, standard dwell weighting was arranged to decrease the rectal dose. HDR-ICBT was performed once a week with a daily dose of 6 Gy at point A. Figure 2 shows the calculated dose distribution curves.
The rectal, bladder, and vaginal reference points were determined according to the guidelines in ICRU Report 38 (21). We commenced calculation of the ICRU bladder reference dose routinely from May 1995. The ICRU rectal reference and vaginal reference dose were also calculated routinely from November 1995. When the vaginal cylinder was used, the rectal reference dose was not calculated. As a result, rectal reference dose was calculated in 189 treatments (70%), the bladder reference dose in 226 (78%), and the vagina in 193 (72%). Point dose calculations were done for all ICBT sessions in 49 cases (64%) for the rectal dose, 66 cases (75%) for the bladder dose, and 48 cases (62%) for the vaginal dose.
Fig. 2. Isodose curves generated with fixed relative dwell weighting based on the Manchester System.
1348
I. J. Radiation Oncology
● Biology ● Physics
Table 3. Treatment schedules EBRT (Gy) WP
WP with MB
HDR-ICBT (Gy/fr.)
No. of pts. treated
0 0 20 30 30 40 40 40 50
50 50 30 20 20 10 10 10 0
24/4 30/5 24/4 18/3 24/4 12/2 18/3 24/4 6/1
1 5 7 2 9 4 55 4 1
Abbreviations: EBRT ⫽ external beam radiation therapy; WP ⫽ whole pelvic irradiation; MB ⫽ midline block; HDR-ICBT ⫽ high-dose-rate intracavitary therapy; fr. ⫽ fraction; pts. ⫽ patients.
Dose and BED The median cumulative point A dose (sum of the EBRT midline dose and point A dose of ICBT) was 58 Gy (range: 24 – 64 Gy). The biologic effective dose (BED) was calculated to the tumor (␣/ ⫽ 10) and late-responding tissues (␣/ ⫽ 3) for both EBRT and HDR-ICBT (22). The total BED values to point A and ICRU reference points (rectum, bladder, vagina) were the summation of those of EBRT and HDR-ICBT. Median cumulative values of point A BED were 76.8 Gy10 (range: 38.4 – 86.4 Gy10) for early-responding tissue, and 120.8 Gy3 (range: 72–138.8 Gy3) for lateresponding tissue. Median calculated doses at ICRU reference points from single ICBT were 4.5 Gy (range: 2.6 – 8.3 Gy) for the rectum, 4.3 Gy (range: 2.0 – 8.0 Gy) for the bladder, and 13.2 Gy (range: 8.1–25 Gy) for the vagina. Dose at reference points exceeded point A doses in 17 treatments (7.5%) for the bladder dose and in 15 treatments (7.9%) for the rectal dose. The sum of the EBRT midline dose and the reference point dose from ICBT were 23.1– 61.8 Gy (median: 50.1 Gy) at the rectum, 25– 61.4 Gy (median: 52.3 Gy) at the bladder, and 36.6 –105.9 Gy (median: 81.2 Gy) at the vagina. Total BED at reference points from EBRT and ICBT was 59.1–134.4 Gy3 (median: 97.7 Gy3) at the rectum, 54.6 –130.4 Gy3 (median: 97.8 Gy3) at the bladder, and 185.5– 618 Gy3 (median: 324 Gy3) at the vagina.
Treatment schedules Table 3 shows treatment schedules for patients with uterine cervical cancer in our department. EBRT preceded ICBT, except for patients with small Stage IB disease. A midline block was inserted after the first application of ICBT. The most frequent schedule in our department was a whole-pelvic EBRT of 40 Gy followed by HDR-ICBT of 18 Gy and whole-pelvic EBRT of 10 Gy with a midline block (55 patients, 63%). No EBRT was given on the same day as ICBT. Median overall treatment time was 48 days with a range of 36 to 84 days.
Volume 56, Number 5, 2003
Statistics and follow-up Statistical analyses were performed using SPSS version 6.1J for Macintosh computers. Actuarial pelvic control rate, survival rate, and late complication rates were calculated using the Kaplan-Meier method. Statistical significance was determined using the log–rank test. Both radiation and gynecologic oncologists followed the patients treated. The patients were followed every month for the first year, every 2–3 months for the next 2 years, and 3–12 months afterward. Patients without regular visits were followed by telephone. No patients were lost to follow-up. Follow-up procedures included pelvic examination, palpation of supraclavicular nodes, cervical Papanicolaou smear, and review of serum squamous cell carcinoma antigen value. When central and/or parametrial recurrence was suspected by pelvic examination and/or Papanicolaou smears, a biopsy was taken for confirmation. A chest radiograph was taken annually. Other imaging studies, such as CT, MRI, ultrasonography, and bone scintigraphy, were not routinely performed. Late complications were graded according to the RTOG/EORTC late radiation morbidity scoring scheme (23). Patients with symptoms defined in the scheme were counted as having radiation-related complications. Confirmation by endoscopic examinations (cystoscopy/proctoscopy) was encouraged but optional. RESULTS Median follow-up time for all 88 patients was 48 months (range: 8 – 88 months), and 58 months (range: 27– 88 months) for the surviving 65 patients. No patients were lost to follow-up. Treatment results The actuarial 3-year pelvic control rate (PC), disease-free survival rate, and overall survival rate were 82%, 73%, and 77%, respectively. Table 4 shows the actuarial 3-year outcomes according to several factors. FIGO stage, ABS stage, and pelvic nodal status significantly affected both PC and disease-free survival. Overall treatment time was not the significant predictor on both PC and disease-free survival. Table 5 shows PC according to treatment schedules. No significant difference in PC was observed by treatment schedules or cumulative point A BED for patients with both ABS early and advanced disease. Late complications Twenty-four late complications were observed in 19 patients (22%). Proctitis occurred in 10 patients (Grade 2: 9, Grade 3: 1), enterocolitis in 5 (Grade 1: 1, Grade 2: 3, Grade 3: 1), and cystitis in 9 (all Grade 2). Two patients who suffered proctitis also had central recurrence. Median intervals from start of treatment to onset of complications were 20 months for proctitis, 21 months for cystitis, and 10 months for enterocolitis. One patient suffered both Grade 3 proctitis and enterocolitis and required surgery. No vaginal ulceration was observed. No patient suffered rectovaginal
High-dose-rate brachytherapy for cervical cancer
Table 4. The 3-year actuarial outcomes Parameters
No.
DFS (%)
PC (%)
88
73
82
13 34 39 2
100 79 63 0 p ⫽ 0.0012
100 85 74 – p ⫽ 0.031
ABS stage Early Advanced
25 63
91 66 p ⫽ 0.0087
96 76 p ⫽ 0.022
Maximum tumor diameter* ⬍ 40 ⬎ ⱖ 40 ⬍
33 55
84 67 p ⫽ 0.087
88 78 p ⫽ 0.34
Enlarged pelvic nodes* Negative Positive
69 19
85 28 p ⬍ 0.0001
90 53 p ⬍ 0.0001
Overall treatment time (days) ⬍ 49 ⬎ ⱖ 49 ⱕ
35 52
68 77 p ⫽ 0.63
83 81 p ⫽ 0.89
All FIGO stage I II III IVA
Abbreviations: DFS ⫽ disease-free survival; ABS ⫽ American Brachytherapy Society; PC ⫽ pelvic control rate. * Assessed by MRI.
fistula or vesicovaginal fistula. No Grade 4 complication was observed. The actuarial 3-year complication rates (ⱖGrade 1) were 12% for proctitis, 11% for cystitis, and 14% for enterocolitis. Table 6 shows the complication rates according to the treatment schedules. The treatment schedules showed significant differences in the incidence of proctitis and enterocolitis, but not cystitis. No patient treated with total point A BED less than 70 Gy10 experienced proctitis. On the other hand, all 4 patients treated with total point A BED of 86.4 Gy10 (EBRT 40 Gy ⫹ ICBT 24 Gy/4 fractions) suffered Table 5. The 3-year actuarial pelvic control rate by treatment schedule Treatment schedule
Pelvic control rate (%)
WP (Gy)
HDR-ICBT (Gy/fr.)
Point A BED (Gy10)
0 20 30 40 30 40 40
30/5 24/4 18/3 12/2 24/4 18/3 24/4
48 62.4 64.8 67.2 74.4 76.8 86.4
ABS early
ABS advanced
100 (n ⫽ 5) 100 (n ⫽ 7) 100 (n ⫽ 2) – 100 (n ⫽ 2) 89 (n ⫽ 9) – NS
– – – 75 (n ⫽ 4) 100 (n ⫽ 7) 74 (n ⫽ 46) 75 (n ⫽ 4) NS
Abbreviations: WP ⫽ whole pelvic irradiation; HDR-ICBT ⫽ high-dose-rate intracavitary therapy; fr. ⫽ fraction; BED ⫽ biological effective dose; ABS ⫽ American Brachytherapy Society.
● T. TOITA et al.
1349
both proctitis and enterocolitis. Whereas cumulative dose at rectal point did not significantly affect the incidence of proctitis (⬍50 Gy: 6% vs. ⱖ50 Gy: 21%, p ⫽ 0.18), total BED had a significant effect (⬍100 Gy3: 4% vs. ⱖ100 Gy3: 31%, p ⫽ 0.013). Total BED at rectal reference point for 10 patients who suffered proctitis ranged from 97.5 to 134 Gy3 (mean: 115 Gy3). The incidence of cystitis was not significantly affected by cumulative dose (⬍50 Gy: 10% vs. ⱖ50 Gy: 13%, p ⫽ 0.68) or total BED at the bladder reference point (⬍100 Gy3: 11% vs. ⱖ100 Gy3: 12%, p ⫽ 0.94). Other factors, such as previous abdominal/pelvic surgery, cardiovascular disease, history of inflammatory bowel disease, renal failure, overall treatment time, use of vaginal cylinder, use of boost EBRT to lymph nodes/parametrium, did not significantly affect the incidence of late complications. DISCUSSION An appropriate dose and schedule of radiotherapy have been determined mainly on the bases of retrospective dose– response analyses of local control and late complications for several malignant tumors. Dose–response analysis on local control has usually been performed as a function of tumor size and total dose delivered (23). However, there are some difficulties in retrospective evaluation of these two factors for uterine cervical cancer. Table 7 shows the data from previously published literatures on HDR-ICBT for uterine cervical cancer. Although FIGO stage was stated, tumor size in diameter was not described in most series. Without tumor size data, meaningful comparison of local control is difficult. In the present analyses, we adapted the patient categorization scheme defined by the ABS (19). This is determined by the combination of FIGO stage and tumor size (19). Whereas an assessment method of tumor size was not defined, we used MRI for this purpose. MRI is considered to be an excellent modality to objectively assess tumor size in uterine cervical cancer (24). Several investigators have shown that tumor size measured by MRI is a significant prognostic indicator in radiotherapy for uterine cervical cancer (12, 25–27). Therefore, we consider our present results to be reference data for future comparisons. A wide variation of treatment schedules, both for EBRT and ICBT, also makes it difficult to analyze dose–response in uterine cervical cancer. The linear-quadratic model has been considered to be one of the most reliable methods for converting each dose into a comparable value (22). Several investigators have tried dose–response analysis using BED based on a linear-quadratic model of HDR-ICRT of uterine cervical cancer (6, 9, 16 –18, 28, 29). In this study, we adapted BED in addition to cumulative total dose. Pelvic control rate for ABS early disease was excellent in this study. No dose–response relationship was observed within total point A BED range 48 –76.8 Gy10 for early disease. Some investigators have also shown excellent PC for early-stage uterine cervical cancer treated with similar point A BED 46 –79 Gy10 (2, 4, 6, 30). Petereit and Pearcey
1350
I. J. Radiation Oncology
● Biology ● Physics
Volume 56, Number 5, 2003
Table 6. The 3-year actuarial incidence of late complications (Grade ⬎1) by treatment schedules Treatment schedule
Complication rate (%)
WP (Gy)
HDR-ICBT (Gy/fr.)
Point A BED (Gy10)
n
Proctitis
Enterocolitis
Cystitis
0 20 30 40 30 40 40
30/5 24/4 18/3 12/2 24/4 18/3 24/4
48 62.4 64.8 67.2 74.4 76.8 86.4
5 7 2 4 9 55 4
0 0 0 0 11 10 100 p ⬍ 0.0001
0 0 50 0 22 10 100 p ⫽ 0.0001
20 0 0 0 22 13 0 NS
Abbreviations: WP ⫽ whole pelvic irradiation; HDR-ICBT ⫽ high-dose-rate intracavitary therapy; fr. ⫽ fraction; BED ⫽ biological effective dose.
claimed that a BED 46 –79 Gy10 was less than tumoricidal (18). This is equivalent to 38 – 66 Gy for conventional fractionation. We consider also that these doses are inadequate to eradicate malignant epithelial tumors of 40 mm in diameter. However, a high-dose gradient of ICBT should be
taken into account to evaluate dose–response in cervical cancer. After administration of some amount of EBRT, the tumor would shrink in size to be delivered a larger dose than that of point A. As a result, the tumor could be controlled even with a low prescribed total dose at point A. This
Table 7. Treatment schedules and results of HDR intracavitary therapy for uterine cervical cancer Pelvic control Point A dose Total BED of HDR-ICBT at point A (Gy/fr.) (Gy10)
Authors
No. of cases
Central dose of EBRT (Gy)
North America Clark et al. (16)
43
46
30/3
115
Petereit et al. (17)
173
Early 20–30 Advanced 51–60 (1.7 Gy/fr.)
36–41/5 18.5–24.5/5 (point M)
85.3–109.7 85–100.9 (point M)
Japan/Asia Shigematsu et al. (1) Arai et al. (2)
143
20
30/3
1022
Early 0–20 Advance 20–40 Early 30–40 Advance 40 18–22 20 (median) 30 0–51.4 26.6 (mean) 40–44
Kataoka et al. (3) Teshima et al. (4) Ito et al. (5)
220
Ogino et al. (6)
253
Wang et al. (8)
173
Kodaira et al. (11) Lee et al. (29)
265
Hareyama et al. (30) Other Kapp et al. (9)
61
Ferrigno et al. (28) Present study
430 659
162
25.2–53 44.2 (mean) 20–45 40 (median) IIA 0 IIB, III 30–40
Total BED at rectum ICRU 38 (Gy3)
Early Advanced (I–IIB) (IIIⱕ)
Late complication All
Rectum Bladder Grade
26% (crude) NA
118–366 (median 169) NA
NA
NA
NA
NA
NA
71% (3 yr)
84
NA
NA
NA
23–29/4–5 15–24/3–4 24–30/4–5 24–30/4–5 24–35/3–5 30/4 (median) 20–34/4.5–6
45.8–59.9 58.5–86.4 74.4–96 86.4–96 – 76.5 65–90.4
NA NA
86–95% (crude) NA
67% (crude) NA
90% (1 yr) NA
NA
NA
NA
NA
15–35/5–8 29/5 (mean) 21.6/3
– 77.7 85.2–90
NA 147 (mean) NA NA
88–94% (crude) 88–100% (5-yr) 94% (5 yr) –
68% (crude) 74% (5-yr) 72% (5 yr) –
NA
–
–
NA
89% (5 yr)
69% (5 yr)
10–32 – 19.8 (mean) 82 24–51/8–17 – 39/13 (median) 98.7 29/5 45.8 17.3–23/2–4 72.2–80.3
161
24–54 46 (median)
8.5–38.7/2–6 18.5/3 (mean)
138
45
24/4
88
Early 0–40
18–30/3–5
Advance 30–40
12–24/2–4
– 85
93.9–175.8 60–93% (5 yr) (median 137) not ICRU 91.5 83.7–138.3 NA (median 108.5) 48–76.8 59.1–134.4 96% (med 64.8) (3 yr) 38.4–86.4 (med 97.7) (med 76.8)
36% (crude) 18% (crude) NA 18% (crude) NA 4% (crude) 76% 28% (crude) (crude) NA 52% (crude) 83% 38% (5 yr) (5 yr) 71% (5 26% yr) (crude) 95% 19% (crude) (crude) NA 3.5% (5 yr)
Comment
NA
3–4
NA
–
Chemoradiotherapy –
2% (crude) 15% (crude) NA
All?
–
All
–
2–3
–
3% (crude) 7% (crude) NA
2–3
–
All
–
All
–
9% (5 yr) 9% (crude) 5% (crude) 4% (5 yr)
All
–
All
Stage III
All
Stage IB
ⱖ3
Stage II, III
47% (5 yr)
NA
16% (5 yr)
13% (5 yr)
All
–
NA
62% (5 yr) 82% (3 yr)
16% (5 yr) 11% (3 yr)
11% (5 yr) 12% (3 yr)
All
–
All
–
76% (3 yr)
Abbreviations: EBRT ⫽ external beam radiation therapy; HDR-ICBT ⫽ high-dose-rate intracavitary therapy; fr. ⫽ fraction; BED ⫽ biologically effective dose; NA ⫽ not available; med ⫽ median.
High-dose-rate brachytherapy for cervical cancer
suggested that dose schedules with a total point A BED of about 50 – 80 Gy10 are adequate to achieve local control for patients with a tumor less than 40 mm in diameter without massive parametrial disease. Favorable outcome was presented also for patients with ABS advanced disease (IIIB/IVA or ⱖ40 mm) in this series. As for patients with ABS early disease, no dose–response relationship was observed within total point A BED range 67.2– 86.4 Gy10 (One patient received only 38.4 Gy10). This result suggested that these BED values might be adequate, and further dose escalation would not be necessary to control advanced cervical cancer. However, this should be evaluated with caution. First, patient selection bias might exist in our present analysis. Most patients in this category were candidates for chemotherapy-combined treatment and were not included in this study. Second, uncertainty in diagnosis of Stage IIIB should be taken into account. Only 10% of Stage IIIB patients had hydronephrosis in our series. This incidence was lower than that found in other series (31), suggesting that some patients with Stage IIB might migrate into Stage IIIB in our series (32). Therefore, we cannot make a firm conclusion regarding the adequate dose for advanced disease. We consider the actuarial 3-year late complication rates (Grade 1/2) of about 10% in our series to be acceptable. Although bladder complication was not dose dependent, rectal and small intestinal complications significantly correlated to total BED at point A. Only 1 patient treated with less than 70 Gy10 at point A developed enterocolitis. On the other hand, all 4 patients treated with a total point A BED of 86.4 Gy10 suffered both proctitis and enterocolitis. Several investigators have reported the comparable incidence of late complication for patients treated with almost the same BED schedule as ours (2, 4, 30). Other investigators (3, 5, 8, 11) have reported a slightly higher incidence of rectal complications for patients treated with higher point A BED (about 80 –100 Gy10). This suggested that total point A BED less than 80 Gy10 might be appropriate in view of rectal complication. However, Ferrigno et al. have reported a late complication rate similar to that of ours in patients treated with a higher BED of 91.5 Gy10 at point A (28). This discrepancy suggested that an analysis based on only BED at point A might be insufficient. Several investigators have analyzed probability of late complications as a function of total BED at ICRU 38 reference points (6, 9, 16, 28, 29). We also tried to analyze the relationship between the reference point BED and the incidence of late complications. Because grading severity of complications was considered to be highly subjective, we analyzed the incidence of all grade (Grade ⱖ1) complications. Calculated total BED at the bladder reference point had no significant correlation with the incidence of bladder complications in our series. Cystitis occurred even in patients with a lower BED at the bladder reference point. Kapp et al. (9) and Ferrigno et al. (28) also have reported the negative predictive value of BED at the bladder reference point. This indicated the unreliability of the bladder refer-
● T. TOITA et al.
1351
ence point. In contrast to the bladder complication rate, in the rectal complication rate a significant difference was observed between patients who had been delivered ⬍100 Gy3 and ⱖ100 Gy3 at the rectal reference point. Although Ferrigno et al. showed a negative predictive value (28), Ogino et al. demonstrated that total BED values at the rectal point were significantly correlated with the incidence of late rectal complications (6). They have reported the incidence of rectal complication (all grades): 0% for ⬍100 Gy3, 14% for 100 –120 Gy3, 44% for 120 –140 Gy3, and 72% for ⬎140 Gy3 (6). Clark et al. have reported the relationship between BED at the rectal reference point and the incidence of severe late complications (16). They demonstrated a threshold for severe rectal complication of 125 Gy3 at the rectal reference point (16). As mentioned above, Ferrigno et al. reported an incidence of rectal complication similar to ours, even though they used higher dose schedules (28). This could be explained by their acceptable BED (median: 108.5 Gy3) at the rectal reference point. In our series, only one suffered severe (Grade 3) rectal complication. Unfortunately, total BED at the reference point could not be calculated for this case. We consider that the low incidence of severe rectal complication was because of relatively low BED at the rectal reference point in our series. Therefore, total BED at the rectal reference point should be kept less than 100 –120 Gy3. The ABS recommended that attempts should be made to keep the dose to the rectal and bladder points below 80% of the prescribed (point A) ICRT dose (19). In this series, the median dose ratio was 0.75 for the rectal and 0.71 for the bladder points. The dose ratio for the rectal point exceeded 0.8 in 38%, 0.9 in 17%, and 1.0 in 8% of the treatments. This suggested that it is difficult to keep the rectal point dose below 80% of point A dose for many patients using our methods of ICBT. To reduce the dose to rectal and bladder mucosa, two ways can be considered for HDR-ICBT. One is proper displacement from the applicator surface to the rectal/bladder mucosa, and the other is dose optimization. The ABS has recommended some methods to displace the bladder and rectum away from the applicator (19). These included an in-built rectal retractor, radiopaque gauze, a posterior vaginal speculum blade, and an inflatable catheter bulb (19). We applied cotton/gauze packing for this purpose. To do this effectively, we routinely performed i.v. anesthesia. Furthermore, to soften and widen the vaginal cavity, especially for elderly patients, we routinely administered estrogen hormone during the treatment course. Despite these precautions, the dose at the rectal reference point could not be reduced enough to be within 80% of the point A dose. Other than the prescribed value of the point A dose, dose distribution is an important factor that affects doses to both tumor and surrounding normal tissues. For this reason, analysis of the calculation method is important. We have adapted a nonoptimized plan using fixed relative dwell weighting, manually input to the planning system based on the classic Manchester System. Uno et al. have indicated the importance of shape of the dose distribution on the inci-
1352
I. J. Radiation Oncology
● Biology ● Physics
dence of late rectal complication (10). Investigators at the Medical College of Wisconsin have introduced an original dose optimization program (33). They have demonstrated that their optimized calculation offered decreased vaginal wall and rectal doses (33). However, we consider that this optimization should be carefully evaluated in clinical practice. Close observation is necessary to evaluate whether this optimization compromises local control. The ABS has presented recommended dose and fractionation schemes for the combination of EBRT and HDRICBT (19). Calculated total BED at point A was 100 –108 Gy10 for these recommended schedules. A standard treatment schedule for HDR-ICBT has also been stated by the Gynecologic Oncology Group. The standard schedule is EBRT of 45 Gy and HDR-ICBT of 6 Gy ⫻ 5 (unpublished). Its calculated BED value at point A of 101 Gy10 is similar to that of the ABS value. These BED values are quite higher than those we assessed as adequate. Ferrigno et al. tried to calculate rectal BED (Gy3) according to percentage of dose to point A when the Gynecologic Oncology Group HDRICBT fractionation schedule was applied (28). They have demonstrated that the total rectal dose exceeded the threshold of 125 Gy3 for severe rectal complication if rectal dose was over 75% of point A (70%: 122.4 Gy3, 75%: 128.3 Gy3, 80%: 134.4 Gy3) (28). As described in the ABS paper, these
Volume 56, Number 5, 2003
suggested fractionation schemes have not been thoroughly tested (19). To our knowledge, no published clinical data are available concerning both local control and late complication in the sequence of these schedules. In conclusion, we consider that cumulative BED 70 – 80 Gy10 at point A might be appropriate for uterine cervical cancer patients with early disease treated with a combination of EBRT and HDR-ICBT in view of therapeutic ratio. Our two schedules, EBRT 30 Gy followed by HDR-ICBT of 24 Gy and EBRT of 20 Gy with a midline block (point A BED: 74.4 Gy10), and EBRT 40 Gy followed by HDRICBT of 18 Gy and EBRT of 10 Gy with a midline block (point A BED: 76.8 Gy10), correspond to the value. This study also showed that the cumulative BED at the rectal reference point was a significant predictor for late rectal complication. The present data and literature review suggested that this should be kept below 100 –120 Gy3 to prevent rectal complication. To further improve local control for patients with advanced disease, dose escalation is not suitable, because of the risk of increasing the incidence of late complications. However, this was only a retrospective analysis. A prospective study is necessary to clearly determine an optimum dose and fractionation schedules for EBRT and HDR-ICBT.
REFERENCES 1. Shigematsu Y, Nishiyama K, Masaki N, et al. Treatment of carcinoma of the uterine cervix by remotely controlled afterloading intracavitary radiotherapy with high-dose rate: A comparative study with a low-dose rate system. Int J Radiat Oncol Biol Phys 1983;9:351–356. 2. Arai T, Nakano T, Morita S, et al. High-dose-rate remote afterloading intracavitary radiation therapy for cancer of the uterine cervix. A 20-year experience. Cancer 1992;69:175– 180. 3. Kataoka M, Kawamura M, Nishiyama Y, et al. Results of the combination of external-beam and high-dose-rate intracavitary irradiation for patients with cervical carcinoma. Gynecol Oncol 1992;44:48–52. 4. Teshima T, Inoue To, Ikeda H, et al. High-dose rate and low-dose rate intracavitary therapy for carcinoma of the uterine cervix. Final results of Osaka University Hospital. Cancer 1993;72:2409–2414. 5. Ito H, Kutuki S, Nishiguchi I, et al. Radiotherapy for cervical cancer with high-dose rate brachytherapy— correlation between tumor size, dose and failure. Radiother Oncol 1994;31: 240–247. 6. Ogino I, Kitamura T, Okamoto N, et al. Late rectal complication following high dose rate intracavitary brachytherapy in cancer of the cervix. Int J Radiat Oncol Biol Phys 1995;31: 725–734. 7. Pechoux CL, Akine Y, Sumi M, et al. High dose rate brachytherapy for carcinoma of the uterine cervix: Comparison of two different fraction regimens. Int J Radiat Oncol Biol Phys 1995;31:735–741. 8. Wang C-J, Leung SW, Chen H-C, et al. High-dose-rate intracavitary brachytherapy (HDR-IC) in treatment of cervical carcinoma: 5-year results and implication of increased lowgrade rectal complication on initiation of an HDR-IC fractionation scheme. Int J Radiat Oncol Biol Phys 1997;38:391–398.
9. Kapp KS, Stuecklschweiger GF, Kapp DS, et al. Carcinoma of the cervix: Analysis of complications after primary external beam radiation and 192Ir HDR brachytherapy. Radiother Oncol 1997;42:143–153. 10. Uno T, Itami J, Aruga M, et al. High dose rate brachytherapy for carcinoma of the cervix: Risk factors for late rectal complications. Int J Radiat Oncol Biol Phys 1998;40:615–621. 11. Kodaira T, Karasawa K, Tanaka Y, et al. Definitive radiotherapy combined with high-dose-rate brachytherapy for stage III carcinoma of the uterine cervix: Retrospective analysis of prognostic factors concerning patient characteristics and treatment parameters. Int J Radiat Oncol Biol Phys 1998;41:319– 327. 12. Toita T, Kakinohana Y, Shinzato S, et al. Tumor diameter/ volume and pelvic node status assessed by magnetic resonance imaging (MRI) for uterine cervical cancer treated with irradiation. Int J Radiat Oncol Biol Phys 1999;43:777–782. 13. Eifel PJ. High-dose-rate brachytherapy for carcinoma of the cervix: High tech or high risk? Int J Radiat Oncol Biol Phys 1992;24:383–386. 14. Eifel PJ, Moughan J, Owen J, et al. Patterns of radiotherapy practice for patients with squamous carcinoma of the uterine cervix: Patterns of Care Study. Int J Radiat Oncol Biol Phys 1999;43:351–358. 15. Nag S, Orton C, Young D, et al. The American Brachytherapy Society survey of brachytherapy practice for carcinoma of the cervix in the United States. Gynecol Oncol 1999;73:111–118. 16. Clark BG, Souhami L, Roman TN, et al. The prediction of late rectal complications in patients treated with high dose-rate brachytherapy for carcinoma of the cervix. Int J Radiat Oncol Biol Phys 1997;38:989–993. 17. Petereit DG, Sarkaria JN, Potter DM, et al. High-dose-rate versus low-dose-rate brachytherapy in the treatment of cervical cancer: Analysis of tumor recurrence—the University of
High-dose-rate brachytherapy for cervical cancer
18.
19.
20. 21. 22. 23.
24.
25.
26.
Wisconsin experience. Int J Radiat Oncol Biol Phys 1999;45: 1267–1274. Petereit DG, Pearcey R. Literature analysis of high dose rate brachytherapy fractionation schedules in the treatment of cervical cancer: Is there an optimal fractionation schedule? Int J Radiat Oncol Biol Phys 1999;43:359–366. Nag S, Erickson B, Thomadsen B, et al., for the American Brachytherapy Society. The American Brachytherapy Society recommendations for high-dose-rate brachytherapy for carcinoma of the cervix. Int J Radiat Oncol Biol Phys 2000;48: 201–211. Tod M, Meredith W. A dosage system for use in the treatment of cancer of the uterine cervix. Br J Radiol 1938;11:809–824. Dose and volume specification for reporting intracavitary therapy in gynecology. ICRU report 38. Washington: ICRU,1985. Fowler JF. The linear-quadratic formula and progress in fractionated radiotherapy. Br J Radiol 1989;62:679–694. Withers HR, McBride WH. Biologic basis of radiation therapy. In: Perez CA, Brady LW. Principles and practice of radiation oncology. 3rd ed. Philadelphia: Lippincott-Raven; 1998. p. 79–118. Hawnaur JM, Johnson RJ, Buckley CH, et al. Staging, volume estimation, and assessment of nodal status in carcinoma of the cervix: Comparison of magnetic resonance imaging with surgical findings. Clin Radiol 1994;49:443–452. Hricak H, Quivey JM, Campos Z, et al. Carcinoma of the cervix: Predictive value of clinical and magnetic resonance (MR) imaging assessment of prognostic factors. Int J Radiat Oncol Biol Phys 1993;27:791–801. Mayr NA, Yuh WT, Zheng J, et al. Tumor size evaluated by
27.
28.
29.
30.
31.
32. 33.
● T. TOITA et al.
1353
pelvic examination compared with 3-D quantitative analysis in the prediction of outcome for cervical cancer. Int J Radiat Oncol Biol Phys 1997;39:395–404. Kodaira T, Fuwa N, Kamata M, et al. Clinical assessment by MRI for patients with stage II cervical carcinoma treated by radiation alone in multicenter analysis: Are all patients with stage II disease suitable candidates for chemoradiotherapy? Int J Radiat Oncol Biol Phys 2002;52:627–636. Ferrigno R, Novaes PERDS, Pellizzon ACA, et al. High-doserate brachytherapy in the treatment of uterine cervix cancer. Analysis of dose effectiveness and late complications. Int J Radiat Oncol Biol Phys 2001;50:1123–1135. Lee S-W, Suh CO, Chung EJ, et al. Dose optimization of fractionated external radiation and high-dose-rate intracavitary brachytherapy for FIGO stage IB uterine cervical carcinoma. Int J Radiat Oncol Biol Phys 2002;52:1338–1344. Hareyama M, Sakata K, Oouchi A, et al. High-dose-rate versus low-dose-rate intracavitary therapy for carcinoma of the uterine cervix. A randomized trial. Cancer 2002;94:117– 124. Logsdon MD, Eifel PJ. FIGO IIIB squamous cell carcinoma of the cervix: An analysis of prognostic factors emphasizing the balance between external beam and intracavitary radiation therapy. Int J Radiat Oncol Biol Phys 1999;43:763–775. Eifel PJ. Problems with the clinical staging of carcinoma of the cervix. Semin Radiat Oncol 1994;4:1–8. Mai J, Erickson B, Rownd J, et al. Comparison of four different dose specification methods for high-dose-rate intracavitary radiation for treatment of cervical cancer. Int J Radiat Oncol Biol Phys 2001;51:1131–1141.