| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Radiation Oncology |
1 From the Department of Radiology, Division of Radiation Oncology (Y.H., M.U., A. Shioda, A. Suda, M.K., T.T., S.K.), Obstetrics and Gynecology (I.N.), and Public Health (Y.S.), National Defense Medical College, 3-2 Namiki, Tokorozawa, Saitama 359-0042, Japan. From the 1999 RSNA scientific assembly. Received April 12, 2000; revision requested June 6; revision received July 21; accepted August 29. Address correspondence to M.U.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: From 1980 to 1997, 124 consecutive previously untreated patients with cervical cancer were treated with external-beam irradiation (50 Gy) and HDR brachytherapy. Clinical stages were I, 4 (3%) patients; II, 51 (41%); III, 64 (52%); and IV, 5 (4%). From 1980 to 1992, 74 patients (group A) were treated with HDR brachytherapy once weekly (about three fractions of 7 Gy each to point A [2 cm superior and 2 cm lateral to the inferior end of the intrauterine radioactive source]), while from 1992 to 1997, 50 patients (group B) were treated twice weekly (about six fractions of 4.5 Gy each to point A).
RESULTS: Overall survival rate at 5 years was 65.2% in group A and 65.3% in group B (P = .96). Local recurrence-free survival rate at 5 years was 69% (51 of 74 patients) in group A and 90% (45 of 50 patients) in group B (P < .001). The rate of grade 2 (moderate) and grade 3 (severe) complications was significantly lower in group B (6% vs 32% in group A, P < .001). At multivariate analysis, the variables significantly associated with increased local-regional recurrence rates were having stage III–IV lesions (P = .04) and with fewer than six sessions of HDR brachytherapy (P = .02).
CONCLUSION: The twice-weekly HDR regimen may improve the local control rate with fewer complications.
Index terms: Radiations, injurious effects, complications of therapeutic radiology, 854.1269 • Radiobiology, time-dose studies • Therapeutic radiology, interstitial and intracavitary, 854.1269 • Uterine neoplasms, 854.32
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
The external-beam irradiation scheme, including central shielding, was the same in both groups. The distribution of age, clinical stage (9), and histopathologic findings is presented in Table 1. Clinical stages in the two groups A and B, respectively, were the following: stage I, two (3%) and two (4%) patients; stage II, 33 (45%) and 18 (36%); stage III, 38 (51%) and 26 (52%); and stage IV, 1 (1%) and 4 (8%). The age distribution in the two groups was comparable, with most of the patients in each group within the range of 46 to 75 years. The extent of the distribution of the disease within each group was not different, nor was the distribution of the histopathologic type. The median follow-up was longer in group A (group A, 64 months; group B, 38 months), because the treatment had been finished before the start of the study of group B.
|
HDR Brachytherapy
Between August 1980 and March 1992, HDR brachytherapy was administered at a dose of 7 Gy per session once weekly (group A). Between April 1992 and March 1997, the dose administered was 4.5 Gy per session twice weekly (group B). Intracavitary brachytherapy was carried out by using a cobalt 60 HDR remote-controlled afterloading system. Brachytherapy was performed in all patients with both an intrauterine tube and Henschke-type vaginal applicators. All patients were treatable with this system. Brachytherapy technique, including the use of intrauterine tubes, vaginal applicators, and the remote-controlled afterloading system, was the same among all patients. Several gauze sponges were packed into the vagina to displace the rectum and bladder in all patients.
HDR brachytherapy was initiated after completion of 30-Gy external-beam irradiation in both groups. At the beginning of the study, 74 patients (group A) were treated once weekly with a dose of 7 Gy to point A (2 cm superior and 2 cm lateral to the inferior end of the intrauterine radioactive source) delivered in approximately 10 minutes per session. Patients with stage Ib, II, or III disease received three sessions of HDR brachytherapy to a total dose of 21 Gy. For patients with more advanced disease, four sessions of brachytherapy to a total dose of 28 Gy were administered.
Between April 1992 and March 1997, 50 patients (group B) were treated twice weekly with a dose of 4.5 Gy to point A. Patients with stage II or III disease typically received six sessions of HDR brachytherapy to a total dose of 27 Gy. Patients with stage Ib disease received five sessions of brachytherapy to a total dose of 22.5 Gy. For patients with stage IV disease and those with stage III bulky lesions, seven sessions of brachytherapy to a total dose of 31.5 Gy were administered. HDR brachytherapy and external-beam irradiation were never performed on the same day. HDR brachytherapy was started in the 4th week for patients in both groups.
Follow-up and Statistical Analysis
Regular follow-up in each group was performed every 1–2 months during the 1st year after radiation therapy, every 2–3 months during the 2nd year, and then every 4–6 months after 2 years. Both radiation oncologists (Y.H., M.U., M.K., T.T.) and gynecologists (I.N.) saw the patients separately. Complete gynecologic examination and cytologic smears were repeated at each follow-up. Follow-up ranged from 5 to 221 months, with a median of 64 months in group A; in group B, follow-up ranged from 7 to 84 months, with a median of 38 months. Censored patients (those for whom follow-up results were uncertain or unknown) were those who died of causes other than cervical cancer, who were lost to follow-up, or whose follow-up was less than 5 years. Six patients in group A and 13 patients in group B were censored within 5 years of follow-up. We included the data of all censored patients in the data collection and calculations.
The criteria for assessing the effectiveness of treatment were local-regional recurrence, overall survival, and complications. Recurrence was determined on the basis of clinical and radiographic findings. Pap smears were routinely performed at least twice a year. Chest radiography, blood biochemical analysis, abdominal sonography, bone scanning, pelvic computed tomography, and magnetic resonance imaging were performed as indicated. Serum squamous cell carcinoma antigen level was routinely assayed since 1989. Any extrapelvic malignancy found at physical examination, radiography, or biopsy was regarded as distant metastasis. The complications were graded into three levels, as reported by Landoni et al (10). They classified complications as follows: grade 1, mild symptoms not affecting the patient’s health and easily cured; grade 2, symptoms that can be resolved with long-term medical therapies; or grade 3, fatal complications or major symptoms that require surgery or invasive procedures and that affect the performance status of the patient.
The rates of local-regional recurrences and survival were estimated with the Kaplan-Meier product-limited method (11) and were compared by using the generalized Wilcoxon test. The survival time was calculated from the last day of therapy. Test of prevalence of grade 2 and grade 3 complications at a specified time of interest was performed by using the generalized Wilcoxon test or 
2 test, when appropriate. A multivariate analysis was performed to analyze the rates of both survival and local-regional recurrences and independent predictors for grade 2 and 3 complications by using a forward step-wise procedure with the Cox proportional hazards model (12). The following parameters were studied: age at diagnosis, clinical stage, histopathologic finding, brachytherapy total dose, brachytherapy dose per fraction to point A, number of fractions, local-regional recurrence, and distant metastasis. A significance level of .05 was used throughout.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Six patients were lost to follow-up less than 5 years after treatment. All were in group A. The overall crude survival rate was 62.2% for group A and 68.0% for group B. However, the overall 5-year survival rate was 65.2% (95% CI: 53.8%, 76.6%) in group A and 65.3% (95% CI: 51.4%, 79.2%) in group B (P = .96) (Fig 1, Table 2). Five-year actuarial progression-free survival rate was 62.7% (95% CI: 51.2%, 74.3%) for group A and 66.4% (95% CI: 52.8%, 79.9%) for group B (P = .63) (Table 3). The survival rates for the two groups were not significantly different.
|
|
|
In the total patient group, 46 (37.1%) of 124 patients developed recurrent disease at 5-year follow-up (Table 4). Among those with recurrence, 29 patients had local-regional recurrences. Twenty-four (32.4%) of 74 patients were in group A, and five (10.0%) of 50 were in group B. The rate of local-regional recurrences at 5 years was significantly different between group A and group B (P = .004). The local recurrence-free survival at 5 years was 69% (51 of 74 patients) in group A and 90% (45 of 50 patients) in group B (P < .01) (Fig 2). The rates of local-regional recurrences at 5 years according to stage were 18% (seven of 40 patients) for patients with stage IIb disease and 27% (17 of 63 patients) for patients with stage IIIb disease (P = .35). In stage IIIb patients, local-regional recurrences occurred in 14 (38%) of 37 patients in group A, but it occurred in only three (12%) of 26 patients in group B (P = .04). The same tendency was demonstrated in patients with stage IIb disease, but the difference between group A and group B was not statistically significant (31% [eight of 26 patients) vs 7% (one of 14 patients), P = .10) because the number of patients with stage IIb disease was small. The low local-regional recurrence rates in group B patients occurred not because follow-up was inadequate but because most local recurrences occurred within the first 2 years of follow-up.
|
|
According to the Cox proportional hazards model, stages III and IV (P < .01), histologic findings of adenocarcinoma and adenosquamous cell carcinoma (P < .01), and distant metastasis (P < .01) were associated with a lower 5-year overall survival rate. However, local-regional recurrence was not identified as a significant prognostic factor (P = .67). The significant parameters for the local-regional recurrence rate were stage III and IV disease (P = .04) and fewer than six HDR brachytherapy sessions (P = .02), whereas brachytherapy total dose to point A was not identified as a significant predictor of local-regional recurrence.
Complications
There was no treatment-related death. The types and frequencies of complications are shown in Table 5. Sixteen (22%) of 74 patients in group A had grade 2 complications, compared with two (4%) of 50 patients in group B (P < .01). Only one (2%) of 50 patients in group B developed grade 3 complications, compared with eight (11%) of 74 in group A. However, the grade 3 complication rates were not significantly different (P = .06), presumably because the number of patients with grade 3 complications was small. The rate of grade 2 and 3 complications was significantly different between group A and group B (32% vs 6%; P < .01). Grade 1 complications were not precisely evaluated because of the slightness of the signs and symptoms; in addition, they mostly depended on subjective symptoms. It should be recalled, though, that patients in group A had longer follow-up. Most, that is, 21 [95%] of 22 rectal complications and cases of ileus occurred during the 1st year, whereas four (80%) of five bladder complications manifested after the 1st year. There was a median delay of 6 months for grade 2 complications and 11 months for grade 3 complications, but 26 (96%) of 27 grade 2 and 3 complications had become evident within 2 years after the primary treatment.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
In this study, the twice-weekly HDR brachytherapy regimen was more effective for the treatment of cervical cancer than was the once-weekly regimen. The twice-weekly regimen substantially improved local control (P < .01) and reduced moderate and severe complications (P < .01). However, despite improvements in local control and a reduction of severe complications, overall survival was not significantly improved, because 27 (93%) of 29 patients who developed local-regional recurrences had also distant metastases, and most deaths were due to distant metastasis and multiorgan failure. The data should be interpreted with caution, since the median follow-up in group A was longer than that in group B.
We are not the first to compare the two HDR brachytherapy fractionation schemes. Le Pechoux et al (1) also have compared two therapeutic regimens consisting of external-beam irradiation to the pelvis (mean dose, 50 Gy) combined with HDR brachytherapy (mean dose, 20 Gy to point A) administered at 5 Gy per session twice weekly or 6 Gy once weekly. They reported that the complication rate was significantly lower in the once-weekly group, while the overall 5-year survival, distant failure, and local recurrence rates between the two groups were not different. The study has been criticized for the comparison of two similar dosages of 5 and 6 Gy per fraction and for the different use of central shielding (1,4). In that study, the patients in the twice-weekly group experienced significantly more complications than did patients in the once-weekly group because the total dose to the central pelvis was much higher.
Our study differed from previous retrospective studies in several ways. During the treatment period in each group, all consecutive patients underwent the same HDR dose and fractionation scheme. This method allowed us to avoid selection bias, although the follow-up periods differed between the two groups. In addition, the dose per fraction to point A differed between the two groups (7.0 vs 4.5 Gy per fraction), while the field sizes and radiation doses of the external-beam irradiation including central shielding and total length of treatment were almost identical. It is unclear as to whether there was any confounding effect from the fact that the once- and twice-weekly groups of patients were treated during different periods.
The differences in outcome could be attributable to the change in the dose per fraction, not necessarily the twice-weekly aspect of the schedule. HDR brachytherapy should be administered in several small fractions to keep the dose per fraction to normal tissue within acceptable limits (14,16), and prolongation of overall treatment time should be avoided because of the risk of tumor repopulation (17–19). Previous investigators (1–7,20) achieved good results with HDR brachytherapy with widely differing doses per fraction ranging from 4.0 to 9.5 Gy. To reduce repopulation, overall treatment time should be shortened either by increasing the dose per fraction or by administering more fractions per week. The former approach increases the incidence of severe complications (1–5,7,20). If the number of fractions is increased from one to two a week, the dose per fraction to point A can be reduced. However, to our knowledge, no randomized study to clarify whether twice weekly regimens are superior or the optimal dose per fraction has been performed.
Recently, several investigators (21–23) have found the combination of pelvic radiation therapy and concomitant chemotherapy to be more effective for locally advanced cervical cancer than radiation therapy alone. The inclusion of chemotherapy substantially reduces both local recurrences and distant metastasis, leading to higher overall and progression-free survival rates (21–23). Although little is known about the optimal dose and fraction of HDR brachytherapy alone, it seems necessary to modify the dose and fraction while the chemotherapeutic agents are used concurrently. However, there are still many unknown parameters to consider (1–5).
Despite the improvement in local control and the reduction of severe complications, our regimen did not improve survival. A possible reason is that local control was not identified as a significant prognostic factor, although the distant metastasis at presentation and/or subsequent development of distant metastasis was the most important factor that correlated with poor prognosis (P < .01) in our study. We found in our experience that the improvement in local control and reduction of severe complications was not associated with a demonstrable improvement in survival. As is suggested in recent articles (21–23), concurrent chemotherapy and radiation therapy might be more effective than radiation therapy alone in terms of distant metastasis and survival rate. Although, on the basis of findings of this study, the twice-weekly HDR treatment seems to be safer than the once-weekly treatment.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, Y.H., M.U.; study concepts, Y.H., M.U.; study design, M.U., I.N.; literature research, Y.H., T.T., S.K.; clinical studies, M.U., A. Shioda, A. Suda, M.K., Y.H.; data acquisition, Y.H.; data analysis/interpretation, Y.H., Y.S.; statistical analysis, Y.S., Y.H.; manuscript preparation, Y.H.; manuscript definition of intellectual content, Y.H., S.K.; manuscript editing, M.U.; manuscript revision/review, S.K.; manuscript final version approval, Y.H., M.U.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S.-W. Chen, J.-A. Liang, S.-N. Yang, and F.-J. Lin High Dose-rate Brachytherapy for Elderly Patients with Uterine Cervical Cancer Jpn. J. Clin. Oncol., May 1, 2003; 33(5): 221 - 228. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
