Toxicities and Recurrences after Co-60 High-Dose-Rate Brachytherapy for Cervical Cancer in a Tertiary Government Hospital in the Philippines

Introduction

The current standard of care in the management of most cases of cervical cancer is external beam radiotherapy (EBRT) concurrent with chemotherapy and intracavitary brachytherapy.^1,2 Brachytherapy allows for tumor dose escalation while limiting the dose received by organs-at-risk (OARs), such as the bladder and the rectum.³ When added to EBRT, brachytherapy improves primary complete remission, survival, and recurrence rates for patients with cervical cancer.⁴

High-dose-rate (HDR) brachytherapy has superseded low-dose-rate (LDR) brachytherapy over the last decade since the former permits treatment in the outpatient setting, limits radiation exposure to the staff, and allows for optimization for a more conformal dose distribution, which is appropriate in the era of image guidance, among many other advantages.⁵ However, the optimal fractionation regimen for HDR brachytherapy, which employs the use of multiple hypofractionated radiotherapy fractions, remains unsettled. From a radiobiological perspective, more hypofractionated regimens are expected to produce more late effects. In 1991, Orton et al concluded that fractionation in HDR brachytherapy significantly influenced toxicities. Complication rates were significantly higher in regimens that utilized >7 Gy per fraction as compared with those that used otherwise.⁶ Some of the regimens listed by the American Brachytherapy Society (ABS) guidelines are 7 Gy × 4, 6 Gy × 5, 5 Gy × 6, and 5.5 Gy × 5 fractions,⁷ although a wide practice variation exists internationally.⁸

Toxicities arising from chemoradiation (CRT), including brachytherapy, are primarily gastrointestinal (GI) and genitourinary (GU) by virtue of proximity of these organs to the cervix. To lessen the risk of GI and GU toxicities, the ABS recommends a cumulative dose limit of ≤75 Gy to the rectum and the sigmoid, and ≤90 Gy to the bladder in the delivery of a 2-Gy-radiobiologically equivalent (EQD2) cumulative radiotherapeutic dose of 80 to 90 Gy to the tumor.⁷ In a pooled analysis of clinical outcomes for HDR brachytherapy for cervical cancer by the American Brachytherapy Task Group, the range of toxicities after CRT was 1 to 11% for late GI and 2 to 20% for late GU toxicities. For the late gynecologic toxicity, only 1 of the 16 prospective trials reported 9% for CRT effects.⁹

Iridium 192 (Ir-192) is among the most commonly used radioisotopes in HDR brachytherapy for cervical cancer due to its small size and high specific activity. Most trials on the safety and efficacy of HDR brachytherapy in the treatment of cervical cancer utilized Ir-192 as the brachytherapy radioisotope. One of the disadvantages in using Ir-192, however, is its relatively short half-life, which warrants replacement every 74.3 days.³ The recent availability of cobalt 60 (Co-60) sources with the same geometric and dosimetric properties as the more traditional Ir-192 sources makes the former an option in HDR brachytherapy and is more appropriate in low-resource brachytherapy facilities where frequent and timely procurement of radioisotopes could be a problem. Limited studies on the use of Co-60 sources in HDR brachytherapy for cervical cancer are, however, available in the literature.¹⁰

Objective

The study aimed to review and report the incidences of GI and GU toxicities and recurrences in patients with cervical cancer who were treated with intracavitary HDR brachytherapy using a cobalt-60 remote afterloading system between September 2016 and September 2018 in a tertiary hospital in the Philippines.

Specifically, the study aimed to:

Determine the incidence of GI and GU toxicities.

Determine the incidence of recurrences after brachyt-herapy.

Determine the overall treatment time (OTT) of patients with cervical cancer in the brachytherapy unit of our institution, defined as days elapsed from first day of EBRT to last fraction of brachytherapy.

Determine the institutional treatment gap between the last day of EBRT and first fraction of brachytherapy, if any.

Analyze the relationship between patient-related (i.e., smoking history, body mass index) and treatment-related (i.e., OTT, EBRT machine use, use of midline shielding, brachytherapy fractionation scheme used) factors, and occurrence of recurrences and toxicities.

Significance

This study was undertaken to provide baseline institutional data on the incidence of GI and GU toxicities and recurrences in patients with cervical cancer treated with intracavitary HDR brachytherapy using a cobalt-60 remote afterloading system. The results can establish the safety of using cobalt-60 instead of the more popularly used iridium-192. This is important in low-resource, high-volume radiotherapy facilities since cobalt-60 has a longer half-life and will thus require less frequent procurement of sources. The results can also guide future institutional policies to streamline brachytherapy services and to improve the oncologic outcomes of patients with cervical cancer.

Methods, Study Design, and Sampling

A case–control study design was employed to clarify any association between the treatment (HDR brachytherapy) and the outcomes, as well as other potential risk factors (e.g., age, duration of treatment). Prior to its implementation, the study secured approval from the institutional research ethics board.

All women with biopsy-confirmed cervical cancer who underwent HDR brachytherapy from September 2016 until September 2018 (period of 2 years) in our institution were included in the study. The inclusion and exclusion criteria are listed in Table 1. Selecting all patients within the period of observation was done to ensure the representativeness of the sample and reduce the possibility of selection bias, since there were no local data available at that time to ascertain the association between the exposure and outcomes. A chart review was conducted to collect data on clinical parameters listed in Table 2. Toxicities were graded using Common Terminology Criteria for Adverse Events (CTCAE)¹¹ version 5.0 (Supplementary Material 1, available in the online version).

After extracting the data from the patient charts, information was manually entered into an electronic spreadsheet file, and subsequent data processing and analysis were performed using the software Stata 13. Descriptive statistics was used for continuous variables, such as age and body mass index,¹² and median and interquartile ranges were reported to describe follow-up duration, dose of midline shielding, and duration of treatment. Reporting frequencies and percentages was done for categorical data variables, such as disease stage, smoking status, EBRT machine used, brachytherapy fractionation scheme, and planning technique.

Proportions per categories of the qualitative variables, such as presence of GU and/or GI toxicities, local or distant recurrences, and deaths, were computed. Point and interval estimates of these proportions were also determined. To determine exposure-specific proportions, participants who previously had clinical and treatment risk factors were further divided between those who developed the toxicities and otherwise. Odds ratios were computed for association of HDR brachytherapy exposure and outcomes of interest. A series of multiple logistic regression modeling procedures were done mainly with the adjustment for probable confounders to be conducted using select clinical variables based on the literature through the backward elimination process. An arbitrary cut-off of a change in the p-value less than 0.25 was used to screen for probable confounders. The level of significance for all sets of analysis was set at p < 0.05 using two-tailed comparisons. Significance levels were adjusted for multiple comparisons performed, when necessary, using the Fisher–Hayter post hoc method.

Results

A total of 174 patients who fit the inclusion criteria of the study received HDR intracavitary brachytherapy as part of their management for cervical cancer. Eleven patient charts were missing and not retrievable. The remaining 163 patient charts were reviewed for baseline patient-related (Table 3) and treatment-related (Table 4) characteristics. Although this represented only 82% of the computed sample size, which could be attributed to missing records and treatment variations, post hoc analyses revealed that the power accrued remained at 80%.

The median age at diagnosis was 47 years (range: 20–73 years). Majority of the patients were nonsmokers (n = 133, 81.6%). Most patients had normal baseline body mass index (n = 73, 45.34%). Majority of the patients had squamous cell carcinoma histology (n = 122, 74.85%). The average pretreatment size of the cervical mass was 6 ± 1.66 cm. The most common stage was IIB (n = 93, 57.06%), followed by IIIB (n = 50, 30.67%).

Almost two-thirds of the patients received external beam irradiation via cobalt teletherapy (n = 107, 66.05%), with the rest treated via a linear accelerator (n = 55, 33.95%) (Table 4). The median central dose was 4,600 cGy (range: 4,000–5,400 cGy). For brachytherapy, most patients underwent a 7 Gy × 4 fractionation scheme (n = 109, 82.58%), while the rest underwent an 8 Gy × 3 fractionation scheme (n = 23, 17.42%). The median number of days elapsed during EBRT was 49 days (range: 21–160), while the median number of days between EBRT and brachytherapy was 57 days (range: 0–447). The median number of days elapsed to complete the brachytherapy course was 18 days (range: 10–59). The median OTT was 129 days (range: 63–541).

Patients who were unable to complete the institutionally mandated 90-day follow-up were excluded from analysis of outcomes (n = 31). Among the remaining 132 patients (Table 5), median follow-up was 389 days (range: 95–900). A total of 26 patients (19.70%) experienced GI toxicities, while 2 patients (1.52%) experienced GU toxicities (Table 5). Recurrence occurred in 42 (31.82%) of these patients, with 14 (10.61%) recurring locoregionally, 17 (12.88%) recurring distantly, and 11 (8.3%) occurring both locoregionally and distantly. Nine patients (6.52%) had tumor persistence despite treatment.

Sites of locoregional and distant failure are shown in Table 6. The uterocervical region was the most common site of locoregional recurrence (n = 16, 59.26%), followed by pelvic nodal (n = 6, 22.22%) and vulvovaginal recurrence (n = 2, 7.41%). The para-aortic nodal region was the most common site of distant failure (n = 7, 26.92%). When combined with those with simultaneous failure at other organ sites, the total incidence of para-aortic involvement was 42.31% (n = 11) of all cases of distant recurrence. The lungs were the second most commonly involved distant site (n = 9, 34.62%).

Based on unadjusted logistic regression, there were no notable associations between the clinically important variables and the occurrence of GI and/or GU toxicities. On adjusted measures, a high central dose was significantly associated with the presence of toxicities (p = 0.03) (Table 7).

There were also no notable associations between the clinically important variables and treatment failure, although a longer treatment time seemed to be weakly associated (p = 0.06) with higher treatment failure (Table 8).

Discussion

Conclusion and Recommendations

We report institutional outcomes on toxicities and recurrences after HDR brachytherapy in patients with cervical cancer. Our study suggests that central tumor dose, as a result of midline shielding, is significantly associated with toxicities, while OTT is weakly associated with recurrence. This study also demonstrates the safety of using cobalt-60 sources for HDR brachytherapy.

To our knowledge, this is the first study to formally report on local brachytherapy outcomes in the Philippines. Efforts should be made to manage outcome biases, such as prolonged OTT, in future researches and policies in low-resource, high-volume brachytherapy centers in the country.