This article was shown as a poster presentation at the 2014 Annual Meeting of the American Society for Radiation Oncology (ASTRO).
Association between body mass index (BMI) and doses in organs at risk during postoperative vaginal cuff brachytherapy (VCB) treatment has not been evaluated. The aim of this study was to analyse the impact of BMI on the dose delivered to bladder and rectum during high-dose-rate VCB using computed tomography (CT) scans at every fraction.
A retrospective analysis of 220 planning CT sets derived from 59 patients was conducted. Every planning CT was re-segmented and re-planned under the same parameters. Rectum and bladder dose-volume histogram values (D0.1cc, D1cc, and D2cc) were extracted and evaluated. The mean values for all applications per patient were calculated and correlated with BMI, as well as other factors influencing rectal and bladder doses. Multiple regression analysis performed to model organ at risk dose-volume parameters.
According to World Health Organization (WHO), 6.8% of patients were normal, 35.6% were overweight, and 57.6% were class I obese. Median rectal doses were 133.5%, 110.9%, and 99.3% for D0.1cc, D1cc, and D2cc, respectively. The corresponding median bladder doses were 96.2%, 80.6%, and 73.3%, respectively. BMI did not show significant association with rectal doses. However, BMI did show a significant association with evaluated bladder dose metrics (D0.1cc, r=–0.366, p=0.004; D1cc, r=–0.454, p < 0.001; D2cc, r=–0.451, p < 0.001). BMI was retained in the multivariate regression models (D0.1cc, p=0.004; D1cc, p < 0.001; D2cc, p=0.001).
In this group of Mediterranean, overweight, and moderately obese patients, BMI showed association with lower bladder dose values, but not with rectal doses.
Obesity is not only a risk factor for cancer development but also a factor affecting treatment outcomes. Endometrial cancer (EC) was the first malignancy to be recognized as related to obesity. One study suggested that up to 90% of type 1 EC patients are obese [
Early stage EC treatment involves surgery (total abdominal hysterectomy and bilateral salpingo-oopherectomy with or without pelvic lymph node dissection) followed by adjuvant radiotherapy in selected cases. Randomized studies have shown that radiotherapy (RT) reduces the risk of pelvic relapse. Vaginal cuff is the most common site of relapse and the Postoperative Radiation Therapy for Endometrial Carcinoma 2 (PORTEC-2) trial demonstrated that patients with intermediate-risk EC can be treated safely with postoperative vaginal cuff brachytherapy (VCB) in the absence of whole pelvic external beam radiotherapy, thus decreasing toxicity.
Although VCB is one most commonly used adjuvant gynaecological treatments, there is a lack of studies analysing whether dosimetric factors are influenced by overweight. Our aim, therefore, was to evaluate the effect of BMI and perivaginal fat on dose in organs at risk during fractionated VCB.
retrospective analysis of 220 consecutive brachytherapy fractions derived from 59 patients who underwent postoperative VCB for gynecological cancer was conducted. Fiftysix patients underwent VCB due to endometrial cancer and three patients due to cervical cancer. Twenty-six endometrial cancer patients underwent postoperative VCB alone using six fractions; the remaining patients were treated with whole pelvis external beam radiotherapy (WPRT) followed by three or four VCB fractions (15 patients underwent each fractionation). All cervical cancer patients underwent VCB as a boost after WPRT. Characteristics of patients, tumour, and treatment are shown in
All patients underwent a CT planning scan at every application, with 2-mm thick slice at 2-mm overlapping intervals in the supine position using a Foley bladder catheter instilling dilute contrast medium (5 mL of Omnipaque350 [GE Healthcare Bio-Sciences, Madrid, Spain] into 45 mL of saline solution) in order to increase bladder visibility during segmentation and volume reproducibility during treatment.
CTs were transferred to a 3D treatment planning system (Oncentra v.4.1, Nucletron, an Elekta company, Elekta AB, Stockholm, Sweden). To improve comparisons, the same physician re-contoured and re-planned every image set under the same conditions for an iridium-192 remote afterloading unit (MicroHDR Nucletron, Nucletron, an Elekta company), regardless of the treatment administered. The entire bladder volume was delineated, and the rectum was defined from 1 cm above the cylinder tip to 1.5 cm below the last activated source dwell position. An active length of 2.5 cm was used to deliver a fraction dose of 5 Gy at 5 mm depth to the vaginal surface. Dose-volume histograms (DVHs) were generated. D0.1cc, D1cc, and D2cc for bladder and rectum were assessed from DVHs for each fraction. The angle of the vaginal cylinder applicator related to the horizontal plane parallel to the craneo-caudal patient axis was calculated, positive values indicating a tip directed to the bladder and negative values indicating a posterior displacement of the cylinder tip toward the rectum. Cylinder size and rectum and bladder volumes were also noted.
BMI was calculated and categorized using the World Health Organization (WHO) definitions. The BMI formula was weight in kilograms divided by the square of the height in metres. Categories were underweight < 18.5 kg/m2; normal 18.5-24.9 kg/mm2; overweight 25.0-29.9 kg/mm2; obese I 30.0-34.9 kg/mm2; obese II 35.0-39.9 kg/mm2; and obese III > 40.0 kg/mm2. We considered fat that was confined within the 100% isodose on the first CT planning scan as the perivaginal fat. It was segmented by thresholding on the Hounsfield units (HU; –190 to –30 HU) and the volume was recorded. Perivaginal fat analysis was performed only in the first CT planning scan because the most common practice for VCB is to perform only one CT and to translate this plan to the next fractions.
Results are shown as median (interquantile range, IQR). Dose-volume metrics were described as a percentage of the prescription dose. Patient mean dose-volume metrics, volumes, and cylinder angle were calculated from the different fractions. Chi-square tests, Kruskal-Wallis analysis, and univariate regression analysis were performed. A stepwise multiple regression analysis was used to model organs at risk (OARs) DVH parameters as a function of other variables (the significance levels for addition to and removal from the model were set at 0.05 and 0.10, respectively). Analysis of residuals was performed to examine model fit and adherence to regression assumptions. Multicollinearity was assessed using the variance inflation factor. Differences were considered statistically significant at p < 0.05 (two-sided). Analyses were performed using Stata ver. 12 (StataCorp LP, College Station, TX).
Patient and dosimetric characteristics stratified by WHO BMI classification are shown in
Multivariate regression was performed to determine the effect of variables analysed on the bladder DVH metrics (
Patients developed grade 1 or 2 acute toxicity, no higher acute toxicity levels were observed. Late toxicity was uncommon, only one patient developed G3 bladder toxicity and two patients developed late rectal toxicity, one of them G1 and the other G2.
The main finding from this study is that higher BMI values showed an association with lower bladder dose parameters but not with rectal dose. Adipose tissues are commonly classified as subcutaneous or visceral fat, the latter being associated with metabolic disease. Visceral fat located in the periprostatic area has been associated with prostate cancer aggressiveness and its volume showed direct correlation with BMI [
Obesity and overweight have been associated with incidence of cancer and overall mortality. Obesity class I patients are strongly associated (hazard ratio, 3.93) with EC [
Although obesity is described as a risk factor for adverse outcomes after treatment for many malignancies, the effect of BMI on EC treatment outcomes has been a subject of debate. While analysis of the National Institutes of Health-AARP Diet and Health Study demonstrated an increased risk of overall and disease-specific mortality among women with higher pre-diagnosed BMI [
Vaginal brachytherapy (VBT) has been one of the main adjuvant treatments for EC since the results of randomized trials that showed VBT was not inferior to pelvic irradiation in high intermediate risk patients, and its use is extended after external beam radiotherapy in high risk patients [
A limitation of our study is the lack of a complete pelvic visceral fat segmentation, which was not possible due to the limited pelvic volume scanned in some insertions. Strong correlation was observed between perivaginal fat volume, which was used as a surrogate for visceral fat volume, and BMI. However, visceral fat volume has not been validated. This measure has not been validated, abdominal adiposity at the umbilicus level being the usual measure of visceral and subcutaneous fat. To study these relationships we are planning visceral fat segmentation at different abdominal and pelvic levels. The fact that there were no severe or extremely obese patients (BMI obese class II and III) in this report is noteworthy, and reflects the BMI distribution among the general population in Mediterranean countries compared with the population in the United States. In addition, the population attributable fraction is greater for the US population (56.8%) than for the European population (45.2%) [
In conclusion, the results reported in this study demonstrate a significant inverse correlation between BMI and bladder dose deposition during VCB. The impact of these findings on late toxicity needs to be evaluated in clinical practice.
Conflict of interest relevant to this article was not reported.
Scatter plot showing the relationship between body mass index (BMI) and perivaginal fat volume.
Scatter plot showing the relationship between perivaginal fat and cylinder angle.
Relationship between fat distribution, bladder, and 100% isodose after a fat pixel intensity transformation. Fat pixels (Hounsfield units, –190 to –30) are shown as white dots. The shaded circle around the vaginal cylinder depicts the 100% isodose. Computed tomography images depict patients with a similar perivaginal fat volume but with different bladder doses (D1cc values as the percent of the prescribed dose). (A) Perivaginal fat volume 4.3 cc, bladder D1cc=74.58%. (B) Perivaginal fat volume 4.4 cc, bladder D1cc=98.3%.
Patient, tumor, and treatment characteristics
Variable | Endometrial cancer (n=56) | Cervical cancer (n=3) |
---|---|---|
Age (yr) | 65.1±10.4 | 52.9±13.1 |
Stage | ||
Ia | 1 | 0 |
Ib | 19 | 2 |
Ic | 21 | 0 |
II | 5 | 0 |
III | 4 | 1 |
IV | 5 | 0 |
Histology | ||
Endometroid | 42 | 0 |
Other | 14 | 0 |
Squamous | 3 | |
Grade | ||
1 | 31 | 1 |
2 | 10 | 1 |
3 | 14 | 1 |
Treatment | ||
VCB alone | 26 | 0 |
WPRT+VCB | 30 | 3 |
Values are presented as mean±standard deviation or number. VCB, vaginal cuff brachytherapy; WPRT, whole pelvis radiotherapy.
Patients' characteristics and rectum and bladder dosimetric parameters according to the World Health Organization body mass index classification
Variable | All (n=59) | Normal weight (n=4) | Overweight (n=21) | Obese I (n=34) | p-value |
---|---|---|---|---|---|
Age (yr) | 64.3 (18.4) | 54 (8.3) | 65.3 (7.9) | 67.9 (20.6) | 0.043 |
Cylinder diameter (%) | |||||
3.0 cm | 16 | 75 | 33.3 | 17.6 | 0.037 |
3.5 cm | 43 | 25 | 66.7 | 82.4 | |
Cylinder angle (°) | –3.6 (9.2) | –2.8 (8.8) | –5.8 (8.1) | –2.3 (8.4) | 0.285 |
Rectum volume (cc) | 47.1 (25) | 57 (32.1) | 55 (28.4) | 42 (20.5) | 0.050 |
Bladder volume (cc) | 79.6 (12.4) | 92 (23.7) | 80 (11.3) | 79 (12.2) | 0.115 |
Rectal DVH metrics (%) | |||||
D0.1cc | 133.5 (14.2) | 140.8 (14.9) | 134.1 (10.2) | 133 (20.3) | 0.476 |
D1cc | 110.9 (14.2) | 116.6 (13.1) | 109.8 (12.1) | 110.9 (16.3) | 0.783 |
D2cc | 99.3 (14.3) | 104.2 (12.6) | 98.04 (12.3) | 99.4 (16.4) | 0.889 |
Bladder DVH metrics (%) | |||||
D0.1cc | 96.2 (18.4) | 100.4 (12.4) | 102.9 (18.4) | 93.3 (13.7) | 0.019 |
D1cc | 80.6 (13.9) | 85.1 (11.1) | 86.9 (11.8) | 76.1 (11.7) | 0.001 |
D2cc | 73.3 (12.7) | 78.7 (10.2) | 78.9 (7.8) | 70.7 (12.1) | 0.002 |
Perivaginal fat (cc) | 8 (8.3) | 4.1 (3) | 5.1 (5.5) | 9.4 (3.8) | < 0.001 |
Values are presented as median (interquartile range). Rectum and bladder dose-volume histogram (DVH) values as the percent dose prescribed to 5 mm.
Multiple linear regression predictors for bladder dose-volume metrics
Variable | Bladder |
||
---|---|---|---|
D0.1cc | D1cc | D2cc | |
Body mass index | –0.924 |
–0.725 |
–0.658 |
Rectum volume | - | 0.132 |
0.145 |
Age | - | - | –0.252 |
3.5-cm cylinder | - | - | 5.310 |
Constant | 126.6 |
96.80 |
99.69 |
N | 59 | 59 | 59 |
r2 | 0.134 | 0.262 | 0.363 |
Adjusted r2 | 0.119 | 0.235 | 0.315 |
p-value | 0.004 | < 0.001 | < 0.001 |
Standard errors in parentheses.
p < 0.05,
p < 0.01,
p < 0.001.