CT-Planned, High-Dose-Rate Brachytherapy Effective in Cervical Cancer

Akila N. Viswanathan, MD, executive vice chair, Johns Hopkins Radiation Oncology and Molecular Radiation Sciences

Akila N. Viswanathan, MD

The use of computed-tomography (CT)-planned high-dose-rate (HDR) intracavitary brachytherapy (BT) demonstrated excellent local control (LC) and survival for patients with stage I/II cervical carcinoma, according to the results of a recent study published in Gynecologic Oncology.

With a median follow-up of 30 months, no patients experienced local-only recurrence.

“We found that, of 9 patient characteristics and various RT prognostic factors tested in a univariate analysis, tumor size and total external beam radiation therapy (EBRT) plus BT dose were initially the most significant prognostic factors for predicting any recurrence (AR),” Akila N. Viswanathan, MD, executive vice chair, Johns Hopkins Radiation Oncology and Molecular Radiation Sciences, and her colleagues wrote in the study.

However, a bivariate model showed that total EBRT plus BT dose were no longer significant when adjusted for tumor size, the authors wrote. The authors of the study found large tumor size to be a prognostic factor for increased risk of recurrence outside the radiation field and worse progression-free survival (PFS) and overall survival (OS). Patients with tumors larger than 4 cm were 3 times more likely to develop regional or distant recurrence, compared with patients who had tumors ≤4 cm.

Moreover, the researchers also found that a volume-optimized plan treated a smaller area than a standard, Point A plan for these patients.

A total of 150 patients were treated for stage I/II cervical cancer using HDR CT-planned BT between April 2004 and October 2014 at Brigham and Women’s Hospital. One-hundred twenty-eight women met eligibility criteria for the current study, as they had stage I/II disease, were treated with CT-based BT, and were not treated with interstitial BT.

Patients were staged according to the International Federation of Gynecology and Obstetrics (FIGO) guidelines. Clinical examination reports provided information about tumor size, which was also subsequently confirmed using CT or MR. Small and large tumor size were defined as ≤4 cm and >4 cm, respectively.

For the BT procedure, patients were placed under general anesthesia and immobilized for applicator insertion, CT scanning, and treatment. A physician assessed the location, size, and firmness of residual disease at the time of brachytherapy. All contouring was performed by the treating physician, who also carried out the final treatment plan evaluation to maximize dose coverage and minimize dose to the organs at risk (OAR) (bladder, rectum, and sigmoid). Contouring per previously published guidelines included a clinical tumor volume (CTV) that encompassed the palpable disease, the contoured cervix to at least a height of approximately 3 cm along the tandem and laterally to the edge of the visualized cervix and parametrial regions. Point A was recorded, but not used for dose specification.

Pelvic recurrence sites were classified as either “true pelvis” (cervix, uterine corpus, vagina, and parametria) or “pelvic nodes” (external and internal iliac, obturator, and perivesical nodes). Regional recurrence (RR) was defined as failure in the para-aortic and inguinal nodes. Distant recurrence (DR) was defined as metastasis outside of the loco-regional sites as evidenced through PET-CT or MRI.

The primary endpoints included PFS—defined as no evidence of disease progression after tumor biopsy, and subdivided into PFS true pelvis (local control) and PFS pelvic nodes (local and regional control, or pelvic control [PC])—and OS, which was defined as the period from the date of biopsy until the date of death from any cause.

Thirty-eight patients (30%) had stage I disease, while 90 women (70%) had stage II cervical cancer. The overall median tumor size at the time of diagnosis was 3.8 cm (range, 0-6 cm), with stage I tumors yielding a smaller median size (2.9 cm; range, 0-7.6 cm) than stage II tumors (4.0 cm; range, 0-10 cm). Lymph node involvement was recorded in 40 patients (31%), 32 of whom (80%) had stage II disease.

Chemotherapy was administered to 114 patients (89%) and mainly included a platinum-based regimen of concurrent cisplatin or carboplatin. At the earliest, BT began in the fourth week of EBRT and was delivered in a dedicated BT suite with CT imaging done immediately after insertion. A total of 630 fractions were analyzed, with each patient receiving a median of 5 fractions (range, 3-5). The median per-fraction prescription dose was 5.5 Gy (range, 5.0-7.0), and the median average of Point A (fraction 1) was 5.0 Gy (range, 2.9-9.0). The cumulative EBRT plus BT dose to 90% of the tumor volume (D90) in EQD2 was 83.7 (range, 66.7-114.7), and the median cumulative doses (Gy3) to 2cm³ (D2cc) were 75.1 Gy for bladder, 63.9 Gy for rectum, and 59.4 Gy for sigmoid.

The median follow-up time was 30 months (range, 3.8-129.7). Eighteen of 128 evaluable patients (14%) had documented disease recurrence or progression, 15 of whom had stage II disease at diagnosis. No patients relapsed in the true pelvis (LR) only. Six patients (5%) demonstrated LR plus PR, RR, and/or DR. Of these, 3 experienced concomitant LR and PR (pelvic sidewall, pelvic node, and pelvic mass), 1 had LR plus PR plus DR, 1 had LR plus PR plus RR plus DR, and 1 patient with a 6-cm primary lesion at diagnosis had residual disease at the end of the treatment.

One patient (1%) had PR plus RR, and 3 women (2%) had RR plus DR. Eight patients (6%) had DR only, 5 of whom had tumors larger than 4 cm.

Overall, of the 18 patients who experienced a relapse, 14 (78%) had a DR. The overall 2-year rate of LC was 96%, PFS was 88%, and OS was 88%.

When these results were broken down by tumor size (≤4 cm vs. >4 cm), the 2-year LC rates were 98% versus 93% (P = NS), the PFS rates were 95% versus 76% (P = .01), and the OS rates were 91% versus 84% (P = NS), respectively.

There were 28 deaths overall in the study (22%). Causes of death included recurrence or metastasis (n = 14; 11%), respiratory failure (n = 1; 1%), metastasis of a secondary cancer (n = 6; 5%), sepsis (n = 1; 1%), and cardiac arrest (n = 1; 1%). The causes of 5 patients’ deaths (4%) were unknown.

A univariate analysis showed no differences in median age at diagnosis, median length of follow-up, median gravid status, stage, lymphovascular invasion, lymph node positivity (biopsy or PET), tumor grade, use of chemotherapy, histology, race, and smoking history between patients with AR and those with no recurrence (NR; all P > 0.1). Large tumor size was significantly associated with AR (unadjusted HR = 3.3; 95% CI, 1.28-9.47). There were no significant prognostic factors associated with OS (all P > 0.7).

The authors also looked at outcomes at 2 years by receipt of chemotherapy. For patients treated with chemotherapy (n = 114) or without chemotherapy (n = 14), LC rates were 97% versus 91%, PFS rates were 88% versus 85%, and OS rates were 93% versus 87%, respectively (all P = NS).

Among the 14 women treated without chemotherapy, 3 (21%) experienced a relapse, 2 (14%) of whom had combined LR (cervix and vagina) and PR (pelvic nodes and sidewall), and 1 (7%) who had DR to the spine.

Among the rest of the patients who were treated with chemoradiation, no differences were identified between those with NR (n = 99; 87%) and those with AR (n = 15; 13%) in either baseline characteristics or clinical outcomes (P > 0.1), with the one exception of median age at diagnosis. For those with NR, the median age at diagnosis was 53 years, and for those with AR, the median age was 47 years (P = .03).

No significant differences were found in terms of median total brachytherapy D90 dose (EQD2), BT dose (EQD2), V100 target coverage, cumulative D90, or cumulative EBRT and BT dose (EQD2) (all P >0.1), between patients who had NR and those who had AR.

The median fraction-1 Point A dose was significantly lower for those with NR than for those with AR (5.0 ± 1.0 vs 5.5 ± 1.0 Gy, respectively; P = .001).

Although a univariate analysis of radiation therapy doses showed that tumor volume receiving 100% of prescribed dose (V100) target coverage (P = .049), point A dose (P = .016; HR = 3.71; 95% CI, 1.27-13.45), and total EBRT plus BT dose (P = .025) were predictors of AR, the effect of V100 target coverage on AR lost its significance (P = .07) after the authors adjusted for tumor size. Thus, tumor size remained as the only predictor for AR (P = .016; HR = 3.71; 95% CI, 1.27-13.45).

Moreover, after tumor size was accounted for, the association between EBRT plus BT (EQD2) dose and AR became less significant (P = .07), and tumor size was, again, the only predictor for AR (P = .038; HR = 2.80; 95% CI, 1.06-8.16). Also, Point A dose was a significant predictor for AR (P = .01; HR = 1.005; 95% CI, 1.001-1.009) after adjustment for tumor size.

Genitourinary toxicities (GU) were observed in this study, including bladder spasm, incontinence, stenosis, urinary frequency and urgency, and dysuria. Gastrointestinal (GI) toxicities included diarrhea without prior colostomy, hemorrhoid pain, constipation, nausea, and ulcer.

Twenty-one patients (16%) experienced grade ≥2 acute toxicity after receiving BT. Of those women, 6 (5%) had GI toxicities, while 12 (9%) had GU adverse events, and 3 patients (2%) experienced vaginal toxicities. Two patients (2%) also experienced grade ≥2 late toxicity, including 1 (1%) with GI and 1 (1%) with GU adverse events. There were no acute or late grade ≥3 toxicities.

Cho LP, Manuel M, Catalano P, et al. Outcomes with volume-based dose specification in CT-planned high-dose-rate brachytherapy for stage I-II cervical carcinoma: a 10-year institutional experience. Gynecol Oncol. 2016;143(3):545-551.