Comparative Survival Assessment of Two-Dimension (2D) vs Three-Dimension (3D) Brachytherapy Treatment in Cervical Cancer: A Retrospective Case Control Study

This retrospective cohort study included cervical cancer patients stage IB2-IVA in figo staging 2008 or IB3, IIA2-IVA in figo staging 2018. These patients were all treated with concurrent chemoradiation to pelvis (+/- paraaortic lymph nodes) followed by brachytherapy(low dose rate (LDR) and high dose rate(HDR)) between March 1990 and August 2018. Every cervical patient had to complete radiation therapy and chemotherapy sessions. All patients were Biopsy-proven to have cervical cancer, including: squamous cell carcinoma, adenosquamous cell carcinoma or adenocarcinoma. All patients were treated with conventional EBRT technique.

Patients were divided into two groups based on imaging type during brachytherapy, 2D or 3D technique brachytherapy data. Brachytherapy treatment was done with plain film in the 2D technique and CT scan in the 3D technique. The brachytherapy applicators were all without interstitial needles. Patient characteristics, cancer characteristics, cancer staging, treatments, outcomes and complications of treatment were collected retrospectively. This study was approved by the Phramongkutklao Hospital Ethics Committee.

Locally advanced cervical cancer stage IB2-IV in figo staging 2008 or IB3, IIA2 -IVA in figo staging 2018 was diagnosed based on physical examination and imaging data. Whole pelvic external beam radiation therapy was used concurrent with chemotherapy. Four field techniques with 45-50 Gray in 25 fractions and 10 MV photons were prescribed. After external beam radiation, high dose rate brachytherapy was performed once weekly. Radioactive use in low dose rate brachytherapy was cesium-137 during 1990-2006 and high dose rate Iridium-192 during 2007-2018.Almost all patients received platinum based chemotherapy such as Cisplatin and Carboplatin. The minority received Fluorouracil(FU) and Mitomycin. Radiation and chemotherapy were all completed followed by standard treatments. All patients had to complete the total treatment time within 8 weeks.

The whole pelvic radiation therapy was all used with four field conventional techniques. 10 MV photon was prescribed. Radiation dose was between 45-50 Gy. Field design for pelvic radiation therapy was AP-PA and lateral fields.

AP-PA field borders were defined as

Superior at L4-L5 vertebral interspace

Inferior at 2 cm below the obturator foramen or 3 cm inferior to distal disease, whichever is lower.

lateral at 1.5-2 cm alteral to the pelvic brim.

Lateral field borders were defined as

Superior at same as AP-PA field

Inferior at same as AP-PA field

Anterior at anterior to pubic symphysis

Posterior at 0.5 cm posterior to the anterior border of the S2/3 vertebral junction, may include the entire sacrum to cover the disease extent

If paraaortic LN field was added, the field would be up to T12-L1 interphase.

Almost all patients received platinum based chemotherapy such as 40mg/m2 of Cisplatin weekly (N=92, 90%) or 100 mg/m2 of Carboplatin Auc 2 weekly (N=7, 6.86%). The minority received 1000 mg/m2/ Fluorouracil(FU) a day on days 1-4 and 29-32 and 10 mg/m2 of Mitomycin a day on day 1 and day 29 (N=3, 2.94%).

High dose rate brachytherapy using tandem and ovoids application without interstitial needles was used in both 2D and 3D techniques.

In the 2D technique, plain film of the lower abdomen was done after application insertion. According to ICRU 38⁵, dose prescription was set 100% at point A. Dose fractionation such as 6.5Gy x 4 times,7.5 Gy x 3 times and 8.3 x 2 times was prescribed. Normal organ dose constraint was described as bladder and rectal point from plain film.

In the 3D technique, a CT scan without contrast media was done followed by applicators insertion. Pre-brachytherapy MRI scan was done in every case since 2017 to evaluate the disease extension. According to GEC-ESTRO guidelines ⁴, target volume, High risk clinical target volume (HR-CTV) and intermediate target volume (IR-CTV) were used in dose prescription instead of point A. HR-CTV was covered the gross tumor at the time of brachytherapy determined by both pre-brachytherapy MRI image and at the exam, the entire cervix and regions of indeterminate T2- weighted MRI signal and the gray zone. IR-CTV was determined of HR-CTV with an asymmetric expansion, not extending into OARs and including sites of initial involvement.

Dose distribution depended on the standard loading system of tandem and ovoids. Organs at risk (OAR) such as rectal and bladder radiation dose was measured using D2cc (minimal dose to the most irradiated 2 cc of OAR). A target dose volume histogram was used for dose to 90% and 95% of the target volume. The total dose was calculated including external beam radiation dose and brachytherapy dose using tumor equivalent dose (EQD210) and OAR equivalent dose (EQD2 3). Dose fractionation such as 6.5Gy x 4 times,7.5 Gy x 3 times and 8.3 x 2 times was prescribed.

Low dose rate brachytherapy delivers radiation at a dose of LDR brachytherapy at 0.55–0.65 Gy/h. Point A doses were 75 Gy in two fractions for LDR. Patients who received the intrauterine tandem and ovoids were hospitalized after placement of the applicator for 24–72 hours to allow radiation therapy treatment. Cs-137 low dose rate radioactive source was applied for intracavitary brachytherapy treatment has been determined using the Manchester System from 1990 to 2006.

High dose rate brachytherapy delivered a dose >12 Gy/hour and was in the outpatient setting. Short duration of brachytherapy treatment time, patient convenience were the advantages over low dose rate brachytherapy. ¹⁹² Ir HDR sources were used in high dose rate brachytherapy from 2007 to 2019.

At our institute, brachytherapy treatment planning system (Varian Medical Systems, Inc., Palo Alto, CA) has been used. The optimization process was equal times, geometric optimization (GO) and volume optimization (VO). In 3D brachytherapy, contouring of target volume and normal organs at risk were input. The GO was performed followed by isodose reshaping by manual adjustments of isodose lines. Adjusting the isodose lines was done to optimize between target volume dose and critical structures dose.

The primary endpoint of this retrospective cohort was overall survival, defined as from the first date of treatment to death from any cause. The secondary endpoints were local control (defined as absence of the local disease evidenced from physical examination, Imaging and biopsy (if performed), progression free survival (defined as no evidence of disease at any site of the body from physical examination, imaging or biopsy (If performed) , acute toxicities at least grade 2(defined toxicities during radiation therapy session to 3 months after radiation therapy and late toxicities at least grade 2 (defined toxicities more than 3 months after radiation therapy).Minimal follow-up required at least 2 years.

Patient and disease characteristics were described with percentage, mean and standard deviation with differences using a chi-square test. Overall survival, local control and progression free survival were all calculated from the date of treatment to the date of events or loss follow up. Kaplan-Meier survival method and log rank test were used to calculate the time to event. Acute and late toxicities were measured using chi-square tests. P-values of less than 0.05 were defined significantly. Analysis using SPSS version 21.0 was performed.

All 102 cervical cancer patients were retrospectively registered in this study. All patients completed the standard treatment (concurrent chemoradiation and brachytherapy). Of the total patients, the 2D technique was used on 52 patients and the 3D technique was used on the other 50 patients. All patients used high dose rate brachytherapy without interstitial needles insertion (Table 1). Patient characteristics (Figo staging, pathology, underlying disease and age at diagnosis) were described and compared between 2D and 3D techniques. All patient characteristics had no statistical significant difference between 2D and 3D technique (P value < 0.05) (Table 1).

The follow up time in the 2D technique was a range of 6-30 years and the 3D was 3-5 years. Loss follow up was measured in 13.7% of all populations (14 of 102 patients).

96 of the 102 patients were alive at the time of analysis. The overall survival rate was 96% at the 3-year follow-up. 2 patients (1.96%) died from cervical cancer while 4 patients (3.9%) died from other conditions such as UTI septicemia, or other underlying disease conditions.

The overall survival rate of the 2D technique was 95.65 % (83.58 - 98.90) and 3D technique was 91.79 (79.57 - 96.84) without statistical significant difference (Figure 1).

Figure 1.Kaplan-Meier survival estimates for overall survival rate (OS) between 2D technique conventional brachytherapy (blue line) and 3D technique Image guided brachytherapy using CT simulation ( red line) showed no statistical significant difference.

94 of the 102 patients achieved the 3-year local control. Local recurrence was observed in 8 patients. Overall local control was 92.16 % at the 3-year mark. Between the 2D and the 3D technique, there was no difference of 3 year local control rate (2D - 88.59% (74.53 - 95.13) and 3D - 93.33% (80.67 - 97.81) (P value = 0.571). (Figure 2)

Figure 2.Kaplan-Meier survival estimates for local control rate (LC) between 2D technique conventional brachytherapy (blue line) and 3D technique Image guided brachytherapy using CT simulation ( red line) showed no statistical significant difference.

Overall progression free survival was 85.29% (87 of the 102 patients). The 3-year progression free survival rate was 86.13% in 2D and 27.14% in 3D ( P value =0.006). In this analysis, 2D technique seemed to have better progression free survival than 3D technique (Figure 3).

Figure 3.Kaplan-Meier survival estimates for three-year progression free survival (PFS) between 2D technique conventional brachytherapy (blue line) and 3D technique Image guided brachytherapy using CT simulation (red line) show better outcome in 2D technique compared to 3D technique.

The overall survival was 96 % at the 3-year follow-up comparable with other data from previous studies which range from 68% to 94 %. Between the 2D and the 3D techniques, there seemed to be no overall difference in the survival rate (range from 79.6 to 98.9 %). Completion of concurrent chemoradiation with acceptable overall treatment time in all patients in this study might relate to a good outcome of overall survival. Death from non-cancer causes were twice as much as cancer causes as referred to in the standard treatment of cervical cancer.

Overall local control was 92.16 % for 3 years. Between the 2D and the 3D techniques, there was no difference of the 3-year local control rate (2D - 88.59% (74.53 - 95.13) and 3D - 93.33% (80.67 - 97.81). Other studies showed a 3-year local control range from 78.5 - 97.9 % which were comparable to our study. Compared to other interstitial needles using brachytherapy studies, no difference in outcome was seen in local control rate. Ongoing results from prospective studies using MRI and interstitial needles brachytherapy are still being waited for.

Surprisingly, progression free survival (PFS) of the 2D technique from this retrospective study was statistically better than the 3D technique (86.13% vs 27.4 %, P value= 0.006). Because of low event rate and low sample sizes from this retrospective study, the PFS might not represent the exact survival. Using the 2D brachytherapy technique also showed excellent results in progression free survival.

The 3D brachytherapy technique showed significant reduction of acute and late grade-3 toxicities compare to 2D technique (acute gastrointestinal toxicities, 2D 1.92 % vs 3D 0%), (Late gastrointestinal toxicities 2D 7.7% vs 3D 0%, genitourinary toxicities 2D - 5.77% vs 3D - 0%). Moreover, reduction of acute grade 1-2 gastrointestinal toxicities and late genitourinary toxicities were statistically observed in 3D technique compared to 2D (acute 4% in 3D vs 23.08 % in 2D, late 16% in 3D vs 50 % in 2D).

Compared to the study from the Netherlands ⁷, which compared between 2D and 3D techniques. The difference between the studies was that some patients of the Netherland study used interstitial brachytherapy and some patients underwent MRI rather than CT scan in image guided radiation therapy. Although, the results were almost similar to our studies in aspects of overall survival, local control and toxicities.

RetroEMBRACE study ⁸ was the image guided brachytherapy study included IA-IVB cervical cancer patients. Using definitive EBRT +/- concurrent chemotherapy followed by IGBT was all used. The 3/5-year local control, pelvic control, cancer specific survival, overall survival were 91%/89%, 87%/84%, 79%/73%, 74%/65%.

According to the oncologic results, local control was similar to our study. Our study showed better survival which may be from including only completing definitive CCRT and IGBT patients. In Retro EMBRACE, IGBT consisted of using CT or MRI scan which resulted in 5-year G3–G5 morbidity was 5%, 7%, 5% for bladder, gastrointestinal tract, vagina. In comparison, 3D based brachytherapy in our study used pre-brachytherapy MRI scan which might help identify the exact volume of HR-CTV and IR-CTV and reduce the toxicities but longer follow up time are needed in our study’s 3D technique.

The 2D technique represented good results in overall survival, local control and progression free survival but had more toxicities compared to the 3D brachytherapy technique. Acute and late toxicities might be related to the destruction of contemporary or permanent quality of life. Some patients had disease free conditions but may suffer for the rest of their lifetime from the side effects.

However, our results should be interpreted with caution because of the retrospective manner and shorter follow up time of the 3D technique, compared to the 2D technique. While figo staging was different between the era of the 2D and the 3D techniques which have some variations on the cancer evaluation such as imaging modalities, staging. Low event rate was the limitation of the survival and local control comparison between the 2D and the 3D technique. A larger number of patients and a longer follow up time were needed to investigate the 2 techniques. Further ongoing multicenter randomized EMBRACE II study which uses the most advanced techniques is being used to further exploit the results.