Introduction
In India, cancer of cervix (CC) is the third most frequently diagnosed cancer and the third leading cause of death. 1 The standard of care for advanced CC is concurrent chemotherapy and radiotherapy. Reference Haie-Meder, Pötter and Van Limbergen2–Reference Viswanathan, Beriwal and De Los Santos5 The latter is delivered by a combination of external beam radiation therapy (EBRT) and brachytherapy (BT) techniques. 3–Reference Viswanathan, Beriwal and De Los Santos5 Image guided brachytherapy (IGBT) in combination with EBRT offers the advantage of providing highly localized and conformal dose to the tumor with a satisfactory OAR sparing. Reference Viswanathan, Beriwal and De Los Santos5,Reference Kirchheiner, Nout and Lindegaard6 The two main brachytherapy techniques namely intracavitary brachytherapy (ICBT) and interstitial brachytherapy (ISBT) are in common practice for CC treatment with EBRT. For achieving acceptable dose coverage in the case of bulky infiltrative disease, asymmetric tumor growth, vaginal spread, and unfavourable anatomy without exceeding the OAR doses, ISBT is the preferable brachytherapy technique. Reference Martinez, Cox and Edmundson7,Reference Demanes, Rodriguez and Bendre8
The correlation between volumetric doses and the resultant toxicities are well established for the OARs namely rectum, bladder, sigmoid, and bowel in radiotherapy of CC. As a result, the ICRU (International Commission on Radiation Units and Measurements) report no. 89 on brachytherapy of CC has laid down guidelines for the tolerance doses. 3 But for the OARs like vaginal wall, setting a definite tolerance protocol has been found to be challenging as the vaginal wall acts partly as target and partly as OAR, and thus needs more attention. Also, vagina’s different regions respond differently to radiation. Reference Hintz, Kagan and Chan9,Reference Au and Grigsby10 Further, no correlation has been found between the vaginal D2cc (the minimum dose to the most irradiated 2cc volume) and vaginal toxicities. Reference Fidarova, Berger and Schüssler11 Also, there are challenges in vaginal wall contouring with current treatment planning systems (TPS) for reporting volume based doses. Westerveld et al. introduced a novel method for reporting the vaginal wall doses in terms of anatomical surrogate points based on the anatomical reference of posterior inferior border of symphysis (PIBS) which was later adopted in ICRU-89 as Level-2 advanced standard for reporting. 3,Reference Westerveld, Pötter and Berger12 The vaginal reference length (VRL) has been shown to play a major role in influencing the PIBS based doses and related toxicities. Reference Westerveld, Pötter and Berger12–Reference Wang, Zhang and Liu16
There are very few studies reporting vaginal doses, especially PIBS based doses, in CC brachytherapy and most of these studies are for ICBT cases. Westerveld et al., Limkin et al., and Wang et al. investigated PIBS point based vaginal doses with the ICBT/ICBT + ISBT applicators. Reference Westerveld, Pötter and Berger12–Reference Wang, Zhang and Liu16 Murakami et al. demonstrated vaginal tolerance for ISBT with Syed-Neblett perineal template for gynecological malignancies. Reference Murakami, Kasamatsu and Sumi17 In a recent study by A. Singh et al. showed that PIBS+2 dose was associated with vaginal strictures (VS) in patients treated with ICBT and ISBT applicators. Reference Singh, Mani and Aggarwal18 In a study by Bajwa et al. it was shown that the Manchester applicator based ICBT reported higher PIBS point doses except PIBS-2 as compared to the Fletcher Suit Delcos applicator based ICBT Reference Bajwa, Talluri and Maqbool19 indicating PIBS dose dependence on types of applicators used for brachytherapy. Studies on PIBS doses and also their correlation with VRL in specifically for ISBT with the Martinez Universal Perineal Interstitial Template (MUPIT) applicator and its dosimetric differences with the Fletcher Williamson (FW) applicator are lacking in published literature.
We carried out a retrospective estimation of PIBS based point doses and VRL on 36 patients who had undergone ICBT and ISBT using the FW (n = 18) and MUPIT (n = 18) applicator respectively for locally advanced CC at our center. We also calculated the correlation of PIBS doses with VRL as the latter was reported to influence the PIBS point doses. Reference Westerveld, Pötter and Berger12–Reference Wang, Zhang and Liu16 Our objective is to find out the applicator based differences in PIBS point doses and VRL. We believe the PIBS point dose values estimated by us would prove to be valuable to the practitioners of this form of brachytherapy, especially in the context of Indian subcontinent.
Material & Methods
Patients
A retrospective cohort of 36 consecutive patients treated with FW (n = 18) and MUPIT (n = 18) applicator at our center was included in this study. All the patients were confirmed cases of advanced CC staged as per the International Federation of Gynecology and Obstetrics (FIGO) system. Two (out of the 18) of the patients were re-irradiation cases treated using MUPIT applicator. Patient data was collected during the year of 2024 to 2025 for the ICBT and ISBT patients treated between 2020–2022 and 2017–2024, respectively.
FW based ICBT procedure
The Fletcher Williamson applicator system (Nucletron B.V, The Netherlands) was used for ICBT which consists intrauterine tube (tandem) and tilted cylindrical vaginal ovoids.
A patient taken up for ICBT was kept nil-orally overnight followed by an enema in the morning before the implant procedure. The implant procedure was carried out in the operation theatre (OT) under spinal/epidural anesthesia with the patient in the lithotomy position. A Foley’s catheter was inserted and 7 cc contrast was pushed in the balloon before starting the implant procedure. The central tandem placement was performed under trans-abdominal ultra-sonographic guidance. The length of the tandem and ovoid size was determined based on uterine sound and anatomy of patient, respectively. Median tandem length was 5 cm (4–6 cm) and median tandem angle 30 degrees (range 15–45 degrees) used. After completion of the implant procedure, the patient was carefully transported for computed tomography (CT) imaging.
MUPIT based ISBT implant procedure
The Martinez Universal Perineal Interstitial Template (Nucletron B.V, The Netherlands) applicator was used for performing the ISBT. We used 20 cm long and 2 mm external diameter stainless steel needles with the template. The central tandem was also used in most of the cases (n = 16). The median tandem angle was 15 degrees (range 15 & 30 degrees) and median tandem length was 5.8 cm (range 5–9 cm). Generally, a median number of 16 needles (range: 13–19) were placed along with the tandem. Patient preparation for ISBT was same as mentioned for ICBT cases prior OT to CT room. The average needle length beyond the external os (EOS) was 4.0 cm depending upon the extent of the disease assessed by magnetic resonance imaging (MRI).
Treatment planning
For treatment planning, CT imaging with 3 mm and 2 mm slice thickness was acquired on Brilliance 16 CT scanner (Philips, The Netherlands) in the contiguous scanning mode for the FW and MUPIT based applicators, respectively. The images were transferred to Oncentra Brachy TPS version 4.6.0.16 (Nucletron B.V, The Netherlands) through a Digital Imaging and Communications in Medicine (DICOM) network. The contouring of OARs was followed by implant reconstruction. For ICBT planning, standard manual loading pattern was followed for dwell position activation with dose normalisation at ICRU point ‘A’ as prescription point. For MUPIT based ISBT, manual dwell position placement was performed for each channel and dose normalization was carried out on a 5 mm box surface around the implant. The plan was further optimized using the graphical optimization (GrO) feature and/or manual dwell time adjustment to achieve the plan objectives.
Posterior-inferior border of the pubic symphysis (PIBS)
Westerveld et al. defined the PIBS as an anatomic landmark used as a surrogate to assess vaginal doses that could be easily identified on 2D radiographs, CT and MR images. Reference Westerveld, Pötter and Berger12 By considering PIBS as a reference, points along the vagina such as PIBS, PIBS+2, and PIBS-2 points could be placed. The PIBS point represented the transition zone between the upper and the mid vagina. The PIBS+2 point was an indicator of the anatomical mid of the vagina whereas the PIBS-2 point was regarded as an indicator of the vaginal introitus. Reference Westerveld, Pötter and Berger12 The VRL was considered as a surrogate for vaginal length. Reference Westerveld, Pötter and Berger12 The PIBS points could be used to quantify vaginal doses received from EBRT as well as from BT. In BT, the PIBS point was defined at the level of the PIBS where the anterior-posterior (AP) line drawn from it crossed the tandem applicator in the vaginal vault. The tandem was assumed as the central axis of the vagina. The vaginal dose points which were placed 2 cm superior to PIBS point and 2 cm inferior to PIBS point were known as PIBS+2 and PIBS-2, respectively. The VRL was the length measured from the EOS/vaginal tip/mid of vaginal source to the PIBS point in the cranio-caudal body axis using the applicator point in the patient coordinate system. 3,Reference Westerveld, Pötter and Berger12–Reference Westerveld, de Leeuw, Kirchheiner, Dankulchai, Oosterveld and Oinam15
In the present study, for the patients with the FW applicator PIBS, PIBS+2 and PIBS-2 points were kept as above procedure. And, TVL was measured from superior vaginal mucosal to the inferior extent of 100% isodose curve.
Treatment delivery
The prescription dose (PD) and fractionation pattern was as per the institution protocol. All patients underwent EBRT of 45–50 Gy in 25 fractions (1.8–2 Gy/fraction) with concurrent chemotherapy and completed it within 5–6 weeks. This was followed by ICBT in 4 fractions (range: 3–4 fractions) with a median dose of 6 Gy (range: 6–8 Gy) per fraction, each fraction 4–7 days apart. In case of ISBT, a total of four fractions (range: 3–6 fractions) with a median dose of 6 Gy (range: 5–8 Gy) per fraction and two fractions per day at least 06 hrs apart on consecutive days were delivered. The ICBT and ISBT dose equivalent to 2 Gy per fraction (EQD2) was estimated to be 29.75–32 Gy and 18.75–48 Gy using linear quadratic model taking α/β = 10 Gy for tumor and α/β = 3 Gy for OARs. The treated volume ranged from 75–121 cc and 69–179 cc for ICBT and ISBT, respectively. The treatment (EBRT plus BT) was planned to be completed within 08 weeks aiming for achieving a total EQD2 of 80–90 Gy for the target, D2ccof less than 85 Gy for bladder and less than 75 Gy each for rectum and sigmoid (2–5) except for the re-irradiation cases. Approved BT plans were delivered on Microselectron-HDR Version V3 machine (Nucletron B.V, The Netherlands).
Statistical analysis
The OriginPro2024b graphing and analysis software version 10.1.5.132 (OriginLab Corporation, USA) was used for statistical analyses. Since the data was not normally distributed, the PIBS doses and VRLs were expressed as median values. Mann Whitney Test was performed to find the statistical significance of the differences observed in values between the two kinds of brachytherapy applications i.e. ICBT and ISBT with FW and MUPIT applicators. The correlation between PIBS point doses and VRL values were obtained by calculating Spearman rank correlation coefficients. Also, for the comparison of our data with the published literature-the latter available as mean absolute dose values, we first converted them (published values) into percentage of PD and then compared to obtain corresponding p-values employing the t-test. The assumption used was that VRL ≤ 4.5 cm to be considered as short vagina type and VRL > 4.5 cm as long vagina type for our study. Reference Wang, Zhang and Wang14
Results
Figure 1 (a) and (b) show the transverse and sagittal CT views for a patient with an in-situ FW and MUPIT applicator for ICBT and ISBT patients, respectively. The PIBS, PIBS+2, and PIBS-2 points are also indicated in the figure. Table 1 provides the measured PIBS point doses in terms of percentage of the PD received per ICBT and ISBT fraction with the mention of significant difference between them. Table 2 shows the PIBS point doses as percentage of PD for comparison with published literature.
(a) Axial and sagittal CT views of an in-situ FW implant showing PIBS, PIBS+2 and PIBS-2 points. FW: Fletcher Williamson; PIBS = Posterior inferior border of the pubic symphysis; PIBS+2, 2 cm superior to PIBS point; PIBS-2 = 2 cm inferior to PIBS point. (b) Axial and sagittal CT views of an in-situ MUPIT implant showing PIBS, PIBS+2 and PIBS-2 points. MUPIT: Martinez Universal Perineal Implant Template; PIBS = Posterior inferior border of the pubic symphysis; PIBS+2 = 2 cm superior to PIBS point; PIBS-2 = 2 cm inferior to PIBS point.
Estimated PIBS point doses in terms of percentage of prescription dose for FW and MUPIT applicator
PIBS, Posterior inferior border of pubic symphysis; PIBS+2, 2 cm superior to PIBS point; PIBS-2, 2 cm inferior to PIBS point; VRL, Vaginal reference length; FW, Fletcher Williamson; MUPIT, MUPIT: Martinez Universal Perineal Implant Template. PIBS point doses are shown as percentage of prescription dose per fraction.
Estimated dose to PIBS points and VRL. For comparison Limkin et al. Reference Limkin, Dumas and Rivin del Campo13 values are also shown
PIBS, Posterior inferior border of the pubic symphysis; PIBS+2, 2 cm superior to PIBS point; PIBS-2, 2 cm inferior to PIBS point; VRL, Vaginal reference length; SD, Standard deviation; FW, Fletcher Williamson; MUPIT, MUPIT: Martinez Universal Perineal Implant Template. PIBS point doses are shown as percentage of prescription dose per fraction.
Figure 2 (a) and (b) shows the scatter plot depicting the correlation between VRL and PIBS doses in terms of percentage of the PD. As evident from the plot, the correlation seems to be strong with FW applicator and poor with MUPIT applicator. The Spearman rank correlation coefficients (ρ) values were estimated to be −0.97, −0.89 and −0.83 between PIBS, PIBS+2, PIBS-2 and VRL, respectively for FW applicator. With MUPIT applicator the corresponding values were −0.31, −0.04 and −0.46, respectively.
(a) PIBS point doses in terms of percentage of PD with respect to VRL for ICBT cases. PIBS = Posterior inferior border of the pubic symphysis; PIBS+2 = 2 cm superior to PIBS point; PIBS-2 = 2 cm inferior to PIBS point; VRL = Vaginal reference length; ICBT = Intracavitary Brachytherapy; PD = Prescription Dose. (b) PIBS point doses in terms of percentage of PD with respect to VRL for ISBT cases. PIBS = Posterior inferior border of the pubic symphysis; PIBS+2 = 2 cm superior to PIBS point; PIBS-2 = 2 cm inferior to PIBS point;VRL = Vaginal reference length;ISBT = Interstitial Brachytherapy; PD = Prescription Dose.
Discussion
PIBS point dose reporting was considered as a surrogate for vaginal wall doses. It is interesting to note that despite being a relatively new concept it remains a point based reporting as against the general trend of volume based prescription and reporting in IGBT. The estimated PIBS point doses in our case follow a general pattern similar to the ones reported in the literature for different types of applicators. The PIBS+2, receiving the maximum dose, being in the high dose region and PIBS, PIBS-2 representing the intermediate and low dose regions in the vagina, respectively. Reference Westerveld, Pötter and Berger12–Reference Westerveld, de Leeuw, Kirchheiner, Dankulchai, Oosterveld and Oinam15,Reference Singh, Mani and Aggarwal18,Reference Bajwa, Talluri and Maqbool19 As predicted, the doses in terms of PD were higher for MUPIT than FW applicator due to the close proximity to dwell position of the needles. And, the difference between them is significant (p < 0.05) in the present study. In terms of dose as a percentage of the PD, the differences are large for MUPIT applicator but statistically insignificant as compared to the values reported by Limkin et al. Reference Limkin, Dumas and Rivin del Campo13 for ICBT cases (p > 0.05) whereas for FW applicator it was statistically significant (p < 0.05) for PIBS, PIBS+2 respectively except PIBS-2. The statistical insignificance of the comparison could be due to a small number of cases having wider variations in estimated values. A recent study on ICBT and ISBT by A. Singh et al. (2023) showed that dose at PIBS+2 was a strong predictor for vaginal stricture. Reference Singh, Mani and Aggarwal18 We have compared PIBS+2 (%) in terms of percentage of prescription dose values with A. Singh et al. (2023), which indicated that ICBT (FW) and ISBT (MUPIT) patients may have grade 2 and grade 3 VS, respectively. The PIBS-2 point reporting was crucial for patients with distal infiltrations of the vagina (i.e. FIGO-IIIA lesions) Reference Limkin, Dumas and Rivin del Campo13 and its dose should be restricted to <5 Gy as per EMBRACE II protocol in cases where the vagina was not involved. Reference Pötter, Tanderup and Kirisits20 And, especially for younger patients who were sexually active. These findings reinforce the need to estimate these values at the planning stage enabling corrective measures to avoid late vaginal toxicities.
The large variations in PIBS (PIBS, PIBS+2, PIBS-2) for MUPIT applicator doses from patient to patient as shown in Figure 2 (b) could be due to the large VRL variations from patient to patient. Also, it was observed that PIBS doses were greatly influenced by the treated length i.e. the extent of the disease. In cases where the disease extended into vagina the PIBS dose was excessively high because of additional tandem loading towards the vagina of median length 3 cm in order to achieve better coverage in case of MUPIT implant. This often led to PIBS points coinciding with the source dwell positions.
As shown in Table 2, the mean VRL value for our FW and MUPIT study population was 3.79 and 2.89 cm respectively. These values of VRL were shorter compared to some of the published studies. Limkin et al. Reference Limkin, Dumas and Rivin del Campo13 reported a mean VRL of 5.44 cm for patients treated with ICBT. The mean differences were 1.65 and 2.55 cm which were statistically significant too. The reasons for the difference in VRL could be (i) actual shorter VRL in the patients recruited for the study (ii) distortion in patient anatomy caused by the extensive disease in the patients chosen for MUPIT based ISBT and hence the resultant underestimation of VRL. Patients chosen for ISBT(MUPIT) always have extensive disease and distorted anatomy of the cervico-vaginal region as compared to patients undergoing ICBT (iii) the difference in methods of measuring VRL in MUPIT and ICBT cases. In ICBT the superior reference point for VRL measurement is mid vaginal source. This position may vary depending on the positions of the ovoids/ring in the vaginal fornices. 3 In the case of MUPIT, the superior reference point we used for VRL measurement was the tip of the cylinder which might not be always flushed with EOS resulting in underestimation of VRL. This could especially happen when due to clinical or anatomical reasons such as narrow vagina or vaginal stenosis (in re-irradiation cases) the cylinder could fully not reach the EOS. With MUPIT, the anatomical VRL measurement was not feasible using CT imaging as EOS delineation became extremely challenging owing to artefacts created by the applicator and/or persistent disease. The method suggested by Westerveld et al. using placement of a radiopaque marker at the EOS to identify the vaginal tip in the case of EBRT may not always work in the case of ISBT. Reference Westerveld, Pötter and Berger12 Therefore, for the sake of ease and consistency of measurement of VRL, the tip of the vaginal cylinder was used as a surrogate for the EOS or vaginal tip. As reported by Wang et al., the VRL for the Chinese population was <4.5 cm which was considered short as compared to the western population. Reference Wang, Zhang and Wang14 It was difficult to attribute shorter VRL values in our study group to any one specific reason. However, results from A. Singh et al. and Bajwa et al. indicated that the Indian population had shorter vagina as compared to populations of many other countries. Reference Singh, Mani and Aggarwal18,Reference Bajwa, Talluri and Maqbool19
Spearman’s rank correlation coefficient (ρ) was calculated to test the correlation between PIBS point doses (in terms percentage of PD) and VRL for our patient population. The calculated ρ value indicates a strong correlation for FW applicator, which was in agreement with the published literature. Reference Westerveld, de Leeuw, Kirchheiner, Dankulchai, Oosterveld and Oinam15 And, for MUPIT applicator the PIBS, and PIBS-2 doses had negative low (ρ = −0.31) and moderate (ρ = −0.46) correlation with the VRL, respectively. Whereas, the PIBS+2 dose was not observed to be correlated (ρ = −0.04) with VRL unlike the previous findings of a multicentre evaluation of PIBS points doses and VRL by Westerveld et al. Reference Westerveld, de Leeuw, Kirchheiner, Dankulchai, Oosterveld and Oinam15 The above correlations can be appreciated well from the Figure 2 (a) and (b). We are not in a position to offer any explanation for this difference with the present data. However, further studies over a larger population of patients, including multicentric studies, with MUPIT implants may be needed to validate or negate our results.
VRL plays an important role in influencing all the PIBS point doses. 3,Reference Westerveld, Pötter and Berger12–Reference Wang, Zhang and Liu16 Wang et al. recently reported that the PIBS, PIBS+2 doses had strong correlation with VRL as compared to PIBS-2 doses. Reference Wang, Zhang and Liu16 It was also shown that the incidence rate of vaginal stenosis Grade (G) ≥ 2 was 2.3 times higher in patients with VRL ≤ 4.6 cm as compared to VRL ≥ 4.6 cm. Reference Wang, Zhang and Liu16 In our case the PIBS point doses were on lower side even for VRL ≤ 4.5 cm for FW applicator. We may conclude that ISBT (MUPIT) will have higher morbidities as compared to ICBT (FW) patients based on VRL.
Conclusion
We estimated the PIBS point doses in ICBT and ISBT with FW and MUPIT based applicators in brachytherapy of cancer of cervix. We observed a large inter-patient variation in the dose values with MUPIT applicator as compared to FW applicators. The high values of PIBS point doses in some cases underscores the need to estimate PIBS point doses at the planning stage enabling corrective measures, if required, to avoid late vaginal toxicities. Also keeping in mind the fact that in cases where the dose values are excessively high, a careful analysis was needed to rule out the possibility of the PIBS point coinciding with source dwell positions. Some of our results deviated from published studies, especially associated with MUPIT applicator. We believe further studies with MUPIT based implants for a larger population of patients, including multicentric studies, are needed to validate our results.
Acknowledgements
None.
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Competing interests
The author(s) declare none.
Open access
