Introduction
Cervical cancer is a global public health problem and the fourth most common cancer among women worldwide. 1 About 604,000 women were diagnosed with cervical cancer in 2020 and approximately 342,000 of them succumbed to the disease. 1
The standard regimen for locally advanced cervical cancer includes concomitant external beam radiotherapy (EBRT) and weekly cisplatin chemotherapy, followed by a brachytherapy boost and intracavitary brachytherapy (IBT). 2 EBRT usually delivers 45–50 Gy to treat the tumour bed, parametria, and regional lymph nodes, while IBT escalates the dose to 80 Gy to 95 Gy to provide a boost to the gross tumour to prevent recurrences. Reference Campitelli, Lazzari and Piccolo3 The inclusion of IBT in the treatment of locally advanced cervical cancer has shown a superior overall survival compared to EBRT alone. Reference Han, Milosevic, Fyles, Pintilie and Viswanathan4,Reference Korenaga, Yoshida and Pierson5
The IBT boost is an invasive technique that requires an applicator insertion into the body cavity. Its superiority is characterised by dose inhomogeneity, high tumour target coverage and the sharp dose gradient, which maximises the sparing of the organ at risk (OAR). Reference Mahmoud, Kilic, Khan, Beriwal and Small6 In IBT boost, an inhomogeneity pear-shaped distribution is created, leading to a higher central dose of upto 300% of the prescribed dose, while sparing the nearby tissues. Reference Al Feghali and Elshaikh7 However, IBT boost has a few drawbacks. Firstly, it is highly dependent on the physician’s skill. For example, improper applicator placement has been reported to reduce both local control (LC) and disease-free survival rates. Reference Viswanathan, Moughan and Small8 Secondly, IBT boost is very labour intense as it requires applicator insertion with general anaesthesia and extensive quality assurance measures. Reference Al Feghali and Elshaikh7 Thirdly, IBT boost might associated with side effects such as pain and uterine perforation. Reference Kirchheiner, Nout and Czajka-Pepl9 Moreover, patient refusal and unflavoured patient conditions such as large tumours, older age and anatomical configuration, have contributed to the issues. Reference Mahmoud, Kilic, Khan, Beriwal and Small6 As a result, the IBT boost has gradually been replaced by EBRT boost techniques using intensity-modulated radiotherapy (IMRT), volumetric modulated arc therapy (VMAT) and stereotactic body radiotherapy (SBRT) techniques in the past decades. Reference Campitelli, Lazzari and Piccolo3
In our institution, the VMAT technique has been used as an alternative to the IBT boost to treat locally advanced cervical cancer because it can achieve highly conformal dose distribution while sparing the normal surrounding tissues. Reference Rashid, Ahmad, Memon and Hashim10 Nevertheless, VMAT boost has inferior target coverage and OAR sparing, and it does not have the substantial dose inhomogeneity seen with IBT boost. Reference Campitelli, Lazzari and Piccolo3 The homogenous dose distribution in VMAT boost has delivered 95–107% of the prescription dose to the planning target volume (PTV) without a high central dose. Reference Tanderup, Fokdal and Sturdza11 Moreover, the radiation in the VMAT boost does not directly target the tumour as in IBT but must pass through the skin to reach the tumour. Furthermore, there is a splash of low dose to surrounding normal tissue in VMAT boost due to the utilisation of a higher monitor unit, which potentially increases the risk of secondary malignancies. Reference Teoh, Clark, Wood, Whitaker and Nisbet12 VMAT boost also changes the target volume position due to the internal target movement, resulting in the normal tissues being irradiated. Reference Assenholt, Petersen, Nielsen, Lindegaard and Tanderup13 Hence, an applicator-guided VMAT (AG-VMAT) boost, with an applicator inserted in the vagina and uterus, is introduced to localise the gynaecological organs and other surrounding organs. Reference Pedicini, Strigari and Caivano14
Previous publications have undoubtedly rendered IBT the ultimate form of boost for locally advanced cervical cancer due to its superior target coverage and OAR sparing. Reference Sharma, Gandhi, Sharma, Rath, Jagadesan and Julka15–Reference Hegazy, Hegazy, Soliman and Elsaid17 Hence, the current major issue in VMAT boost is to achieve a tumour dose coverage that is comparable to the IBT plans while sustaining OAR sparing. In this study, a pear-shaped dose distribution was proposed to be generated in the AG-VMAT boost technique on the existing IBT boost patient.
Materials and Methods
Patient selection
This study was a retrospective review of eighteen locally advanced cervical cancer patients who were referred to our institute between the years 2020 and 2023. All the patients received whole pelvis EBRT with a prescription dose of 45 Gy/25 fractions or 48·6 Gy/27 fractions, followed by 3–5 cycles of cisplatin or carboplatin. Ten patients underwent IBT boost using a Fletcher (consists of tandem and ovoid) applicator with the prescription dose of 21 Gy/3 fractions, 22·5 Gy/3 fractions or 18 Gy/2 fractions due to the different dose protocols practised in our institute over time. Meanwhile, eight patients received a 20 Gy/10 fractions of VMAT boost. The patient characteristics are shown in Table 1. Before conducting this study, approval was sought from the Human Research Ethics Committee of our institution.
Table 1. Patient and disease characteristics
Evaluation of IBT boost treatment plan
Firstly, 10 patients were inserted with Fletcher (consists of tandem and ovoid) applicator in the operation theatre with general anaesthesia. Then, a series of IBT boost computed tomography (CT) images were acquired using the Phillips Brilliance CT Big Bore 32 Slices (version 3·6 oncology) (Phillips Medical System, Cleveland, OH, USA) with a 2·5 mm slice thickness. The images were then transferred to the Oncentra treatment planning system (TPS) (V4·3·0·410) (Nucletron BV, Veenendaal, The Netherlands). The OARs, including the bladder, rectum, sigmoid, and left and right femoral heads, were delineated by the medical physicist. Following this, an IBT boost treatment plan was generated using a 5 mm step size and the Manchester system, where the dose was prescribed to point A. Point A is defined as 2 cm lateral to the tandem and 2 cm superior to the mucosa of the lateral fornix of the vagina, along the plane of the tandem. Reference Matsukawa, Sasaki, Hirayama, Hirose and Fukunaga18 The planned dose was then analysed. The 100% isodose line is defined as the ‘high risk (HR)-clinical target volume (CTV)’, which represented the tumour target for the IBT boost as shown in Figure 1.
Figure 1. Pear-shaped outline was produced using the 100% isodose line in the existing intracavitary brachytherapy boost treatment plan.
AG-VMAT boost planning in eclipse TPS
For these patients, two plans were generated: IBT boost and AG-VMAT boost plans. The IBT boost CT image with the contouring was exported to the Aria Workstation for AG-VMAT boost treatment planning using Eclipse TPS V.13·6 (Varian Medical System, Inc. Palo Alto, CA, USA). In Eclipse TPS, the mass density of the Fetcher applicator was overwritten to density of water (1·00 g/cm3) to exclude the attenuation of the applicator. The AG-VMAT boost treatment plan was created using one full arc (179°–181°) in a counterclockwise direction, with 6 MV photon energy, and a prescribed dose of 20 Gy/10 fractions. The HR-CTV generated in Oncentra TPS was designated as PTV and the collimator angle was set at 45°. After setting all the parameters, optimisation was done for each plan.
For all the treatment plans, the PTV coverage had to exceed 95% of the prescribed dose (D95 ≥ 95%), and the hotspots were limited to 110% or less. Since no standard protocol exists for OAR dosing in AG-VMAT, our institution protocol was referred to in Table 2.
Table 2. Dose constraint to organ at risk (OAR) for applicator-guided volumetric modulated arc therapy boost planning
AG-VMAT boost planning analysis using hypofractionation
In the planned AG-VMAT boost for ten patients, different hypofractionation dose prescriptions, which were 20 Gy/5 fraction, 25 Gy/5 fraction, 20 Gy/4 fraction, 19·5 Gy/3 fraction, 24 Gy/3 fraction, 25·5Gy/3 fraction, were applied. The dose constraint for each treatment plan is shown in Table 3.
Table 3. Dose constraint to organ at risk (OAR) for hypofractionation schemes of applicator-guided volumetric modulated arc therapyboost based on Bisello et al. Reference Bisello, Cilla and Benini19
Evaluation of VMAT boost treatment plan
On the other hand, CT simulations were performed for eight patients (without the applicator in situ). The CT images were exported to the Aria Workstation using Eclipse TPS V.13 for VMAT boost treatment planning. The treatment plan was created using one full arc (179°–181°) in a counterclockwise direction, with 6 MV photons and a prescribed dose of 20 Gy/10 fractions. The planned dose was then analysed.
Dosimetric analysis in IBT, AG-VMAT, and VMAT boosts
IBT, AG-VMAT and VMAT boosts were evaluated using dosimetric parameters for the tumour target and OAR. For the tumour target, the dose to 90% of the target volume (D90), conformity index (CI) and homogeneity index (HI) were used for comparison. For OAR, minimum doses of the most exposed 2 cm3 volume (D2cc) among the three boost techniques were compared. The formula of CI and HI is as follows:
$$CI = {{{V_{95\%}}}\over{{{volume\;of\;PTV}}}}$$
where V95 is the target volume that encompasses a 95% isodose line.
$$HI = {{{D_2} - {D_{98}}}\over{{D_{50}}}}$$
where Dx is the dose received by x% of the target volume.
Then, Statistical Package for Social Science software version 26 was used to carry out the Kruskal-Wallis statistical test to analyse the dose differences for the target and OAR across the three boost techniques. The analysis indicated a significant difference when the p-value ≤ 0·05. Moreover, the accumulated dose from the EBRT and each boost technique was evaluated in terms of equivalent dose in 2 Gy per fraction (EQD2) using α/β = 3 Gy for OAR and α/β = 10 Gy for tumour target. The dose-volume constraints for the accumulated dose to the tumour target were D90 = 80–90 Gy EQD2, D2cc ≤ 90 Gy EQD2 for the bladder, and D2cc ≤ 75 Gy EQD2 for rectum and sigmoid as recommended by the American Brachytherapy Society (ABS). The formula for EQD2 is as follows:
$$EQ{D_2} = {\;D.\;{{{d + \alpha /\beta }}\over{{2 + \alpha /\beta }}}}$$
where D is the total dose (Gy) and d is the dose per fraction (Gy).
Result
Target delineation in IBT, AG-VMAT, and VMAT boosts
Figure 2(a) and (b) show pear-shaped HR-CTV and PTV in the IBT and AG-VMAT boosts, respectively. They had the same size and volume as they were appointed from the 100% isodose line in the IBT. However, in the real clinical setting, the shape of the PTV in the VMAT boost was irregular, conforming to the shape of the cervix (Figure 2(c)) rather than a pear-shaped distribution.
Figure 2. Pear-shaped target tumour in (a) intracavitary brachytherapy boost, (b) applicator-guided volumetric modulated arc therapy (AG-VMAT) boost, as well as an irregular planning target volume in (c) VMAT boost.
Dosimetry analysis of tumour target and OAR in IBT, AG-VMAT, and VMAT boosts
The dose to target tumour and OAR between IBT, AG-VMAT, and VMAT boosts was compared, as shown in Table 4. In terms of target coverage, the study found that the D90 and HI of the target in the IBT boost were significantly higher than those in the AG-VMAT boost (p-value = 0·023 and <0·001) and VMAT (p-value = 0·034 and <0·001) boosts. The D90 of the target dose showed no significant difference (p-value = 0·897) between the AG-VMAT and VMAT boosts, but the HI in the VMAT boost was significantly higher than the AG-VMAT boost (p-value = 0·009). Regarding the CI of the target, the results indicated a significant difference between the IBT and AG-VMAT boosts.
Table 4. Comparison of dose to target and OAR between IBT, AG-VMAT, and VMAT boosts using the Kruskal–Wallis statistical test
a IBT boost versus AG-VMAT boost.
b IBT boost versus VMAT boost.
c VMAT boost versus AG-VMAT boost.
Abbreviations: OAR, organ at risk; SD, standard deviation; IBT, intracavitary brachytherapy; AG-VMAT, applicator-guided volumetric modulated arc therapy; D90, dose received by 90% of the target; CI, conformity index; HI, homogeneity index; D2cc; minimum doses of the most exposed 2 cm3 volume of the OAR; EBRT, external beam radiotherapy; EQD2, minimum doses of the most exposed 2 cm3 volume of the OAR.
While assessing the OAR dose, the D2cc of the bladder, rectum and both femoral heads in the VMAT boost was in significantly higher doses than the IBT boost and AG-VMAT boost. Nevertheless, there were no significant differences in D2cc doses for the bladder (p-value = 0·912), rectum (p-value = 0·075) and sigmoid (p-value = 1·000) between the IBT boost and AG-VMAT boost. The D2cc of the left (p-value = 0·002) and right femoral heads (p-value = 0·009) in the AG-VMAT boost were significantly higher than that in the IBT boost.
In terms of EQD2, as shown in Table 5, the cumulative EQD2 for D90 of the target tumour in the IBT boost fell within the dose constraint. However, the mean D90 in the AG-VMAT boost and VMAT boost were 65·35 Gy and 65·10 Gy, respectively, which failed to meet the recommended dose range of 80–90 Gy EQD2. For OAR, D2cc for all the OAR doses in the IBT, VMAT and AG-VMAT boosts fell within the EQD2 constraint set by the ABS.
Table 5. Comparison of cumulative equivalent dose in 2 Gy per fraction (EQD2) for target and organ at risk (OAR) between external beam radiotherapy (EBRT) + intracavitary brachytherapy boost, EBRT + applicator-guided volumetric modulated arc therapy (AG-VMAT) boost, and EBRT + VMAT boost using constraint set by American Brachytherapy Society (ABS) Reference Georg, Kirisits and Goldner20
Dosimetric evaluation for different hypofractionation schemes of AG-VMAT boosts
Table 6 represents the evaluation of cumulative EQD2 between different hypofractionation schemes of AG-VMAT boost. Among the six hypofractionation schemes, only the cumulative EQD2 for D90 of the target (PTV) dose in 24 Gy/3 fractions (mean: 81·81 Gy) and 25·5 Gy/3 fractions (mean: 84·88 Gy) fell within the tolerance limit of 80–90 Gy EQD2. As for OAR, the cumulative EQD2 for D2cc of the bladder, sigmoid and rectum in all hypofractionation schemes were within the tolerance limit set by ABS, except for the D2cc of the rectum in 25·5 Gy/3 fraction. In short, a dose prescription of 24 Gy/3 fractions for AG-VMAT boost technique achieved the target and OAR dose within the ABS tolerance limit.
Table 6. Evaluation of cumulative equivalent dose in 2 Gy per fraction (EQD2) between different hypofractionation schemes of applicator-guided volumetric modulated arc therapy (AG-VMAT) boost using dose constraint set by American Brachytherapy Society (ABS)
Discussion
The current study was undertaken to mimic the pear-shaped distribution of the IBT boost in the AG-VMAT boost. Pear-shaped distribution of IBT boost is crucial in the treatment of locally advanced cervical cancer because it includes the cervix, uterus and vagina, promoting a heterogenous dose distribution that spares the nearby OAR. However, the CTV in the VMAT boost only includes the gross tumour volume and the entire cervix. Reference Lee, Song and Kim21 This allows a more homogeneous dose distribution around the target volume, which is inferior to a boost treatment. Reference Al Feghali and Elshaikh7
Table 4 shows that the D90 of the target was significantly better and in favour of the IBT boost. The result followed the study done by Khosla et al., Reference Khosla, Patel, Oinam, Tomar and Sharma16 Pinzi et al., Reference Pinzi, Landoni, Cattani, Lazzari, Jereczek-fossa and Orecchia22 and Hegazy et al., Reference Hegazy, Hegazy, Soliman and Elsaid17 but in contrast with the findings by Dobelbower et al. Reference Dobelbower, Shen, Popple and Kim23 and Wali et al. Reference Wali, Helal, Darwesh and Attar24 who reported a superior target coverage in the VMAT boost compared to the IBT boost. This advantage in the IBT boost was attributed to the heterogeneity within the target volume, as evidenced by the significant increase in the HI. This unique characteristic allowed the IBT boost to deliver a high central dose to the target volume and hypoxic cells while sparing nearby OARs, thus a better LC. Reference Sharma, Gandhi, Sharma, Rath, Jagadesan and Julka15 Meanwhile, both VMAT and AG-VMAT boosts had a more homogenous dose distribution as their HI value was closer to 0. On the other hand, the CI difference between the three boost techniques was due to the small sample size and planning factors. Reference Brennan, Thirion, Buckney, Shea and Armstrong25
In assessing the OAR sparing, the VMAT boost demonstrated the poorest OAR sparing as indicated by the significant increases in the D2cc dose for all OARs compared to the IBT and AG-VMAT boosts, except for the sigmoid. This was because it did not involve the insertion of the applicator and gauze packing of the vagina that displaces OAR. Reference Sud, Roth and Jones26 Nevertheless, the D2cc dose for the bladder, rectum and sigmoid in IBT and AG-VMAT boosts were comparable because they were planned to use the same CT images where the patient’s anatomy and position were the same. The findings complied with the study done by Georg et al. Reference Georg, Kirisits, Hillbrand, Dimopoulos and Pötter27 but contrasted with the study performed by Wali et al. Reference Wali, Helal, Darwesh and Attar24 who found that the AG-VMAT boost had caused significant reductions to D2cc doses for the bladder and rectum compared to the IBT boost. Furthermore, this study highlighted significant increments in the D2cc dose to both femoral heads’ doses in the VMAT and AG-VMAT boosts in comparison to the IBT boost. This was because these two VMAT boost techniques exposed surrounding normal tissues to more low-dose radiation, leading to a potential risk of having a secondary malignancy. Reference Al Feghali and Elshaikh7 The result was supported by the findings from Sethi et al., Reference Sethi, Jozsef and Grew28 Merrow et al. Reference Merrow, deBoer and Podgorsak29 and Wali et al.. Reference Wali, Helal, Darwesh and Attar24
In short, the IBT boost achieved the best target coverage and provided better OAR sparing. VMAT boost showed inferior target coverage and the poorest OAR sparing. This signified the importance of this research to help the VMAT boost reproduce the dosimetry as in the IBT boost by duplicating the pear-shaped distribution in the AG-VMAT boost. However, the AG-VMAT boost still had inferior target coverage but had achieved similar OAR sparing as in the IBT boost, except for higher doses to the femoral heads. Although a pear-shaped PTV had been created in AG-VMAT boost, the reproducibility of dose distribution remained undetermined and must be confirmed by performing EQD2 calculation.
According to Table 5, only the D90 dose of the target and the D2cc dose of the OAR in the IBT boost were within the recommended tolerance limit by ABS. Meanwhile, the cumulative EQD2 for D90 dose of PTV in VMAT and AG-VMAT boosts did not fall within the tolerance limit although their OAR doses were within the ABS limit. The result of the current study was inconsistent with the study conducted Pinzi et al. Reference Pinzi, Landoni, Cattani, Lazzari, Jereczek-fossa and Orecchia22 in which the cumulative EQD2 for the D90 dose of the target in the IBT boost was similar to IMRT boost and both boost techniques did not meet the tolerance limit. However, the result of the current study aligned with Khosla et al. Reference Khosla, Patel, Oinam, Tomar and Sharma16 who found that the cumulative EQD2 for D2cc of bladder and rectum in both IBT boost and IMRT boost was within the dose constraint suggested by ABS.
The IBT boost practised at our institution was able to provide sufficient target coverage and OAR sparing. Nevertheless, the target doses in both AG-VMAT and VMAT boosts were not achievable to reproduce the dosimetry of the IBT boost, which was within 80–90 Gy EQD2. This was due to the different dose fractionation schemes in IBT (18 Gy/2 fractions, 21Gy/3 fractions, or 22·5 Gy/3 fractions), AG-VMAT (20 Gy/10 fractions), and VMAT boost even though they had the same total prescription dose. The smaller dose fractionation in AG-VMAT and VMAT boosts resulted in a significant EQD2 reduction. Reference Wang, Li, D’Souza and Stewart30,Reference Shaikh, Burmeister, Joiner, Pandya, Zhao and Liu31 Hence, we concluded that the 20 Gy/10 fraction prescription dose for VMAT and AG-VMAT boosts at our institution was insufficient to provide good target coverage. Thus, the prescription dose for the VMAT boost treatment delivery at our institution should be changed to a hypofractionated regimen.
To identify a new prescription dose for VMAT and AG-VMAT boost, six different hypofractionation schemes were applied to see which dose fractionation could reproduce the target coverage of the IBT boost while maintaining the OAR sparing. Table 6 showed that only the cumulative EQD2 for D90 of the target and D2cc of all the OAR doses in 24 Gy/3 fractions AG-VMAT boost complied with the dose constraint set by ABS. Thus, the 24 Gy/3 fractions AG-VMAT boost reproduced the pear-shaped distribution as in the IBT boost.
AG-VMAT boost with 24 Gy/3 fraction used an ultra-hypofractionation scheme which was believed to significantly lessen the sublethal damage repair in the tumour, leading to a reduction in tumour repopulation, thus better tumour control. Nevertheless, the large dose fractionation might worsen the late toxicity and radiobiological effect on surrounding normal tissues and toxicity rates. Reference Small, Mell and Anderson32 For example, Albuquerque et al. Reference Albuquerque, Tumati and Lea33 reported a high rate of late toxicity (26·7%) and a lower 70% LC rate using ultra-hypofractionation schemes of SBRT boost in locally advanced cervical cancer. Furthermore, the ultra-hypofractionation caused a larger reduction in the redistribution and reoxygenation of the tumour, which may eventually reduce the efficacy of the treatment. Reference Brand, Kirby, Yarnold and Somaiah34 In the 24 Gy/3 fractions treatment, the reoxygenation utilisation rates were only 66%, which were lower than 75% and 80% in the 4 and 5-fraction treatments, respectively. Hence, for better reoxygenation, the interfraction intervals of at least 3–7 days may be applied in 24 Gy/3 fractions of AG-VMAT boost. Reference Tanderup, Fokdal and Sturdza11,Reference Shibamoto, Miyakawa, Otsuka and Iwata35 Given the biological effects, further research is needed to validate the application of the 24 Gy/3 fractions AG-VMAT boost in clinical settings.
There were several limitations in this study. Firstly, the target delineation in the IBT boost was a key issue. Most of the previously published studies followed the GEC ESTRO recommendations for HR-CTV delineation. Reference Hegazy, Hegazy, Soliman and Elsaid17,Reference Wali, Helal, Darwesh and Attar24 However, in this study, HR-CTV was defined using the 100% isodose line in the IBT boost, which was not accurate. Moreover, the dose was prescribed to point A instead of HR-CTV in this study. This was detrimental as point A did not correspond to specific anatomical structures, leading to uncertainty in target coverage. Reference Kehwar and Goyal36 Thirdly, the small sample size and lack of clinical follow-up data, especially for the application of different hypofractionation schemes were also one of the limitations. Additionally, the AG-VMAT boost was planned to use an IBT CT image to maintain consistent contours, rather than using two separate patient groups. This meant that the AG-VMAT boost treatment plan was always virtual and the patient’s anatomy and position were not the same as those used for VMAT boost. Lastly, the LQ model was inappropriate for quantitative predictions of dose fractionation as the LQ model might overestimate the biological effect of the dose per fraction exceeding 6 Gy. Reference Tornero-López and Guirado37
Future research should focus on the radiobiological evaluation of 24 Gy/3 fractions VMAT boost. Furthermore, comparisons between both IBT and VMAT boost techniques should not only address dosimetry but also consider logistic, clinical, and quality assurance comparisons. Despite these limitations, this study provides valuable data on the dosimetry difference between IBT boost and VMAT boost. It also provides significant data on the potential of VMAT boost to replace the IBT boost in locally advanced cervical cancer as this study had successfully reproduced the pear-shaped distribution.
Conclusion
The pear-shaped outline was successfully reproduced in the AG-VMAT boost by using the 100% isodose line from the IBT boost. However, both AG-VMAT and VMAT boosts failed to replicate the dosimetry of the IBT boost in terms of target coverage due to the insufficient prescription dose. Hypofractionation implemented for AG-VMAT boost using 24 Gy/3 fraction was able to reproduce the dosimetry in IBT boost by providing sufficient target coverage and OAR sparing. However, further research is needed to validate the application of a 24 Gy/3 fractions of VMAT boost in clinical settings.
Acknowledgements
We would like to thank School of Health Sciences and Hospital Universiti Sains Malaysia for the equipments and facility provided to conduct this research.
Financial Support
This work was supported by the Universiti Sains Malaysia [304/PPSK/6315497].
Competing Interests
The authors declare that there are no competing interests.
Ethics Approval and Consent
Ethical approval for this study has been reviewed and is granted approval for implementation by the Jawatankuasa Etika Penyelidikan Manusia Universiti Sains Malaysia (JEPeM-USM) and assigned study protocol code USM/JEPeM/KK/23010086.
Open access
