Myeloablative Total Body Irradiation With Volumetric Modulated Arc Therapy for Children—a Single Center Experience

Abstract

Purpose

Total body irradiation (TBI) is a critical component of conditioning therapy prior to allogeneic hematopoietic stem cell transplantation in pediatric acute lymphoblastic leukemia. Our institution implemented volumetric modulated arc therapy-TBI (VMAT-TBI) in 2023, aiming to improve target conformity and organ-at-risk sparing using standard linear accelerators. This study describes our initial 2-year institutional experience with VMAT-TBI, focusing on treatment planning and delivery metrics, as well as early clinical outcomes and acute toxicities in pediatric patients.

Methods and Materials

We conducted a review of 10 pediatric patients treated with VMAT-TBI at our institution from October 2023 to June 2025. All patients received myeloablative conditioning with 12 Gy in 6 fractions over 3 days, followed by intravenous etoposide preceding allogeneic hematopoietic stem cell transplantation. Data collected included treatment planning parameters, planning duration, delivery times, overall survival, relapse-free survival, and acute toxicities.

Results

Treatment plans achieved robust target coverage (V₁₀₀ > 90%) and met organ-at-risk constraints, with mean lung, kidney, and lens doses of 8.8 Gy, 8.7 Gy, and 8.5 Gy, respectively. Mean beam-on time per fraction was 21.3 minutes, and in-room time was 77 minutes. Image guided verification showed that online translation was within 3 mm and <1° rotation. The median follow-up was 19 months, with 1-year overall survival and relapse-free survival rates of 90%. Nonrelapse mortality at 1 year was 10%. The most common acute toxicity was mucositis, affecting 70% of patients (grades 2–3 in 40%), with a median onset of 2 days. Gastrointestinal toxicities, including vomiting (50%) and diarrhea (40%), occurred early, with a median onset of 2–3 days.

Conclusions

Our early institutional experience demonstrates that VMAT-TBI is safe, feasible, and effective in pediatric patients, achieving excellent target coverage and acceptable acute toxicity profiles. These findings support the viability of implementing VMAT-TBI even in resource-constrained settings.

Introduction

Total body irradiation (TBI) remains a fundamental component of conditioning regimens prior to allogeneic hematopoietic stem cell transplantation (HSCT) in pediatric patients with leukemia.¹ TBI is primarily used to eradicate residual malignant cells, provide effective immunosuppression, and create marrow niches to facilitate successful engraftment, particularly in cases of acute lymphoblastic leukemia (ALL).^2,3

Conventional TBI is delivered using large static fields at extended source-to-surface distances.² However, this method often results in suboptimal dose coverage and significant radiation exposure to critical organs at risk (OARs), leading to substantial long-term adverse effects. Children are vulnerable to these late toxicities, including growth retardation, endocrine dysfunction, cognitive impairment, and increased risk of secondary malignancies.⁴

Advanced radiation therapy techniques such as intensity-modulated radiation therapy, helical tomotherapy, and volumetric modulated arc therapy (VMAT)² improve dose distribution, conformity, and sparing of normal tissues compared to conventional TBI.⁵ Studies report low rates of severe acute toxicity with VMAT-TBI, with reduced mean lung and kidney doses.⁶ Despite these benefits, VMAT-TBI is complex, requiring multi-isocenter planning, precise junction management, and meticulous immobilization.⁷ Sedation or anesthesia may be required for pediatric patients to ensure treatment accuracy and reproducibility, further prolonging overall treatment duration and increasing resources.

TBI practices vary widely, with only 17% of centers using advanced planning techniques.^8,9 In Asia, VMAT-TBI implementation, particularly for pediatric patients, remains limited.^10,11 This study describes our institutional experience with myeloablative VMAT-TBI for pediatric patients undergoing HSCT conditioning. We examined the technical aspects, dosimetric outcomes, clinical feasibility, and early treatment-related toxicities and survival outcomes. Using the Monaco treatment planning system (TPS) for whole-body VMAT plans without a rotatable tabletop, our study provides practical insights for centers with standard linear accelerators considering this technique.

Methods

Patient selection

This retrospective cohort study included children aged <18 who received TBI using VMAT at our institution between October 2023 and June 2025. All patients underwent myeloablative conditioning prior to allogeneic HSCT. Patients who received nonablative TBI regimens for nonmalignant indications or did not complete therapy were excluded.

Each patient received 12 Gy in 6 fractions over 3 consecutive days, delivered twice daily with at least 6 hours separating each fraction following the recommendations of the For Omitting Radiation Under Majority Age study (NCT 01949129).¹ Clinical data, treatment parameters, dosimetric indices, toxicities, and outcomes were obtained from electronic medical records, the MOSAIQ oncology information system, Monaco version 5.51.10 TPS (Elekta AB), and MIM Maestro® version 7.0.6 (MIM Software Inc) image processing workstation. Additional treatment planning and delivery times were recorded manually to assess workflow efficiency and resource requirements. The study was approved by the institutional ethics review board (MRECID No: 2025116-14609). The database was locked on September 12, 2025, for analysis.

Simulation and immobilization

The planning process is summarized in Supplementary Figure 1. All patients underwent CT simulation on a SOMATOM go.Sim CT simulator (Siemens Healthineers) in the supine position using a standard 3-point thermoplastic mask and a full-body vacuum cushion (Vacloc 100 × 200 cm) for immobilization. Hands were positioned close to the body to maximize scan coverage and dose uniformity. A bent-knee position was adopted to improve patient comfort and reproducibility. For patients shorter than 110 cm, a single head-first supine (HFS) scan was sufficient. Patients taller than 110 cm underwent 2 CT scan orientations: (1) HFS (vertex to midthigh) and (2) feet-first supine (FFS) (feet to pelvis). To aid image stitching, radiopaque fiducial markers were placed at predefined anatomic landmarks (eg, pelvis for HFS, midthigh for FFS).

Our center does not use a rotatable couch-top, as has been reported in other centers.12, 13, 14 Patients were manually rotated on the CT simulation and treatment couch in the treatment position to minimize positional error between HFS and FFS scans. Scans were acquired with 5-mm slice thickness using 120 kV and 300 mAs and the largest available field-of-view under an institutional pediatric TBI simulation protocol.

Image processing and contouring

The CT data sets were imported into MIM for image fusion and stitching. The FFS scan was coregistered to the HFS data set to generate a continuous data set reconstructed in HFS orientation. Target delineation was contoured on the composite data set and verified by the treating radiation oncologist.

Contouring followed institutional protocol and adapted the European Society for Radiotherapy and Oncology—European Society for Paediatric Oncology (ESTRO-SIOPE) consensus guidelines for pediatric TBI.² OARs included the lungs, kidneys, and lenses. The planning target volume (PTV) includes the entire external body contour, excluding 3 mm contractions around the lungs and kidneys to form planning organ-at-risk volumes (PRVs). The contracted lung and kidneys are as per the ESTRO ACROP SIOPE recommendation to ensure adequate coverage of the surrounding lymphatic and medullary structures. Hence, body PTV = (Body-3 mm) – (Lungs-3 mm) – (Kidneys-3 mm) – lenses (Supplementary Figure 2). In patients with definite central nervous system (CNS) involvement (CNS3) and CNS recurrence, lenses are included in the PTV volume.

Treatment planning

Treatment planning was performed by medical physicists using Monaco TPS. Whole-body VMAT plans were generated using multiple isocenters to ensure adequate coverage of the entire body, with overlap regions of at least 2 cm between adjacent fields to improve junction robustness. Junction robustness refers to the stability of the dose distribution at the HFS-FFS junction region when minor setup or positioning variations occur during treatment. Dose optimization used template-based inverse planning, adjusted iteratively to fulfill institutional dose objectives. Bias dose optimization between HFS and FFS fields was employed to ensure homogeneous dose distributions across junctions (Supplementary Figure 3). Dose constraints were applied as detailed in Table 2.

Table 2.

Dosimetric analysis for target volumes and organs at risk

Target volume
Target volume	Optimization parameters (%)	Mean (%)	Range (%)	Standard deviation
V90	>98%	99.16	98.61-99.7	0.38
V95	>90%	97.90	95.96-99.05	0.94
V100	>90%	90.00	89.41-90.22	0.23
V110	<3-5%	1.15	0.02-3.37	1.06
V120	<1%	0.002	0-0.02	0.006
D0.03cc (Gy)		14.17	13.66-15.15	0.46
Organs at risk
OAR (Dmean)	Optimization parameters (Gy)	Mean dose (Gy)	Range (Gy)	Standard deviation
Lungs	≤10	8.8	8.25-9.49	0.43
Lungs-3 mm	<8	7.93	7.50-8.60	0.41
Kidneys	≤9	8.74	8.15-9.33	0.35
Lenses	≤12	8.54	8.10-9.20	0.40
Plan characteristics
	Mean	Range		SD
Monitor units	4611	2972-6491		1118
Isocenters	6	4-7		0.82
Planning time (d)	3.8	2-7		1.99
Beam-on time per fraction (min)	21	14-29		5.2
In-room time per fraction (min)	77	53-96		14.2

Dose rate was not restricted in this cohort, with instantaneous dose rates of >200 monitor units (MU)/min, based on prior literature indicating limited clinical impact of dose-rate effects in fractionated TBI.^2,15 Treatment plans were reviewed and approved by radiation oncologists and medical physicists prior to treatment delivery.

Patient-specific quality assurance

All treatment plans underwent patient-specific quality assurance prior to initiation. Fluence maps for each arc were measured using the I’mRT MatriXX 2D detector array (IBA Dosimetry) and compared with TPS-calculated fluence using gamma analysis (3%/3 mm criteria, passing threshold ≥ 90%) (Supplementary Figure 4). Gafchromic EBT3 films (Ashland Advanced Materials) were used for additional verification in junction regions.

Treatment delivery and verification

TBI was delivered using an Elekta VersaHD linear accelerator (Elekta AB) equipped with kilovoltage cone beam CT (CBCT) (Elekta AB). Image guided verification for each fraction was performed using fast CBCT for each HFS isocenter (cranium, thorax, abdomen, pelvis) and the FFS isocenters (upper leg, mid leg, and lower leg).

Online corrections in the longitudinal, vertical, and lateral directions were applied at the cranial isocenter, while translational adjustments in the vertical and lateral directions only were made at other isocenters based on CBCT alignment. Setup accuracy was verified with a 0-mm no-action level protocol. No rotational correction was done for all isocenters. In vivo dosimetry was performed using Gafchromic EBT3 film placed at the HFS-FFS junction, assessing cumulative dose and junction homogeneity.

Patient outcomes and statistical analysis

Planning parameters, treatment times, and dosimetric indices were collected and summarized using descriptive statistics. Overall survival (OS) was defined as the time from the end of TBI to death. Relapse-free survival (RFS) was defined as the time from the end of TBI to relapse or death from any cause.¹⁶ Nonrelapse mortality was defined as death without prior relapse and considered a competing risk for relapse.¹⁷ Outcomes of patients who were followed up by their primary referring center were updated in our study database.

Toxicities were retrospectively collected during inpatient stay and on monthly outpatient visits based on electronic medical records. Acute toxicities were defined as toxicities occurring 90 days after transplant and were graded using the Common Terminology Criteria for Adverse Events version 5.0. Toxicities were categorized as likely TBI-related if no alternative cause was documented. Acute graft-versus-host disease (GVHD) was graded using the Modified Glucksberg criteria.¹⁸ Chronic GVHD was graded using National Institutes of Health criteria.¹⁹

Statistical analysis was performed using IBM SPSS version 29 (IBM Corp). Descriptive analysis was performed using Fisher’s exact test and independent t tests. Survival outcomes were analyzed with Kaplan-Meier curves and life tables. Correlation analysis between variables affecting treatment times was performed with Pearson’s correlation test.

Results

Patient characteristics

A total of 10 patients were treated with VMAT-TBI between October 2023 and June 2025. The median age at treatment was 9 years (range, 4-13 years). Baseline characteristics are described in Table 1. All patients received 12 Gy in 6 fractions without boost or dose escalation, and no patient received previous radiation. No patient had CNS involvement or CNS recurrence in our cohort. Five children required sedation, and 2 required general anesthesia.

Table 1.

Baseline demographics for pediatric patients treated with myeloablative TBI

Characteristics	N = 10 n (%)
Mean age at RT, years (range)	8.6 (4-13)
Mean height, cm (range)	128 (100-158)
Gender
Male	8 (80)
Female	2 (20)
ECOG PS
0-1	10 (100)
Puberty
Pre	9 (90)
Post	1 (10)
Diagnosis
ALL	9 (90)
AUL	1 (10)
TBI indicated for relapsed leukemia
Yes	6 (60)
No	4 (40)
Anesthesia requirement
Sedation	3 (30)
General anesthesia	2 (20)
None	5 (50)
Transplant types
Haploidentical	7 (70)
Matched-sibling donor	3 (30)
Number of transplants
1	9 (90)
2	1 (10)
Outcomes
1-year overall survival	90%
1-year relapse-free survival	90%
1-year nonrelapse mortality	10%

ALL = acute lymphoblastic leukemia; AUL = acute undifferentiated leukemia; ECOG PS = Eastern Cooperative Oncology Group Performance Status; RT = radiation therapy; TBI = total body irradiation.

Treatment planning

The treatment planning time ranged from 2 to 7 days, with a median duration of 3.8 days. The mean total MU for TBI plans were 4611 (SD 1118; range, 2972-6491). Patient height ranged from 100 to 158 cm, and weight from 13 to 39 kg. Taller patients required higher MUs and additional isocenters (r = 0.915, P < .001) (Fig. 1) which resulted in a longer time for treatment delivery (r = 0.982, P < .001) (Fig. 2). Treatment plans incorporated a mean of 6 isocenters per patient (range, 5-7), reflecting the complexity of full-body VMAT planning in a pediatric population.

Figure 1.

Correlation between patient height and monitor units for total body irradiation (TBI) plans.

Figure 2.

Correlation between monitor units and beam-on times for total body irradiation (TBI) plans.

Target coverage goals were consistently met (Table 2). The average target volume coverage V₉₀, V_95, and V₁₀₀ was 99.16%, 97.90%, and 90.00%, respectively. Dose hotspots remained within acceptable limits (V₁₁₀ 1.15%, V₁₂₀ 0.002%). Mean doses to OARs were within institutional guidelines: 7.93 Gy to the lung PRV (lungs-3 mm), 8.74 Gy to the kidneys, and 8.54 Gy to the lenses. The mean lung dose was 8.80 Gy and was not used as a plan optimization structure in our early experience.

Treatment verification and delivery

The patient-specific quality assurance test for all treatment plans was above 90%, and no cold or hot spot was observed for the junction verification. The mean treatment beam-on time was 21 minutes per fraction (range, 14-29 minutes), while the average in-room time was 77 minutes (range, 53-96 minutes) (Supplementary Figure 5), with no obvious reduction of in-room time with each successive fraction delivered. The 2 patients who completed treatment under 60 minutes were aged 4 and 5 years old and required sedation or anesthesia. Treatment plans for these patients were with 4 and 6 isocenters, respectively.

Table 3 tabulates the magnitude of correction for online verification performed at different sites. On average, the correction applied was within the PRV margin of 0.3 cm. Lateral and vertical corrections ranged from 0.1 to 0.3 cm, and no longitudinal correction was applied, with a minor shift of up to 0.2 cm. In addition, rotational corrections were not performed with the recorded values ranging from 0.3° to 0.6°, well within 2° tolerance for advanced radiation therapy techniques.

Table 3.

Mean online correction during verification for 6 fractions of TBI based on the isocenter

Fraction	Cranial				Thorax				Abdomen				Pelvis				Upper leg				Lower leg
	x	y	z	Ro	x	y	z	Ro	x	y	z	Ro	x	y	z	Ro	x	y	z	Ro	x	y	z	Ro
1	0.1	0.2	0.2	0.6	0.1	0.2	0.1	0.4	0.1	0.1	0.1	0.4	0.1	0.1	0.0	0.6	0.1	0.2	0.1	0.4	0.2	0.3	0.0	0.3
2	0.3	0.3	0.3	0.8	0.2	0.4	0.0	0.4	0.1	0.3	0.1	0.4	0.1	0.1	0.1	0.4	0.1	0.3	0.2	0.5	0.1	0.2	0.0	0.0
3	0.2	0.2	0.2	0.8	0.2	0.3	0.0	0.4	0.1	0.2	0.0	0.4	0.1	0.1	0.1	0.6	0.1	0.4	0.2	0.6	0.2	0.3	0.1	0.0
4	0.2	0.2	0.2	0.8	0.1	0.3	0.0	0.5	0.1	0.2	0.1	0.4	0.3	0.1	0.0	0.6	0.2	0.3	0.2	0.4	0.1	0.2	0.1	1.0
5	0.1	0.2	0.2	0.4	0.1	0.3	0.0	0.5	0.2	0.2	0.0	0.5	0.2	0.1	0.0	0.7	0.2	0.3	0.1	1.0	0.1	0.3	0.1	0.6
6	0.2	0.3	0.3	0.2	0.2	0.3	0.0	0.3	0.2	0.2	0.0	0.4	0.2	0.1	0.1	0.4	0.2	0.2	0.3	0.4	0.2	0.1	0.0	0.1
Mean	0.2	0.2	0.2	0.6	0.2	0.3	0.0	0.4	0.1	0.2	0.1	0.4	0.2	0.1	0.1	0.6	0.2	0.3	0.2	0.6	0.2	0.2	0.1	0.3

Ro = rotation; x = lateral; y = vertical; z = longitudinal.

Survival

The median follow-up was 19 months from the initial date of diagnosis or relapse if TBI was indicated for relapsed leukemia (range, 4-28 months). At last review, 9 patients (90%) remained alive, all in morphologic remission. The 1-year nonrelapse mortality rate was 10% at the time of analysis. The estimated 1-year OS was 90%, with 1 death occurring at day +24 posttransplant due to hemophagocytic lymphohistiocytosis and Pseudomonas sepsis. The estimated 1-year RFS is 90%. Median OS and RFS were not reached.

Toxicities

The most common acute toxicity observed was mucositis, occurring in 7 patients (70%), with grades 2-3 severity in 4 patients (40%) (Table 4). The median time to onset of mucositis was 2 days post-TBI (range, 1-3 days). Gastrointestinal toxicities were also frequent, including vomiting in 5 patients (50%) with a median onset of 2 days, diarrhea in 4 patients (40%) with a median onset of 3 days, and mild pancreatitis in 1 patient (10%) with onset at 3 days, which may have been multifactorial and attributable in part to concurrent chemotherapy.

Table 4.

Toxicity profiles and grading as per CTCAE version 5

	Grade 0 n (%)	Grade 1 n (%)	Grade 2 n (%)	Grade 3 n (%)	Onset, days Median (range)
Headache	9 (90)	1 (10)	0	0	1
Mucositis	3 (30)	4 (40)	2 (20)	1 (10)	4 (1-13)
Skin reaction	9 (90)	1 (10)	0 (0)	0 (0)	10
Parotitis	9 (90)	1 (10)	0	0	2
Vomiting	5 (50)	5 (50)	0	0	2 (1-8)
Pancreatitis	9 (90)	1 (10)	0	0	3
Diarrhea	6 (60)	4 (40)	0	0	3 (1-3)
Nephritis	9 (90)	1 (10)	0	0	17
Hormonal disturbance	1 (10)	1	0	0	30
Veno-occlusive disease	6 (60)	3 (30)	0	1 (10)	9.5 (6-20)
Cytomegalovirus reactivation	8 (80)	2 (20)
Hemophagocytic Lymphohistiocytosis	8 (80)	2 (20)
Acute GVHD*	3 (30)	7 (70)
Skin		2 (20)	0	2 (20)
Lung		4 (40)	0	0
Gut		4 (40)	0	0
Chronic GVHD†		Score 1	Score 2	Score 3
	7 (70)	0	3 (30)	0
Skin			2 (20)
Musculoskeletal			1 (10)

^⁎Glucksberg grading system.

^†NIH criteria.

There were no cases of pneumonitis reported in our cohort. Two patients developed cough on days 4 and 6 post-TBI and were treated as pneumonia with sputum cultures positive for bacterial infection. Both patients were treated with intravenous antibiotics, and their respiratory symptoms were not deemed due to radiation pneumonitis. Nephritis, presenting as hemorrhagic cystitis, was observed in 1 patient (10%) at 17 days post-TBI. Parotitis occurred in 1 patient (10%) at 2 days post-TBI. A headache was reported in 1 patient (10%) on day 1, while a skin reaction was observed in 1 patient (10%) with an onset at 10 days. Four patients (40%) developed veno-occlusive disease at a median time of 9.5 days post-TBI; the majority of cases were mild.²⁰ All patients were managed conservatively with diuretics and fluid restriction.

A single case of hormonal disturbance (10%) was noted, characterized by secondary amenorrhea approximately 30 days post-TBI in a patient; this represented the latest toxicity observed. This patient remained amenorrhoeic until hormone replacement therapy was commenced on follow-up.

GVHD

GVHD occurred in 7 of 10 patients (70%), with 5 patients experiencing Glucksberg grades 1 to 2 acute GVHD and 2 patients experiencing grade 3 acute GVHD. The patient who died of septicemia and hemophagocytic lymphohistiocytosis also had concomitant grade 3 GVHD (grade 3 skin and grade 1 gut involvement). Chronic GVHD developed in 3 patients (30%). Two had skin GVHD (score 2), which was managed with steroids and ruxolitinib, while 1 patient developed GVHD of the musculoskeletal system (score 2), which required imatinib therapy.

Additional posttransplant complications included graft failure in 1 patient (10%) and hemophagocytic lymphohistiocytosis in 2 patients (20%). One patient who had graft failure received a haploidentical transplant post-VMAT-TBI and required a salvage transplant after initial graft failure. Cytomegalovirus reactivation was observed in 2 patients (20%), 1 of whom also developed cytomegalovirus retinitis and required both systemic and intravitreal ganciclovir.

Discussion

This study describes our initial institutional experience implementing VMAT-TBI in pediatric patients undergoing HSCT at a tertiary center in Malaysia. Our results demonstrate that VMAT-TBI is feasible, achieves high-quality treatment plans with excellent target coverage and acceptable OAR sparing, even in a resource-limited setting with standard C-arm linear accelerators without a rotatable tabletop.

Outcomes in our cohort were similar to those reported previously with VMAT-TBI experiences in other centers. OS at 1 year was 90%, and RFS was 90%. For pediatric patients, the Stanford experience reported a 1-year OS of 89% and 1-year RFS of 88% for myeloablative TBI.²¹ Our reported graft failure rate was 10%, which is similar to local data for haploidentical transplants prior to the introduction of VMAT-TBI, with lower rates being reported for matched-sibling donors of 2%-3%.

Our dosimetric findings are consistent with those reported in the literature. Zhang-Velten et al²¹ reported their 6-year experience using Eclipse TPS for planning, which used VMAT-TBI plans for HFS and AP-PA opposing plans for FFS orientation. The mean treatment delivery time was 72 minutes, and an average of 5 to 7 isocenters was used for pediatric and adult patients. Our center used full-body VMAT planning with Monaco TPS for pediatric patients only; mean treatment times were comparable at 77 minutes. Losert et al¹³ reported mean treatment times of 57 minutes in adult patients, and this was reduced to 38 minutes when no image guidance was used. In contrast, Tas et al¹¹ used 3 isocenters with full-body VMAT, and mean treatment times were 35 minutes for pediatric patients. These durations reflect the complexity of multi-isocenter VMAT-TBI delivery in pediatric patients, including patient cooperation, image guidance, immobilization adjustments, and treatment verification steps.

Target volume coverage was achieved within our cohort, and we report comparable mean lung and kidney doses with other studies. Mean lung doses reported by Zhang-Velten et al²¹ and Tas et al¹¹ was 8.3 Gy and 9.7 Gy, respectively, while patients in our cohort received a mean dose of 8.8 Gy to the lungs. In our initial experience, we used Lung-3 mm as an optimization structure and restricted Lungs PRV to receive <8 Gy, and we were able to achieve a mean Lungs PRV of 7.9 Gy and a mean whole lung dose of 8.8 Gy. Previous studies reported improved OS in pediatric patients with mean lung doses <8 Gy, and in the adult population, mean lung doses >8 Gy were associated with a higher risk of pneumonitis and infection. As a result, we have lowered our lung PRV dose constraints to achieve <6.5 Gy and were able to reduce the mean lung dose to <8 Gy while maintaining similar PTV coverage. Mean kidney dose was 8.7 Gy in our study, comparable with Tas et al¹¹ which reported 9.6 Gy. Lower doses reported by Keit et al²² (6.3 Gy), which did not contract the kidneys for treatment planning, accepted target coverage of V₁₀₀ >85% for PTV.

There was no dose-rate control for TBI planning in our center for all isocenters, and we used highly fluctuating instantaneous dose rates of >200 MU/min. This controversial topic has been discussed previously in the literature.¹⁵ Radiobiologically, dose rates have less effect on toxicity when using fractionated TBI; however, this was in the conventional treatment era.²³ VMAT treatments deliver continuous radiation with the linac, simultaneously selecting the best combination of dose rate, gantry, and multileaf collimator speed.²⁴ Different reports have published their experience using low instantaneous dose rates of <20 MU/min²¹ and higher rates of 100-200 MU/min⁶ and >200 MU/min¹¹ when treating the lungs. We did not experience any cases of pneumonitis in our cohort, similar to Tas et al¹¹, who also used high dose rates of >200 MU/min with no grade 3+ pneumonitis. Our findings are similar to other studies that used lower dose rates of 20 MU/min (17% in pediatric patients).²¹ However, VMAT-TBI had a lower risk of pneumonitis when compared to conventional TBI when using low dose rates to treat the lung (0% vs 17%).⁶

We also report the image-guided verification for each treatment fraction and found that online correction in all directions was acceptable despite having no rotatable tabletop. All translation discrepancies were within 3 mm, and rotation was less than 2°. This shows that VMAT-TBI is feasible and reproducible even in the pediatric population without additional equipment. Studies using a rotational tabletop top found that HFS isocenter shifts were >5 mm in 26% of treatment fractions¹⁴ and was 90% in another study that did not utilize a rotatable tabletop.²⁵ In a robustness study by Sandt et al,²⁶ a maximum loss of PTV V₉₅ coverage by 1.2% was recorded when applying 5-mm shifts in the longitudinal direction.

Acute toxicities were predominantly mild to moderate, with early onset and spontaneous resolution in most cases. Hui et al⁶ further demonstrated that VMAT-TBI reduces the incidence of pneumonitis and nephrotoxicity compared to conventional 2D TBI. Our cohort’s toxicity profile mirrors this observation, with acute toxicities predominantly mild to moderate, comprising mostly mucositis and gastrointestinal side effects. Grade 3 mucositis was lower in our cohort compared to Hui et al⁶ (10% vs 37%). Grade 1 to 2 toxicities were comparable with the previous study commonly vomiting (50% vs 81%) and diarrhea (40% vs 66%).⁶

Hui et al⁶ reported lower rates of GVHD with VMAT compared to conventional TBI (51% vs 70%). In our study, 70% of patients experienced acute GVHD, the majority being grades 1 to 2. Importantly, our observed rate of GVHD remains within the expected range reported in transplant literature,^27,28 although longer follow-up will be needed to assess for late toxicities and chronic GVHD with the use of VMAT for TBI.

Several limitations of our study should be acknowledged. First, our cohort is small, reflecting our early institutional experience; however, it is comparable with the reported literature. Additional data from other centers and longer-term follow-up will be required to assess late effects and confirm our early observations. Our median follow-up is short, and this requires ongoing surveillance for late effects and relapse with the use of TBI with VMAT techniques. However, given the previous reports with advanced techniques, we do not anticipate any increase in toxicity compared to conventional TBI. Third, due to the retrospective nature of this study, documentation of toxicity was not collected in a systematic way and may be subject to bias. Finally, despite having reported the image-guided verifications for all isocenters, robustness analysis was limited to a single patient. Comprehensive robustness evaluations should be done to better understand the impact of target volume and OAR dosimetry and will be included in future work.

In Malaysia, pediatric radiation therapy is delivered in all centers with radiation facilities, and TBI is focused in centers that have a bone marrow transplant service. Historically, TBI was delivered in a few centers with conventional techniques; however, there has been increased use of TBI tomotherapy in Malaysia, with the acquisition of tomotherapy machines, and our center is the first to deliver VMAT-TBI with standard linear accelerators. Implementing TBI in our middle-income setting posed several challenges. Implementation of VMAT-TBI is resource intensive, requiring time-consuming planning, robust quality assurance, and significant multidisciplinary collaboration. Limited machines, staffing, and planning capacity hinder application. Treating pediatric patients poses additional resource demands, including anesthesia requirements and extended treatment sessions. Extensive QA required ensures safety but increases the overall resource burden of this advanced technique.

Despite these challenges, the added resources incurred are justified as conventional TBI was not feasible in our center due to limited bunker dimensions and a lack of positioning equipment. We can now deliver VMAT-TBI with our standard linear accelerators, avoiding costly referrals to private hospitals. To optimize departmental efficiency, we are streamlining treatment delivery through treatment arc and CBCT acquisition directions, surface-guided verification, and using autoplanning scripts on Monaco TPS. For pediatric patients, audio-visual distraction and sedation-free approaches have reduced in-room times and enhanced patient experience.

Emerging techniques, such as total marrow irradiation and total marrow and lymphoid irradiation, offer further improvements by targeting bone marrow and lymphoid tissues while sparing nontarget organs. Although these modalities exhibit considerable potential to reduce late toxicities, their complexity, resource demands, and limited application in pediatric populations necessitate further investigation before widespread adoption.¹⁵

Conclusion

These results add to the growing body of evidence supporting VMAT-TBI as a safe, effective, and feasible method for delivering conditioning regimens prior to HSCT and demonstrate that it can be successfully implemented even in centers with resource constraints. However, our experience also highlights some of the practical challenges associated with establishing VMAT-TBI as a service.

Disclosures

None.