The research team, based primarily at the Istituto Nazionale Tumori IRCCS Fondazione G. Pascale in Naples, Italy, retrospectively replanned 22 consecutive adult patients who had originally received external beam radiotherapy for bulky tumors between April 2019 and December 2025. The cohort spanned a deliberately heterogeneous clinical spectrum: soft tissue sarcomas of the extremities and abdomen or pelvis, mediastinal and thoracic carcinomas, gynecological malignancies, and one melanoma. More than half of the cases showed local tissue infiltration, and nearly two-thirds carried high-grade histology. Treatment intent split evenly between neoadjuvant and palliative settings, with palliative courses ranging from 20 to 30 Gy in 4 to 10 fractions down to a single 8 Gy fraction for acute bleeding control, while neoadjuvant patients uniformly received standard 50 Gy in 25 fractions.

To make the paired comparison meaningful, every case was replanned twice using an identical lattice design: a staggered three-dimensional triangular grid with a uniform center-to-center vertex spacing of 5 centimeters, each vertex modeled as a 1-centimeter-diameter sphere. Vertices were confined within an inner volume produced by eroding the gross tumor volume by 2 centimeters, guaranteeing a safety margin between the ablative hot spots and the tumor surface. For the dosimetric test itself, both techniques received the same prescription: 21 Gy in three fractions to the vertex target volume, with peaks allowed up to 120 percent of prescription. The same planner generated all plans from standardized templates, and two experienced medical physicists independently reviewed each one, limiting inter-planner variability.

The central metric of the study was the valley-to-peak dose ratio, or VPDR — the ratio of the dose deposited in the cool inter-vertex valleys to the ablative dose in the peaks. Lower VPDR values mean steeper gradients and cooler valleys, which is the dosimetric goal of LRT. Because the published literature computes VPDR in wildly different ways, the authors applied three distinct methods to identical data: a pairwise midpoint approach described by Prado and colleagues, a dose-volume-histogram-based D90/D10 ratio from Borzov and colleagues, and Wu’s method using the mean dose in the coldest 5 percent of the lattice volume relative to the prescription dose. All three methods told the same story: helical tomotherapy achieved significantly lower VPDR than VMAT, with large effect sizes. Median VPDR by Wu’s method was 0.29 for HT versus 0.37 for VMAT; by the D90/D10 method, 0.31 versus 0.42; and by the pairwise median method, 0.46 versus 0.54, all with p values below 0.001.

The most methodologically interesting contribution, however, goes beyond the scalar numbers. In any staggered three-dimensional lattice, inter-vertex distances differ systematically by direction: pairs within the same transverse plane sit at different distances than pairs connecting adjacent planes, and the resulting valley doses differ accordingly. Rather than averaging this anisotropy away, the team decomposed VPDR into six directional categories — the three principal anatomical axes, the intra-plane radial direction, and the coronal and sagittal diagonals. The result was striking: the helical tomotherapy advantage was direction-selective. HT produced significantly lower VPDR in the radial (0.43 versus 0.59), superior-inferior (0.47 versus 0.53), left-right (0.44 versus 0.56), and anterior-posterior (0.46 versus 0.56) directions, but showed no significant difference from VMAT along either diagonal. Scalar summary metrics, the authors argue, obscure this clinically relevant directional structure entirely.

Even more intriguing was the within-technique analysis. VMAT showed its greatest dosimetric weakness in the radial direction, where valley doses ran significantly warmer than along the diagonals. Helical tomotherapy inverted that hierarchy: its best performance was precisely in the radial direction, with its only significant directional limitation appearing along the sagittal diagonal. The geometry explains this. Continuous gantry rotation combined with couch translation gives helical delivery near-isotropic angular sampling, which excels at shaping gradients in the transverse plane, while coplanar arc arrangements are structurally weaker there. The direction in which one technique is most limited is exactly the direction in which the other performs best.

Dose-volume histogram endpoints reinforced the picture. Vertex coverage was statistically equivalent between techniques, with median D95% of 20.03 Gy for VMAT versus 20.23 Gy for HT, and no significant difference in GTV prescription-isodose coverage — meaning HT’s cooler valleys were not bought at the cost of underdosing the peaks. But in the inter-vertex tissue compartments, VMAT delivered substantially warmer valley doses: median D50% in the gross tumor volume was 11.03 Gy with VMAT versus 7.45 Gy with HT. Normal tissue sparing also favored HT at low dose levels, with only 2.29 percent of the body volume outside the target receiving at least 5 Gy versus 5.10 percent for VMAT, and near-maximum external dose of 6.03 versus 7.62 Gy. The advantage concentrated at the low end of the dose spectrum, consistent with the highly modulated rotational delivery of tomotherapy.

The study also connected dose modulation to tumor shape. Using principal component analysis on the segmented targets, the team derived aspect ratio, elongation, flatness, compactness, and Wadell sphericity for each tumor. Target elongation correlated positively with radial VPDR in both techniques — a physically coherent finding, because in elongated targets accommodating multiple vertex planes, radial valley points receive cumulative dose contributions from inter-plane vertices above and below them, a dose summation effect intrinsic to the staggered triangular geometry. Compactness influenced radial VPDR only in HT, and larger tumors achieved more favorable anterior-posterior VPDR with both techniques, though the authors flag those particular correlations as exploratory because the anterior-posterior direction was evaluable in only five cases.

Practical considerations temper the dosimetric enthusiasm. Helical tomotherapy required a median beam-on time of 961.5 seconds per fraction versus 389 seconds for VMAT — roughly 2.5 times longer — and a median of 13,813.8 monitor units per fraction versus 1,969.9, a sevenfold difference. Longer delivery raises concerns for intrafraction motion management, particularly since previous work suggests lattice dose distributions degrade once motion exceeds about half a centimeter. The authors also acknowledge that their conservative lattice geometry produced very low vertex filling fractions of 0.1 to 0.3 percent, below the 1 to 2 percent typically reported, and that relatively warm valley doses — with the 50 percent isodose bridging adjacent vertices in some cases — partly reflect deliberate planning choices to enforce uniform vertex coverage and a spatially uniform valley floor. Dedicated VMAT optimization with tighter valley objectives, vertex-specific ring structures, or non-coplanar arcs could plausibly narrow the gap, and the different treatment planning systems used for each technique mean optimizer performance cannot be fully separated from delivery technique.

The study’s limitations are clearly stated: the retrospective paired design permits no randomization or clinical endpoint collection, the results concern dosimetry rather than tumor response or immune-mediated effects, and the findings may not transfer to denser lattice configurations. Still, the work marks a meaningful step toward standardization in a field where VPDR definitions, lattice geometries, and reporting conventions vary so widely that cross-study comparisons risk being misleading. By applying a uniform planning template and a multi-method, directionally resolved analytical framework to a paired patient cohort, the STRYKE investigators have translated phantom-based observations into clinical reality and made a case that helical tomotherapy deserves a place in the lattice radiotherapy toolkit — especially for palliative patients, for whom compressing treatment time while maximizing tumor kill is the overriding priority. Prospective studies with clinical outcomes will ultimately determine whether the dosimetric advantage translates into better care.

Subject of Research: Paired dosimetric comparison of helical tomotherapy versus VMAT for lattice radiotherapy in bulky tumors

Article Title: Dosimetric feasibility of lattice radiotherapy for bulky tumors: a paired VMAT vs. helical tomotherapy comparison – the STRYKE trial

Article References: R, M., M, S., D, P., S, C., V, B., E, S., G, S., G, D. P., G, A., F, B., C, A., V, D., MA, D. F., S, M., A, C., E, C., E, M., C, D., F, C., … R, D. F. (2026). Dosimetric feasibility of lattice radiotherapy for bulky tumors: a paired VMAT vs. helical tomotherapy comparison – the STRYKE trial. Frontiers in Oncology, 16, Article 1986573. https://doi.org/10.3389/fonc.2026.1986573

Image Credits: AI Generated

DOI: 10.3389/fonc.2026.1986573

Keywords: lattice radiotherapy, spatially fractionated radiation therapy, helical tomotherapy, VMAT, bulky tumors, dosimetry, valley-to-peak dose ratio, STRYKE trial, radiation oncology, palliative radiotherapy, dose modulation, treatment planning