Towards an adequate description of the dose-response relationship in BNCT of glioblastoma multiforme

Abstract

Background: Boron Neutron Capture Therapy (BNCT) is a binary radiotherapy based on the intravenous administration of a borated drug to the patient and the subsequent irradiation with a low-energy neutron beam. The borated formulation accumulates in the tumor cells, and when neutrons interact with boron, a nuclear capture reaction occurs, releasing high-linear energy transfer, short-range particles that cause lethal damage to the cancer cells. Due to its selectivity, BNCT has the potential to treat aggressive brain tumors such as glioblastoma multiforme (GBM), minimizing the side effects. GBM is a brain neoplasia that poses significant treatment challenges due to its invasiveness and resistance to conventional treatments.

Purpose: This work aims to find a suitable model for calculating the photon isoeffective dose for GBM, producing ad hoc radiobiological data to feed the model.

Methods: To describe adequately the dose-effect relation of BNCT for GBM, the following strategy has been applied 1.We studied the impact of choosing two different photon radiation types (x- or gamma- rays) 2.We assumed that the correct description of the photon-equivalent dose is obtained with the photon isoeffective dose model. This model calculates the photon dose that equals the cell survival obtained with BNCT, taking into account synergism and sub-lethal damage (SLD). 3.Survival curves as a function of the dose for the human GBM U87 cell line were constructed using the clonogenic assays for irradiation with photons (reference), neutron beam, and BNCT. 4.Survival curves were fitted with the modified linear quadratic model, using SLD repair times derived for U87. The radiobiological parameters were determined for the photon isoeffective dose model. 5.The model was applied to a clinical case that received BNCT in Taiwan. Treatment planning has been simulated using an accelerator-based designed neutron beam following the real treatment process and parameters. The results were discussed and compared to the current method, which employs relative biological effectiveness (RBE) factors to obtain BNCT dosimetry in photon-equivalent units.

Results: The dose-survival curves have been obtained with two different photon radiation sources as the reference with a thermal neutron beam and neutrons in the presence of boron. The fitted parameters have been obtained as the input for the photon isoeffective dose and the traditional RBE model. For the first time, the radiobiological parameters of a photon isoeffective dose model were produced for BNCT of GBM. Photon isoeffective dose value can differ up to 32% using gamma photons and low-energy x-rays. Photon isoeffective dose values are lower (17%) than the RBE model currently employed in clinical trials.

Conclusion: The results highlight the impact of the reference radiation chosen for the isoeffective dose calculation and the importance of feeding the model with the appropriate radiobiological parameters.The dosimetry obtained with the new radiobiological data is consistent with the dose delivered in modern stereotactic radiotherapy, enabling tumor control predictions.

Keywords: BNCT; cell survival curves; dosimetry; glioblastoma multiforme.

FIGURE 1

Model of the x‐ray irradiator set‐up. The sample into the tissue‐equivalent material (PMMA) is irradiated from the top and the bottom by two x‐ray beam of average energy 60–80 keV (arrows). Left: the central rectangle represents the calibration position where the detector is irradiated in CPE conditions. Right: The upper part of the phantom is removed and the flask is irradiated in the upper part of the drawer. CPE, charged particles equilibrium; PMMA, polymethylmethacrylate.

FIGURE 2

Left: vertical section of the facility for Co‐60 irradiation, right: horizontal section of the facility.

FIGURE 3

Kinetic repair time data (dots) from and bi‐exponential fit (solid line). The dash‐dot lines represent the upper and lower bounds of the 68% confidence interval. The dash‐dot orange lines represent the upper and lower bounds of the 68% confidence interval. The fit has a mean squared error (MSE) of 0.86.

FIGURE 4

Medical images of the patient (axial, coronal, sagittal) with the GTV ROI marked by the red polygon. GTV, gross tumor volume; ROI, region of interest.

FIGURE 5

Experimental data and fit of U87 cell survival as a function of the dose of photons emitted by 60‐Co (line) and x‐rays (dash line). The dash‐dot lines represent the confidence intervals of the parameters obtained from the fit of the two curves.

FIGURE 6

Survival curves of U87 cells as a function of the absorbed dose. Left: using as the reference radiation. Right: using x‐ray as the reference radiation. The blue dots and line represent the data from the reference radiation and the fit, respectively. The red stars and the solid curve are the experimental data and the fit for the beam‐only irradiation, respectively. The red dashed curve is the isolated contribution of protons (i.e. subtracting the effect of photons). The solid black curve shows the fit for the BPA‐BNCT experimental points (black triangles). The dotted black line is the model considering only the boron component (i.e., subtracting photon, proton and contributions). BNCT, Boron neutron capture therapy; BPA, Boronophenylalanine.

FIGURE 7

(a) DVH of the healthy brain; the black vertical line highlights the prescription limit of $2.5\mathrm{Gy}(\mathrm{RBE})$ or less to 50% of the brain volume. (b) DVH for the tumor (GTV) calculated using the photon isoeffective dose model with parameters derived from glioblastoma (Blue) and from GSM (Red). DVH, dose‐volume histogram; GSM, gliosarcoma; GTV, gross tumor volume.