Patients and specific characteristics

The study included 25 female patients with non-metastatic breast cancer who underwent adjuvant radiotherapy.

The mean age at treatment was 47 years (range: 35–62 years). Sixteen patients were classified as premenopausal and nine as postmenopausal according to available clinical records. Perimenopausal status was not specifically documented, and therefore, patients were categorized based on recorded menopausal status. The tumour was located in the right breast in 15 patients and in the left breast in 10 patients.

Regarding tumour stage, 19 patients presented with stage T1 disease and 6 patients with stage T2; all patients were node-negative (N0). The predominant histological subtype was invasive breast carcinoma, which accounted for 22 cases. Additionally, there was one case of ductal carcinoma in situ and two cases of mucinous adenocarcinoma.

On a molecular level, most tumours were positive for hormone receptors (ER and PR) and negative for HER. The Ki-67 proliferation index ranged from 1% to 59%. Molecular subtype distribution included 14 Luminal A, 5 Luminal B, 3 HER2-positive, and 3 triple-negative breast cancer cases.

Finally, 10 patients received adjuvant chemotherapy, whereas the remaining 15 patients were treated with radiotherapy alone. The molecular and clinical characteristics of the patients are summarized in Table 1.

Table 1.

Clinical, tumour, and molecular characteristics of the 25 patients included in the study

Patient positioning and computed tomography acquisition

Patients were positioned in the supine position on an inclined breast board specifically designed for breast cancer treatment. This setup elevates the thorax, thereby optimizing exposure of the breast region. A knee support was used to enhance patient comfort and to minimize involuntary movements, both during computed tomography (CT) image acquisition and throughout the course of radiotherapy. These immobilization devices, integrated into the standardized protocol of the International Oncology Center, ensure optimal reproducibility of patient positioning, which is essential for the accuracy of highly conformal techniques such as IMRT and VMAT.

CT simulation was performed using a multi-slice CT scanner (General Electric, GE), enabling the acquisition of thin axial slices to achieve precise anatomical reconstruction and reliable dose calculation. This scanner is equipped with a wide bore, which facilitates patient positioning in treatment conditions, and a flat carbon fibre couch identical to that of the linear accelerator, thereby ensuring strict reproducibility between the simulation and treatment phases.

The laser positioning system, identical to that of the linear accelerator, includes two lateral transverse lasers (right and left), two frontal lasers and a sagittal laser. This configuration guarantees optimal geometric consistency between simulation and treatment and contributes to the accurate delineation of target volumes and OARs during the planning process. Such precise alignment helps reduce setup uncertainties and ensures reproducibility throughout the treatment course. This consistency is particularly important in advanced techniques such as IMRT and VMAT, where even small deviations may affect dose distribution.

Target volume delineation and dose constraints

A radiation oncologist specialized in breast irradiation performed meticulous delineation of the planning target volume (PTV) and OARs, including the ipsilateral, contralateral and bilateral lungs, the whole heart, the contralateral breast, the spinal cord and the skin. Target and OAR delineation followed standardized international contouring recommendations. The spinal cord was contoured from the cervical to thoracic region according to RTOG contouring guidelines to allow evaluation of dose constraints.

The gross tumour volume (GTV) corresponds to any visible residual tumour, when present. The clinical target volume (CTV) encompasses the operated breast gland, with or without regional lymphatic areas (axillary, supraclavicular, etc.), depending on the clinical characteristics and therapeutic recommendations. The PTV is defined by adding appropriate margins around the CTV to account for uncertainties related to respiratory motion, setup variations between fractions and the inherent limitations of immobilization systems (breast board, knee support, arms raised, etc.). ^12,13

Subsequently, the radiation oncologist established specific dosimetric constraints for the OARs detailed in Table 2, to ensure treatment safety while maintaining therapeutic efficacy. This anatomoclinical approach allows precise targeting of the tumour region while optimally sparing surrounding healthy structures and ensuring dose homogeneity. Such accuracy is made possible through the use of highly conformal treatment techniques such as intensity-IMRT and VMAT.

Table 2.

Clinical goals for organs at risk (OARs) ^{Reference Balaji21,Reference Noël, Antoni, Barthelemy and Chauvet22}

Treatment planning

In the Eclipse treatment planning system, a virtual bolus is used to optimize the dose distribution in superficial regions such as the chest wall. This structure mimics tissue on the skin surface, compensating for dose loss due to respiratory motion and ensuring a homogeneous target coverage, particularly in cases of tumour or chest wall displacement. Considered as part of the patient’s anatomy, this virtual bolus directly influences dose calculation and plan optimization. Using the Acuros 18.1 algorithm, which provides advanced heterogeneity corrections and highly accurate dose computations in complex geometries, further enhances the precision of dose calculations in the presence of a virtual bolus.

Modern techniques such as IMRT and VMAT allow for precise modulation of the dose.

For each patient, both IMRT and VMAT plans were generated under identical planning objectives and constraints, allowing intra-patient comparison and minimizing anatomical variability.

IMRT is based on the use of four static fields (dynamic MLC modulation) for both left and right breast cases. Gantry and collimator angles are adjusted to optimize dose distribution while minimizing exposure to OARs. For left-sided breast IMRT, gantry angles typically range from 125° to 325°, whereas for right-sided cases they extend from 35° to 235°, with adapted collimator angles (10° or 350°).

VMAT, on the other hand, employs partial arcs with simultaneous modulation of beam intensity, gantry rotation speed and MLC position. Four arcs are used for each breast, with specific configurations: for left-sided cases, arcs range from 300° to 179° and from 179° to 300°, with collimator angles set at 0° and 90°; a similar configuration is applied for the right breast, but with arcs ranging from 181° to 60° and from 60° to 181°. This technique enables fast and homogeneous dose delivery, optimizing both treatment time and organ-at-risk sparing. Figure 1 illustrates the beam arrangements for IMRT and VMAT techniques.

Figure 1.

Beam arrangements for IMRT and VMAT treatment techniques.

Evaluation parameters

The treatment plans were evaluated based on dosimetric parameters, including the dose homogeneity index (DHI) and the conformity index (CI) for the PTV. Dosimetric parameters such as monitor units (MUs), minimum dose (D2%) and maximum dose (D98%) delivered to the PTV were recorded and analysed using dose–volume histograms (DVHs). For the OARs, the volumes receiving 5Gy, 20Gy and 30Gy (V5Gy, V20Gy and V30Gy) were extracted and documented for each structure. In addition, the normal tissue complication probability (NTCP) was calculated for the lungs and whole heart, the secondary cancer complication probability (SCCP) was estimated for the contralateral breast and both lungs, and TCP was calculated for the PTV for each treatment plan.

The DHI ^{Reference Sylla14} and CI ¹⁵ indices were calculated according to the following equations:

(1) $DHI = {{{D_{{\rm{max}}}} - {D_{{\rm{min}}}}} \over {{D_{{\rm{mean}}}}}}$

(2) $CI = {{{V_{PTV}}} \over {{V_{RI}}}}$

The lower the DHI, the more homogeneous the dose distribution within the target volume. Similarly, the closer the CI is to 1, the more accurately the dose distribution conforms to the shape of the target volume, thus reflecting a high-quality treatment plan. In this context, VRI refers to the volume receiving the prescription dose (typically corresponding to the 95% isodose line), while VPTV represents the total volume of the PTV. The Lyman–Kutcher–Burman (LKB) model ^{Reference Purdy16,Reference Landberg, Chavaudra, Dobbs, Gerard, Hanks, Horiot, Johansson, Möller, Purdy, Suntharalingam and Svensson17} was used to calculate the NTCP for both the ipsilateral and contralateral lungs.

The definitions and parameter values used in this model are presented in Table 3. The equations used are as follow:

(3) $NTCP = {1 \over {\sqrt {2\pi } }}\int_\infty ^t {{e^{ - {{{x^2}} \over 2}}}dx}$

(4) $t = {{EUD - T{D_{50}}} \over {m \cdot T{D_{50}}}}$

(5) $EUD = {\left( {\sum\limits_i {D_i^{{1 \over n}}} \cdot {{{v_i}} \over {{v_{{\rm{tot}}}}}}} \right)^n}$

Table 3.

Definitions and parameter values in the LKB model ^{Reference Alqahtani, Alshammari and Alhussain20}

The relative seriality model ^{Reference Aldosary, Alnajjar, Alhussain and Alshammari19} was used to calculate the NTCP for the whole heart. The definitions and parameter values used in this model are summarized in Table 4. The equations applied are as follows:

(6) $NTCP = {\left\{ {1 - \mathop {\mathop \prod \limits_i }\limits^n {{\left( {1 - P{{({D_i})}^s}} \right)}^{{v_i}}}} \right\}^{{1 \over s}}}$

(7) $P\left( {{D_i}} \right) = {2^{ - {e^{\gamma \left( {1 - {{{D_i}} \over {{D_{50}}}}} \right)}}}}$

Table 4.

Parameter definitions for NTCP calculation in relative seriality model ^{Reference Alqahtani, Alshammari and Alhussain20}

The Schneider model was used to calculate the SCCP for the contralateral breast and both lungs. ^{Reference Landberg, Chavaudra, Dobbs, Gerard, Hanks, Horiot, Johansson, Möller, Purdy, Suntharalingam and Svensson17,18} The definitions and parameter values used in this model are presented in Table 5. The equations applied are as follows:

(8) $OE{D_{{\it{org}}}} = {1 \over {{V_T}}}\sum\limits_i {\left( {{v_i} \cdot {{1 - {e^{ - {\delta _{{\it{org}}}}{D_i}}}} \over {{\delta _{{\it{org}}}}}}} \right)}$

(9) $SCCP = {I_{{\it{org}}}} \cdot OE{D_{{\it{org}}}}$

Table 5.

Secondary cancer risk parameter in the relative seriality model framework ^{Reference Alqahtani, Alshammari and Alhussain20}

The LQ-Poisson model was used to calculate TCP for the PTV. The definitions and parameter values used in this model are summarized in Table 6. The equation applied is as follows:

(10) $TCP = exp\left\{ { - {N_0} \cdot exp\left[ { - n \cdot \left( {\alpha d + \beta {d^2}} \right)} \right]} \right\}$

Table 6.

Definitions and parameter values in the LQ-Poisson model ^{Reference Alqahtani, Alshammari and Alhussain20}

The DVHs were converted from their cumulative form to differential histograms in order to enable a more detailed analysis of the local dose distribution within the target structures and OARs.

Although NTCP, TCP and SCCP models provide useful estimations, they remain predictive mathematical models and may not fully represent clinical outcomes. Long-term clinical follow-up would be necessary to validate these predictions.

Statistical analysis

Dose differences between the IMRT and VMAT techniques were analysed using Microsoft Excel and MATLAB, taking into account the paired nature of the data. The normality of the differences was assessed using the Shapiro–Wilk test, which indicated a non-normal distribution (p < 0.05). Consequently, the non-parametric Wilcoxon signed-rank test was applied. A p-value of less than 0.05 was considered statistically significant.

Ethics statement

This retrospective planning study was conducted in accordance with institutional ethical standards and approved by the local ethics committee. Patient data were anonymized.