Assessing the Ompact of Neutral Pelvic Positioning on Bone Repositioning Quality in Pelvic Radiotherapy Patients

Radiotherapy (RT) is an integral part of curative treatment for pelvic cancers. Intensity-Modulated Radiotherapy (IMRT) has demonstrated benefits in target volume (TV) coverage and sparing healthy tissues. IMRT achieves a highly conformal dose distribution, with steep dose gradients between the TVs and Organs at Risk (OARs). However, its effectiveness depends on the quality of patient repositioning, ensured through positioning imaging, immobilization devices, and skin markers. Image-Guided Radiotherapy (IGRT) involves using imaging systems to verify the correct positioning of the patient, TVs, and OARs. IGRT has shown substantial benefits in enhancing treatment precision and reducing side effects. Translational adjustments can be applied before treatment, but, except for TomoTherapy®, which can correct roll, other rotations (pitch and yaw) require a 6D robotic table, a technology that is not widely available and limited to 3°.

The immobilization systems available and used in RT departments vary widely, classified into customizable and non-customizable systems. Two publications have shown that using a customizable immobilization system does not improve bone repositioning in these patients. Two teams have compared non-customizable immobilizations consisting of a knee support alone or combined with a foot support. The team of Cartal et al. (2015) found no significant difference in translations. However, using a knee support alone significantly reduced rotations over 3°. Laaksomaa et al. found no difference in performance between the two setups.

A pilot study in our department with 32 patients and nearly 700 images revealed an RX rotation greater than 3° in 19% of the images on average, ranging from 0% to 75% of the positioning images affected by an RX rotation over 3°. These patients were positioned supine with a knee support according to Cartal et al. Data, with a headrest or arm support added if lomboaortic irradiation was needed. Accessories were indexed, and their positions marked during the planning CT for reproducibility across sessions, with skin markers as additional setup aids.

Physical therapists recommend a maneuver called the "glute bridge," where patients bend their legs and lift their pelvis three times before straightening their legs without voluntary movement. For the study, the legs will rest on a reproducibly positioned knee support. This maneuver ensures a neutral pelvic position for each patient, relieving contractures in the lumbopelvic region. The "neutral zone" or "balance point theory," as defined by physiology and biomechanics, describes a state that relieves contractures in the lumbopelvic region and ensures balanced baseline tension across muscle groups. The pelvis lift creates co-contraction of the hamstrings, quadriceps, abdominals, and lumbar spinal muscles, promoting this balance.

In this randomized study, we propose evaluating the benefit of a neutral pelvic positioning setup achieved by the patient, in addition to immobilization devices and skin markers, on the quality of bone repositioning.

• The median, mean, and standard deviation of rotations (RX, RY, and RZ) obtained from the positioning imaging; [Time frame: At to 3 months]
• The median, mean, and standard deviation of translations (TX, TY, and TZ) obtained from the positioning imaging [Time frame: At to 3 months]
• The calculation of the dose delivered to nodal CTVs and organs at risk, recalculated by simulating the actual rotations observed during treatment [Time frame: At to 24 months]
• Gastrointestinal, urinary, and hematological toxicities graded according to NCI-CTCAE v5 [Time frame: Up to 24 months]
• Systematic error estimation calculated using the Van Herk et al. formula [Time frame: At to 24 months]
• Random error estimation calculated using the Van Herk et al. formula [Time frame: At to 24 months]
• Estimation of nodal CTV-PTV margins calculated using the Van Herk et al. formula [Time frame: At to 24 months]
• Evaluation of repositioning error in relation to L4, without applying rotations, will be obtained using the mean and standard deviation of measurements between the anterosuperior part of the L4 vertebral body on the reference image and on the daily image [Time frame: At to 3 months]
• Evaluation of the RX rotation discrepancy observed after registering the daily image to the reference image, comparing two regions of interest: one defined by L1-L5 and the other defined from the iliac crests to the ischial tuberosities [Time frame: At to 24 months]