Clinical Investigation
Inter- and Intrafractional Tumor and Organ Movement in Patients With Cervical Cancer Undergoing Radiotherapy: A Cinematic-MRI Point-of-Interest Study

https://doi.org/10.1016/j.ijrobp.2007.08.055Get rights and content

Purpose

Internal tumor and organ movement is important when considering intensity-modulated radiotherapy for patients with cancer of the cervix because of the tight margins and steep dose gradients. In this study, the internal movement of the tumor, cervix, and uterus were examined using serial cinematic magnetic resonance imaging scans and point-of-interest analysis.

Methods and Materials

Twenty patients with Stage IB-IVA cervical cancer underwent pelvic magnetic resonance imaging before treatment and then weekly during external beam radiotherapy. In each 30-min session, sequential T2-sagittal magnetic resonance imaging scans were obtained. The points of interest (cervical os, uterine canal, and uterine fundus) were traced on each image frame, allowing the craniocaudal and anteroposterior displacements to be measured. The mean displacements and trends were analyzed using mixed linear models. Prediction intervals were calculated to determine the internal target margins.

Results

Large interscan motion was found for all three points of interest that was only partially explained by the variations in bladder and rectal filling. The intrascan motion was much smaller. Both inter- and intrascan motion was greatest at the fundus of the uterus, less along the canal, and least at the cervical os. The isotropic internal target margins required to encompass 90% of the interscan motion were 4 cm at the fundus and 1.5 cm at the os. In contrast, smaller margins of 1 cm and 0.45 cm, respectively, were adequate to encompass the intrascan motion alone.

Conclusion

Daily soft-tissue imaging with correction for interfractional motion or adaptive replanning will be important if the benefits of intensity-modulated radiotherapy are to be maximized in women with cervical cancer.

Introduction

The development of high-precision radiotherapy (RT), such as intensity-modulated RT (IMRT), during the past decade offers an opportunity to escalate to a greater tumoricidal radiation dose while minimizing treatment-related toxicity. This has been best exemplified with stereotactic radiotherapy for intracranial tumors for which immobilization of the patient's head to the treatment couch allows for accurate radiation delivery to a relatively immobile target (1). However, in extracranial sites, tumor and organ mobility need to be considered to maximize the benefit of IMRT, given the tight margins and steep dose gradients that are characteristic of this approach (2). Setup variability and tumor and organ motion have been studied most extensively in prostate cancer 3, 4, 5, 6 and lung cancer 7, 8, 9 patients.

Interest is increasing in the use of IMRT to treat cervical cancer. High-precision techniques have been described that encompass the primary tumor and regional lymph nodes, with relative sparing of normal tissues such as small bowel and bone marrow 10, 11, 12, 13. In addition, IMRT has been advocated as a boost to the primary tumor alone in patients unable to undergo intracavitary brachytherapy (14). However, little is currently known about internal target movement in these patients 15, 16, 17, 18, 19. The goal of this study was to objectively describe the internal inter- and intrafractional tumor and organ movement during a course of fractionated external beam RT for cervical cancer, using surrogate inter- and intrascan parameters derived from weekly cinematic magnetic resonance imaging (cine-MRI).

Section snippets

Study design and patient eligibility

Twenty-three patients with biopsy-proven cervical cancer undergoing radical RT with or without chemotherapy participated in this trial. The local ethics review board approved the study. Three patients withdrew because of claustrophobia, leaving 20 eligible patients. Their characteristics are summarized in Table 1. Each was asked to undergo a baseline MRI scan before beginning external beam RT and then weekly scans for 5 weeks during external beam RT.

Scanning protocol

The same MRI protocol was used for each of

POI reproducibility

The intraobserver reproducibility of the POI tracking method was evaluated by comparing all possible differences among the three sets of repeated data (1 vs. 2, 1 vs. 3, and 2 vs. 3). The mean differences ranged from −0.66 to 0.25 mm. The differences in the CC localization of the promontory point and in the AP localization of the fundus point were statistically significant, but neither was deemed to be clinically significant (−0.25 and −0.35 mm, respectively). Thus, the POI method appeared to

Discussion

Knowledge of interfractional and intrafractional organ movement is critical to the success of high-precision conformal RT. Improved external immobilization and laser-assisted setup have reduced day-to-day variability in patient positioning to the point at which internal tumor and organ movement and deformation have emerged as major contributors to the uncertainties of radiation delivery 23, 24, 25, 26. Our results have provided a comprehensive description of the internal movement of the cervix

Conclusion

The results of this study have provided evidence that the tumor, cervix, and uterus may move substantially during the course of fractionated RT in women with cervical cancer. This is of minimal consequence when large conventional pelvic fields are used. However, it has the potential to contribute to geographic miss and treatment failure when high-precision pelvic RT is used to treat the primary tumor and regional lymph nodes or as a boost in patients who cannot undergo brachytherapy. Daily

Acknowledgments

The authors gratefully knowledge the important contributions of Anna Kirilova, Ami Syed, Michael Sharpe, and Graham Wilson.

References (31)

Cited by (0)

Supported by the Giovanni and Concetta Guglietti Family Trust.

Conflict of interest: none.

View full text