CHICAGO – Radiation therapy for cervical cancer resulted in pelvic insufficiency fractures more frequently than previously thought, and many fractures were slow to heal, according to research presented at the annual meeting of the Radiological Society of North America.
“Pelvic insufficiency fractures had a prevalence of 38% on MRI follow-up” after chemoradiotherapy for locally advanced cervical cancer, said Alina Dragan, MD. This figure is more than double the previously reported prevalence of about 14%.
Dr. Dragan, a radiology resident at London North West Healthcare, National Health Service Trust, and coinvestigators also tracked the natural history of these fractures over time, to fill a knowledge gap about whether, and at what rate, these pelvic insufficiency fractures healed.
In the single-center retrospective study, the investigators found that just 14% of sacral fractures healed during the period of observation. For acetabular and pubic fractures, roughly one in three fractures had healed by the last MRI scan. About a third of all fractures remained stable across scans, while just over 10% of fractures were either fluctuant or worsened.
The study included 115 women with locally advanced cervical cancer who were treated with radical or adjuvant concurrent chemoradiotherapy over a 5-year period, and had MRI scans performed in-house; the follow-up protocol had patients receiving scans at 3, 12, and 24 months post treatment. From an initial pool of 197 patients, those who had previously had pelvic radiation or were receiving palliative treatment, as well as those with incomplete imaging follow-up and those with metal implants or prostheses that could affect radiation therapy delivery or imaging quality were excluded.
The chemoradiotherapy protocol involved five doses of weekly cisplatin at 400 mg/m2 of body surface area, as well as high–dose rate cervix brachytherapy. In practice, all but six participants received these treatments. Patients also received external beam radiotherapy with or without a simultaneous integrated boost to target affected lymph nodes, as clinically indicated.
The fractures were graded as mild, moderate, or severe by the interpreting radiologist according to the course of the fracture line and corresponding bone edema.
Patients were aged a median of 54 years, and 64 (56%) were postmenopausal. Most patients (n = 84; 73%) had never used tobacco. Participants’ median body mass index was 26 kg/m2.
Most patients (n = 73; 64%) were International Federation of Gynecology and Obstetrics stage 2b, and almost half (n = 55; 48%) had pelvic nodal involvement.
Patients were followed for a median of 12 months, with patients receiving a median of two MRIs curing that period. In all, 105 fractures were identified in 44 patients. A median of two fractures were identified among the group of patients who had pelvic insufficiency fractures.
The fractures were graded as mild, moderate, or severe by the interpreting radiologist according to the course of the fracture line and corresponding bone edema. In this schema, 41% of identified fractures were considered mild, while 32% were moderate and 12% were severe.
Although just over two-thirds of fractures (70%) were identified within 6 months of beginning surveillance, a quarter were not identified until 9-13 months post therapy, and 5% were found after more than 13 months.
Sacral fractures accounted for 72% of those identified, in keeping with previous findings, said Dr. Dragan. Acetabular and pubic fractures made up 16% and 10% of fractures, respectively. One fracture was seen at the ilium and one at the ischium.
Dr. Dragan and colleagues turned to multivariable analysis to look for risk factors for pelvic insufficiency fractures in this cohort of cervical cancer patients. Younger patients had a hazard ratio of 0.30 for fracture, compared with those over the age of 50 years (P less than .01). Similarly, being menopausal carried a hazard ratio of 2.25 for fracture. Higher radiation doses to the sacrum also boosted fracture risk (HR, 2.00; P = .03). Neither sacral volume and slope nor the receipt of simultaneous integrated boost were associated with increased fracture risk.
Dr. Dragan reported that she had no relevant conflicts of interest. She reported no outside sources of funding.
SOURCE: Dragan A et al. RSNA 2019, Presentation SSE25-03.