A single institution retrospective study of pelvic insufficiency fractures following curative-intent pelvic intensity-modulated radiation therapy for gynecologic, gastrointestinal, and genitourinary cancers
Menée à partir de données portant sur 658 patients atteints d'un cancer gynécologique, génito-urinaire ou gastrointestinal, cette étude rétrospective analyse l'association entre une radiothérapie pelvienne avec modulation d'intensité à visée curative et le risque de fracture de la région pelvienne
Purpose: We aimed to determine the incidence of and explore risk factors for pelvic insufficiency fractures (PIFs) among patients that received curative-intent pelvic intensity-modulated radiation therapy (IMRT) for gynecologic (GYN), gastrointestinal (GI), and genitourinary (GU) cancers.
Methods and materials: We audited records of adult patients with GYN, GI, and GU cancers who received neoadjuvant, definitive, or adjuvant pelvic IMRT from 2011 to 2015 with a treatment volume that included the primary tumor or post-operative bed and at least one pelvic lymph node region, and a prescription dose of at least 40 Gy (if conventionally fractionated at 1.8–2 Gy per fraction) or a prescription dose of 25 Gy in 5 fractions for patients with rectal cancer treated with hypo-fractionated neoadjuvant short course IMRT. All baseline and follow-up pelvic imaging reports were reviewed to identify PIF diagnoses. Demographic and relevant clinical, treatment, and dosimetric factors were analyzed to explore risk factors for PIFs.
Results: Among 658 audited patients, 46 (7%) developed 86 PIFs. The incidence for GYN, GI, and GU patients was 8% (13/159), 11% (30/276), and 1% (3/223), respectively, with anal and endometrial subsites having the highest incidence at 14% each. IMRT was delivered as neoadjuvant therapy for 16/46 (35%), definitive therapy for 22/46 (48%), and adjuvant therapy for 8/46 (17%). The median time to PIF diagnosis was 14 months (range 3–53 months), and 26/46 (57%) were symptomatic at diagnosis. The most common PIF location was the sacrum (67/86 [78%]). Multivariable logistic regression analysis found female gender (odds ratio [OR] 2.74, 95% CI 1.36–5.80; p = 0.006), osteoporosis (OR 6.91, 95% CI 2.43–18.8; p = 0.0002), and a dose fractionation of 25 Gy in 5 fractions (compared to schedules of > 25 to < 50 Gy (OR 5.34, 95% CI 1.87–15.6; p = 0.0019) or ≥ 50 Gy (OR 9.53, 95% CI 1.82–50.1; p = 0.0077)) to be significant PIF risk factors.
Conclusions: In our cohort, PIFs were a common complication for patients with GYN and GI cancers, but not GU cancers, in the IMRT era. Most PIFs occurred within 2 years of treatment, and most occurred in the sacrum. Female patients and patients with osteoporosis appeared to be at higher risk. Prospective studies using validated PIF diagnostic criteria should further examine the relationships between PIFs and dosimetric variables, including hypo-fractionated regimens such as 25 Gy in 5 fractions.
Supportive Care in Cancer , résumé, 2025