Radiation-induced plexopathy and fibrosis. Is magnetic resonance imaging the adequate diagnostic tool?

PURPOSE: To investigate magnetic resonance imaging (MRI) features of radiation-induced plexopathy (RIP) and radiation-induced fibrosis frequently associated with RIP.

PATIENTS AND

METHODS

Seven patients with late radiation sequelae in the supraclavicular region were examined with MRI after a median interval of 7 years (range, 5-18 years) following radiotherapy and 4-7 years after the onset of RIP. Four patients had RIP plus severe soft-tissue fibrosis, two RIP without soft-tissue fibrosis (n = 2/6), and one patient fibrosis without RIP. Patients underwent surgery of breast cancer (n = 6) or chest wall relapse (n = 1) and radiotherapy to the supraclavicular fossa with cobalt with an anterior portal in fractions of 1.7-2.6 Gy to 43-51.6 Gy in 3 cm depth. All patients were relapse-free at the time of MRI. Fibrosis and RIP were scored clinically (RTOG classification). Fibrosis of the supraclavicular and/or axillary region was marked in three and mild in two patients. RIP was mild, marked and severe in two patients each. MRI was performed with a 1.5-T unit including coronal STIR, coronal and transversal T2-weighted, transversal T1-weighted and fat-saturated post-contrast (gadolinium-DTPA) spin echo sequences.

RESULTS

The brachial plexus appeared normal in all patients, but subtle changes of adjoining tissue (slight, linear signal intensity in T2-weighted images or contrast enhancement surrounding the plexus) were detected in patients with RIP (n = 4/6) and the patient without RIP (n = 1). However, alterations of the soft tissue (marked signal intensity in T2-weighted sequences) correlated well with the clinical degree of fibrosis and were restricted to areas of marked to severe fibrosis (n = 3/3).

CONCLUSION

Reliable MRI signs of RIP could not be identified. The severity of fibrosis closely corresponded to MRI features. The role of MRI in the diagnostic work-up of RIP is, therefore, the exclusion of tumor relapse.