Can nonenhancing white matter lesions in cancer patients be disregarded?

PURPOSE To assess the risks and implications of assuming that white matter lesions in cancer patients that do not enhance with gadopentetate dimeglumine (Gd-DTPA) can be considered to be benign.

METHODS Gd-DTPA was administered prospectively to 131 consecutive patients with biopsy-proved extracranial malignancies referred for cranial MR imaging to exclude cerebral metastases over a 21/2-year period. From this initial group, 50 patients were identified who had focal nonenhancing lesions of the white matter on T2-weighted images, but no other findings to suggest metastatic disease. This cohort of 50 patients was then followed for at least 1 year to determine the risk and clinical implications of assuming these nonenhancing white matter lesions were benign.

RESULTS Thirty patients (60%) were alive and clinically free of cranial metastatic disease at least 1 year following their initial MR study (median follow-up time, 17 months). Twenty of the 50 patients (40%) died within 1 year of their study (median survival, 4.1 months). Review of clinic notes and hospital charts revealed no evidence for deterioration of neurologic status in any of these patients before death, and the cause of death in each case was ascribed to extracranial complications of their systemic malignancies. Eight of these 20 patients who expired had at least one follow-up cranial CT or MR scan before death showing no new cerebral metastases or change in the nonenhancing white matter lesions previously identified. In a single patient, however, follow-up MR scan revealed conversion of one of her several white matter lesions from nonenhancing to enhancing without appreciable change in its size on T2-weighted images. Unfortunately, this patient died 4 months later from surgical complications without interval change in her neurologic status nor pathologic proof of the nature of this lesion.

CONCLUSIONS White matter lesions in cancer patients that do not enhance with Gd-DTPA at the time of the initial MR study have a low probability of representing metastatic disease. Clinical management or final outcome will not likely be altered by assuming such lesions are benign.