Can Nonenhancing White Matter Lesions Be Disregarded?

In this issue of AJNR (1), Drs Elster and Chen pose a provocative question that has certainly given us pause as we routinely dismiss a large nonenhancing white matter lesion in a cancer patient. It was with relief that we read the conclusion that white matter lesions in cancer patients that do not enhance with gadolinium-DTPA at the time of the initial MR study have a low probability of representing metastatic disease. However, as the authors remark, a low probability does not mean zero and we feel our continuing paranoia is still justified, although reduced. Within the limits of clinical practice the authors have performed an excellent prospective study. The study would have been more reassuring if the authors had selected patients with cancers most likely to have cerebral metastases, and those most likely to have white matter disease. While over 50% of the 50 patients with nonenhancing white matter disease did have cancers with a predilection for CNS seeding (eg, lung, breast, lymphoma, head and neck), the remaining patients have primaries (GI, urinary tract, and reproductive system) less likely to result in metastatic brain disease . Ideally, only patients with focal neurologic deficits and seizures would have been included in the cohort. There are only eight of these patients, and there were six patients who were asymptomatic but referred for MR due to the highly aggressive nature of their primary malignancies with a propensity for cerebral metastases. The remaining 36 patients presented with nonfocal neurologic signs and symptoms that have low probabilities of being due to metastatic disease. Selecting for the elderly, ie, the over 60 age group, would have increased the incidence of nonmalignant white matter lesions. A white matter lesion, whether enhancing or not, is less likely to occur in a 23-year-old than in an 85-year-old, 23 to 85 years being the age range of patients in this study. More complete follow-up would also have been helpful. Only 13 of 30 patients who were alive without clinical evidence of metastatic disease a year after their initial MR exam had follow-up CT or MR studies. Of the 20 patients who died, only nine had follow-up imaging and no autopsy data are presented. We feel the conclusion this paper reaches is valid, but it would have been more compelling if all patients had been followed up and if autopsy data were available on the 20 who died, confirming that they had no metastases in the white matter disease. The authors used the conventional dose of gadolinium-DTPA (0.1 mmol/kg) and began imaging after a 5-10 minute delay. Recent work by Yuh et al (2) indicates that an additional dose of 0.2 mmol/kg at 30 minutes is more efficacious, inasmuch as 46 new lesions were detected in 19 of 27 patients. How many of these lesions were white matter foci that only enhanced with the higher dose of gadolinium-DTPA is not detailed but a case could be made for evaluating cancer patients in the future with higher doses of gadolinium-DTP A. In our practice we use additional imaging criteria to evaluate white matter disease beside enhancement. The authors only characterize the white matter lesions as being discrete without contrast enhancement. The anatomic location of the lesions and their signal characteristics are not described. We must admit we have a lower suspicion for lesions closer to the ependyma and have a proportionately higher suspicion for lesions that involve the subcortical U fibers. We find a large lesion in the subcortical U fibers hard to dismiss even if it does not enhance-but this is the whole point of this paper. We assume that the white matter lesions in the 50 patients ending up in the cohort were nonspecific and of the usual