Naschitz and colleagues suggest that fasciitis-panniculitis would be a more appropriate term than fasciitis to qualify the lesions we described recently in graft-versus-host disease. Both terms refer to pathologic entities. As a pathologist, I insist that panniculitis, in the strict sense, applies to inflammation of fat lobes rather than to inflammation of interlobular septa or of large cutaneous vessels. The fetal development makes each fat lobe a separate organ with a nourishing artery and two draining veins, and this separation explains the sharp limitation of pannicular inflammation [1]. We did not find fat cell necrosis in any of our patients' biopsy specimens. The earliest histologic changes in our patients were indeed localized in the muscular fascia and the intermediate septa separating the fat lobules, but the changes did not involve fat cells. The initial change is not, therefore, a panniculitis.

Because one of the biopsy specimens was obtained only 17 days after the onset of the symptoms, the fascia was involved very early. Moreover, the different biopsy specimens, obtained 17 to 1015 days after the onset of the symptoms, showed that the fascia was constantly involved and that inflammation and fibrosis predominated within the fascia. Finally, although we agree that simultaneous involvement of the fascia and the subcutaneous septa can occur from the onset in fasciitis in patients with graft-versus-host disease, as in patients with eosinophilic fasciitis [2], we do not believe that "panniculitis" is the proper term for the associated inflammation and fibrosis limited to the subcutaneous septa.