Search

High resolution agarose isoelectric focusing was used to compare the galactosyltransferase isoenzyme patterns of serum from 9 healthy controls with those from 38 patients with either breast, lung, ovarian, stomach or colonic cancer. At least 12 peaks of enzyme activity were found in every sample, the healthy controls having major forms with isoelectric points of 4.74, 4.87, 4.96, 5.16 and 5.23. Thirty patients (79%) had elevated levels of at least one isoenzyme and 23 (61%) had at least 3 isoenzymes elevated compared to only 10 (26%) patients who had elevated total serum galactosyltransferase activity. The isoenzymes which were most often elevated in the cancer patient group had isoelectric points of 4.93, 5.16 and 4.61. One isoenzyme with an isoelectric point of 4.43 was preferentially elevated in patients with ovarian cancer. Those isoenzymes containing little or no sialic acid were rarely elevated in cancer patients. Although no cancer-associated isoenzyme was detected the quantitative differences observed in the cancer patient group were striking.

A series of 32 cases in which crush preparations were used in addition to frozen sections for the rapid diagnosis of lesions of the central nervous system (CNS) is presented. The cytopathologic features in crush preparations of astrocytomas, glioblastomas multiforme and a pituitary adenoma are described. Excellent preservation of cellular detail was seen in the crush preparations. Frozen sections lacked cytologic detail but provided a better view of the tissue architecture. The crush preparations yielded the correct diagnosis in 29 of the 32 cases. In the other three, a secondary component of the neoplasms (oligodendroglioma and fibrosarcoma) was identified only in the paraffin sections. Use of both frozen sections and crush preparations is recommended for all cases in which an immediate diagnosis of a CNS lesion is required.

The serum level of galactosyltransferase was measured in a group of 218 patients with a variety of solid tumors and most with advanced disease. The pretreatment enzyme level showed little potential as a diagnostic tumor marker, and its change with treatment did not reflect the initial response. There was, however, a significant correlation between the length of survival and the pretreatment enzyme level. Patients with normal levels survived over twice as long as those with elevated levels. When Cox's proportional hazards regression analysis was used to compare the prognostic potential of galactosyltransferase with a number of known clinical indicators of prognosis, the variable most related to survival was performance status (P less than 10(-4) followed by galactosyltransferase (P = 0.01) and then the extent of disease (P = 0.03). The other variables, such as previous therapy, the type, site, and size of primary tumor, did not contribute significantly to the relationship with survival. The pretreatment level of galactosyltransferase is therefore a relatively independent prognosticator of survival and, as such, could be potentially useful in patient management by increasing the accuracy of the initial assessment of prognosis.