S. Dionisi

BACKGROUND: A new commercially available (so-called second-generation) IRMA for parathyroid hormone (PTH) separately detects intact PTH and its N-truncated fragments; however, no studies have compared the first- and second-generation IRMAs for PTH in patients with primary hyperparathyroidism (PHPT) to assess their respective diagnostic accuracies. METHODS: We concomitantly investigated 39 postmenopausal patients with PHPT and a control group of 70 healthy postmenopausal women matched for age, renal function, and vitamin D status. In all individuals, PTH was measured with a classic IRMA (PTH-S; DiaSorin Inc.), which uses antibodies directed against epitopes 1-34 and 39-84, and a new method (Scantibodies Laboratory. Inc.), which uses antibodies against epitopes 1-4 and 39-84 (PTH-W) and epitopes 7-34 and 39-84 (PTH-T). We also assayed serum PTH in 10 PHPT patients every 24 h for 5 days after successful surgery. RESULTS: The different assays gave serum PTH values that were >2 SD higher than values for the control population in 59% (PTH-S), 77% (PTH-W), and 82% (PTH-T) of patients with PHPT. However, ROC curve analysis showed no significant differences among the three PTH assays, demonstrating overlapping diagnostic sensitivities. In PHPT patients, the correlation among the assays was highly significant (r = 0.91-0.92; P <0.001). The ratio PTH-W:PTH-T x 100 showed a gaussian distribution in both PHPT patients and controls, whose mean (SD) values [63.4 (13.3)% vs 64.5 (9.5)%, respectively] did not differ significantly. After parathyroidectomy, the mean percentages of variation in PTH detected with all of the assays were quite similar. CONCLUSIONS: The distribution of the PTH-W:PTH-T ratio in patients and controls suggests that PHPT does not markedly influence the rate at which biologically inactive fragments are generated by central or peripheral cleavage of PTH. The similar postoperative curves seem to contradict the hypothesized effect of acute hypocalcemia in modulating the central secretion of hormonal fragments. Our results indicate that the three investigated assays have similar diagnostic sensitivities in PHPT.

To the Editor: The report by Walsh et al. (Jan.24 issue) 1 of a trial of voriconazole, as compared with liposomal amphotericin B, for patients with neutropenia and persistent fever and the accompanying editorial and letter to the editor 2,3 leave many questions about what the trial actually demonstrated and how the findings should be applied.The abstract implies that the study showed equivalence and concludes, "Voriconazole is a suitable alternative to amphotericin B preparations."Yet the text and the accompanying editorial point out that the prespecified statistical criterion for equivalence was not met.To confuse the issue further, correspondents from the Food and Drug Administration provide alternative data, indicating that voriconazole was actually statistically inferior to liposomal amphotericin B with respect to overall success rates. 3Thus, whether the study showed equivalence, near equivalence, or frank inferiority remains unclear.Even if equivalence had been demonstrated, the clinical implications of the findings would remain uncertain.Therapeutic decision making must take into account not just efficacy and safety, as emphasized in the editorial, 2 but also cost.Avoidance of toxic effects may allow a net cost savings despite higher drug acquisition costs, 4 but that possibility requires empirical demonstration.