Use of the Pastorex aspergillus antigen latex agglutination test for the diagnosis of invasive aspergillosis.

AIMS To evaluate the Pastorex aspergillus antigen latex agglutination test for the diagnosis of invasive aspergillosis in patients undergoing liver or bone marrow transplantation. METHODS Serum samples were taken at least twice weekly post-transplant and tested for Aspergillus antigen. Latex agglutination test results were compared with microbiological examination of respiratory, urine and bile specimens. Serum samples from liver transplant patients were also tested for antibodies to Aspergillus fumigatus by counter immunoelectrophoresis. RESULTS Eight of the 91 patients studied developed invasive aspergillosis. Positive latex agglutination tests were obtained in eight of 187 (4.3%) serum samples from four of these eight patients. The other four patients with invasive aspergillosis gave consistently negative latex agglutination tests. A positive latex agglutination test was the first indication of invasive aspergillosis in two patients; these patients were already on amphotericin B. Positive latex agglutination tests were the only evidence of invasive aspergillosis in one patient who subsequently died of the infection. False positive latex agglutination tests were obtained in five of 83 (6%) patients with no evidence of invasive aspergillosis and misleading positive cultures seen in nine of 83 (10.8%). No antibodies were detected in three of four liver transplant patients with invasive aspergillosis. Conversely, antibodies were detected in 63 of 262 (24%) serum samples from 43 liver transplant patients with no evidence of invasive aspergillosis; one of these patients had an antibody titre of 1:2 on four separate occasions. CONCLUSIONS The Pastorex aspergillus antigen latex agglutination test, when used alone, lacks sensitivity and specificity for the early diagnosis of invasive aspergillosis. A diagnosis was made in all patients with invasive aspergillosis when both culture and antigen tests were performed although using these criteria a false positive diagnosis would have been made in 13 of 83 (15.6%) patients. Microbiological and serial serological investigations for antigen should both be performed and the results considered in conjunction with radiological and clinical evidence.