A multi-institutional study was conducted for prevalence, diagnosis and treatment of deep
fungal infection in gastrointestinal surgery. Forty five cases including thirty six complete cases were entered basing on the diagnostic criteria and the mortality of complete cases was 17%. By the administration of
miconazole (MCZ), clinical response rate was obtained in 87%, mycological radiation rate in 82% and valuable rate in 71%, with occurrence of side effects in 20%. Fungal
endophthalmitis was found in 41% (12/29 cases).
beta-glucan [the toxinometer method (T), the Seikagaku-Kogyo method (S)] and candida
antigen [the C and-Tec method (C)] were monitored as sero-diagnostic methods, the positive rates for all cases,
fungemia, and mycosis of which were 63, 83, 62% (T), 88, 100, 92% (S) and 52, 83, 50% (C), respectively. After mycological re-classification, the positive rates of the T method and
endophthalmitis, and the mortality rates were 86, 60, 38% in
fungemia, 63, 43, 23% in
urinary tract infection, 64, 80, 7% in TPN
infection, and 33, 17, 33% in drain
infection, respectively. High association of
endophthalmitis was found in TPN
infection. The
beta-glucan assay and a prospective ophthalmoscopic examination are useful for the early diagnosis of deep
fungal infection and it is reasonable to treat patients in gastrointestinal surgery by MCZ when mycosis is confirmed or suspected.