Evaluation of hematoxylin and eosin and special stains for the detection of acanthamoeba keratitis in penetrating keratoplasties.

PURPOSE

To compare the efficacy of a battery of routine and special histologic stains for the detection of acanthamoeba keratitis.

DESIGN

Observational study.

METHODS

Nine patients with culture-proven infectious keratitis whose clinical differential diagnosis included acanthamoeba and who had undergone penetrating keratoplasty were identified. Three cases each of culture-proven acanthamoeba, fungal, and herpes simplex keratitis were reviewed. Serial sections of the keratoplasty specimens were stained with hematoxylin and eosin, periodic acid-Schiff (PAS), Gomori methanamine silver (GMS), giemsa, Gram, calcofluor white, and acridine orange. Sections were reviewed in a random order and a masked fashion and classified as positive or negative for acanthamoeba, fungus, or herpes.

RESULTS

The correct diagnosis was made by examination of the hematoxylin and eosin stained slides in all cases. Correct diagnoses in decreasing order of frequency were made for slides stained with PAS, GMS, acridine orange, calcofluor white, giemsa, and Gram. There were false-positive diagnoses made only with calcofluor white and acridine orange stained slides because of staining of extracellular debris and other material.

CONCLUSIONS

When sections are examined by an experienced observer, hematoxylin and eosin is useful compared with calcofluor white, acridine orange, GMS, PAS, giemsa, and Gram stains for the detection of acanthamoeba keratitis.