False-Positive Culture Results from Patient Tissue Specimens Due to Contamination of RPMI Medium with Cryptococcus albidus

Abstract

Cryptococcus albidus, a rare opportunist, was isolated from biopsy specimens from three patients over 4 days. An investigation showed that the specimens had been contaminated by placement in RPMI medium. The importance of rapid communication between the microbiology laboratory, the infectious diseases/infection control division, and other involved parties in the event of unusual occurrences is highlighted.

Cryptococcus albidus is a fungus that can be isolated from the environment. It is a rare pathogen, with fewer than 30 cases of human infection reported. The microbiology laboratory at our institution isolated C. albidus from biopsy specimens obtained from three different patients over a 4-day period in August 2006. This highly unusual occurrence prompted an investigation into possible sources of contamination of the specimens.

Patient 1 was a 63-year-old man with hemoptysis of several months ’ duration. A computed tomography scan of his chest revealed an upper lobe mass suspected of being a primary bronchogenic carcinoma. On 22 August 2006, the interventional radiology department performed a computed-tomography-guided aspiration of the mass, and specimens were sent both to the pathology department for examination and to the microbiology laboratory for fungal and mycobacterial cultures. Gram staining of the specimen was not performed since a bacterial culture was not ordered. Acid-fast staining of the tissue performed by the microbiology laboratory showed no organisms, and cultures were subsequently negative for mycobacterial growth. However, a fungal culture grew C. albidus, which was definitively identified by the microbiology laboratory on 31 August 2006. The C. albidus isolate was germ tube negative, and there were no hyphal/pseudohyphal elements seen on cornmeal agar with trypan blue stain (Remel, Lenexa, KS). The negative reaction on birdseed agar indicated that the yeast was not Cryptococcus neoformans. C. albidus was definitively identified using the API 20 C AUX identification strip (bioMérieux, Marcy-l'Etoile, France). Cytology, reported on 25 August 2006 by the pathology department, revealed squamous cell carcinoma and necrosis without evidence of infection. The patient was not treated for cryptococcal infection.

Patient 2 was a 25-year-old man with a history of intravenous drug use. He had been diagnosed with pan-sensitive tuberculosis of the thoracic spine. He was admitted with fever despite ongoing treatment with appropriate antituberculous drugs. The patient refused testing for human immunodeficiency virus. On 19 August 2006, computed tomography of the abdomen and pelvis was performed because of abdominal pain and showed rim-enhancing lesions in the liver and spleen, a lytic lesion in the T12 vertebral body, and enhancing paraspinal soft tissue. On 24 August 2006, interventional radiology performed percutaneous needle biopsies of the liver and paraspinal lesions. Specimens were sent to the pathology department and the microbiology laboratory. Gram staining of the liver and paraspinal specimens sent to the microbiology laboratory showed many round and oval budding yeasts. Staining for acid-fast bacilli by the microbiology laboratory was negative. Cultures from both specimens grew C. albidus, which was reported on 5 September 2006. The culture and identification methods were the same as those described for patient 1. Pathological examination of both the liver and paraspinal specimens using Fite staining showed organisms consistent with mycobacteria. Given the clinical presentation and possible risk factors (injection drug use and the question of being immunocompromised), this patient was initially treated with amphotericin B deoxycholate until an investigation indicated that the specimens had been contaminated during collection.

Patient 3 was a 31-year-old man with human immunodeficiency virus infection. He presented to the clinic with a large axillary lymph node. A fine-needle aspiration was performed in the clinic on 25 August 2006 by the cytopathology department. Tissue was sent to the microbiology laboratory and to the pathology department. In the microbiology laboratory, acid-fast staining revealed few acid-fast bacilli. A culture for mycobacteria subsequently grew Mycobacterium avium complex organisms. Gram staining of the tissue revealed few oval budding yeasts. A culture grew C. albidus, as reported on 5 September 2006. Again, culture and organism identification were performed as described for patient 1. Cytological examination and Fite staining performed by the pathology department showed acute necrotizing inflammation and numerous organisms consistent with mycobacteria. The patient was not treated with antifungal therapy.

On 6 September 2006, the infectious diseases/infection control division became aware that specimens from three patients taken within a 4-day period had grown C. albidus. An investigation had been undertaken in the microbiology laboratory to look for possible mechanisms of contamination after the specimens reached the laboratory. This investigation included a check of the labeling on the specimen tubes and testing to demonstrate the sterility of media used in the microbiology laboratory for specimen processing and culture inoculation. No evidence of contamination within the microbiology laboratory was discovered. On 7 September 2006, review of the original laboratory requisitions revealed that one cytopathologist had collected the specimens obtained by interventional radiology from the first two patients and had collected the specimen from the third patient in the clinic. On 8 September 2006, representatives from the microbiology laboratory and the infectious diseases/infection control division met with the cytopathologist to examine the pathology slides and to review the procedures used when the specimens were collected and transported to the pathology and microbiology departments.

All the specimens had been obtained by needle aspiration using a fresh sterile needle and syringe. None of the specimens submitted to the pathology department showed the presence of yeasts. Special fungal stains were not used on the lung biopsy specimen from patient 1; however, Gomori methenamine silver staining was performed on both pathology specimens from patient 2 (liver and paraspinal lesions). Yeast forms were not identified in the pathology specimens. The lymph node biopsy specimen from patient 3 was also negative for yeasts as indicated by Gomori methenamine silver staining.

Specimens submitted to the microbiology laboratory were collected with a needle and syringe and then rinsed into tubes of RPMI cell culture medium (named for Roswell Park Memorial Institute) before being transported to the laboratory. No RPMI medium was used for any of the specimens submitted to the pathology department. The original bottle of RPMI medium was stored at 4°C in the pathology department laboratory. Aliquots of RPMI medium were periodically poured into sterile capped tubes, which were stored at room temperature in the biopsy tray carried by the cytopathologist to the site of specimen collection.

The microbiology laboratory obtained a capped tube of RPMI medium from the fine-needle-aspirate tray that had been poured from the refrigerated bottle. Gram staining of this medium on 8 September 2006 revealed many round budding yeasts consistent with Cryptococcus spp. Clinicians caring for all three patients were notified of this finding on 8 September 2006, as this provided evidence that the patient specimens were contaminated by use of contaminated RPMI medium. Subsequent culture of both the capped tube of RPMI medium and the original refrigerated bottle grew C. albidus, which was identified by use of the API 20 C AUX identification strip on 14 September 2006.

The microbiology laboratory also obtained an unopened bottle of RPMI medium of the same lot number from the pathology department and cultured it. The RPMI medium from the sealed bottle was sterile, with no yeasts or bacteria recovered. The contaminated RPMI medium was discarded. Policies and procedures have been created that eliminate the use of RPMI medium for collection of specimens for microbiology cultures and institute the use of single-use sterile saline tubes for collection of these specimens.

This report describes three clinical cases in which C. albidus was isolated from patient specimens as a result of contamination of RPMI medium. Cryptococcus neoformans has been reported previously as a laboratory solution contaminant in a pseudo-outbreak of cryptococcal meningitis (3). However, we did not find any previous cases in which C. albidus was found in patient specimens as a result of laboratory contamination. True infections with C. albidus are extremely rare. Systemic infections have been reported largely in immunocompromised patients (4-6, 8, 10). Ocular infections (1, 2) and cutaneous infections (7, 9) have also been reported in patients with underlying risk factors for opportunistic infections.

Due to the rarity of finding C. albidus in clinical specimens and the proximity of the cases in time, contamination was quickly suspected. However, it was not immediately apparent where the contamination had occurred or whether one of the patients could have true disease and that patient's specimen had contaminated the laboratory environment. Patient 2 was the only patient clinically suspected of having disease due to an environmental cryptococcal species after culture results were reported. It was thought that he might have become infected through frequent injection drug use. The finding of large numbers of yeast cells through Gram staining also suggested to his clinicians the possibility that true infection was present. However, the presence of both C. albidus cultured by the microbiology laboratory and acid-fast bacilli seen on staining by the pathology department in the same specimen was puzzling.

Discussion with the cytopathologist who obtained the specimens pointed to a likely mechanism of contamination, which then was rapidly confirmed. An RPMI medium solution is known to sustain growth of bacteria and fungi, and it is not needed for transport of tissue specimens to the microbiology laboratory. In addition, diluting the specimens in a fairly large and nonstandardized volume of fluid may lead to a decreased culture yield. In order to decrease the risk of specimen contamination and needlestick exposures from recapping needles, our institution now uses single-use sterile saline tubes for specimens collected for microbiological culture with a needle and syringe.

The false-positive cultures described in this report demonstrate how important it is for unusual occurrences in the microbiology laboratory to be immediately reported to a hospital's infectious diseases/infection control department and to be investigated thoroughly so corrective actions can be addressed and inappropriate therapy can be avoided.