Open Access Open Badges Research article

Infrequent detection of Pneumocystis jirovecii by PCR in oral wash specimens from TB patients with or without HIV and healthy contacts in Tanzania

Lotte Jensen1, Andreas V Jensen1, George Praygod3, Jeremiah Kidola3, Daniel Faurholt-Jepsen2, John Changalucha3, Nyagosya Range4, Henrik Friis2, Jannik Helweg-Larsen1, Jorgen S Jensen5 and Aase B Andersen1*

Author Affiliations

1 Department of Infectious Diseases, Rigshospitalet, Denmark

2 Department of Human Nutrition, University of Copenhagen, Denmark

3 National Institute for Medical Research, Mwanza, Tanzania

4 National Institute for Medical Research, Muhimbili Centre, Dar es Salaam, Tanzania

5 Department of Bacteriology, Mycology, and Parasitology, Statens Serum Institut, Copenhagen, Denmark

For all author emails, please log on.

BMC Infectious Diseases 2010, 10:140  doi:10.1186/1471-2334-10-140

Published: 28 May 2010



In tuberculosis (TB) endemic parts of the world, patients with pulmonary symptoms are managed as "smear-negative TB patients" if they do not improve on a two-week presumptive, broad-spectrum course of antibiotic treatment even if they are TB microscopy smear negative. These patients are frequently HIV positive and have a higher mortality than smear-positive TB patients. Lack of access to diagnose Pneumocystis jirovecii pneumonia might be a contributing reason. We therefore assessed the prevalence of P. jirovecii by PCR in oral wash specimens among TB patients and healthy individuals in an HIV- and TB-endemic area of sub-Saharan Africa.


A prospective study of 384 patients initiating treatment for sputum smear-positive and smear-negative TB and 100 healthy household contacts and neighbourhood controls.

DNA from oral wash specimens was examined by PCR for P. jirovecii. All patients delivered sputum for TB microscopy and culture. Healthy contacts and community controls were clinically assessed and all study subjects were HIV tested and had CD4 cell counts determined. Clinical status and mortality was assessed after a follow-up period of 5 months.


384 patients and 100 controls were included, 53% and 8% HIV positive respectively. A total number of 65 patients and controls (13.6%) were at definitive risk for PCP based on CD4 counts <200 cells per mm3 and no specific PCP prophylaxis. Only a single patient (0.3% of the patients) was PCR positive for P. jirovecii. None of the healthy household contacts or neighbourhood controls had PCR-detectable P. jirovecii DNA in their oral wash specimens regardless of HIV-status.


The prevalence of P. jirovecii as detected by PCR on oral wash specimens was very low among TB patients with or without HIV and healthy individuals in Tanzania. Colonisation by P. jirovecii was not detected among healthy controls. The present findings may encourage diagnostic use of this non-invasive method.