It is important to obtain either tissue samples or body fluid in which the organism can be identified (category III recommendation). If induced sputum (IS) is routinely available, this can be performed initially (sensitivity 50–90%). If IS results are negative or inconclusive, then the patient should be assessed for bronchoscopy with broncho-alveolar lavage (BAL; diagnostic sensitivity >90%) [24–27]. Some may choose BAL as the first-line investigation employed. Open lung biopsy (diagnostic sensitivity 95–98%) is reserved for the occasional patient, with negative initial tests, and who is not improving

on empirical treatment [28,29]. Spontaneously expectorated sputum is not an adequate alveolar sample and should not be processed. Pneumocystis jirovecii cannot be cultured in vitro; diagnosis relies on visualization of the organism using either histochemical (typically with silver stains such BGB324 as Grocott–Gomori methenamine silver stain) or immunofluorescent stains. Nucleic acid amplification techniques (NAAT) using a variety of primers have been reported with induced sputum, BAL and oral wash specimens [30–33]. In general NAAT-based tests have increased sensitivity but reduced specificity compared to visualization; and the specificity varies by protocol. In one study comparing two NAAT-based assays the sensitivities were 97–98% but specificities ranged from 68% to 96% [30]. The sensitivity is lower using samples that are obtained

from more easily obtained specimens such as sputum or oral washes. NAAT-based assays, although not widely available, can provide information on the molecular Erlotinib epidemiology of PCP and the frequency of mutations in Pneumocystis’ dihydropteroate synthase gene (which is associated with previous exposure to sulpha- drugs – see later). Currently, no definitive PLEK2 recommendations concerning NAAT-based assays can be made. Where centres use them as part

of a diagnostic algorithm they must be interpreted with input from the testing laboratory in the light of their sensitivity and specificity. They should be combined with a definitive visualization technique (category IV recommendation). Treatment should not be delayed in a presumed case, by having to wait for a diagnostic procedure, as adequate pulmonary samples can be obtained up to 7–10 days after starting specific anti-pneumocystis therapy [34]. First-line treatment for moderate–severe PCP [PaO2≤9.3 kPa (≤70 mmHg)] is with high-dose intravenous (iv) trimethoprim-sulphamethoxazole for 21 days (co-trimoxazole, TMP-SMX) (category Ib recommendation). Co-trimoxazole is a highly effective agent when given for 21 days. It has an efficacy of at least 90% in mild disease and around 70% in more severe cases [35–38]. It has shown similar or better outcomes when compared to iv pentamidine in randomized clinical trials of treatment of PCP [35–37]. Dosing for moderate–severe PCP [PaO2≤9.3 kPa (≤70 mmHg), see Table 3.