March 2019: ID diagnostics

Antibiotic treatment of infective endocarditis prior to valve surgery decreases the yield of both microbiologic cultures and PCR, but when cultures are negative PCR is often diagnostically valuable.  Such is the takeaway from this single-center retrospective study out of the University of Helsinki in Finland.  The authors reviewed patients with infective endocarditis (possible or definite by modified Duke criteria) who underwent cardiac surgery between 2011 and 2016.  They stratified the patients by duration of antibiotics received before surgery. The primary outcomes of interest were the yields of microbiologic culture and PCR of the valve.  Patients were excluded from the study if valve was PCR not sent and if the endocarditis was fungal.

A total 87 patients were included in the analysis; the mean age was 53, 82% were men, 97% of valves were either aortic or mitral, and 88% were native valves.  Blood cultures were positive in 85% of cases; the most common pathogens were S.aureus (n=24), viridans strep (n=22), E.faecalis (n=9), and coagulase-negative staph (n=8).  Table 1 shows the yield of culture and PCR stratified by preoperative antibiotic duration, and shows that the biggest drop in diagnostic yield occurred at 2 weeks (valve culture positivity rate 36-43% with fewer than 2 weeks of antibiotics before surgery vs 0% with more than two weeks; PCR positivity rate 91-92% with fewer than 2 weeks vs 47-59% with more than two weeks).  Among the patients with negative blood cultures, PCR identified a likely pathogen in 77% of cases, meaning that PCR added diagnostically valuable information in 12% of all cases.  The pathogens identified solely by PCR included Bartonella quintana, Coxiella burnetti, Tropheryma whipplei, and a smattering of various streptococci.  When comparing PCR-positive and PCR-negative patients, the latter had received much longer antibiotic durations prior to surgery (24 days vs 8.5 days; p=0.001); in multivariate analysis, receipt of fewer than 2 weeks of antibiotics strongly predicted a positive valve PCR (OR 7.2; 95% CI 2 to 26; p=0.003).

In conclusion: in infective endocarditis, the valve PCR is frequently positive if the patient has received fewer than two weeks of antibiotics before surgery, even when blood and valve cultures are negative. 30680557

Speaking of valve PCR in endocarditis, another group of authors evaluated next-generation sequencing of resected valve tissue as a means of diagnosis for infective endocarditis, comparing the results to those of blood culture, valve culture, and valve staining in 44 patients.  They found that NGS had a sensitivity of 98% and specificity of 86%, versus 46% and 100% for blood culture, 17% and 100% for valve culture, and 51% and 100% for valve gram staining.  As with the previous study, NGS added greatest value in cases of unculturable or difficult to culture organisms. 30926543

BAL Aspergillus PCR increases both the probability and certainty of invasive pulmonary aspergillosis diagnoses in immunocompromised patients.  This retrospective study examined immunocompromised patients who underwent bronchoscopy and BAL to evaluate pulmonary infiltrates.  In addition to serum and BAL galactomannan levels and BAL fungal cultures, BAL specimens underwent Aspergillus PCR testing.  Patients were classified as having proven, probable, possible, or no invasive pulmonary aspergillosis (IPA) based on standard diagnostic criteria.

Over 12 years, a total 1072 patients underwent 1248 bronchoscopies.  Of these, 77% had a hematologic malignancy, 40% had AML, and 36% received a stem cell transplant; 43% received a diagnosis of IPA, among which 7 were proven, 280 probable, and 244 possible.  Aspergillus PCR had a true positive rate of 80% and a false positive rate of 20% (sensitivity 40%, specificity 90%, PPV 80%, NPV 68%), and including PCR in the diagnostic criteria would have reclassified 80 cases from possible to probable.  The authors conclude that Aspergillus PCR increases both the rate and certainty of IPA diagnosis, which suggests this could be a valuable addition to current mycological criteria. 30926540

BAL testing also has diagnostic value for patients with suspected TB and negative sputum testing. Another retrospective study out this month looked at the value of BAL testing for tuberculosis in patients who’ve already had a negative sputum stain and culture.  This study, from Qatar, looked at 150 such patients, finding a BAL cultures and PCR were positive in 32% and 30%, respectively.  BAL was most likely to yield a positive result among patients with a positive quantiferon or PPD (go figure), radiographic pathology in the upper lung fields, or Indian origin.  So yes, if you have a patient in whom you strongly suspect tuberculosis despite negative sputum stains and cultures, a bronchoscopy with BAL is warranted.  Just warn the pulmonologists beforehand. 30904678