Severe acute respiratory syndrome-associated coronavirus in lung tissue

Efforts to contain severe acute respiratory syndrome (SARS) have been limited by the lack of a standardized, sensitive, and specific test for SARS-associated coronavirus (CoV). We used a standardized reverse transcription-polymerase chain reaction assay to detect SARS-CoV in lung samples obtained from well-characterized patients who died of SARS and from those who died of other reasons. SARS-CoV was detected in all 22 postmortem lung tissues (to 10(9) viral copies/g) from 11 patients with probable SARS but was not detected in any of the 23 lung control samples (sample analysis was blinded). The sensitivity and specificity (95% confidence interval) were 100% (84.6% to 100%) and 100% (85.1% to 100%), respectively. Viral loads were significantly associated with a shorter course of illness but not with the use of ribavirin or steroids. CoV was consistently identified in the lungs of all patients who died of SARS but not in control patients, supporting a primary role for CoV in deaths.