C-Reactive Protein May Miss Infections Caused by Weaker Organisms

March 7, 2016

C-reactive protein levels may not capture the real incidence of infection with less-virulent organisms, AAOS 2016 study shows.

C-reactive protein in the blood serum and the synovial fluid is used as a biomarker for periprosthetic joint infection (PJI). But C-reactive protein levels may not capture the real incidence of infection with less-virulent organisms, leading to false negative results, new research finds.

In a study presented on March 2 at the annual meeting of the American Association of Orthopaedic Surgeons (AAOS) in Orlando, Carl Deirmengian, M.D., an orthopedic surgeon at the Rothman Institute in Philadelphia, reviewed 12,144 synovial fluid samples from joint arthroplasties for virulent and "less virulent" organisms. Staphylococcus aureus, S. lugdunensis, S. agalactiae, enterobacteriaceae and non-fermenting Gram-negative bacilli, were all classified as virulent; any other organism found was classified as non-virulent.

Of 909 samples with a positive culture, 325 contained virulent organisms, and 584 contained less-virulent organisms. Deirmengian and his colleagues tested all of these samples for C-reactive protein level, and then compared those levels by type of organism found.

The results revealed that C-reactive protein levels are lower in the presence of less-virulent organisms than virulent organisms, and that false-negative results are higher when less-virulent organisms are present than when virulent organisms are present. For example, the median synovial fluid c-reactive protein level in the presence of typically non-pathogenic S. epidermidis was 13.4 mg/L, the researchers reported. That was significantly lower (p<0.015) than c-reactive protein levels in the presence of S. aureus (41.6 mg/L), E. coli (53.7 mg/L), and S. mitis (32.9 mg/L).

Overall, the median C-reactive protein level in synovial fluid was 14.3 mg/L on average in the presence of less-virulent organisms compared with 33.9 mg/L on average in the presence of virulent organisms, the researchers reported. The rate of false negatives, as defined by a cutoff of 6.6 mg/L c-reactive protein, was 28 percent for less-virulent organisms, compared with 9 percent for virulent organisms.

C-reactive protein levels remain an important marker of periprosthetic joint infection, the researchers concluded, but might miss infection by less-virulent organisms.

“This large study demonstrated a high dependence of the synovial fluid (SF) CRP on the identity of the infecting organism, and supports recent research on the serum CRP suggesting the same,” Dr. Deirmengian wrote. “Certain organisms, specifically those generally considered less-virulent, do not induce a robust CRP response and have an unexpectedly higher chance of yielding a false-negative CRP result. Although the use of a CRP level is an important part of the workup for PJI, surgeons must beware that this protein may yield a false-negative result in the setting of less-virulent.”



“The C-Reactive Protein Test May Not Detect PJIs Caused by Less-Virulent Organisms,” Carl A. Deirmengian, MD, Rothman Institute in Philadelphia. American Academy of Orthopaedic Surgeons, 2016 Annual Meeting.  March 2, 2016.