(Pertussis outbreaks are occurring in the U.S. and elsewhere, yet many healthcare professionals seem reluctant to test for it. We asked Dr. James D. Cherry to explain when testing should be done and we extend our thanks to him for this post on diagnosing whooping cough. Dr. Cherry is a member of the Global Pertussis Initiative (GPI) and author of previous papers on pertussis Dx. Please feel free to share this post with your healthcare provider.)
In pertussis the site of infections is on ciliated epithelial cells in the nasopharynx (NP). In primary infections (infants and young child not previously vaccinated) the bacterial load is high and is present in the nasopharynx from the onset of illness (coryza) the second week of the paroxysmal stage and often longer. In children who are vaccine failures the bacterial load in the NP is less than in primary infections and the bacteria are present for a shorter period of time (ie onset of coryza through the second week of cough).
In adults (all of whom have had previous infections unknown to them) the bacterial load is less than in previously vaccinated children and the duration of presence of bacteria is also less. Also, adolescents and adults rarely seek care for their pertussis cough illness until the third or fourth week from illness onset. Nevertheless adults with unrecognized pertussis are the most common source of infection in infants who are unimmunized or only partially immunized.
Culture is 100% specific whereas all other tests are not. Culture in children is a much more sensitive test than generally believed. However, today in the U.S., for the most part, culture is a lost art in most diagnostic laboratories because of lack of fresh media and technicians with little experience. With a good laboratory the main reason for failure to isolate Bordetella pertussis is that the specimen was not collected properly or that it was collected too late in the illness.
To obtain an adequate sample the ciliated cells in the NP must be touched by the dacron tip of the NP swab or the catheter used in a NP aspirate must touch the ciliated cells. Nasal wash is frequently done but this is much less sensitive than either NP swab or NP aspirate. For PCR the same facts apply regarding specimen collection.
For children (during the first 3 weeks of illness) and adults (during the first week of illness) PCR is the method of choice because it is much more sensitive than culture. Unfortunately, there has been much misinformation disseminated about PCR results. PCR is readily available in the U.S. in hospital labs and several commercial labs.
The test that is universally available in the U.S. uses primers that identify insertion sequence (IS) 481 for B. pertussis and IS1001 for B. parapertussis. Because B. pertussis contains ~238 copies of IS481 this test is exceedingly sensitive. It is so sensitive that it can pick up examination room contamination because of a previous patient with pertussis or the immunization with DTaP of a previous patient in the room. Therefore NP specimens should not be collected in rooms where DTaP immunization is being carried out or in rooms that have been occupied by previous patients with pertussis.
Today in the U.S. real time PCR is the method most often used and the number of cycles necessary to obtain a positive result reflects the concentration of B. pertussis in the sample. The lower the cycle the greater the number of bacteria. With high cycle detection the possibility of contamination at the collection site is a likely possibility. However positives are positives regardless of the cycle. It has been suggested that labs not report high cycle positives as positive results. This is wrong; these results represent infections or contamination and the physician who obtained the specimen must decide if the findings are consistent with the patient’s illness. The lab should not call high cycle positives as negative or indeterminate because this relays false information to the physician.
In situations in which both IS481 and IS1001 are positive this may be due to infection with B. holmesii (also a cause of clinical pertussis) or a mixed infection with B. pertussis and B. parapertussis.
PCR should only be performed on patients with cough illnesses. During pertussis outbreaks asymptomatic infections are very common in previous vaccinees so that you will get positive PCR results from people who are well and these results just confuse the picture (except in planned surveillance studies).
All persons who have been previously vaccinated or who have had previous infection will have a rapid rise in antibody to various B. pertussis antigens so that pertussis illness can be diagnosed by single serum serology. The most useful antibody to determine if a cough illness is pertussis is that to pertussis toxin (PT) because this antigen is exclusive of B. pertussis. Some tests also determine antibody to filamentous hemagglutinin (FHA) but since this antigen is not exclusive to B. pertussis high titers could be due to B. pertussis infection, other Bordetella spp and M. pneumoniae and perhaps other microorganisms.
Single serum serologic Dx has been used successfully in Massachusetts for over 20 years. Commercial laboratories also perform single serum serology but unfortunately many of these tests are poor. Specifically any test that uses the whole B. pertussis bacterium is virtually useless as are tests that don’t express results in units. Tests that say they are measuring IgM antibody are also useless. To my knowledge the only commercial test available in the U.S. that is acceptable is that offered by Focus Laboratories. This test has specificity of ~95%.
Serologic diagnosis will be affected by recent immunization with either DTaP and Tdap so it should not be attempted if the patient has been vaccinated within the previous year. In general single serum serology should be used for the diagnosis of pertussis in adolescents and adults who have not been recently vaccinated.
White Blood Cell (WBC) Count
Primary infections of pertussis universally have high WBC counts with absolute lymphocytosis. This is seen in all infants who have not been immunized and who have not received antibody to PT from the mother transplacentally. Therefore in young infants with afebrile cough illnesses the WBC count with differential can be diagnostic. Because the WBC count has prognostic implications it should be performed on all infants who might have pertussis at the time of first physician encounter. A WBC count of > 20,000 cells/mm3 with a lymphocyte count of > 10,000 cells/mm3 should be diagnosed as pertussis and immediately treated with azithromycin.
- The onset of cough illnesses in afebrile or minimal febrile persons of all ages should be suspected of having pertussis.
- Lab confirmation tests should be performed and specific tests employed relate to patient age, vaccine history and duration of cough.
- For infants and children PCR is the most sensitive and specific test.
- Whenever possible cultures should also be performed so that epidemic trends can be followed (ie antibiotic resistance, genetic changes in the organism)
- For adolescents and adults single serum serology to determine antibody to PT is the most sensitive and specific test (unless the patient is seen during the first week of cough).
- For young infants the WBC count with differential is often diagnostic.
James D. Cherry, MD, MSc
Distinguished Professor of Pediatrics
David Geffen School of Medicine at UCLA
Pediatric Infectious Diseases
Mattel Children’s Hospital UCLA