Conference Coverage

In Zambia, PCR tracks pertussis



In the periurban slum of Lusaka, Zambia, asymptomatic pertussis infections were common among both mothers and infants, a surprising finding since asymptomatic infections are assumed to be rare in infants. The findings suggested that pertussis should be considered in cases of chronic cough, and that current standards of treating pertussis infections in low-resource settings may need to be reexamined.

The results come from testing of 1,320 infant-mother pairs who were first enrolled at a public health clinic, then followed over at least four visits. The researchers tracked pertussis infection using quantitative PCR (qPCR) on nasopharyngeal swabs. Over the course of the study, 8.9% tested positive, although only one infant developed clinical pertussis during the study.

The study was presented by Christian Gunning, a postdoctoral researcher at the University of Georgia, at the annual meeting of the European Society for Paediatric Infectious Diseases, held virtually this year. The group also included researchers at Boston University and the University of Zambia, where PCR tests were conducted.

“That was amazing,” said session moderator Vana Spoulou, MD, PhD, professor of pediatric infectious diseases at National and Kapodistrian University of Athens, who is associated with Aghia Sofia Children’s Hospital of Athens. She noted that the study found that many physicians misdiagnosed coughs, believing them to be caused by another agent. “It was very interesting that there was so much pertussis spreading around in that community, and that nobody knew that it was around,” said Dr. Spoulou.

It’s important that physicians provide appropriate treatment, since ampicillin, which is typically prescribed for childhood upper respiratory illnesses, is believed to be ineffective against pertussis, while macrolides are effective and can prevent transmission.

Dr. Spoulou also noted that Zambia uses a whole cell vaccine, which is contraindicated in pregnant women because of potential side effects. “The good thing, despite that there was [a lot of] infection, there were no deaths, which means that maybe because the mother was infected, maybe some antibodies of the mother had passed to the child and could help the child to develop milder symptoms. So these are the pros and cons of natural infection,” said Dr. Spoulou.

The study took place in 2015, and participants were seen at the Chawama Public Health Clinic from about age 1 week to 4 months (with a target of seven clinic visits). Researchers recorded respiratory symptoms and antibiotics use at each visit, and collected a nasopharyngeal swab that was tested retrospectively using qPCR for Bordetella pertussis.

Real-time PCR analysis of the samples yields the CT value, which represents the number of amplification cycles that the PCR test must complete before Bordetella pertussis is detectable. The fewer the cycles (and the lower the CT value), the more infectious particles must have been present in the sample. For pertussis testing, a value below 35 is considered a clinically positive result. Tests that come back with higher CT values are increasingly likely to be false positives.

The researchers plotted a value called evidence for infection (EFI), which combined a range of CT values with the number of positive tests over the seven clinic visits to group patients into none, weak, or strong EFI. Among infants with no symptoms, 77% were in the no EFI category, 16% were in the weak category, and 7% were in the strong EFI group. Of infants with minimal respiratory symptoms, 18% were in the strong group, and 20% with moderate to severe symptoms were in the strong EFI group. Among mothers, 13% with no symptoms were in the strong group. 19% in the minimal symptom group were categorized as strong EFI, as were 11% in the moderate to severe symptom group.

The study used a full range of CT, not just positive test results (for pertussis, CT ≤ 35). Beyond contributing to composite measures such as EFI, CT values can serve as leading indicators of infectious disease outbreaks in a population, according to Dr. Gunning. That’s because weaker qPCR signals (CT > 35) can provide additional information within a large sample population. Higher CT values are successively more prone to false positives, but that’s less important for disease surveillance where sensitivity is of the highest importance. The false positive “noise” tends to cancel out over time. “It may be the case that you don’t make that call (correctly) 100% of the time for 100% of the people, but if you get it right in 80 out of 100 people, that’s sufficient to say we see this pathogen circulating in the population,” said Dr. Gunning.

The study was funded by the National Institute of Allergy and Infectious Diseases. Dr. Gunning and Dr. Spoulou have no relevant financial disclosures.

Next Article: