International circumpolar surveillance: update on the interlaboratory quality control program for Streptococcus pneumoniae, 2009 to 2020

Arctic populations face a series of unique challenges, both social and physical, that lead to the increased spread and incidence of invasive diseases. The harsh northern climate forces residents indoors; however, small isolated communities often have inadequate housing; crowded households and poor ventilation can lead to increased person-to-person transmission of infectious diseases (1, 2). Improper or overuse of antimicrobial agents in some remote communities with poorly developed or poorly supported public health systems has led to, and may continue to support, the emergence of resistant bacterial clones (1). There are also significant health inequities between Indigenous and non-Indigenous people living in circumpolar regions, where Indigenous populations tend to fare worse for a large number of health indicators (3, 4).

To improve surveillance, prevention, and disease control in Arctic populations experiencing health disparities, the International Circumpolar Surveillance (ICS) program was implemented in 1999, to monitor invasive pneumococcal disease (IPD) in Alaska and across the Canadian Arctic (2). Initially, participating laboratories for IPD surveillance included the National Centre for Streptococcus at the Provincial Laboratory for Public Health (PLPH) in Edmonton, Alberta; Laboratoire de santé publique du Québec (LSPQ) in Sainte-Anne-de-Bellevue, Québec; and the Centers for Disease Control’s Arctic Investigations Program (AIP) in Anchorage, Alaska. This collaboration provided an opportunity to expand Canada’s previously existing Interlaboratory Quality Control (QC) program for Streptococcus pneumoniae to include AIP.

In August 2004, the ICS QC program for S. pneumoniae was expanded to include the Neisseria and Streptococcus Reference Laboratory (previously the WHO Collaborating Centre for Reference and Research on Pneumococci) at Statens Serum Institut (SSI) in Copenhagen, Denmark. In September 2006, the Department of Clinical Microbiology at Landspitali—the National University Hospital of Iceland in Reykjavik joined the ICS program. On 1 April 2010, national streptococcus laboratory services in Canada were transferred from PLPH to the National Microbiology Laboratory (NML) in Winnipeg, Manitoba, and NML was added to the interlaboratory QC program bringing the number of participating laboratories to six.

The purpose of the QC program is to provide an external proficiency testing mechanism for pneumococcal serotyping and antibiotic susceptibility testing. The S. pneumoniae QC panel serves to identify and correct potential problems in reagents and equipment, test interpretation, and assess technical precision and accuracy. A previous publication by Reasonover et al. detailed the first 10 years of the S. pneumoniae QC program, from 1999 to 2008 (5). The current report describes the results of the S. pneumoniae ICS Interlaboratory QC program collected from 2009 to 2020.

The ICS QC program for S. pneumoniae is critical to ensure that accurate testing results are being provided to clinical and laboratory partners in vulnerable Arctic communities. Our QC data demonstrate excellent concordance for serotype and antimicrobial susceptibilities for 140 S. pneumoniae isolates distributed across six laboratories in North America and Europe. Considering the antisera available at each laboratory, serotype concordance was 96.9%, with 99.3% concordance to pool level. For antimicrobial susceptibility testing, overall concordance by modal MIC was 99.4% (>98% for all compounds tested) and overall categorical concordance using CLSI and EUCAST interpretive criteria was 98.4% and 98.5%, respectively. For two participating laboratories performing a portion of testing by disc diffusion, rates of concordance in comparison to the modal MIC were >97% for most antimicrobials.

New, higher-valency pneumococcal conjugate vaccines (15-valent, 20-valent formulations), which would replace existing 10- and 13-valent formulations, have been approved for use around the world, including in Canada, United States, and European Union member states (11–13). During this transitionary time, it will be increasingly important to monitor regional serotype distributions, assess vaccine coverage, and track the emergence of non-vaccine types. External quality assurance (EQA) programs will remain a crucial tool to assess the quality of pneumococcal serotyping provided by reference laboratories. An early EQA for pneumococcal serotyping was developed for the Sistema Regional de Vacunas (SIREVA) project in Central and South America, as described by Lovgren et al. (14). Overall serotyping accuracy for phase II of the EQA program development (1999–2005) was 93.8% for 130 isolates (23 vaccine-related and 26 non-vaccine-related) over 13 distributions to 20 participating countries (14). In a more recent study, Slotved et al. describe an EQA scheme performed by eight reference laboratories in countries across Europe and the Middle East, consisting of 22 distributions spread across 11 years (2005–2016) (15). Of 154 isolates (representing 49 serotypes), discrepant results for one or more participants were only identified for 7.1% of isolates. Their results showed that comparable serotyping results can be obtained across various typing methods, including latex agglutination, Quellung reaction, whole-genome sequencing, and various PCR methodologies (15). Our results compare favorably to those from these two manuscripts, as well as to the results of the previous decade of ICS QC results, of 95.8% and 97.4% concordance by serotype and serogroup, respectively (5). Lastly, the World Health Organization’s Global Invasive Bacterial Vaccine Preventable Disease (IB-VPD) Surveillance Network (GISN) provides a yearly EQA panel consisting of Haemophilus influenzae, Neisseria meningitidis, and S. pneumoniae. Litt et al. describe the results from the 2014–2019 exercise, where each panel contained 2–3 S. pneumoniae for identification and serotyping per year; the serotyping exercise was mandatory for regional reference laboratories (RRLs) and optional for sentinel site laboratories and national laboratories (SSLs/NLs), depending on resources and capability (16). GISN’s nine RRLs had high serotype concordance, with only a single error across six distributions. SSLs/NLs had low participation (16%–33%), and the serotyping success rate varied between 33% and 100%, depending on the serotype. However, there was a significant upward trend over time in the number of correct serotyping results for these laboratories, as resources and training improved during the reported 6-year time span (16).

One aspect noted by Slotved et al.’s and Litt et al.’s studies was difficulty with concordance when the expected serotype had known cross-reaction of common factors, specifically for serotypes 35B and 29 (15, 16). Our report describes two similar discrepancies, where some participating laboratories identified a serotype 29 isolate as serotype 35B. A second, related discrepancy in our report surrounded the recently described novel serotype 35D (17). Serotype 35D differs from 35B due to the presence of inactivating mutations in wciG; as 35D does not react with group 35 antiserum, the type may be missed if all relevant antisera are not used for screening. Serotype 35D can also easily be mistaken for serotype 29, as it differs in serological profile by only one factor serum (35b). It is crucial to serotype Pool G pneumococci with caution, due to the numerous antiserum cross-reactions that may lead to incorrect results. If an isolate tests positive by Pool G antisera, all pertinent group, type and factor sera should be tested each time.

There are few other reports describing external proficiency results for antimicrobial susceptibility testing of S. pneumoniae. The ICS QC report for the 1999 to 2008 S. pneumoniae distributions describes >96% concordance for all antimicrobials except erythromycin and clindamycin; however, the testing methods used during this time period varied substantially in comparison to the current summary (5). The previously discussed Lovgren et al.’s study, reporting the establishment of the SIREVA EQA, notes MIC concordance of 91.0% by modal MIC and 95.3% by categorical agreement for phase II of the program development (1999–2005) (14). Similarly to the current ICS QC program results, there were a handful of major errors for chloramphenicol and erythromycin. The European Antimicrobial Resistance Surveillance Network (EARS-Net) offers a yearly EQA panel for participating laboratories. The EARS-Net panel focuses on a variety of key antimicrobial-resistant pathogens and recently has included up to one S. pneumoniae isolate per panel. From 2017–2019, there was one S. pneumoniae isolate on each yearly panel; very good concordance (>90%) with intended results was reported for most compounds during this time span; however, authors described ongoing issues with concordance for β-lactam-intermediate S. pneumoniae isolates (10, 18, 19). Though the number varied depending on the breakpoints applied to our ICS QC results, >20 and >15 errors were identified for penicillin and ceftriaxone/cefotaxime, respectively. The majority of these errors resulted from a difference in one doubling dilution on either side of the breakpoint for susceptible and intermediate, regardless of applied breakpoints.

A limitation of this program, particularly the antimicrobial susceptibility testing portion, is the different testing methods used by each participating laboratory. Though most laboratories use broth microdilution for this exercise, the differing test ranges for each antimicrobial make it increasingly difficult to calculate modal MIC values. Though the concordance for all antimicrobials was high, this underscores the importance of standard protocols to be used across submitting laboratories when calculating values for the final report. Moving forward, many participating laboratories are adopting routine whole-genome sequencing-based identification of serotypes and antimicrobial susceptibilities for S. pneumoniae. In this event, the ICS QC program will need to implement minimum quality standards for genomic data to ensure that submitted typing results are of high quality.

The ICS QC program for S. pneumoniae has been a successful collaboration for over 20 years, in addition to QC programs for another vaccine-preventable disease such as Haemophilus influenzae and Neisseria meningitidis (20). Following the success of these programs, an interlaboratory QC program for emm typing of Group A Streptococcus (Streptococcus pyogenes) began in 2011 (21). More recently, a QC program has been developed for serotyping of Group B Streptococcus (Streptococcus agalactiae), beginning with three participating laboratories in 2021 and expanding to five by the 2023 distribution. Ongoing participation in ICS QC programs will ensure the continuation of quality surveillance systems within Arctic populations that experience health disparities.

We thank the laboratory staff at all six participating laboratories for their technical expertise and support of the ICS S. pneumoniae quality control program.

This project was supported by internal funding from the Public Health Agency of Canada.

The findings and conclusions in this report are those of the author(s) and do not necessarily represent the official position of the Centers for Disease Control and Prevention.