Molecular epidemiology of tuberculosis in selected sites across Papua New Guinea

With an estimated one third of the global population being infected with latent tuberculosis (TB) and 8.6 million people developing active TB in 2012, this infectious disease remains a major global health concern. Increasing drug resistance (DR) and the HIV pandemic are further challenges to the control of the disease.
Mycobacterium tuberculosis (Mtb) is responsible for most of the TB cases in humans. For a long time it was thought that only environmental factors and the host immune status are the driving forces of TB transmission. Recently, evidence of the influence of the bacterial genetic background on transmission and disease outcome is increasing. Clinical samples from around the globe are required to further analyze the impact of the complex interactions of drug resistance, bacterial and host genetics, as well as environmental and social factors on TB epidemiology and individual patient management.
Even though Papua New Guinea (PNG) is one of the high TB burden countries in the South Pacific, not much data on the local epidemiology of TB exists. Apart from the urgent need to fill such evidence gaps, the country also provides an interesting platform for TB research, considering its population genetic diversity and its isolation in the past. The presented project aimed at providing baseline data about the molecular epidemiology of tuberculosis from specific sites in PNG, including drug resistance and the population structure of Mtb.
Between July and December 2010, active TB case detection surveys were conducted in the catchment area of two health centres in PNG: around Sausi health centre in Madang Province and around East-Cape health centre in Milne Bay Province. Each household in the catchment area was screened for people with chronic productive cough aged 15 years and above. Of household members with chronic productive cough not having received TB treatment yet, three sputum samples were collected. Subsequently, samples were analysed by light microscopy to diagnose pulmonary TB by detecting the presence of acid fast bacilli. Around Sausi, 24 so far undetected pulmonary TB cases were identified, whereas in East-Cape only one additional case was found, reflecting the differences in the performance of the control program between different sites in PNG. Active case detection as a complementary case detection approach turned out to be a useful tool to increase the case detection rate in certain areas, but appeared unsuitable to investigating the prevalence of drug resistance and the genetic background of Mtb. However, operational limitations did not allow obtaining better estimates on the real burden of TB in the country in the frame of our study.
From November 2010 to July 2012, passive case detection was conducted in three provincial hospitals of PNG: Modilon General Hospital in Madang Province, Goroka General Hospital in Eastern Highlands Province and Alotau Provincial Hospital in Milne Bay Province. Three sputum samples were collected from TB suspects aged 15 years and above and subsequently analysed by light and fluorescent microscopy. Furthermore, the level of drug resistance as well as the genetic background of M. tuberculosis strains was determined and findings compared between sites. Of 225 passive case detection samples grown in culture, 212 samples could successfully be tested for drug susceptibility. Overall, 10.8% (23/212) strains were found to be resistant to at least one of the first-line drugs streptomycin, rifampicin, isoniazid, pyrazinamide or ethambutol. Differences between study sites in any type of DR were marginal and ranged from 10% to 12%. Multi-drug resistant (MDR) TB was found in 2.8% (6/212) of cases, the highest percentage of MDR TB being found in Alotau (4.6% compared to 2.2% in Madang and 1.8% in Goroka). These results show a significant amount of DR TB being present in all three sites investigated. It is therefore crucial to make diagnosis of DR TB and second-line treatment more widely available in the country to decrease the delay of diagnosing DR TB as well as the duration of possible transmission and to avoid further DR development.
Genotyping of Mtb could successfully be conducted of 147 samples. These strains could be classified into three of the so far seven known lineages of Mtb: 75/147 (51.0%) of samples belonged to lineage 4 (European-American lineage), 67/147 (45.6%) to lineage 2 (East-Asian lineage) and 5/147 (3.4%) to lineage 1 (Indo-Oceanic lineage). All three lineages were detected in all three sites, but the individual lineage compositions varied significantly between sites (p<0.001). In Madang, lineage 4 was the most prevalent (76.6%), whereas in Alotau lineage 2 was dominant (84.4%). In Goroka, a trend towards a higher prevalence of lineage 2 (60.5%) was found, but the difference between lineage 2 and lineage 4 was not statistically significant and not as high as in Alotau. Lineage 1 was generally only rarely found (5/147). The overall lineage composition found in PNG is similar to what can be observed globally: modern lineages (e.g. lineages 2 and 4) have more successfully spread around the globe and are more prevalent than ancient lineages (e.g. lineage 1). Further molecular subtyping by large sequence polymorphisms, Luminex based SNP-typing and whole genome sequencing revealed that a single introduction of lineage 2 into PNG with a subsequent clonal expansion is most probable, whereas for lineages 4 and 1, several introductions are likely. All three lineages appear to have undergone to a certain degree PNG-specific evolution.
The present study is the first directly comparing DR and Mtb genotyping data between different sites of PNG, discovering the presence of significant differences in DR prevalence and Mtb lineages. However, the reason for these observed differences has yet to be determined. The questions about how and from where TB was introduced into PNG in the first place, and about details on transmission dynamics of TB remain to be answered.
Besides Mtb, also non-tuberculous mycobacteria (NTM) could be detected in a few sputum samples of study patients. NTM were detected in 4% (9/225) of sputum samples grown in culture. Five of these turned out to be samples containing NTM only, the detected species being Mycobacterium fortuitum, Mycobacterium terrae and Mycobacterium intracellulare. Four isolates contained both, Mtb and Mycobacterium avium or Mtb and Mycobacterium intracellulare, respectively. To our knowledge this is the first study describing the presence of NTM in PNG.
A key component of the National TB Program should be the detection and continuous monitoring of DR TB to stop transmission. Our data emphasizes the need of a GeneXpert system for DR diagnosis and monitoring in each province of PNG. Priority should be given to those provinces with an increased proportion of DR TB such as Milne Bay Province. In addition, an in-country capacity to perform TB culturing and DST is urgently required. Implementing both recommendations could assist in achieving a reduction of time to diagnosis of DR TB and consequently decrease the risk of MDR TB transmission.