Pyrazinamide resistance in Mycobacterium tuberculosis

Whitfield, Michael Gordon (2016-12)

Thesis (PhD)--Stellenbosch University, 2016.

Thesis

ENGLISH SUMMARY: Pyrazinamide (PZA) is one of the drugs included in the standardized TB treatment regimen for both drug susceptible and resistant TB. PZA has a unique ability to target persister bacilli and since its inclusion has reduced TB treatment duration to 6 months. However, drug susceptible testing (DST) for PZA is not routinely performed due to the challenges associated with the DST (acidic media; false resistance and inoculum size). Significant evidence indicates that mutations in pncA (gene encodes for pyrazinamidase (PZase)) are the primary mechanism of resistance to PZA. Sequencing studies reveal that mutations occur across the entire length of the gene, thus sequencing of the entire gene is required to capture all possible resistance conferring mutations. A rapid molecular diagnostic is required to routinely test for PZA resistance, especially given the continued recommendation of its inclusion in the shortened MTB treatment regimen. PZA is currently being considered for inclusion of new treatment regimens which also include the newer anti-TB drugs. However, we need to better understand the prevalence of PZA resistance globally and across the spectrum of drug resistant groups to determine the utility of this drug in current and future treatment regimens. This PhD thesis aims to further the current knowledge of PZA resistance. The prevalence of PZA resistance regionally and globally is largely unknown. Our systematic review collates the PZA resistance data in the literature from all TB cases (16.2%); in high-risk MDR-TB (41.3%) and in MDR-TB (60.5%). These findings caution against relying on PZA in current and future TB drug regimens, especially in MDR-TB patients. This review also identified more than 600 unique SNPs across the entire length of the pncA gene as the causal mechanism of resistance. This highlights the complexity of the challenge to develop future genetic based PZA resistance tests. From this review it was evident that the prevalence of PZA resistance across a spectrum of drug resistant isolates is unknown. To address this knowledge gap 775 clinical isolates from South Africa were classified into groups according to resistance profile (ranging from pan-susceptible to XDR-TB) and their association with PZA resistance was determined. The prevalence of PZA resistance in each group was: 0% in the pan-susceptible; 2% in the INH-mono; 7.5% in the RIF-mono; 39.3% in the MDR-TB and 96.8% in the XDR-TB group. A statistically significant increase in PZA resistance was observed as the number of resistance markers increases. We concluded that PZA DST should be performed when considering its inclusion in treatment of patients with rifampicin-resistant TB or MDR-TB. This study also showed an excellent correlation between pncA genotype and PZA phenotype. The resulting sensitivity and specificity was 95% and 99%, respectively, thereby confirming the utility of pncA sequencing as a rapid PZA susceptibility test. In an attempt to understanding discordance between genotype and phenotype we characterised pncA mutations which did not confer PZA resistance at a susceptibility breakpoint of 100 μg/ml pyrazinamide in MGIT. A total of 10 non-resistance conferring mutations were identified in collaboration with the U.S. Centers for Disease Control and Prevention. These results will be essential for the interpretation of pncA sequencing results to guide treatment. In a recent paper the authors suggested that pncA mutation have a fitness cost which would impact on the transmission of drug resistant M. tuberculosis strains. To test this hypothesis we compared the growth rates of clinical isolates and in vitro mutants (wild type and SNPs) and clinical isolates which have a large pncA deletion. No difference was observed in the time to positivity in MGIT media when comparing wild type and strains harbouring pncA mutations, thereby suggesting that pncA mutation do not have a significant fitness cost. However, strains harbouring large pncA mutations showed a significant growth deficit (p-value: <0.001) in vitro. However, this fitness cost did not prevent these strains from transmitting in the community. In an attempt to simplify the diagnosis of PZA susceptibility, we investigated the diagnostic utility of a LATE-PCR based technique in combination with fluorescence probes to identify mutations along the entire length of the pncA gene. Using this method the sensitivity and specificity was 98.7% and 98.4%, respective, suggesting that this method could be using in routine laboratories due to the single tube format. Together this body of research has addressed critical knowledge gaps and has introduced a methodology that could simplify the diagnosis of PZA resistance.

AFRIKAANSE OPSOMMING: Pyrazinamide (PZA) is ‘n middel wat gebruik word in die standaard behandeling van beide sensitiewe en weerstandige tuberkulose (TB). PZA beskik oor die unieke vermoë om langlewende bakterieë te teiken en het sedert die insluiting daarvan in die behandelingsregime die duur van TB behandeling verminder tot 6 maande. Middel weerstandige toetsing (MWT) vir PZA word selde uitgevoer as gevolg van uitdagings wat geassosieer word met MWT (suur media, vals positiewe weerstandigheid en inokulum grootte). Betekenisvolle resultate het gewys dat mutasies in pncA (die geen wat kodeer vir pyrazinamidase (PZase)) die primêre meganisme is wat lei tot PZA weerstandigheid. Volgordebepaling het gewys dat mutasies oor die hele lengte van die geen voorkom en gevolglik moet die hele geen se volgorde bepaal word om al die moontlike mutasies wat kan bydra tot weerstandigheid te identifiseer. ‘n Vinnige molekulêre diagnostiese toets word benodig om op ‘n gereelde basis te kan toets vir PZA weerstandigheid, veral gegewe die volgehoue aanbeveling dat PZA ingesluit moet word in die verkorte TB behandeling. PZA word tans oorweeg vir die insluiting in nuwe behandelingsregime wat ook nuwer anti-TB middels sal insluit. Ons benodig egter verbeterde kennis omtrent die globale voorkoms van PZA weerstandigheid oor die spektrum van weerstandige groepe om vas te stel wat die verbruik van die middel in die huidige en toekomstige behandeling is. Die doel van hierdie PhD tesis is om die huidige kennis van PZA weerstandigheid uit te brei. Die plaaslike en globale voorkoms van PZA weerstandigheid is grotendeels onbekend. Ons sistematiese oorsig artikel vat die beskikbare PZA weerstandige data in die literatuur oor alle TB gevalle (16.2%), in hoë risiko multiweerstandige TB (MDR-TB) (41.3%) en in MDR-TB (60.5%) gevalle saam. Hierdie bevindinge waarsku teen die betroubaarheid van PZA in huidige en toekomstige TB behandeling, veral in MDR-TB pasiënte. Die studie het meer as 600 unieke enkelnukleotied polimorfismes oor die lengte van die pncA gene geïdentifiseer as die oorsaak van weerstandigheid. Dit beklemtoon die kompleksiteit van die uitdaging om toekomstige geneties gebaseerde PZA weerstandige toetse te ontwikkel. Hierdie oorsig artikel het ook daarop gedui dat die voorkoms van PZA weerstandigheid oor ‘n spektrum van middel weerstandige isolate onbekend is. Om hierdie gebrek in kennis aan te spreek is 775 kliniese isolate van Suid-Afrika geklassifiseer in groepe volgens hul weerstandige profiele (wat strek van middel vatbare to XDR-TB) en is hul assosiasie met PZA weerstandigheid bepaal. Die voorkoms van PZA weerstandigheid in elke groep was soos volg: 0% in die middel vatbare; 2% in die INH-enkelweerstandig; 7.5% in die RIF-enkelweerstandig; 39.3% in die MDR-TB en 96.8% in die XDR-TB groep. Soos die aantal weerstandige merkers toeneem is ‘n beduidende verhoging in PZA weerstandigheid waargeneem. Ons het beveel aan dat PZA MWT uitgevoer moet word wanneer dit oorweeg word om pasiënte met rifampisien-weerstandige TB of MDR-TB ook met die middel PZA te behandel. Hierdie studie dui ‘n uitstekende korrelasie tussen pncA genotipes en PZA fenotipes aan. Die sensitiwiteit en spesifisiteit was onderskeidelik 95% en 99%, wat die bruikbaarheid van pncA volgordebepaling bevestig as ‘n vinnige PZA sensitiewe toets. In ‘n poging om die teenstrydigheid tussen genotipe en fenotipe te verstaan, het ons pncA mutasies wat nie PZA weerstandigheid veroorsaak by ʼn konsentrasie van 100 μg/ml nie, in meer diepte ondersoek. ‘n Totaal van 10 mutasies wat nie bydra tot weerstandigheid nie is geidentifiseer in samewerking met die VSA se Sentrum vir Gesondheidsbeheer en Voorkoming. Laasgenoemde data is krities vir die interpretasie van pncA volgordebepaling resultate om behandeling voor te kan skryf. In ‘n onlangse artikel het die outeurs voorgestel dat die fiksheid koste van pncA mutasies ‘n impak het op die oordrag van middel weerstandige M. tuberculosis stamme. Om hierdie hipotese te toets het ons die groei tempo van kliniese isolate en in vitro mutasies (wilde tipe en enkelnukleotied polimorfismes) en kliniese isolate wat ‘n groot pncA delesie bevat vergelyk. Geen verskil is opgemerk in die tyd tot positiwiteit in MGIT media wanneer wilde tipe met die stam wat pncA mutasies bevat vergelyk is nie, wat daarop dui dat die pncA mutasies nie ‘n merkwaardige effek op fiksheid het nie. Stamme wat egter groot pncA mutasies bevat, het ‘n merkwaardige groei vertraging (p-waarde: <0.001) in vitro getoon. Hierdie effek het egter het nie die vermoë van hierdie stamme verander om oorgedra te word aan die gemeenskap nie. In ‘n poging om die diagnose van PZA weerstandigheid te vereenvoudig het ons die diagnostiese gebruik gemaak van ‘n “LATE-PCR” gebasseerde tegniek kombineer met fluoresserende merkers om mutasies oor die hele pncA geen te identifiseer. Hierdie metode se sensitiwiteit en spesifisiteit was onderskeidelik 98.7% en 98.4%, wat daarop dui dat hierdie metode gebruik kan word in roetine laboratoriums, veral omdat daar van ‘n enkel-buis formaat gebruik gemaak word. Tesame het hierdie navorsing die kritieke gapings in kennis aangevul en het dit ‘n metode voorgestel wat die diagnose van PZA weerstandigheid kan vereenvoudig.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/100389
This item appears in the following collections: