|FULL LENGTH ARTICLE
|Year : 2016 | Volume
| Issue : 4 | Page : 412-416
Fluoroquinolone consumption and -resistance trends in Mycobacterium tuberculosis and other respiratory pathogens: Ecological antibiotic pressure and consequences in Pakistan, 2009–2015
S Shakoor1, S Tahseen2, K Jabeen1, R Fatima3, FR Malik1, AH Rizvi2, R Hasan1
1 Aga Khan University, Karachi, Pakistan
2 National Reference Laboratory (NRL), National TB Control Programme, Pakistan
3 National TB Control Programme, Pakistan
|Date of Web Publication||14-Feb-2017|
Department of Pathology, Laboratory Medicine, and Paediatrics, Aga Khan University, Stadium Road, P.O. Box 3500, Karachi 74800
Source of Support: None, Conflict of Interest: None
Objective/background: Fluoroquinolones (FQs) are important anti-tuberculous drugs for the treatment of multidrug-resistant (MDR) tuberculosis. Resistance to FQs leads to fewer options for treatment of tuberculosis (TB), and infection with such strains may also require longer treatment duration. Trends of resistance in Mycobacterium tuberculosis (MTB) are indicators of MTB-resistance evolution. Drivers of such resistance need to be understood and studied to inform preventive strategies.
Methods: Here, we present FQ-resistance rates and trends in Pakistan from 2010 to 2015 and compare rates with FQ-consumption data and rates in other community pathogens.
Results: Our results reveal a recent decrease in FQ-resistance rates in MTB, but an increase in resistance for Haemophilus influenzae and Shigella spp. Correlation of FQ resistance with FQ consumption at the population level was weak for MTB, although strong associations were noted for H. influenzae and Shigella spp.
Conclusion: We discuss the possible reasons for the decrease in resistance rates in TB, putative drivers of resistance other than volume of FQ consumption, and the possible impact of the National Tuberculosis Programme and drug regulatory activities.
Keywords: Consumption, Defined daily dose, Fluoroquinolone, Resistance, Selective pressure, Tuberculosis
|How to cite this article:|
Shakoor S, Tahseen S, Jabeen K, Fatima R, Malik F R, Rizvi A H, Hasan R. Fluoroquinolone consumption and -resistance trends in Mycobacterium tuberculosis and other respiratory pathogens: Ecological antibiotic pressure and consequences in Pakistan, 2009–2015. Int J Mycobacteriol 2016;5:412-6
|How to cite this URL:|
Shakoor S, Tahseen S, Jabeen K, Fatima R, Malik F R, Rizvi A H, Hasan R. Fluoroquinolone consumption and -resistance trends in Mycobacterium tuberculosis and other respiratory pathogens: Ecological antibiotic pressure and consequences in Pakistan, 2009–2015. Int J Mycobacteriol [serial online] 2016 [cited 2020 Aug 8];5:412-6. Available from: http://www.ijmyco.org/text.asp?2016/5/4/412/200124
| Introduction|| |
Fluoroquinolones (FQs) are essential drugs for the treatment of multidrug-resistant (MDR) tuberculosis (TB). Increasing rates of resistance worldwide have sparked concerns over the driving factors of FQ resistance in vivo and in the environment . Environmental factors contributing to resistance include residual concentration of FQs in food, water, animal feed, and also subinhibitory concentrations of FQs used to treat various community acquired infections . This phenomenon may be especially pronounced in countries with weak healthcare systems. Such environmental stressors have also led to increases in FQ-resistance rates in other community pathogens, such as Salmonella spp., Shigella spp., Neisseria gonorrhoeae, and Streptococcus pneumoniae . A directly proportional relationship is expected between the volume of FQ use in the community at the population level (selective pressure in the environment) and emergence, maintenance, and rise in resistance rates of many organisms . This was demonstrated for S. pneumoniae , but has not been studied in Mycobacterium tuberculosis (MTB).
We previously reported high and increasing FQ-resistance rates among MTB isolates from all over Pakistan from 2005 to 2009 ,. Here, we report subsequent resistance data and correlate population-level FQ-consumption rates in Pakistan with resistance trends in MTB and other pathogens.
| Methods|| |
For resistance rates and trends, laboratory records were examined from the Aga Khan University (AKU) clinical microbiology laboratory from 2010 to 2015 for S. pneumoniae and Haemophilus influenzae among respiratory pathogens, N. gonorrhoeae (a venereal disease pathogen), Salmonella Typhi, Salmonella Paratyphi, Shigella spp., and Vibrio cholerae among gastrointestinal pathogens, where the use of FQ is common, and for the MTB complex. The AKU clinical laboratories received samples on physician request from all over Pakistan, with more than 200 collection units located throughout the country. Moreover, the mycobacteriology laboratory served as the WHO Supranational Laboratory. Following inception in 2009 of the National TB-reference Laboratory (NRL), and later in 2013 as a result of its expansion to perform culture and drug-susceptibility testing services for clinical management of the patients, the number of samples received and processed at the NRL increased substantially. Therefore, from 2013 to 2015, data from the NRL was also included for MTB.
Laboratory records were retrieved and duplicates removed for FQ resistance reported among pathogens of interest for the years 2010–2015. FQ-susceptibility testing for MTB was performed with the proportion method using ofloxacin (2 μg/mL) against MTB as recommended by the Clinical Laboratory Institute Standards (CLSI) . MTB H37Rv was used as a control for each batch of susceptibility testing. Susceptibility testing of community acquired bacterial pathogens was performed in accordance with CLSI guidelines by disk diffusion (the Kirby-Bauer method) . Staphylococcus aureus ATCC 25923 and Escherichia coli ATCC 25922 were used as control organisms.
FQ-sales data from Pakistan in total number of grams sold per year was obtained from IMS Health, Pakistan (IMS Health Inc., Karachi, Pakistan). The IMS Health database is a multisourced national database that compiles sales data from registered pharmacies, prescribing physicians, manufacturers, and wholesalers.
Data were analyzed and bar charts created in MS Excel (Microsoft Corp., Redmond, WA, USA). Coefficient-of-regression (r2) values for trends were calculated in MS Excel (Microsoft Corp.), with an r2 ≥0.65 representing a significant trend . Resistance trends were then correlated with FQ-sales data for MTB and for community pathogens to show a significant FQ-resistance trend.
To correlate resistance rates with FQ sales, we converted sales data (from grams sold annually) to defined daily dose (DDD) per 1000 inhabitants for each year from 2009 to 2014. DDD conversion was based on 2015 WHO guidelines for anatomical therapeutic chemical (ATC) classification  and DDD assignment for ciprofloxacin (oral and parenteral), levofloxacin (oral and parenteral), moxifloxacin (oral and parenteral), ofloxacin (oral and parenteral), enoxacin (oral), sparfloxacin (oral), norfloxacin (oral), pefloxacin (oral), and gatifloxacin (oral and parenteral). To convert to 1000 inhabitants per year, estimates for the population of Pakistan for each year were obtained from the World Bank. We also converted resistance data to the logarithm of the odds of resistance (odds=proportion of resistance organisms (p)/1−p; odds converted to the natural logarithm of the odds). DDD-log odds resistant pairs were examined with 1-year gaps (e.g., 2009 DDDs correlated with 2010 resistance rate), given that previous population-level data showed that the effect of change in consumption was observed with a lag of ≥1year . Linear regression (least squares) was then applied to obtain coefficients of determination (r2) and p values.
The study was exempted from ethical review by the AKU Ethical Review Committee.
| Results|| |
Resistance rates in MTB
During the study period, 18,776 MTB strains were isolated, including 8492 (45.2%) that were MDR. The trend in MDR cases showed a non-significant increase (R2 = 0.4) from 2010 to 2015; however, a decrease in MDR rates was observed in 2015 ([Figure S1]).
FQ-resistance trends in MTB indicated variable rates (an increase from 2010 to 2011, followed by a consistent rate and a recent decrease in 2015; all MTB, R2 = 0.1; [Figure 1]). Among non-MDR MTB strains, FQ resistance increased from 10.3% (n = 214/2059) in 2010 to 16.8% (n = 250/1487) in 2015 (linear R2 = 0.1), while an insignificant decrease in FQ resistance in MDR TB was observed from 54.6% (n = 691/1266) in 2010 to 52.3% (n = 591/1129) in 2015 (linear R2 = 0.5).
|Figure 1: Fluoroquinolone-resistance rates and trends in Pakistan from 2010 to 2015. Coefficients of regression (r2) are shown for each linear trend.|
Click here to view
Resistance rates in community pathogens
Over the 6-year study period, FQ-resistance rates remained consistently high in Salmonella Typhi and Paratyphi and N. gonorrhoeae and showed variability with no consistent trends in V. cholerae and S. pneumoniae (linear R2 = 0.3; [Figure S2]).
However, in H. influenza , FQ resistance increased consistently from 9.1% in 2010 to 29% in 2015 (linear R2 = 0.9), while in Shigella spp. FQ resistance increased from no detectable resistance in 2010 to 22% FQ resistance in 2015 (linear R2 = 0.9). [Figure 1] shows resistance rates and trends in MTB (all strains, MDR strains, and non-MDR strains), as well as in H. influenzae and Shigella spp.
Fluoroquinolone DDDs based on sales from registered pharmacies are shown in [Figure 2]. Sales increased steadily from 2009 to 2013, with a 7% decrease in consumption in 2014. Nevertheless, an increasing trend was observed, with R2 = 0.8.
|Figure 2: Fluoroquinolone consumption in adult DDDs per 1000 inhabitant days for Pakistan from 2009 to 2014. DDDs were derived from sales data (in grams sold) by applying the WHO ATC classification for fluoroquinolones. A rising trend was observed, albeit a 7% decrease was apparent for 2014. ATC = anatomical therapeutic chemical; DDD = defined daily dose; WHO =World Health Organization.|
Click here to view
Correlation of resistance rates with consumption
Linear regression of FQ DDDs per 1000 inhabitant-year and FQ-resistant MTB ([Figure 3]) showed a weak correlation between increases in consumption and resistance (R2 = 0.2). In contrast, regression of H. influenzae and Shigella spp. both showed positive increases along with increases in FQ consumption ([Figure 4]).
|Figure 3: Regression of the log odds of fluoroquinolone resistance in MTB shows a weak correlation (p = 0.18). MTB = Mycobacterium tuberculosis.|
Click here to view
|Figure 4: Regression of FQ consumption with (A) Haemophilus influenzae and (B) Shigella spp. resistance rates shows strong and significant positive correlations. FQ = fluoroquinolone.|
Click here to view
| Discussion|| |
Density of antibiotic consumption has long been known to increase drug-resistance rates in a number of microorganisms. In our population, there appeared to be a strong correlation between FQ consumption and increasing resistance rates in H. influenzae and Shigella spp.; however, for MTB, the relationship between FQ resistance and consumption was weak.
While the discrepancy is surprising, there is evidence that at a population level, MTB does not follow the same model of resistance emergence and maintenance as other community organisms . The genetic basis for resistance in MTB is also unique, as it does not possess plasmids . Furthermore, FQ-resistance strains have a fitness cost  that has not been correlated with survival and R0 rates in the population. Moreover, the fitness costs may be different for MDR and drug-sensitive strains . In Taiwan, despite widespread uncontrolled use of FQ in the community, from 1995 to 2003, resistance was only observed in MDR strains, indicating that different triggers (rather than volume of FQ use) are likely to be responsible for maintenance of FQ resistance in the MTB resistome . It is hypothesized that at the individual level, especially in MDR patients, FQ resistance is driven by inappropriate use of FQ rather than consumption of FQ in the population .
We also showed a recent decrease in MTB FQ-resistance rates. Several improvements at the population level are likely to have been responsible for this decrease. The National Tuberculosis Program (NTP) has taken measures to improve the treatment and care of TB patients in NTP-partner clinics by enhanced detection ,, with adequate treatment of TB patients achieving up to 88% treatment-success rates , thereby decreasing inappropriate use of FQ among patients at risk of TB or with active TB. More importantly, increases in Programmatic Management of Drug-resistant TB (PMDT) efforts resulted in streamlined second-line treatment regimens . Since we did not observe a proportionate decrease in FQ resistance in non-MDR strains, the decrease is likely driven by the upscale of PMDT.
The decline in FQ consumption in 2014 is likely related to implementation of the Drug Regulatory Authority of Pakistan (DRAP) Act 2012 , which led to a reduction in unregulated FQ usage; over six brand applications were rejected by DRAP due to concerns about quality ,. Furthermore, FQ-consumption data revealed a recent decrease in sales in 2010. This decrease was driven by a decrease in sales of enoxacin, sparfloxacin, and gatifloxacin, which are more likely to have been used inappropriately, because they do not find applications in the mainstream medical community in Pakistan. This lends credence to the hypothesis that MTB FQ resistance is driven by inappropriate FQ usage in the community rather than the volume of use and overall selective-ecological pressure.
Our study had limitations. MTB strains in this study were recovered from samples from all over Pakistan, while some data from private laboratories was not available for analysis. However, the resistance rates shown were representative of the MDR- and FQ-resistance rates in the country, as data from the two largest reference centers was included. The consumption data was based on adult and not pediatric dosages, and, therefore, was only an estimate and not the actual consumed DDD/1000 inhabitant-year. Moreover, our data represented the population level and, therefore, was prone to ecological bias. Both individual and population-level data are required to overcome the inherent bias when correlating antibiotic consumption with resistance rates .
| Conclusions|| |
We showed that MTB FQ resistance was unrelated to the volume of FQ consumption in the community and was more likely driven by other factors, such as inappropriate use at the individual level. Prospective ecological studies that combine individual DDDs with MTB FQ resistance are required to develop a better understanding of drivers of FQ resistance and ecological antibiotic pressure in MTB.
| Conflicts of interest|| |
All contributing authors declare no conflicts of interest.
| Acknowledgments|| |
FQ-consumption/sales data was obtained from IMS Health. We would also like to acknowledge Mr. Mahmood Qadir from the NRL and Ms. Samreen Shafiq, Mr. Khalid Wahab, and Dr. Joveria Farooqi from the AKU clinical laboratory for their efforts in generating and maintaining resistance data.
| Appendix A. Supplementary data|| |
Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ijmyco.2016.07.008.
| References|| |
D.E. Low, Fluoroquinolones for treatment of community acquired pneumonia and tuberculosis: putting the risk of resistance into perspective, Clin. Infect. Dis. 48 (2009) 1361–1363.
K. Jabeen, S. Shakoor, R. Hasan, Fluoroquinolone-resistant tuberculosis: implications in settings with weak healthcare systems, Int. J. Infect. Dis. 32 (2015) 118–123.
D.J. Austin, K.G. Kristinsson, R.M. Anderson, The relationship between the volume of antimicrobial consumption in human communities and the frequency of resistance, Proc. Natl. Acad. Sci. U.S.A. 96 (1999) 1152–1156.
D.K. Chen, A. McGreer, J.C. de Azavedo, et al, Decreased susceptibility of Streptococcus pneumoniae to fluoroquinolones in Canada, N. Engl. J. Med. 341 (1999) 233–239.
K. Jabeen, S. Shakoor, S. Chishti, et al, Fluoroquinoloneresistant Mycobacterium tuberculosis
, Pakistan, 2005–2009, Emerg. Infect. Dis. 17 (2011) 564–566.
Y. Rafiq, K. Jabeen, R. Hasan, et al, Fluoroquinolone resistance among Mycobacterium tuberculosis
strains from Karachi, Pakistan: data from community-based filed clinics, Antimicrob. Agents Chemother. 55 (2011) 929–930.
Clinical Laboratory Standards Institute (CLSI), Susceptibility testing of Mycobacteria, Nocardiae, and other aerobic Actinomycetes; approved standard – second ed., CLSI Document M24–A2, Wayne, PA, 2011.
Clinical Laboratory Standards Institute (CLSI), Performance standards for antimicrobial-susceptibility testing; twentysecond ed., CLSI Supplement M100–S22, Wayne, PA, 2012.
A.C. Bryhn, P.H. Dimberg, An operational definition of a statistically meaningful trend, PLoS One 6 (2011) 19241. doi: 10.1371/journal.pone.0019241.
M. Lipsitch, M.H. Samore, Antimicrobial use and antimicrobial resistance: a population perspective, Emerg. Infect. Dis. 8 (2002) 347–354.
S.H. Gillespie, Antibiotic resistance in the absence of selective pressure, Int. J. Antimicrob. Agents 17 (2001) 171–176.
B. Eilertson, F. Maruri, A. Blackman, et al, High proportion of heteroresistance in gyrA and gyrB in fluoroquinoloneresistant Mycobacterium tuberculosis
clinical isolates, Antimicrob. Agents Chemother. 58 (2014) 3270–3275.
T. Huang, C.M. Kunin, S.S. Lee, et al, Trends in fluoroquinolone resistance of Mycobacterium tuberculosis
complex in a Taiwanese medical centre: 1995–2003, J. Antimicrob. Chemother. 56 (2005) 1058–1062.
R. Fatima, E. Qadeer, S.G. Hideraker, et al, Can the number of patients with presumptive tuberculosis lost in the general health services in Pakistan be reduced?, Int J. Tuberc. Lung Dis. 19 (2015) 654–656.
R. Fatima, E. Qadeer, D.A. Enarson, et al, Success of active tuberculosis case detection among high-risk groups in urban slums in Pakistan, Int. J. Tuberc. Lung Dis. 18 (2015) 1099–1104.
The Gazette of Pakistan, An act to provide for the establishment of Drug Regulatory Authority of Pakistan, Act No XXI of 2012. <http://www.dra.gov.pk
> 2012 (last accessed 02.07.2016).
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
|This article has been cited by|
||ß-Lactamase-non-producing ampicillin-resistant Haemophilus influenzae is acquiring multidrug resistance
| ||Shogo Yamada,Shoji Seyama,Takeaki Wajima,Yuna Yuzawa,Masumi Saito,Emi Tanaka,Norihisa Noguchi |
| ||Journal of Infection and Public Health. 2019; |
|[Pubmed] | [DOI]|
||Integrating tuberculosis and antimicrobial resistance control programmes
| ||Rumina Hasan,Sadia Shakoor,Johanna Hanefeld,Mishal Khan |
| ||Bulletin of the World Health Organization. 2018; 96(3): 194 |
|[Pubmed] | [DOI]|
||Fluoroquinolone resistance in multidrug-resistant Mycobacterium tuberculosis independent of fluoroquinolone use
| ||Yang Che,Qifa Song,Tianchi Yang,Guohua Ping,Mei Yu |
| ||European Respiratory Journal. 2017; 50(6): 1701633 |
|[Pubmed] | [DOI]|