Leprosy Drug Resistance Surveillance in Colombia: The Experience of a Sentinel Country

An active search for Mycobacterium leprae drug resistance was carried out, 243 multibacillary patients from endemic regions of Colombia were included from 2004 to 2013 in a surveillance program. This program was a World Health Organization initiative for drug resistance surveillance in leprosy, where Colombia is a sentinel country. M. leprae DNA from slit skin smear and/or skin biopsy samples was amplified and sequenced to identify mutations in the drug resistance determining region (DRDR) in rpoB, folP1, gyrA, and gyrB, the genes responsible for rifampicin, dapsone and ofloxacin drug-resistance, respectively. Three isolates exhibited mutations in the DRDR rpoB gene (Asp441Tyr, Ser456Leu, Ser458Met), two in the DRDR folP1 gene (Thr53Ala, Pro55Leu), and one isolate exhibited mutations in both DRDR rpoB (Ser456Met) and DRDR folP1 (Pro55Leu), suggesting multidrug resistance. One isolate had a double mutation in folP1 (Thr53Ala and Thr88Pro). Also, we detected mutations outside of DRDR that required in vivo evaluation of their association or not with drug resistance: rpoB Arg505Trp, folP1 Asp91His, Arg94Trp, and Thr88Pro, and gyrA Ala107Leu. Seventy percent of M. leprae mutations were related to drug resistance and were isolated from relapsed patients; the likelihood of relapse was significantly associated with the presence of confirmed resistance mutations (OR range 20.1–88.7, p < 0.05). Five of these relapsed patients received dapsone monotherapy as a primary treatment. In summary, the current study calls attention to M. leprae resistance in Colombia, especially the significant association between confirmed resistance mutations and relapse in leprosy patients. A high frequency of DRDR mutations for rifampicin was seen in a region where dapsone monotherapy was used extensively.


Introduction
Leprosy is a chronic infectious disease caused by M. leprae that results in disfigurement, exclusion from society and, often, in or from poverty [1]. Prior to the 1940s, leprosy was considered an untreatable disease. However, in the early 1940s, diamino-diphenyl sulfone (DDS; also known as dapsone), was shown to successfully treat leprosy. At that time, dapsone was used for treatment of leprosy until lesions were cured or until the bacillary index (BI) became negative, a process that took decades [2,3] and one that often resulted in patients relapsing after the termination of treatment [3]. In the 1980's, the World Health Organization (WHO) recommended dapsone, rifampicin and clofazimine as the primary agents for multidrug therapy [4]. The frequency of clinical relapse due to dapsone resistance after monotherapy, is higher than the results of dapsone resistance tests in mouse foot pad since few strains exhibited full resistance [5]. Nevertheless, resistance to low levels of dapsone is detected in mouse footpads, indicative of inadequate treatment. Dapsone is a bacteriostatic antibiotic and the persistence of bacilli may be one cause of relapse in dapsone monotherapy [6].
WHO, concerned with clinical relapses, emergence, and transmission of organisms with low levels of resistance to dapsone, recommended the addition of rifampicin, a bactericidal drug taken from the TB treatment stock [4]. Thus, multidrug therapy (MDT) for leprosy was initiated in the early 1980s [3,4]. To treat paucibacillary leprosy (PB), the MDT combination is dapsone and rifampicin. To treat multibacillary (MB) leprosy, clofazimine was included as a third drug [4,6]. Currently, WHO recommends an MDT duration of 1 year for MB leprosy and 6 months for PB leprosy. Globally, these guidelines apply to more than a million patients, having been adapted to a variety of resource-constrained settings over the past decades [3].
In Colombia, MDT was introduced in 1985 when the prevalence of leprosy was 5.5 cases per 10,000 people. In 1997, Colombia achieved the goal of elimination proposed by WHO: a prevalence of less than 1/10,000. Although leprosy treatment in Colombia is based on WHO guidelines, some modifications are made according to the criteria of clinicians [7,8]. For example, the duration of treatment for PB patients may be extended for one year, and in the case of MB the treatment, up to 2 years, if lesions persist and the Bacillary index (BI) is positive [7,9]. Leprosy resistance in Colombia has been associated with relapses, especially in patients undergoing dapsone monotherapy [10,11].
When the molecular targets for DDS and rifampicin were identified, drug resistant bacterial strains with high resistance in mouse footpad systems were found to have mutations in genes encoding dihydropteroate synthase (dhps or folP1) and RNA polymerase B (rpoB) [12][13][14][15]. However, drug resistant strains exist that do not have mutations in known genetic targets [3,4]. While mutations that confer low levels of resistance have not been identified, such low-level resistance is seen in wild-type strains [4,13]. The mechanism of clofazimine resistance has been under investigation for many years. Recently, it was shown that clofazimine is a substrate for bacterial type-2 NADH reductase [16]. However, mutations for genes involved in this redox process have not been ascribed to clinical clofazimine resistance [16]. MDT treatment of new leprosy cases is not routinely accompanied by monitoring for primary or secondary drug resistance [3,6]; surveillance and retreatment of dapsone monotherapy in patients undergoing MDT is practiced in only in a few settings [15,17]. Treatment of returning patients with additional rounds of MDT, without knowledge of eventual underlying drug resistance, is the default course of action in leprosy control programs [7,18,19]. Unfortunately, dwindling resources for leprosy does not allow for individualized treatment regimens and long-term follow-ups [7]. Research has shown that patients with a high pre-MDT bacteriological index (>4) are at risk of relapsing after MDT, with relapses occurring 6-16 years after the completion of treatment [12]. An earlier study showed that a high pre-MDT bacterial index is also a risk factor for relapse. Relapses were not due to drug resistance, as partial and full dapsone resistance was detected in only 3 of 18 relapses [20].
Thus, distinguishing relapses due to incomplete treatment, persistent bacilli, drug resistance and re-infection are challenges for clinicians and health care workers. The definition of relapse differs for those undergoing dapsone monotherapy and those receiving fixed, 2 year MDT (and the now shortened 1 year MDT) [19]. Reduction of BI is a slow process, dropping only one unit each year with MDT. Therefore, no proper end point for the completion of treatment exists and BI determination is not a universally practiced standard of care [19].
For better understanding of the extent of primary and secondary drug resistance, researchers worldwide have independently used molecular epidemiological approaches. In 2006, WHO called for leprosy drug resistance surveillance (DRS) in several sentinel countries [18,19]. This goal of this study was to report the DRDR mutations in leprosy patients from Colombia from a 10 year WHO recommended surveillance program.

Methods Patients
This molecular epidemiology survey was performed using 243 individual M. leprae DNA samples from Colombian leprosy patients from 2004 to 2013. Volunteer patients (each of whom signed a consent form) were recruited from a convenience sample of patients diagnosed with leprosy in fourteen departments of Colombia: Amazonas, Antioquia, Atlántico, Bolívar, Caquetá, Cesar, Chocó, Cundinamarca, Huila, Magdalena, Norte de Santander, Santander, Tolima and Valle. Patients, in various stages of treatment were enrolled with the aim to search for primary and secondary M. leprae drug resistance and included 33 new patients before treatment,136 currently undergoing treatment with not clinical improvement of lesions after three months of treatment, 36 post-treatment with positive bacillary index (BI) persistence, 4 nonadherent to treatment, and 34 relapse cases.
Bacterial DNA was extracted from skin biopsies using the Dneasy Blood & Tissue Qiagen kit (Invitrogen) according to the manufacturer's protocol [21,22]. Slit skin samples were collected from four different sites, with an emphasis on active lesions. Extracted DNA was stored at -20°C until processing.

Ethical considerations
The ethical committee of the Instituto Colombiano de Medicina Tropical-Universidad CES approved this project as the minimal risk, all the data analyzed were anonymized.

Amplification and sequencing analysis
Multiplex PCR was performed using primers suggested by WHO to amplify the DRDR regions of the rpoB, folP1, gyrA and gyrB genes [19]. For each amplification, 2 μl of purified M. leprae DNA (reference strain 4264) was used as a positive control. 2 μl of RNAse-free ultrapure water and 2 μl of DNA of Salmonella enteritidis (ICMT-99 strain) were each used as negative controls. DNA products were electrophoresed on agarose gel in 2.5% agarose gel inTE buffer and stained with ethidium bromide. A 20 pb molecular weight marker was used to determine amplicons size (Biorad, Cat No. 1708351).
PCR products were sequenced using on an ABI PRISM 3130xl genetic analyzer (Applied Bio systems) at the Colorado State University core sequencing facility. Chromatograms were analyzed with Chromas lite software [23] and compared with M. leprae reference sequences of antibiotic sensitive strains (in the leproma database http://genolist.pasteur.fr/Leproma/) by BLAST, as suggested by WHO [24].

Statistical analysis
The odds ratio and 95% CI of the relationship between likelihood of relapse and the presence of confirmed resistance mutations was calculated and significance assessed by Chi square test. P < 0.05 was considered significant.

Results
The geographical distribution of patients included in this study and the origin of M. leprae DRDR mutations are shown in Fig 1. The Andean region had 50% of M. leprae DRDR mutations. Of note, within this Andean region are the departments of Cundinamarca and Santander where two existing Colombian leprosaria, Agua de Dios and Contratación, are located.
The distribution of samples according to patient origin, gender and age is summarized in Table 1. Overall, the male to female ratio was 2.45:1; 72.2% of patients were male. The mean age was 52 years with a range of 9 to 91 years. Sequences showed that three isolates exhibited mutations in DRDR rpoB gene (Asp441Tyr, Ser456Leu, Ser458Met), and two in DRDR folP1 gene (Thr53Ala, Pro55Leu). One isolate exhibited mutations in both DRDR rpoB (Ser456Met) and DRDR folP1 (Pro55Leu). This result in this isolate suggested multidrug resistance (Table 2).
One isolate had a double mutation in folP1: Thr53Ala and Thr88Pro (outside of DRDR). Six mutations outside of DRDR that required an in vivo test for validation were found: rpoB Ala426Thre and Arg505Trp, folP1 Asp91His, Arg94Trp, and Thr88Pro, and gyrA Ala107Leu ( Table 2).
It is remarkable that all mutations found in DRDR for rpoB and folP1 were detected in seven relapsed patients and in one patient currently undergoing treatment (Table 2).
Further, we found isolates from three relapsed patients that exhibited mutations that will require further validation by mouse footpad method, in specialized laboratories-two isolates in rpoB and one isolate in folP1. The two mutations in rpoB (Ala426Thr and Arg505Trp) were close to the rifampicin resistance determining region located between nucleotides 432 and 458  [25][26][27][28][29]. The folP1 mutation (Asp91His) was located outside the dapsone resistance determining region, at positions 53 and 55 [28,30].
Regarding gyrA, no isolate displayed a mutation in the region determining fluoroquinolone resistance (nucleotides 89-91) [28][29][30][31]. However, one patient under treatment exhibited a mutation within gyrA (Ala107Leu) that also will require further validation by mouse footpad method, in specialized laboratories, for ofloxacin resistance.
In terms of frequency, five of the 134 cases with a positive amplification and DNA sequence (3.7%) had mutations in rpoB, five of 215 (2.32%) in folP1, and one of the 200 (0.5%) in gyrA. None contained a mutation in gyrB.
Four of 34 (11.8%) relapsed patients exhibited isolates with mutations in the DRDR rpoB and/or folP1. In contrast, one of 209 (0.47%) non-relapse patients contained isolates with mutations in DRDR folP1mutations.

Discussion
In this report, drug resistant M. leprae strains from 243 patients at various leprosy clinical and treatment stages were determined. The study utilized a patient group assembled over a tenyear period. While the threshold for alarm over drug resistance to MDT, locally and globally, has not been well defined, we demonstrate that single and multi-drug resistant strains are detectable in Colombia, albeit at low frequencies [10,11].
Of notable concern is the detection of a folP1 mutation previously related with drug resistance in one patient under MDT that had not previously undergone DDS treatment, and in two others completing one and two treatments of MDT. Similarly, three patients had rpoB drug resistance mutations-one exposed to rifampicin during the first cycle of MDT and two treated with more than one cycle of MDT. The finding that one patient had dapsone and rifampicin resistance, combined with a previous report concerning two cases of rifampicin resistance [10], suggests that the rifampicin resistance is not trivial in Colombia, a country considered in the post-elimination stage of leprosy (prevalence <1/10,000). Despite the known presence of dapsone resistance mutations, relapsed patients are commonly retreated with an MDT regimen that contains dapsone.
We observed novel mutations in relapsed patients ( Table 2) including folP1 Thr88Pro, rpoB Ala426Thre, Arg505Trp, and gyrA Ala107Leu. Clinical evidence of relapse and mutations DRDR genes, can help the clinicians to choose alternative treatments to the patient, such as a combination of rifampicin, ofloxacin and minocycline [32]. However, identifying drug resistance in leprosy patients implies clinical, logistical and financial challenges. Further, the change of treatment is not a common clinical practice since drug resistance tests are not available in the routine activities of the leprosy control programs in Colombia, much less in the majority of endemic regions [7]. Also, alternative treatments are not freely available in the leprosy control programs [7]. The results of this policy are evidenced in the seven cases we reported in this study which received regular MDT after relapse.
We found that Colombian dapsone monotherapy patients can relapse. Further, in patients carrying M. leprae drug resistant mutations, current results show an association between drug resistance mutations and relapse (OR range 20.1-88.7, p < 0.05). This result suggests that dapsone monotherapy patients require focused attention because they have the potential to act as a reservoir for dapsone resistant strains [10,11,33]. However, it is difficult to identify at-risk individuals for relapse. For example, patients associated with leprosaria have been found to have higher frequency of dapsone resistance in this and prior studies, similar to that reported in Brazil [32,33]. However, no single department or town was a leprosy hot spot in this cross sectional surveillance.
Mutations outside of the genetic regions associated DR had been described in other reports [34,35]. Uncharacterized polymorphisms and novel mutations have to be tested in vivo to test for resistance and add to our knowledge base. Phenotypic tests for drug resistance, such as those reported by Nakata et al [36], could be developed. Perhaps a saturated mutational analysis of target genes could be attempted to distinguish neutral polymorphisms from mutations leading to drug resistance [37,38].
Other targets for DR had been described, Monot et al [39] described rpoT for rifampicin, Singh et al [34] reported not clear explanation about rifampicin resistance phenotype of Airaku-3, a multi-drug resistance strain, the authors found two non-synonimous SNPs in transporter genes, ctpC and ctpl, however, a functional assay is required to determine if these variants confers any degree of rifampicin resistance.
This study had limitations and strengths. Due to sampling restrictions, conclusions regarding primary versus acquired resistance, and distinguishing relapse from reinfection, was not possible in the present study.
Current results are related with the initiative that WHO has formalized, an annual meeting for regions/countries serving as voluntary sentinel sites for leprosy drug resistance surveillance (DRS). The program has grown from its original focus on rifampicin resistance to molecular surveillance for dapsone and quinolone resistance [18]. The gene targets for quinolones are gyrA and gyrB, which encode gyrAses A and B that are responsible for uncoiling DNA [37,40]. Reasons for including these targets are that ofloxacin is a standby drug for some patients and an alternative to the ROM (rifampicin, ofloxacin and minocycline) regimen. Moreover, fluoroquinolones are secondary drugs for TB [41]. An increase in availability and exposure to these drugs should be monitored for their potential to aid in the increase of drug resistant M. leprae [41,42].
The current study calls attention to M. leprae resistance in Colombia, especially the significant association between confirmed resistance mutations and relapse in leprosy patients. 70% of mutations we detected were related to drug resistance and were isolated from relapsed patients, five of which received dapsone monotherapy as a primary treatment. A high frequency of DRDR mutations for rifampicin was seen in a region where dapsone monotherapy was used extensively. The reported drug-resistance M. leprae suggests that the situation needs to be monitored in Colombia and the sentinel surveillance program have to continue in Colombia [32].