Drug Susceptibility Patterns of Mycobacterium tuberculosis Isolates from Tuberculosis Patients in Coastal Kenya

Background: Tuberculosis (TB) is an infectious disease caused by the bacillus Mycobacterium tuberculosis. Anti-tuberculosis drug resistance is an emerging health problem in Kenya and especially in Coastal region. This is a major challenge in tuberculosis control. Diagnosis is based on Ziel-Neelsen staining alone and patients are treated without information on sensitivity patterns. Aim: This study aimed to determine drug susceptibility patterns of Mycobacterium tuberculosis in Coastal Kenya. Study Design: Hospital and laboratory based cross-sectional study was carried between April 2015 and July 2016 at Coast General Referral hospital; Tudor, Port-Reitz, Likoni Sub-County hospitals; Mlaleo, Kongowea and Mikindani health centers. Methodology: Sputum samples from patients with bacteriological confirmed TB on microscopy were cultured on Lowenstein Jensen (LJ) media. Strains of MTB complex from Lowenstein Jensen (LJ) slopes were subjected to drug susceptibility testing (DST) to first-line drugs including isoniazid (H), rifampicin (R), streptomycin (S) and Ethambutol (E) using proportional method on the Mycobacterium Growth Indicator Tube (MGIT) conventional method. Participants were offered diagnostic testing and counselling for HIV testing. Results: Drug sensitivity test was performed for a total of 210 Mycobacterium tuberculosis isolates for the first line anti-TB drugs. About seventy eight percent and twenty nine percent of the strains from new patients and previously treated patients were fully sensitive to all the drugs tested respectively. Prevalence of any resistance to one drug was 102 (48.6%, 95% CI: 20.45 28.23). Any single drug resistance was most frequent in isoniazid 30 (16.0%), Ethambutol 20 (10.0%), Streptomycin 18 (18.3%) and Rifampicin 4 (2.1%) in newly diagnosed patients. How to cite this paper: Yonge, S.A., Otieno, M.F., Sharma, R.R. and Nteka, S.S. (2017) Drug Susceptibility Patterns of Mycobacterium tuberculosis Isolates from Tuberculosis Patients in Coastal Kenya. Journal of Tuberculosis Research, 5, 201-219. https://doi.org/10.4236/jtr.2017.54022 Received: July 26, 2017 Accepted: October 10, 2017 Published: October 13, 2017 Copyright © 2017 by authors and Scientific Research Publishing Inc. This work is licensed under the Creative Commons Attribution International License (CC BY 4.0). http://creativecommons.org/licenses/by/4.0/ Open Access


Introduction
Tuberculosis (TB) is an infectious disease caused by strains belonging to the Mycobacterium tuberculosis complex. Multidrug resistant TB (MDR TB), defined as resistance to at least isoniazid and rifampicin, has been spreading rapidly in recent years. In 2013, 3.8% of newly diagnosed and 20% of retreatment cases were estimated to have MDR-TB globally with noticeable geographical variations in prevalence [1]. In Africa, various reports have demonstrated that resistance to one or more anti-TB and MDR-TB ranges from 3% to 37.3% [2] [3] and 1.4% to 11.6% [4] [5] [6] [7], respectively. Besides, extensively drug resistant TB (XDR-TB) has been reported by 92 countries (including Kenya) and about 10.6% of MDR TB patients have XDRTB globally. Therefore, the rapid spread of MDRTB and XDRTB especially in new TB patients is challenging the effectiveness of TB control programmes in many low income countries [8] [9].  [10]. Drug resistance in M. tuberculosis occurs as a result of random spontaneous chromosomal mutations during natural cell replication. These mutations are neither drug induced nor drug linked. The probability of a drug-resistant mutant occurring is directly proportional to the size of the bacterial population [11]. The frequency of primary resistant organisms varies for each drug; however, it is usually between 10 −6 to 10 −8 . Spontaneous resistance to isoniazid is estimated to occur once in every 106 organisms, and to rifampicin once in every 108 organisms. The probability of spontaneous mutants being simultaneously resistant to two or more drugs is the product of the individual mutants. The development of drug resistance is a man-made amplification of a naturally occurring phenomenon. Previous treatment for tuberculosis predisposes to the selection of multi-drug resistant organisms. Non-compliance is a major factor in allowing the resistant organisms to survive [12]. According to the WHO, an MDR patient infects 10 -15 people every year [USAID, 2009].
Treatment of MDR-TB lasts for 18 months but can extend to two years or more because it is difficult to cure and drugs used for treatment are less potent, more toxic and 50 -200 times more expensive than first-line drugs. If not treated properly, it can result in complications that may require surgical interventions hence increasing period of hospitalization and raising treatment cost. Although an unequal global distribution of drug resistance exists between poor and rich countries, the problem is global [13].

Study Area
The study was conducted in Mombasa County which has a population of 1,031,266 by the year 2012. The population is steadily growing due to rural-urban migration and immigration from unstable countries. The total area Mombasa is 109 km 2 with about 60% of the people living overcrowded informal settlements in the form of shelters. Residents are of mixed ethnicity and are engaged in low-income generating activities, mainly informal sector and small trading. The County has rapid population growth and is characterized by low socio-economic indicator. This creates huge demands on health facilities and inability to keep pace with the environment, continued economic prosperity, public health and quality of life of residents. Tuberculosis and HIVAIDS are the leading causes of deaths in the area representing 50%.

Study Site
The study was done at Coast provincial General hospital (CPGH), Mlaleo Health and Mikindani Health Centers, Likoni, Portreizt and Tudor County hospitals. These facilities were selected because of the fact that they represented the largest centres for TB diagnosis and treatment in their respective regions. These hospitals like all others at their levels have chest clinics where TB patients obtain health care respectively.

Study Design
This was hospital and laboratory based descriptive cross-sectional study carried out between April 2015 and July 2016. New smear positive pulmonary TB patients Kenya, however, due to inadequate laboratory capacity, there is only one public reference centre located in Nairobi performing TB culture and drug susceptibility testing (DST). Therefore, the service is not currently available in other regions in which Coastal region of Kenya is not an exception.

Inclusion and Exclusion Criteria
All adult patients 18 years and above suspected of having TB and resident in Mombasa County for at least six (6) months, not on anti-TB chemotherapy and consented to participate in the study were recruited. Tuberculosis suspects who were below 18 years and unwilling to participate in the study and not meeting the above inclusion criteria were excluded.

Sample Size
The sample size was determined by taking the prevalence of 2.8% from previous study, desired precision of 1%, a 95% confidence interval and nonresponse rate of 15%. The sample size was calculated based on the sampling method recommended by WHO for drug resistance survey in tuberculosis [26]. The final sample size was 500.

Sampling Frame
Coastal region was purposively sampled because of high cases of TB. The sampling frame consisted of all the public health facilities within the study area. In Kenya, majority of TB patients visit public health facilities compared to private wings. In this study, we focused on public health facilities in order to obtain the required number of patients and to minimize duration of the study. After the selection of the study sites, each was allocated a proportionate number of study subjects based on the level of health care delivery system and the average client attendance in the past one month before embarking on the study. To minimize bias in selecting study subjects, consecutive sampling was used hence every alternate TB suspect who satisfied the inclusion criteria was selected for the study.

Collection of Demographic Data
A structured and pre-tested questionnaire was used to obtain socio demographic data, history of previous anti-TB treatment, HIV serostatus, history of hospitalization and other relevant data from each study participant. (spot, early morning, spot) were collected from 500 TB suspects under the supervision of trained and competent medical staff as recommended by WHO [26]. The patients were advised to rinse their mouth twice with water before producing the specimen and this helped to remove food and any contaminating bacteria in the mouth. They were instructed to take two breaths, coughed vigorously and expectorated the material in into the sterile 50 ml blue cap screw-capped bottle. This process allowed sputum to be produced from deep in the lungs. The TB suspects were asked to hold the sputum container close to the lips and spit into it gently after a productive cough. At the peripheral laboratory, the standard Acid-fast (AFB) direct smear microscopy using Ziehl-Neelsen (ZN) staining was done on the initial sputum to confirm TB diagnosis of suspected patients. A second sputum specimen was then collected which was refrigerated at 4˚C and transported to the Central reference Laboratory (CRL) weekly for culture.

Microscopic Examination of Specimens
Sputum smears were examined for acid-fast bacilli (AFB) after staining following ZN method. The degree of ZN smear positivity was quantified as 1+ for 10 -100 AFBs per 100 fields, 2+ for 1 -10 AFBs per field (50 fields) and 3+ for >10 AFBs per field (20 fields). For less than 10 AFBs per 100 fields, the exact number of AFBs was indicated. A suspect was considered to be ZN smear positive if at least one specimen was positive.

Sample Processing, Mycobacterial Culture and Isolation
Sputum specimens were processed for isolation of mycobacteria using standard Petroff's method [16]. They were decontaminated with NAOH solution (40

Quality Control
Standard Operating Procedures (SOPS) were followed for microscopy and culture. Mycobacterium tuberculosis H37Rv reference strain (ATCC 27294) which is susceptible to all anti-TB drugs tested was included in each test batch as positive and Escherichia coli as a negative control. All activities like reagent and media preparation were carried out as described in standard operating procedure by Kent and Kubica [27]. An experienced microscopist read an arbitrary 10% positive and 10% negative slides randomly selected, with concordance of 99% and 97% respectively. For data obtained through interview, the questionnaire was pretested before use and data were collected by trained nurses under close supervision by the investigators.

HIV Testing
Blood samples were tested for HIV antibodies according to the Kenyan national testing algorithm for voluntary counseling and testing by using Determine HIV1/2 (Abott laboratories, Japan co. LTD), Capillus HIV1/2 (The Trinity Biotech, Ireland) and Unigold H1/2 (Trinity Biotech, Ireland) rapid test kits and positives confirmed with the enzyme linked immunosorbent assay (ELISA).

Data Management and Analysis
Demographic data were confidentially obtained from the TB suspects by clini- test was used to compare categorical data and logistics regression to assess the association between drug resistance and independent factors. A significance level of p < 0.05 was considered statistically significant.

Ethical Issues
This study was approved by Kenyatta University Ethical Review Committee.
Clearance was also obtained from respective County health authorities and hospital administrations. The study was conducted in accordance with the declaration of Helsinki. Written consent was obtained from all study participants and code numbers rather than names were used to identify candidates in order to maintain confidentiality. The study did not expose candidates to any unusual risks as competent hospital staff obtained sputum and blood specimens from candidates using standard procedures. Drug susceptibility test results were reported to the respective health facilities for further management of the patients.

Drug Resistance between HIV-Positive and HIV Negative in Diagnosed TB Patients
A total of seventy eight patients (37.1%) were TB-HIV co-infected. Prevalence of any type of drug resistance in TB-HIV co-infection patients was 40 (19.1%) and 15 (7.1%) in HIV negative. The difference was not statistically significant (P > 0.05). In mono drugs, Isoniazid 25 (11.9%) showed the highest resistance pattern in TB-HIV co-infected patients followed by streptomycin 10 (4.8%), Ethambutol 9 (4.3%) and Rifampicin 5 (2.4%). Combined resistance to R + E was 4 (1.9%) in TB-HIV co-infected patients. These differences were statistically significant (p < 0.05). Triple resistance to Isoniazid, Ethambutoland streptomycin was 1 (0.5%) in both HIV positive and negative TB patients respectively. These differences was statistically significant (p < 0.05). The prevalence of other types of drug resistance was also significantly different in HIV positive and HIV negative TB patients (Table 3).

Discussions
In our finding, resistance to one or more first line anti-TB drugs was 26.2%. This is relatively higher than previous reports from other parts of Kenya [28]. This results was comparable with studies done in Ethiopia (25%) [29] and in Uganda (28.6%) [30]. Studies from Nairobi [31], Addis Ababa [32] and Central Asia [33] reported a higher resistance level of (30%, 32.5% and 30.5%), respectively. The variations in overall prevalence of drug resistance among the different study settings could be due to difference in sample size (small sample size could overestimate the proportion), irregular supply of ant-tuberculosis drugs, poor TB case   resistance among the study population also is an indicator that this drug will be completely useless. Therefore, mono resistance to isoniazid should be properly monitored in order to minimize the spread of MDRTB strains in the study area.
In this study, resistance to Rifampicin was 3.8% which was higher than that observed in earlier studies in Kenya where resistance was 0.3% and 1.3% [41] and in India (1.1%) [42]. Rifampicin has several adverse effects such as nausea, vomitting, rashes, GIT upset, flu-like symptoms, fever and jaundice which could result in patient non-adherence and hence may lead the selection of resistant strains. In the DOTs program, Rifampicin is given in the intensive phase under direct observation together with at least three drugs. Moreover in the continuation phase, Rifampicin is spared.
In our study, resistance to Ethambutol was 5.2% and it was higher than a previous study done by Raviglione et al. [43] in Ethiopia (2.7%) and others in Bukina Faso (0.3%) [44] and Uganda 0.2%) [45]. One possible explanation for increased Ethambusol resistance could be increased defaulting rate from TB treatment in the continuation phase when it is administered. It also enhances the  [55]. However, reports in Tanzania [8] and in Central African Republic [56] were not in agreement with our study. Thus, study may be interpreted to directly argue for the hypothesis that drug resistant TB is virulent and causes disease mainly in immunocompromised TB patients (p < 0.05). In this regard, similar findings have been reported by different studies.
Studies in Latvia showed that any resistance and multi-drug resistant tuberculosis (MDR-TB) were significantly associated with HIV infection [38]. A study done in Northern Tanzania showed that among TB-HIV co-infection, patients' resistance to at least one drug was 10.8%. Tuberculosis infections in high incidence countries have been shown to be recently transmitted and failure to contain MDR-TB and XDR reflects inability to diagnose the problem early to prevent transmission of the same while continuing to prescribe an ineffective regimen [57]. Once MDR-TB has developed, further progression to pre-XDR and XDR is only a question of time and will place over few months or even years. The patient remains infectious and transmission of MDR TB and XDR-TB continues particularly in areas with high prevalence of HIV/AIDS and overcrowding. Sarita et al. [13] found out in a study in Texas that TB patients with HIV were more likely to have rifampicin resistance and less likely to have isoniazid resistance. However a study done in Punai Mahararashltra India on anti-TB drug resistance showed that prevalence of drug resistant isolates among HIV seropositive patients was similar to that of seronegative TB patients indicating that HIV infection may not be associated with drug resistant TB [40]. Several factors have been proposed to explain such an association. These may include malabsorption, of anti-TB drugs among HIV patients, poor treatment adherence, lack of access to proper treatment, exposure of HIV/AIDS persons to MDR-TB patients during hospitalization or frequent visits to health facilities which increases the risk for MDR TB nosocomial outbreaks and rapid progression of TB in HIV patients [58].

Conclusion
The current study reveals that the overall resistance to first line anti-TB drugs is high. The highest monodrug resistance was found against Isoniazid (H). Although the rate of MDR-TB was relatively low, this signifies that conditions favouring the spread of MDR-TB are on high rise. Patients who were HIV co-infected were more likely to develop drug resistance compared to HIV negative patients. Therefore, it is essential to address the problems of development of drug resistant strains of TB by establishing good TB programmes (DOTS). Patients adherence to anti-TB drugs and scaling up of drug sensitivity testing (DST) service at County level hospital will help to reduce the development of drug resistance in the study area.