MDR-TB is the tuberculosis with resistant bacilli to isoniazid (INH) and rifampicin (RMP), the two most powerful antituberculosis drugs. The incorrect use of drugs, the inefficient uptake of therapeutic regimes and the premature discontinuation of treatment cause drug-resistances (DR-TB) that are likely to be transmitted to the population, especially in places with a strong promiscuity, such as prisons, hospitals, orphanages and some dwellings [1] .

In contrast to drug-susceptible TB, the diagnosis of MDR-TB requires more complex laboratory investigations than simple microscopy [2] . The resistance profile is obtained either by the sensitivity test after so-called “traditional” methods, such as culture on liquid or solid medium, whose average duration varies from 2 to 8 weeks. Or you can use the rapid diagnosis or molecular diagnostics (RMD), the most common of which is the Xpert MTB/RIF test, which detects M. tuberculosis (MTB) and rifampicin resistance (RR) with greater sensitivity [3] [4] . According to the World Health Organization (WHO) report of 2017, the proportion of MDR-TB cases among new patients is estimated at 4.1% [95% CI: 2.8% - 5.3%] and 19% [95% CI: 9.8% - 27%] among the patients already treated, for the entire world, about 600,000 (540,000 - 660,000) cases [5] [6] . The MDR-TB most affected regions are respectively Central Europe and the countries of the former Soviet Union, Southeast Asia (China and India) and sub-Saharan Africa, and particularly the southernmost [5] .

Under the best conditions, all tuberculosis patients should have a profile of DR-TB as is the case in developed and advanced countries; this allows to prescribe personalized therapy and to obtain real TB prevalence drug-resistant TB [1] [6] [7] [8] .

Africa, which contains only 12% of the world’s population, has 25% of TB cases [5] [7] [9] . The number of MDR-TB cases in Africa is estimated at 93,000 (range: 81,000 - 106,000) cases, or 15% of the global burdens based on estimations with 2.7% [95% CI: 2.0 - 3.5] are cases of MDR-TB among new patients, and 14% [95% CI: 8.4 - 20] are cases of MDR-TB among previously treated cases. In the continent, the burden is generated by the Republic of South Africa (RSA), Nigeria, Ethiopia, the DRC and other countries in the southern region that have a high prevalence of HIV infection [5] [10] .

In 2015 and 2016, the number of TB cases reported in the DRC was respectively 120,508 and 132,515, of which 64% were new patients with bacteriologically confirmed pulmonary tuberculosis [5] [10] [11] . The number of patients in retreatment for these two years was 5399 and 5490, representing around 4% of all the TB cases. The latter group consists of 3 types of patients with a higher risk of developing MDR-TB: relapses (incident cases), failures to initial treatment and retreatment after being defaulters [11] .

For poor countries such as the DRC, where access to the diagnosis of MDR-TB is still insufficient, the MDR-TB identification is done primarily among the at-risk groups [1] [6] [7] . The knowledge of the extent of the disease remains a major challenge. The recent DRC prevalence survey found a prevalence of 2.2% of MDR-TB among new patients with bacteriologically confirmed TB (NP TP +) and 10% among previously treated cases [5] . That kind of survey is actually effective only if it is performed every 3 to 5 years. The estimate of this magnitude can also be deduced by a routine monitoring of at-risk groups [6] [12] . Groups of presumptives MDR- TB (which includes retreatment cases) are defined according to the period when the sputum persistence is detected by microscopy during treatment administered under surveillance. The proportion of MDR-TB and the risk are different from one type to another; it is useful to determine it [4] [7] [13] . National guidelines in the DRC agree with these WHO recommendations [14] [15] . However, previous studies in this area have focused either on the new patient group [6] [10] or on the risk factors [16] [17] . There is currently no long-term study that analyzes the results of MDR-TB surveillance since the integration of MDR-TB control.

The MDR-TB surveillance is a fundamental part of the TB control strategy, or even its elimination. The objective of this study is to determine the proportion of MDR-TB in each group of presumptive MDR-TB patients in the DRC and to identify the associated factors to MDR-TB.

The number of patients received for a laboratory test during the period was 8008 patients, of whom 6512 (81.3%) for a drug resistance study, of whom 5407 (83.0%) had actually performed culture tests. and/or Xpert MTB/RIF with available results. Lab results were not found for 1105 (17.0%) patients (Figure 2).

The present study analyzed the proportion of RR-/MDR-TB among presumptives in the DRC and some factors associated with the diagnosis of MDR-TB during 10 years of surveillance. Eight thousand eight (8008) patients were identified in the laboratory for a diagnosis of TB and MDR-TB, of which 6512 (81.3%) were known to be presumptive of MDR-TB. For 1105/6512 (17.0%) patients, the results were either not found or unrealized. The analysis were not performed for a variety of reasons, the most common being the loss of samples, empty spittoons or unusable samples.

This study counted a total of 13 presumptive groups. Previously treated patients accounted for the majority (95%), new patients were poorly represented (2%), and the remainder was the HIV seropositive group without patient’s history details. Retreatment failures formerly known as chronic cases constituted the majority group (30%). They have long been considered the most at risk and therefore a priority [1] [16] [22] . These failures, which included three-quarters of patients in the first period, accounted for only one-fifth in the last few years of the study period. At the same time as their proportion fell in a linear manner, so was the absolute number. This could be explained by the fact that the patients in retreatment are detected more and earliest with the integration of F1 and R1. Patients failing and relapsing after a first episode did not become actual targets until later, as their registration increased as shown in Figure 2. This situation has also been found in Bangladesh, with time there have been fewer and fewer suspected MDR-TB cases with multiple episodes [8] . Comparing the time of issuance of the guidelines with their application, this suggests a delay that may suggest that many patients have not been screened for MDR-TB. These patients were therefore not spared the prolonged use (of the additional drug pressure) of the RMP and INH (genetic mutation determining factor) responsible for the genesis of drug resistance [2] [22] .

The analysis showed that HIV patients as targets noted in the 2nd period, were recognized at the beginning of the implementation of the response against MDR-TB. But their application in a context of inadequate HIV testing and an under-mobilized population has not been facilitated. This deficiency is known in several reports covering the period. The number of TB patients tested in the DRC has long been below 50% [23] [24] . However, some patients have been registered and have benefited from screening. Improving global HIV-TB and TB-HIV collaborative activities, which advocate for systematic TB-screening of HIV-positive people, provides more comprehensive information [25] . HIV patients have been considered presumptive to carry TB for several years and their analysis in this setting will increase confounders. They are analyzed as a separate group and give some information, although insufficient. The Correct assessment of co-infection indicators would be more accurate with more complete data. The most appropriate research for TB and MDR-TB in HIV patient is the combination of genotypic testing and other methods in accordance with recent guidelines with more elaborate algorithms [25] [26] .

The comparison between the numbers of patients registered for retreatment during the period and those whose samples were sent to the laboratory shows that the proportions are very small (less than 10%) as shown in previous analyze [5] [23] [27] . The limits are numerous as described by Nourzad [28] who showed that in all the countries studied, less than 50% of MDR-TB presumptive had a DST and a little more than half in a study in Peru [29] . The deficit related to sample transportation, cost, logistics and the preparation of the package to be transported guaranteeing safety according to the ISO standards [7] [19] explains why the majority of the samples come from the urban environments or close to the laboratory, the case of Kinshasa, Kongo-Central or near an airport.

The city of Kinshasa is the home of the main screening laboratory. The proportion is more than half of the cases. This situation is encountered in most countries with few laboratory units that can screen for MDR-TB in developing countries [26] [30] . In some African countries, an organized system of sample transport has been put in place using all the ways (road, rail and air) where the infrastructure permits. The resolution of the sample transfer problem in the DRC remains a major challenge despite recent efforts, screening is still low [5] ; this big issue requires multisectoral and permanent measures that should be properly applied and assessed.

This analysis showed an incomplete availability of phenotypic sensitivity tests that decreased with time and the number of samples, and only a part of patients have access to a quality diagnosis before the advent of gambling rapid and molecular test [3] [31] [32] . In the present study more than 80% of patients had a laboratory result; this influences the level of validity because it is an optimal percentage for evaluations of routine monitoring [6] [28] . The spread of the rapid test has the downside of the use of conventional methods, which become reserved for the results that have detected a RR-TB which we necessarily want to have a more complete resistance profile including the detection of extensively drug- resistance (XDR-TB) [26] . This decline was marked during the 3rd period of the study.

The last two periods were dominated by the use of the Xpert MTB/RIF test, although the test detects only the resistance to RMP. These performances have been described in several studies at the global and national level [3] [33] , both for the detection of MTB and for the diagnosis of DR-TB, through the RR-TB recognized as a proxy to MDR-TB. The Xpert-MTB/RIF test detected RR-TB in 56 formerly contaminated specimens and 236 RR-TB in those that were negative for culture on solid media. This is largely due to the high sensitivity of the Xpert MTB/RIF test.

In this study, the resistance pattern to 4 molecules listed in Table 3 has shown a proportion of MDR-TB of 49.6% in a population with a large majority of patients repeatedly exposed to RMP and INH. The number of resistances was associated with the number of previous regimens by amplification phenomenon, which is described in other studies [8] [34] [35] .

This proportion of MDR-TB is divergent from that of 17% found by the drug resistance survey conducted in the DRC between 2015 and 2016 mentioned in the WHO Report [5] . This proportion is also higher than that found among the cases already treated by Van Deun et al. in Bangladesh. These authors had found a percentage of MDR-TB of 11.7% (range 8.1 - 23.5) among the failures and relapses at the beginning of the retreatment [8] . In the context of Bangladesh at the time, Category 1 or the initial treatment was predominantly made of 2 RHEZ/6HT (containing thiacetazone in place of ethambutol) for this sample of patients. This regimen is less inducing MDR-TB than those used in the DRC made of 2 RHEZ/6EH (until 2007) and 2 RHEZ/4RH. An investigation more than 10 years after the introduction of the 2 RHEZ/4RH regimen in Bangladesh showed a higher rate of MDR-TB among the previously treated cases [34] .

The systematic review done in Ethiopia (2017) showed higher proportions than the present study among the cases previous treated, here the prevalence of MDR-TB in new patients is associated with this rate, was 6 times higher than that found in the DRC [35] . In this same review of several studies, they showed that in the same region, the resistance profile and the prevalence of MDR-TB could vary greatly. Several determinants were considered, the previous regime and its duration, regularity of treatment and socioeconomic factors [35] .

Compared to previous studies conducted in the DRC, the difference with this survey can be in the sampling. In the 2015-2016 national survey identified above that reported 17% of RR-/MDR-TB among previously treated cases [5] , values lower than those described in this study. In this prevalence survey the cases already treated represent only 10% of the total sample of patients and the majority of which were retreatment cases. Another element that would support this difference is that the distribution did not take into account the burden of TB disease in the national territory. The selected clusters between rural and urban areas were equitably distributed when they were not. In this analysis, high -burden provinces are the most represented.

The percentage of MDR-TB confirmed is still under discussion as to the question of the denominator for this indicator, all registered presumptives or essentially patients with a laboratory result. These conditions expose the risk of either minimizing or overestimating the magnitude of the problem if an adjustment factor is not applied [6] [8] [22] . The use of the proportions on the present tests is more precise, it is used to extrapolate the results on a larger population if one wants to estimate the burden of acquired MDR-TB [14] [22] . The burden of acquired MDR-TB will consider the rate of MDR-TB to be applied to each group of presumptives, adjusted for factors related to undetected cases, either due to insufficient health coverage or poor system performance [5] .

In the present study patients sensitive to all 4 molecules tested represented 40% of cases; this requires careful interpretation especially when the therapeutic history is known. It is possible that no mutation occurs despite the conditions of therapeutic insufficiency. Indeed, the appearance of a mutation responds to the random mode [36] . Since mutation is a phenomenon that depends on chance, the probability of a mutation occurring is complex and difficult to apply to a single individual. Another way of explanation would be the exogenous reinfection by a strain totally sensitive to the classical test; this notion is well exploited in the hypotheses studying the dissemination of communicable diseases [2] [12] [22] .

The classical test which is based on the comparison of the conventionally established proportion (resistant strains/susceptible strains) shows its limits. It is the dominant strain that leads the pattern. However, there are other strains carrying resistant mutation genes that remain hidden. This is also applies to mono-resistance cases which accounted for 8% to 10% of patients and careful interpretation should applied [8] [18] [37] .

The logistic regression analysis showed that age under 35 was associated with MDR-TB, and the risk is twice that of those over 35 years of age. This corroborates the situation found in a previous study in Kinshasa [1] [7] [38] , Bangladesh [8] [34] and Ethiopia [30] [35] which showed that although the average age of patients with TB and MDR-TB is increasing, it still affects the youngest age groups.

Regarding the history of TB, multivariate logistic regression analysis showed that the possibility of finding MDR-TB was greater for treatment failures, the risk was twice greater than those who default in the past and four times more than those who had therapeutic success, which is consistent with observations from other studies in which the genesis of MDR-TB was first associated with failure and relapse and lost to follow up [17] [22] [39] [40] . For the last two groups the positions could vary. Indeed, in the group of defaulter’s patients, there are patients with sputum that has been negativated or converted to bacillary load. The causes of the interruptions are various, sometimes due to a bad evolution and sometimes follow a good clinical evolution during the continuation phase when the patient resumes his usual activities [41] [42] .

This continuous surveillance based on routine testing also showed that both times of sputum delivery and screening tests have an influence on the diagnosis of drug resistance when using culture; this is not confirmed for molecular testing. The faster a sample is transported and processed in the laboratory, the more satisfactory the results. This has also been described elsewhere [6] [26] [43] . Culture requires living germs whereas molecular tests are based on the presence of genetic material even though germs have lost viability. It is therefore imperative to strengthen the transportation of specimens and imperatively achieve the culture which remains the accurate one. We therefore do not encourage the abandonment of culture and DST at the advent of molecular tests, but rather to move towards culture on liquid media and tests to detect the viability of germs using algorithms combining all these methods [26] .

The proportion of RR-/MDR-TB among the presumptive has represented higher rate than those estimated generally. The NTP should improve patient follow-up to reduce TB treatment failures and defaulting; also, while increasing the use of molecular tests, they should reduce sample delivery times when using culture and DST concomitantly.

SBF, principal investigator;

SBF, JKN, PKK for the design of the study;

PKK, ESK, SBF for data analysis;

PMS, SBF, coordinated the collection of data;

MKB, HSN, JDD, JKN, PKK for proofreading and editing of the final form of article.

Thanks to the officers of the National Laboratory of reference especially to Buteka JF, and Simelo JP, the Head of National Tuberculosis Programme/DRC, provincial coordinators, nurses of health centers and technicians of field laboratories and Professor Georges Mpwate for revising the document.

The authors declare no conflicts of interest regarding the publication of this paper.

Bisuta-Fueza, S., Kayembe-Ntumba, J.M., Kabedi-Bajani, M.-J., Sabwe, P.M., Tuma, H.S.-N., Simelo, J.-P., Sumaili-Kiswaya, E., Ditekemena-Dinanga, J. and Kayembe-Kalambayi, P. (2019) Multidrug-Resistant Tuberculosis in the Democratic Republic of Congo: Analysis of Continuous Surveillance Data from 2007 to 2016. Journal of Tuberculosis Research, 7, 25-44. https://doi.org/10.4236/jtr.2019.71004