Among individuals aged 13 to 17 years, previously non-vaccinated with BCG, and who are in direct contact with TB, is the combination therapy of INH 900mg + Rifapentine 900mg once a week for 3 months as effective when compared to individuals aged 13 to 17 years who are also previously non-vaccinated with BCG and have completed the gold prophylactic standard of INH 300mg daily for 6 months in preventing active tuberculosis?
ARTICLE CITATION:
Sterling TR, Villarino ME, Borisov AS, et al. Three months of rifapentine and isoniazid for latent tuberculosis infection. N Engl J Med. 2011;365(23):2155–2166.
CATEGORY: Infectious Diseases
APPRAISAL TYPE: Therapy
/>
INTRODUCTION
Tuberculosis (TB) is currently a public health issue. This disease is caused by an acidresistant bacillus, the Mycobacterium tuberculosis.1 Nearly one third of the world population is infected with TB.2 It was responsible for 1.3 million deaths worldwide in 2007,2 and has been reported as responsible for nearly two million deaths worldwide each year.1 Although this disease is most common in nonindustrialized countries, where the incidence rate exceeds 100 cases per 100,000 persons, it is still present in industrialized nations such as the USA, where the incidence rate is less than 25 cases per 100,000 persons.2
The main preventing measure for this disease has been the bacillus CalmetteGuérin vaccine (BCG).2 The standardofcare treatment against Mycobacterium tuberculosis consists of 900mg of Isoniazid daily for 9 months.1 While the vaccine is used for primary prevention of TB, a pharmacological approach has been used as secondary prevention, namely to reduce the risk of progression from latent tuberculosis infection to active disease.1 Due to the long duration of treatment, completion rates are between 30 and 64% 1. It is more common for people over 35 years old than adolescents and young adults to complete the treatment.3
The combination of Rifampin and Pyrazinamide for two months have proven to be as effective as isoniazid against acquiring active TB infection, but the main drawback has been high hepatotoxicity rates.4 Moreover, Rifapentine and Isoniazid once weekly for 3 months are effective in patients with low bacteria load.5 Therefore, the chosen study to analyze compares between the combination of Isoniazid (INH) 900mg and Rifapentine 900mg once a week for three months, against INH 900mg daily for 9 months in stopping the progression from latent TB to active disease in adolescents.
PATIENT PRESENTATION
M. is a 17-year-old male who was not vaccinated with BCG as a child. His parents did not want to comply with the vaccination scheme because of personal beliefs. His father had not been vaccinated either. 6 months ago, his father was diagnosed with latent pulmonary tuberculosis, but M. and his siblings resulted negative at the time. A prophylactic scheme of INH once daily for 6 months was initiated, but M. confesses that between school, extracurricular activities, family, and friends, it was hard to remember to take it daily.
CLINICAL QUESTION
Patient / Problem: Non-vaccinated individuals aged 13 to 17 years, direct contact of a TB patient.
Intervention: INH 900mg + Rifapentine 900mg once a week for 3 months.
Comparison: INH 300mg daily for 9 months.
Outcome: Progression from latent tuberculosis into active infection.
PICO Question: Among individuals aged 13 to 17 years, previously non-vaccinated with BCG, and who are in direct contact with TB, is the combination therapy of INH 900mg + Rifapentine 900mg once a week for 3 months as effective when compared to individuals aged 13 to 17 years who are also previously non-vaccinated with BCG and have completed the gold prophylactic standard of INH 300mg daily for 9 months in preventing latent tuberculosis?
LITERATURE SEARCH
The selected databases for this literature search were PubMed (MEDLINE), CINAHL, and ScienceDirect, and they were explored in November 2014. First, the MEDLINE database via PubMed was searched for keywords Latent Tuberculosis Infection, Rifapentine and Isoniazid in All Fields. The only Boolean operator used was AND between the 3 fields. This search yielded 42 results. The search was narrowed down to 34 results taking into account only articles published in the last 5 years (2009-2014). Then, the article type was specified as case reports, clinical trials, controlled clinical trials and randomized controlled clinical trials. This last search yielded only 2 results.
The search in CINAHL was done similarly. Key terms such as Latent Tuberculosis Infection, Rifapentine and Isoniazid in All Fields, using AND as a Boolean operator between them yielded 11 results. When keeping the last 5 years as publication date I obtained 10 results. Finally, when specifying English as the language and clinical trial as the article type, the search yielded 2 results.
Last for ScienceDirect, the same keywords Latent Tuberculosis Infection, Rifapentine and Isoniazid with the Boolean operator AND were used. This search was done in the Journal section of the Expert search, looking for articles published in the last 5 year (2009-2014), coming from all sciences, and yielded 34 results. When selecting medicine and dentistry as the field, the search yielded 32 results. In order to further narrow down the number of results, a manual exploration of each article was done, and 31 were excluded for study design, not being studies on human beigns and having a population with chronic diseases. This process yielded only one result.
ARTICLE SELECTION
The literature search yielded 5 results, and when excluding duplicates, I obtained 3 articles1,3,6 (Martinson et al., 2011; Sterling et al., 2011; Li et al., 2009). These three articles are primary research published in the last 5 years. However, the article by Martinson et al.6 (2011) did not include the study population needed to answer my research question. My study population consists of adolescents age 13 to 17, and the study population of the article by Martinson et al.6 (2011) were patients over 18 years of age. Then, I excluded the article by Li et al.3 (2009) because of the study design. They used a retrospective cohort design, whereas Sterling et al.1 (2011) used a randomized controlled trial, which provides a higher level of evidence. Therefore, the article by Sterling et al.1 (2011) was chosen.
ARTICLE SUMMARY
Research Question:
In the article by Sterling et al.1, although it is not clearly stated, and since the authors designed a non-inferiority trial, it can be infered that the research question explored whether the combined therapy of INH 900mg and Rifapentine 900mg once a week for three months was as effective as the standard-of-care, namely INH 300mg daily for 9 months, for stopping the progression from latent tuberculosis to active disease. The non-inferiority margin was 0.75%, and the primary end-point was confirmed tuberculosis.
Methods:
This was an open label, noninferiority, randomized controlled clinical trial1. The inclusion criteria were at least 12 years old (age range was expanded in 2005 to include children 2 to 11 years old), having been in contact with a confirmed tuberculosis patient, and having being test for the presence of M. tuberculosis with a positive result, without exhibiting clinical manifestations of active tuberculosis. The exclusion criteria were active disease, resistance or sensitivity to the treatment, previous treatment with these drugs (INH and Rifapentine), abnormal liver tests, pregnancy or lactation, and body weight < 10kg.1
After the recruitment period, patients were randomly allocated in a 1:1 ratio to any of the two possible treatment groups using simple and cluster randomization methods. This means that every patient was at first simply randomized, but if the patient shared households with another patient previously enrolled, a cluster method was applied, and therefore all patients in the same cluster received the same treatment. The two possible treatment groups were INH 900mg plus Isoniazid 900mg once a week for the months (3HP) or 9 months of a daily dose of 900mg of INH (9H).1
Following randomization, patients were evaluated monthly during the treatment period. Then, the followed period started, and lasted 33 months. The follow up consisted of visits every 3 months for 21 months, and then every six months until the 33rd month. Loss-to-follow up patients were cross-matched with local and state tuberculosis databases.1
Since this is an open-label trial, no blinding was considered. Both intention-to-treat and a per-protocol analyses were made. Furthermore, a modified intention-to-treat analysis was performed, where all elegible patients were included. Within the scope of this review, only results from the modified intention-to-treat analyses will be taken into account. Baseline characteristics like age, sex, indication for treatment, ethnicity, body mass index (BMI), risk factors, and liver disease were considered, and showed no statistical differences between the groups except for north american ethnicity, subjects enrolled in a cluster, and homelessness as a risk factor. The proportion of subjects in the intervention group was larger in these particular variables.1
Data Presentation and Analysis
During the course of the study, the authors recorded 10,327 person-years of follow-up in the intervention group, and 9,616 person-years in the comparison group. Data were presented also visually by using tables and figures. Figure 1 shows the difference rates in tuberculosis, and it is clear that the difference in rates favor combination therapy. Furthermore, table 3 shows data on the adverse events, and for the combination therapy the data shows significantly differently less rates of adverse events, less rates of permanent drug discontinuation, less rates of hepatotoxity, etc. Table 4 shows the results of the risk analysis (uni and multivariate) and the results are given below.1
According to table 2, the absolute risk reduction (ARR) was 0.24, and deriving other calculations, the relative risk was 56%, and the number needed to treat was 5, which means that at least 5 patients must be treated with the intervention treatment in order to spare 1 person from presenting tuberculosis.1
Results
Tuberculosis developed in 0.19% of the combination-therapy group against 0.43% in the INH-only group, for a difference of 0.24. Higher completion rates were observed in the combined-therapy group against the control group (82.1% vs. 69.0%, P<0.001). Hepatotoxicity was higher in the INH-only group (2.7% vs. 0.4%, P<0.001), as well as the proportion of subjects who discontinued a study drug due to hepatotoxicity (2.0% vs. 0.3%, P<0.001). In the case of hypersensitivity, the combination group had larger discontinuation propotions (2.9% vs. 0.4%, P<0.001). According to the risk analysis, tobacco smoking at time of enrollment, HIV infection, and low BMI were independently associated with an increased risk of acquiring HIV infection, and after adjusting for these variables, the subjects in the combined group were at lower risk of tuberculosis (adjusted hazard ratio 0.38; 95% CI, 0.15 to 0.99; P=0.05).1
Since the proportion of subjects who presented tuberculosis after taking the combination therapy was almost half of the INH-only group, I consider that there is enough evidence to consider non-inferiority. Furthermore, the combination group presented less hepatotoxicity and discontinuation proportions, which could be due to easier dosage and less treatment time. Plus, after taking aside associated risk factors, the risk of tuberculosis was lower in the intervention group. Therefore, the combined therapy is as effective as the standard-of-care, while exhibiting lower short-term toxicity and higher compliance.1
Conclusion
According to this study, the authors conclude that only 3 months of combined treatment with 900mg of Rifapentine and 900mg of INH weekly was as effective as 300mg of INH daily for 9 months, with lower hepatotoxicity, mild to severe adverse effects, and higher compliance rates.1
ARTICLE CRITIQUE
The strengths of this randomized controlled study by Sterling et al.1 are large sample size, 1:1 allocation, detailed description of the values they used to calculate sample size, enough follow-up period to find mild to severe adverse effects, well definition of the study population, intervention and comparison groups as well as primary and secondary outcomes, and recruitment process. They also performed this study as a multicentered trial in different countries, including as much variability as possible within it. The statistical analysis was appropriated, using measures of frequency and association, hypothesis testing as needed, and risk assessment. The authors decided to show the results for both the modified intention-to-treat analysis and the per-protocol analysis.
Drawbacks include lack of a specific research question and set of hypotheses, a high non-inferiority margin, modified intention-to-treat analysis instead of the traditional version, clustered random allocation instead of simple random allocation, handling of missing data from subjects who were lost to follow-up, and open label design. Although the high non-inferiority margin was obtained from prior research results available at the moment, the margin is so high that it made it hard to not reject the null hypothesis. The choice of clustered random allocation may have introduced selection bias, and that would explain the statistically significant differences found between the groups at baseline. The open label design may have introduced information bias, because of patients’ awareness according to the treatment they received. Plus, subjects in the combined therapy were directly observed, which may have risen the compliance rates and may have lowered the permanent discontinuation rates. The INH group was not observed, but subjects were left for self-administration of the drugs. The authors handled loss of follow-up by estimating data from cross-matching, which may have overestimated the effect. Furthermore, modified intention-to-treat analysis has been previously associated with post-randomization exclusions, industry funding, and authors’ conflicts of interest7. In fact, this study was supported by the Centers of Disease Control (CDC), and at least 3 of the authors have received grant fundings from different pharmaceutical companies.1
CLINICAL BOTTOM LINE
In general, and based on the study by Sterling et al.,1 the combination of INH 900mg and Rifapentine 900mg weekly for three months seems to be as effective as the standard therapy of INH 300mg daily for 9 months. The advantages that the intervention therapy presents are easy administration, apparently higher compliance rates, lower rates of adverse effects, and lower hepatotoxicity. When supported by further evidence, these results might influence me to recommend the combined therapy for my patients, in order to avoid tuberculosis.
APPLICATION TO SPECIFIC PATIENT
I would recommend M. to take the combined therapy once weekly for three months to stop his M. tuberculosis infection from progressing to active disease. By only taking a dose every week, it might be easier for M. to comply with the treatment. Furthermore, his parents can help him to remind him to take the pills on the weekends, which might be hard to do every day during weekdays.
REFERENCES
1. Sterling TR, Villarino ME, Borisov AS, et al. Three months of rifapentine and isoniazid for latent tuberculosis infection. N Engl J Med. 2011;365(23):2155–2166.
2. Sia I, Wieland M. Current concepts in the management of tuberculosis. Mayo Clinic Proceedings. 2011;86(4):348-361.
3. Li J, Munsiff SS, Tarantino T, Dorsinville M. Adherence to treatment of latent tuberculosis infection in a clinical population in New York City. Int. J. Infect. Dis. 2010;14(4):e292–7.
4. Gordin F, Chaisson R, Matts J, Miller C. Rifampin and pyrazinamide vs isoniazid for prevention of tuberculosis in HIV-infected persons: an international randomized trial. Jama. 2000;283(11):1445-1450.
5. Benator D, Bhattacharya M, Bozeman L, et al. Rifapentine and isoniazid once a week versus rifampicin and isoniazid twice a week for treatment of drug-susceptible pulmonary tuberculosis in HIV-negative patients: a randomised clinical trial. The Lancet. 2002;360(9332):528–534.
6. Martinson NA, Barnes GL, Moulton LH, et al. New regimens to prevent tuberculosis in adults with HIV infection. N Engl J Med. 2011;365(1):11–20.
7. Montedori A, Bonacini MI, Casazza G, et al. Modified versus standard intention-to-treat reporting: Are there differences in methodological quality, sponsorship, and findings in randomized trials? A cross-sectional study. Trials. 2011;12(1):58.
