Published ahead of print on August 5, 2004, doi:10.1164/rccm.200405-584PP
© 2004 American Thoracic Society doi: 10.1164/rccm.200405-584PP
Considering the Role of Four Months of Rifampin in the Treatment of Latent Tuberculosis InfectionNew Jersey Medical School National Tuberculosis Center, Newark, New Jersey Correspondence and requests for reprints should be addressed to Lee B. Reichman, M.D., M.P.H., New Jersey Medical School National Tuberculosis Center, 225 Warren St., P.O. Box 1709, Newark, NJ 07101-1709. E-mail: reichmlb{at}umdnj.edu Treatment of Latent Tuberculosis Infection (LTBI) is a major priority of the Centers for Disease Control and Prevention (CDC) in its quest for TB elimination (1–3). Isoniazid (INH) has been the mainstay of LTBI treatment for nearly 40 years. Unfortunately, the required lengthy treatment with INH (3), INH's reputation for hepatotoxicity (4) (albeit undeserved [5–8]), the increasing influx of foreign-born persons from countries with a high prevalence of INH resistance, and the virtual proscription of a shortened and effective regimen because of toxicity (9) have compromised treatment of LTBI as an effective tool. In this perspective, we argue for the expanded use of an already approved but not widely used, effective, and safe short-course regimen for the treatment of LTBI. In the American Thoracic Society (ATS), CDC, and Infectious Disease Society of America (ISDA) guidelines published in 2000, 2 months of rifampin and pyrazamide (2RZ) and 4 months of rifampin (4R) were both recommended as acceptable short-course regimens for the treatment of LTBI (3). After the recommendations were published, however, because of severe toxicity and deaths among patients on 2RZ, recommendations were revised to indicate that this regimen should not generally be offered (9), despite its having been shown to be as efficacious as 12 months of INH (12H) in the treatment of patients with LTBI and with HIV (10). The CDC is currently conducting a clinical trial which compares the efficacy of a 9-month regimen of isoniazid and a 12-week regimen of once-weekly isoniazid and rifapentine, a long-acting rifamycin. Although rifapentine has been approved for the treatment of selected patients with TB disease, the final results of the 12-week regimen for the treatment of LTBI are not expected to be available for several more years. Consequently, we have taken a fresh look at widening the use of the currently recommended, but not widely used, 4R regimen. The case for 4R can be justified on several levels: likelihood of high efficacy, low toxicity, increasing levels of INH resistance among new immigrants, cost considerations, adherence, and site-specific considerations. EFFICACY Rifampin is bactericidal for Mycobacterium tuberculosis and has significantly shortened the required duration of chemotherapy for TB disease (11, 12). Rifampin appears to act particularly well on mycobacterial subpopulations with only short bursts of metabolic activity (13). In a well-designed study using mice, Lecoeur and coworkers produced subclinical infection by vaccinating mice with BCG and subsequently challenging with M. tuberculosis (14). Once a stable count of bacilli in the spleen had been achieved, the mice received one of four treatment regimens: 6H, 3R, 3HRZ, or 2RZ. Efficacy was measured by the proportion of mice on each regimen with bacilli remaining in the spleen at the end of treatment and at 6 months after treatment. The proportion of mice with positive spleen cultures, both at the end of treatment and at 6 months after treatment, was significantly lower with the 3R and 2RZ regimens compared with the proportion treated with the 6H and 3HRZ regimens. Furthermore, there was no significant difference in efficacy between the 3R and 2RZ regimens. These findings suggested that 3R may be highly effective in humans. In the only randomized clinical trial that evaluated the efficacy of rifampin alone as treatment of LTBI, rifampin given daily for 3 months (3R) to persons with LTBI and silicosis (a highly potent facilitator of progression from LTBI to active TB) had an efficacy of 63% (compared with 48% for 6 months of INH and 29% for a 3-month regimen of INH and rifampin), which was significantly better than that of placebo (15). An earlier trial conducted by the International Union Against Tuberculosis (IUAT) showed that among adherent patients (i.e., those who completed the regimen and who were compliant during treatment), 6H had an efficacy of 68% and 12H had an efficacy of 93% (16). Through interpolation of data from this and other investigations, Comstock concluded that optimal protection from INH appeared to be obtained by 9 or 10 months of treatment (17). The Statement Committee on LTBI reasoned that because the efficacy of treatment for LTBI would be expected to be less in patients with silicosis, using those ratios and extrapolating to the IUAT study, one could estimate the efficacy of 3R in most patients to approach 90%. And because 4R would be expected to be better than 3R, one could argue that 4R is likely comparable to 9H (Richard O'Brien, M.D., personal communication). Other studies have suggested rifampin to be effective in the treatment of LTBI. One found that of 209 tuberculin skin test convertors exposed to INH-resistant cases, none treated with rifampin-containing LTBI regimens developed TB over a 2-year follow-up period (18). However, those treated with INH alone developed active TB at a rate the same as those refusing any treatment. In another study among 157 adolescents exposed to an INH-resistant case and placed on rifampin for 6 months, none developed TB during a 2-year follow-up period (19). However, there is a problem of rifampin-induced resistance. In a single case report, a close household contact to an INH-resistant patient (diagnosed in 1982) who was placed on rifampin preventive therapy developed rifampin-resistant TB shortly after completing treatment (20). However, adherence to the regimen was questionable and the patient suffered from chronic alcohol abuse, poor dietary habits, and high stress. It should also be noted that in the clinical trial conducted in patients with silicosis in Hong Kong, there was no evidence that giving rifampin alone led to the emergence of rifampin resistance (15). More recently, during a clinical trial in which patients with TB disease received either a once-weekly regimen of INH and rifapentine or a twice-weekly regimen of INH and rifampin during the continuation phase of treatment, all five HIV-infected patients who had received the once-weekly regime relapsed with monoresistance to rifampin (21). None of the three HIV-infected patients who relapsed after completing the twice-weekly regimen had drug-resistant organisms. Furthermore, no rifampin monoresistance had occurred among the 1,004 HIV-seronegative patients enrolled in the study. Although this study involved patients with TB disease, it does suggest the possibility that rifampin resistance might develop more frequently among HIV-infected persons receiving rifampin for the treatment of LTBI. Therefore, we would not recommend 4R for the treatment of LTBI patients who are co-infected with HIV. HEPATOTOXICITY Rifampin is one of the least hepatotoxic drugs with which to treat LTBI. The Hong Kong study among patients with silicosis showed that 3R was the least toxic (no hepatic toxicity and significantly fewer elevations of serum aminotransferase concentrations), compared with 6H and 3HR (15). A randomized, controlled trial in 402 HIV-seronegative participants found no adverse reactions requiring drug discontinuation in the 4R group, compared with 2 in the 6H group and 15 in the 2RZ group (22). Furthermore, there was only one aminotransferase elevation of > 100 IU in the 4R group (n = 132), compared with 5 in the 6H group (n = 131) and 17 in the 2RZ group (n = 139). In a recently published analysis of patients treated for active TB, rifampin caused fewer serious side effects than pyrazinamide and isoniazid and, unlike the other two drugs, caused no cases of hepatitis (23). INH RESISTANCE As noted above, rifampin has been shown to be effective in preventing TB among contacts exposed to INH-resistant cases (18, 19). During 2003, 53% of TB reported in the United States occurred among foreign-born individuals (24). Among TB cases reported in the United States from 1996 to 2000, high levels of INH resistance were found among foreign-born individuals with reported TB from countries which account for the majority of foreign-born individuals with TB in this country: Vietnam 21% resistance, South Korea 17%, Haiti 16%, Philippines 16%, China 16%, India 11%, and Mexico 9% (Marisa Moore, personal communication). A recent WHO publication on the global trends in resistance to anti-TB drugs included reports from areas within two of these countries. The prevalence of INH resistance among individuals with newly reported cases of TB from three areas in India ranged from 15 to 23%; the prevalence from four areas of China ranged from 6 to 25% (25). Consequently, in LTBI-infected individuals from these countries, the use of a rifampin-based regimen would appear to be more appropriate. A recent decision analysis demonstrated the importance of the prevalence of drug resistance in an immigrant population when selecting a drug regimen to treat LTBI (26). At our request, the authors of that publication recalculated their data using $120 as the cost of medication for 4R (the cost to selected public agencies in New Jersey, as opposed to the average wholesale price of $391 used in their model). They found that if the level of INH resistance was greater than 12%, 4R would become the dominant treatment strategy (i.e., be less expensive and more effective than 9H). Furthermore, even if the level of INH resistance were zero, using 4R would still be an overall less expensive strategy than using 9H. On the other hand, the level of rifampin resistance would have to be greater than 23% for 9H to become a dominant treatment strategy over 4R (Kamran Khan, M.D., personal communication). COST CONSIDERATIONS In Khan and colleagues' publication (26), sensitivity analysis demonstrated that if 4R could be purchased for less than $270, it would dominate 9H as a treatment strategy (i.e., less expensive and more effective) (26). Through provisions of Section 340B of the Public Health Service Act, which limits the cost of drugs to federal purchasers and certain grantees of federal agencies, New Jersey paid only $120 in mid-2004 for 4R, a cost which was available to public health programs throughout the United States (Joe Ware, personal communication). Granted, the cost of medication for 4R is higher than 9H ($120 versus $19 in New Jersey). However, if one includes personnel and indirect costs for monthly clinic visits for monitoring of toxicity and adherence, the overall costs would likely be substantially higher for 9H, compared with 4R. In decision analysis by Khan and coworkers noted above (26), if $120 is used as the cost of 4R, the overall cost for the 4R regimen ($384) is 28% less than the overall cost of 9H ($528). 4R would likely also increase the number of high-risk individuals who complete treatment, thus averting more cases and associated costs. ADHERENCE With decreased federal and state funding and, consequently, increasing staff shortages, TB programs are already finding it difficult to treat enough persons with LTBI to accelerate the decline of TB, even with targeting. Shortening the LTBI regimen from 9 to 4 months should allow existing staff to significantly increase the number of persons who complete LTBI treatment. Compared with INH regimens, shorter rifampin-based regimens have been shown to increase adherence levels (10, 27). One recent study followed nearly 4,000 patients on 6H and demonstrated that only 64% completed the 6-month regimen (8). The authors speculated that even fewer patients would have completed a 9H regimen. Using 4R would likely increase the proportion and number of persons completing LTBI therapy. The ATS/CDC guidelines recommend that "Because more than one regimen can be used to treat LTBI, health care providers should discuss options with the patient, and, when possible, help patients make the decision, unless medical indications dictate a specific regimen. Discussion should include the length and complexity of the regimens, possible adverse effects, and potential drug interactions" (3). Due to poor adherence with 9H among persons with LTBI in a county chest clinic in northern New Jersey who are primarily foreign born, working poor, or school children, we have offered 4R since 2002. We find that patients, when given a choice, overwhelmingly choose 4R over 9H. Of 286 persons we treated with 4R between October 2002 and September 2003, 243 (85%) completed therapy, compared with 66% (208/316) completion for persons treated between January 1999 and June 2000, when 9H or 6H was prescribed exclusively (statistically significant; p = 0.0001). There has been no hepatotoxicity among persons on 4R and only 6 (2.6%) had side effects requiring discontinuation of rifampin (1 allergic rash, 2 flu-like complaints, 1 nausea, and 1 joint pain). All were successfully switched to INH. Furthermore, if minor side effects are experienced early in treatment, patients on the shorter regimen are more likely to persist in completing therapy, compared with patients on the longer regimen (Lardizabal and coworkers, unpublished data). SITE/POPULATION-SPECIFIC CONSIDERATIONS Foreign-born persons who have arrived in the United States within the past 5 years from countries with high TB incidence account for an increasing proportion of TB in the United States, and should be targeted for tuberculin testing and treatment of LTBI (3). Many community health centers, some funded by the Health Resources and Services Administration (HRSA), serve large numbers of recent immigrants at high risk of developing active TB and who may harbor INH-resistant organisms. These individuals are often working poor—yet of higher socioeconomic levels than U.S.-born patients, as we have recently shown (28)—and have difficulty adhering to lengthy treatment regimens that require monthly clinic visits. The advantages of 4R suggest its utility when health departments collaborate with these facilities to strengthen targeted testing and treatment of LTBI. It is increasingly recognized that there is a high prevalence of active TB with TB transmission in prisons and jails (29), which may significantly impact community TB rates (30). Bureau of Justice Statistics officials estimate that more than three times as many jail inmates are incarcerated for at least 4 months (7.1% or 900,000), compared with inmates incarcerated for at least 9 months (2.1% or 265,000) (Allen J. Beck, personal communication). Studies show that up to 25% of correctional facility inmates have LTBI (31, 32). Furthermore, persons started on LTBI treatment while incarcerated seldom keep initial referral appointments for continued treatment following release, and even fewer complete a full course of therapy (33–35). Therefore, if treatment is made available during incarceration, many more inmates could complete a 4R regimen than 9H, thus decreasing the number of referrals to the health department for follow-up treatment. We recently reported that, in reviewing the medical records at a county jail in New Jersey, only 1 of 20 inmates with a projected stay of 4 months or more and who were eligible for LTBI treatment according to published guidelines (3, 36) was placed on treatment for LTBI (37). This regimen might also be considered for use in prisons, to minimize problems with tracking inmates who are frequently transferred between facilities during incarceration. Targeted school-based tuberculin testing and on-site, directly observed treatment of LTBI by school nurses with isoniazid has been shown to be successful (38) and cost-effective (39). However, many programs do not start early enough in the school year, and the student must complete treatment on a self-administered basis over the summer. We have shown, even in a well coordinated, effective school-based program that 56% of the students had to complete treatment during summer vacation (38). This tended to disrupt care and burden health departments with follow-up visits. A regimen of 4R might be particularly indicated in schools that enroll large numbers of students from countries with high levels of INH resistance. In addition, 4R would give schools the flexibility of starting the program later in the school year and would ensure that students completed therapy before the summer break. PRECAUTIONS As with any form of monotherapy, including 9H, the most important precaution is that clinicians thoroughly evaluate the patient to exclude the presence of active TB before starting treatment. Patients with unrecognized TB disease can rapidly develop rifampin mono-resistance if treated with rifampin alone. This may especially be a problem in HIV-infected patients in whom establishing a diagnosis of TB is often difficult (40). At a minimum, the process of exclusion requires a good posterior–anterior chest film read by a competent radiologist and an assessment of signs and symptoms for TB. If radiographic or clinical findings are consistent with pulmonary or extrapulmonary TB, further studies (e.g., medical evaluation, bacteriologic examinations, and a comparison of the current and old chest radiographs) should be done before any treatment to determine if treatment for active TB is indicated (3). As noted earlier, due to the increased potential for HIV-infected patients to develop rifampin-resistant disease, we would caution against using 4R in persons known or suspected of having co-infection with HIV. Although 4R appears to be less hepatotoxic than 9H or 2RZ, clinicians should still assess the patient for potential drug–drug interactions before prescribing rifampin (3). Rifampin accelerates clearance of several drugs metabolized by the liver (e.g., methadone, coumadin derivatives, glucocorticoids, hormonal contraceptives, oral hypoglycemic agents, digitalis, anticonvulsants, dapsone, ketoconazole, and cyclosporine). In our experience, we have usually recommended treatment with 9H instead of 4R for patients taking any of these medications. Rifampin also decreases the effectiveness of protease inhibitors and nonnucleoside reverse transcriptase inhibitors in the treatment of HIV infection—another reason to avoid 4R in the treatment of LTBI in HIV-infected patients. In addition, patients should be advised that rifampin is excreted in urine, tears, sweat, and other body fluids and colors them orange. If a patient wears contact lenses, we recommend not wearing them during 4R treatment or switching to a 9H regimen. Finally, it is to be remembered that Rifampin is a broad spectrum antibiotic that when appropriately administered could have some impact on pathogens of public health such as methacillin-resistant Staphylococcus and other pathogens of public health. However, when Rifampin became very widely used as a new, mandatory first line drug in short course chempotherapy of TB disease, such impact was not reported. SUMMARY In summary, with 2RZ having been curtailed due to toxicity, and because of adherence and other difficulties with the 9H regimen, there are several reasons for considering the use of 4R in the treatment of LTBI (Table 1).
We argue that 4R is an effective, relatively nontoxic, affordable strategy that clinicians and program managers should consider for more widespread use in selected populations and settings to effectively and efficiently treat LTBI, thereby accelerating the decline of TB in their communities. Acknowledgments Reynard McDonald M.D., Bonita Mangura M.D., Lillian Pirog R.N., Karen Galanowsky R.N., Rajita Bhavaraju M.P.H., and Eileen Napolitano reviewed the manuscript and made helpful suggestions. FOOTNOTES Conflict of Interest Statement: L.B.R. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; A.L. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript; C.H.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form May 4, 2004; accepted in final form July 28, 2004 REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
