Abstract
Major adverse reactions to antituberculosis drugs can cause significant morbidity, and compromise treatment regimens for tuberculosis (TB). Among patients treated for active TB we estimated the incidence, and risk factors, of major side effects from first-line anti-TB drugs. Side effects, resulting in modification or discontinuation of therapy, or hospitalization, were attributed on the basis of resolution after withdrawal, and/or recurrence with rechallenge. Among 430 patients treated between 1990 and 1999, the incidence of all major adverse effects was 1.48 per 100 person-months of exposure (95% confidence interval [95% CI], 1.31 to 1.61) for pyrazinamide, compared with 0.49 (95% CI, 0.42 to 0.55) for isoniazid, 0.43 (95% CI, 0.37 to 0.49) for rifampin, and 0.07 (95% CI, 0.04 to 0.10) for ethambutol. Occurrence of any major side effect was associated with female sex (adjusted hazard ratio, 2.5; 95% CI, 1.3 to 4.7), age over 60 years (adjusted hazard ratio, 2.9; 95% CI, 1.3 to 6.3), birthplace in Asia (adjusted hazard ratio, 2.5; 95% CI, 1.3 to 5.0), and human immunodeficiency virus–positive status (adjusted hazard ratio, 3.8; 95% CI, 1.05 to 13.4). Pyrazinamide-associated adverse events were associated with age over 60 years (adjusted hazard ratio, 2.6; 95% CI, 1.01 to 6.6) and birthplace in Asia (adjusted hazard ratio, 3.4; 95% CI, 1.4 to 8.3), whereas rifampin-associated adverse events were associated with age over 60 years (adjusted hazard ratio, 3.9; 95% CI, 1.02 to 14.9) and human immunodeficiency virus–positive status (adjusted hazard ratio, 8.0; 95% CI, 1.5 to 43). The incidence of pyrazinamide-induced hepatotoxicity and rash during treatment for active TB was substantially higher than with the other first-line anti-TB drugs, and higher than previously recognized.
A major adverse reaction to one of the first-line antituberculosis drugs, which results in discontinuation of that drug, has several implications. There may be considerable morbidity, even mortality, particularly with drug-induced hepatitis (1–4). These events may incur substantial additional costs because of added outpatient visits, tests, and in more serious instances hospitalizations (3, 4). Alternative agents may have greater problems with toxicity, and are often less effective, so that treatment must be prolonged, with attendant challenges to ensure compliance. As a result the risk of treatment failure and relapse are higher.
The occurrence, risk factors, morbidity, and mortality of adverse events from isoniazid (INH), particularly hepatotoxicity, have been well defined (1, 2, 5). Adverse reactions to rifampin (RIF) and ethambutol (EMB) have been well documented (6–8), although causality of these drugs may be less certain because they are seldom used alone. The incidence of major side effects associated with pyrazinamide (PZA), is somewhat controversial. Authoritative treatment guidelines (9) have stated that “there does not appear to be a significant increase in hepatotoxicity when PZA is added to INH and RIF, based on results from large scale randomized trials” (10–12). However, studies of patients treated for active disease (4, 13–16), or receiving 2 months of RIF and PZA for latent infection (17–19), have reported serious adverse events attributable to PZA. We have estimated the incidence and associated risk factors of serious side effects from first-line antituberculosis drugs among all patients treated for active tuberculosis (TB) at a single center.
All patients who were treated for active TB at the Montreal Chest Institute (Montreal, Canada) between 1990 and 1999 were identified. Active TB was considered confirmed if Mycobacterium tuberculosis was isolated from mycobacterial culture, or DNA probe for M. tuberculosis complex was positive after amplification by polymerase chain reaction. Active TB was considered clinically diagnosed if patients were considered improved by their treating physician after completing a full course of multidrug treatment for active TB.
Patients with active TB were seen at least monthly by the nurse case manager and treating physician. At the time of these visits, patients were questioned specifically regarding occurrence of common side effects to TB drugs. Liver transaminases were checked routinely in all patients after 1 month of therapy, and thereafter if symptoms arose. Patients were encouraged to return at any time if new symptoms or problems arose during therapy. If drug-induced hepatitis was suspected or observed then INH, RIF, and PZA were stopped, and if a rash or drug fever occurred then all anti-TB agents were stopped. Once the side effect improved, drugs were restarted, one by one. When the responsible drug was not known, the timing and order of rechallenge were at the discretion of the treating physician.
From patients' medical and nursing records, information was abstracted regarding age, sex, country of origin, date of arrival in Canada, symptoms, alcohol and intravenous drug use, comorbid conditions, other medications, human immunodeficiency virus (HIV) serology, whether women were pregnant or postpartum, method of detection, site of disease, potential risk factors for active TB, results of acid-fast bacillus smear, cultures, drug sensitivity results, dose plus duration of all anti-TB drugs prescribed, and patients' weight. Detection of TB disease was considered active if the diagnosis was made as a result of contact tracing or immigration screening programs. Records of patients who developed side effects were reviewed in detail for risk factors for side effects, specific investigations such as hepatitis serology or ultrasonic examinations, as well as consequences including hospitalizations, additional visits to clinic by patients, or at patients' homes by nurses.
A major side effect was defined as any adverse reaction that resulted in discontinuation of one or more drugs, and/or directly resulted in hospitalization. Hepatitis was defined as liver transaminases more than three times higher than the upper limit of normal in the presence of symptoms such as anorexia, nausea, vomiting, or abdominal pain, or transaminases more than five times the upper limit of normal without symptoms. Episodes of hepatitis were considered drug induced if transaminases were normal before therapy, increased during therapy, and returned to normal after discontinuation of the responsible drug.
A drug was defined as responsible for the side effect if symptoms and signs resolved after withdrawal, and recurred after rechallenge with that drug. Attribution was also made if a major side effect resolved with discontinuation of the drug, even without rechallenge. If a side effect resolved after discontinuation of two or more drugs, which were not reintroduced, then for the calculation of incidence, each event was divided equally between all possible causative drugs. (For example, if two drugs could have been responsible, each was assigned an incidence of 0.5 for that event.)
The incidences of major side effects, overall, and by drug and type, were calculated in person-months, with 95% confidence intervals estimated for proportions as suggested by Dawson-Saunders and Trapp (20). To account for the variable time, and for multiple factors potentially affecting their occurrence, Cox proportional hazards multivariate regression was used to estimate adjusted hazard ratios, and their 95% confidence intervals for risk factors for side effects (21). SAS for PCs (Statistical Analysis System; SAS, Cary, NC) was used for all statistical analysis.
Of the 430 patients treated for active TB, 23 were transferred to other institutions, mostly to their referring institutions far north of Quebec, and 4 patients were lost to follow-up. Serious side effects were not reported in these 27 patients while they were treated at the Montreal Chest Institute. For the remaining 403 patients, completion of a full course of treatment was documented in their hospital records. A total of 408 patients (95%) took daily self-administered therapy. As shown in Table 1
n | Percentage of Total | |
|---|---|---|
| Age, yr | ||
| 17–34 | 219 | 51 |
| 35–59 | 126 | 29 |
| 60–94 | 85 | 20 |
| Sex | ||
| Male | 278 | 65 |
| Female | 152 | 35 |
| Region of origin | ||
| Canada | 53 | 12 |
| Europe and Central/South America | 125 | 29 |
| Africa | 68 | 16 |
| Asia | 180 | 42 |
| Year of diagnosis | ||
| 1990–1993 | 137 | 32 |
| 1994–1996 | 153 | 36 |
| 1997–1999 | 140 | 33 |
| Method of detection | ||
| Active (screening) | 151 | 35 |
| Passive (symptoms) | 279 | 65 |
| Site of disease | ||
| Pulmonary | 392 | 91 |
| Pleural | 22 | 5 |
| Other | 16 | 4 |
| Microbiologic status | ||
| Smear positive, culture positive | 133 | 31 |
| Smear negative, culture positive | 231 | 54 |
| Smear negative, culture negative | 66 | 15 |
| Drug susceptibilities* | ||
| Sensitive to all drugs | 313 | 88 |
| MDR | 6 | 2 |
| Other resistance, not MDR | 38 | 11 |
| Risk factors for hepatotoxicity | ||
| 1. Known alcohol, viral hepatitis | 72 | 17 |
| 2. Abnormal baseline LFTs | 38 | 9 |
| Either (1) or (2) | 96 | 23 |
| HIV | ||
| Positive | 18 | 4 |
| Negative | 149 | 35 |
| Not available | 263 | 61 |
| Drug dose (mg/kg) | 5.2 | 0.9 |
| Isoniazid, mean | ||
| Rifampin, mean | 10.2 | 1.8 |
| Pyrazinamide, mean | 24.2 | 4.4 |
| Ethambutol, mean | 16.8 | 3.7 |
| Outcomes | ||
| Died | ||
| Total | 18 | 4 |
| Before/during treatment | 6 | 1 |
| After treatment | 12 | 3 |
| Serious side effects | ||
| Total—patients | 37 | 9 |
| Total—events | 46 | |
| Rash/drug fever | 21 | |
| Hepatitis | 12 | |
| Severe gastrointestinal upset | 11 | |
| Visual toxicity | 1 | |
| Arthralgia | 1 |
Thirty-seven patients had major side effects, 9 of whom experienced a second major adverse reaction, for a total of 46 events. Of these, 36 could be attributed to a single drug, and 10 could not. The most common serious adverse event was rash and/or drug fever. Twelve patients developed drug-induced hepatitis. Of these, all had normal pretreatment liver transaminases, none had a history of other comorbid conditions, or of alcohol or drug use, and none were pregnant or postpartum. Of 10 tested for hepatitis B and 7 tested for hepatitis C, none were positive. Three had pretreatment drug resistance; of these, two had prior anti-TB therapy, and one was HIV-positive. Eleven were symptomatic, and in all 12 the transaminases exceeded 5 times the upper limit of normal, yet returned to normal levels while patients were still undergoing rifampin-containing treatment. There was no difference in patient weight, or in drug dosage calculated on a milligram-per-kilogram basis, between all 37 who developed serious side effects, or between the 12 with hepatitis, and the 393 patients without side effects.
On six occasions, severe hepatitis resulted in discontinuation of INH and PZA, and neither was restarted. In three instances (two of rash and one of severe gastrointestinal [GI] intolerance) RIF and PZA were stopped, and not rechallenged, and one patient developed severe vomiting, which resolved only when INH, RIF, and PZA were stopped. These 10 events were divided equally among the potentially responsible drugs. As shown in Table 2
n | Person-months | Side Effect | No. of Persons | No. of Events | Incidence* | 95% CI | |
|---|---|---|---|---|---|---|---|
| All drugs | 429 | 8,488 | All | 37 (9%) | 46 | 0.55 | 0.48 to 0.62 |
| Rash | 18 (4%) | 21 | 0.25 | 0.20 to 0.30 | |||
| Hepatitis | 12 (3%) | 12 | 0.14 | 0.10 to 0.18 | |||
| GI | 8 (2%) | 11 | 0.13 | 0.10 to 0.16 | |||
| Isoniazid | 427 | 2,747 | All | 17 (4%) | 13.3 | 0.49 | 0.42 to 0.55 |
| Rash | 4 (1%) | 4 | 0.15 | 0.11 to 0.18 | |||
| Hepatitis | 8 (2%) | 5 | 0.18 | 0.14 to 0.22 | |||
| GI | 5 (1%) | 4.3 | 0.16 | 0.12 to 0.20 | |||
| Rifampin | 425 | 2,972 | All | 14 (3%) | 12.83 | 0.43 | 0.37 to 0.49 |
| Rash | 10 (2%) | 9 | 0.30 | 0.25 to 0.35 | |||
| Hepatitis | 0 | 0 | 0 | — | |||
| GI | 5 (1%) | 3.8 | 0.13 | 0.10 to 0.16 | |||
| Pyrazinamide | 405 | 1,336 | All | 24 (6%) | 19.83 | 1.48 | 1.36 to 1.60 |
| Rash | 9 (2%) | 8 | 0.60 | 0.52 to 0.68 | |||
| Hepatitis | 10 (2%) | 7 | 0.52 | 0.45 to 0.59 | |||
| GI | 4 (1%) | 2.8 | 0.21 | 0.16 to 0.25 | |||
| Ethambutol | 329 | 1,433 | Visual | 1 (0.3%) | 1 | 0.07 | 0.04 to 0.10 |
Figure 1. Incidence of serious side effects by type and drug. Shaded columns, isoniazid; cross-hatched columns, rifampin; open columns, pyrazinamide; dotted columns, ethambutol.
[More] [Minimize]As shown in Table 3
Any Serious* | Rash/Fever† | Hepatitis‡ | GI Upset§ | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| HR | 95% CI | HR | 95% CI | HR | 95% CI | HR | 95% CI | |||||
| Female sex (versus male) | 2.5 | 1.3 to 4.7 | 1.9 | 0.7 to 4.8 | 2.2 | 0.7 to 6.9 | 3.6 | 0.6 to 11.8 | ||||
| Age, yr | ||||||||||||
| 35–59 (versus < 35) | 1.7 | 0.8 to 3.8 | 1.0 | 0.3 to 3.1 | 4.8 | 0.9 to 25 | 2.1 | 0.3 to 14.9 | ||||
| 60+ (versus < 35) | 2.9 | 1.3 to 6.3 | 1.3 | 0.4 to 4.1 | 7.7 | 1.5 to 40 | 6.4 | 1.2 to 36 | ||||
| From Asia (versus all others) | 2.5 | 1.3 to 5.0 | 2.8 | 1.1 to 7.5 | 2.2 | 0.7 to 6.9 | 3.6 | 0.8 to 15.2 | ||||
| Method of detection passive (versus active) | 2.5 | 0.9 to 6.6 | 2.3 | 0.6 to 8.3 | —‖ | — | 2.6 | 0.6 to 11.7 | ||||
| Smear positive (versus smear negative) | 1.3 | 0.7 to 2.6 | 1.0 | 0.4 to 2.7 | 1.8 | 0.6 to 5.7 | 0.5 | 0.1 to 2.4 | ||||
| Drug resistant (versus pansensitive) | 1.8 | 0.8 to 4.3 | 1.0 | 0.2 to 4.5 | 2.7 | 0.7 to 10.5 | 0.9 | 0.1 to 7.3 | ||||
| Abnormal baseline LFTs (versus normal) ¶ | 1.6 | 0.6 to 4.2 | 2.3 | 0.6 to 8.0 | — | — | 3.9 | 0.8 to 19.5 | ||||
| HIV-positive (versus negative or NA) | 3.8 | 1.05 to 13.4 | 5.1 | 1.02 to 27 | 4.3 | 0.5 to 38 | — | — | ||||
Figure 2. Association of sex with interval from start of therapy until occurrence of any major side effect (using Cox proportional hazards analysis). Solid line, male sex; dotted line, female sex. Lines truncated at time after which no further events occurred.
[More] [Minimize]
Figure 3. Association of age with interval from start of therapy until occurrence of any major side effect (using Cox proportional hazards regression). Solid line, age 17 to 34 years; dotted line, age 35 to 59 years; dashed line, age 60 years and older. Lines truncated at time after which no further events occurred.
[More] [Minimize]INH | RIF | PZA | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Characteristics (Comparison) | HR* | 95% CI | HR* | 95% CI | HR* | 95% CI | |||
| Female sex (versus male) | 2.5 | 0.96 to 6.6 | 2.2 | 0.8 to 6.1 | 2.2 | 0.96 to 4.8 | |||
| Age, yr | |||||||||
| 35–59 (versus < 35) | 3.4 | 1.1 to 10.3 | 2.3 | 0.6 to 8.7 | 1.1 | 0.4 to 3.0 | |||
| 60+ (versus < 35) | 1.9 | 0.5 to 8.1 | 3.9 | 1.02 to 14.9 | 2.6 | 1.01 to 6.6 | |||
| From Asia (versus all others) | 2.1 | 0.8 to 5.6 | 2.8 | 0.9 to 8.4 | 3.4 | 1.4 to 8.3 | |||
| Method of detection passive (versus active) | —† | — | — | — | 1.5 | 0.5 to 4.1 | |||
| Smear positive (versus smear negative) | 1.2 | 0.5 to 3.3 | 0.8 | 0.2 to 2.5 | 1.5 | 0.7 to 3.5 | |||
| Drug resistant (versus pansensitive) | 3.4 | 1.1 to 10.9 | 1.4 | 0.3 to 6.6 | 1.9 | 0.7 to 5.3 | |||
| Abnormal baseline LFTs (versus normal)‡ | 0.5 | 0.1 to 4.1 | 1.7 | 0.4 to 8.1 | 2.2 | 0.7 to 6.6 | |||
| HIV-positive (versus -negative or NA) | 2.4 | 0.3 to 19.5 | 8.0 | 1.5 to 43 | 2.1 | 0.3 to 17.1 | |||
Of the 46 serious adverse events, 20 resulted in hospitalization, for a median of 16 days. One patient was hospitalized twice for adverse drug reactions. In addition, 29 of the 37 patients who developed serious side effects made a total of 91 extra clinic visits, and the TB clinic nurses made an additional 30 home visits to 8 of these patients. Total duration of therapy was 380 (± 209) days for the patients with serious side effects, compared with 228 (± 111) days for the remainder (p < 0.001). Rechallenge with possibly responsible drugs was attempted 21 times. Only 3 of the 12 patients with hepatitis were rechallenged—none successfully, compared with 11 of 21 with rash/drug fever—1 successfully, and 7 of 11 with GI intolerance—2 successfully.
In this study, among 430 patients treated for active TB, 46 serious adverse reactions occurred. The incidence of serious side effects, especially hepatitis and rash, was highest with PZA, and was associated with female sex, older age, birth in Asia, and HIV infection. Rechallenge with responsible drugs was generally unsuccessful. The consequences of these adverse events included hospitalizations, prolonged therapy, and more clinic and home visits.
A strength of the study was the inclusion of all patients with a wide spectrum of disease severity and comorbid illnesses treated at a single center by a small group of physicians who provided a reasonably standardized approach to identification and management of side effects. Serious side effects were defined as those requiring a documented change in therapy or hospitalization, eliminating the need for interpretation of symptoms. However, this could have resulted in an underestimate of all drug-related side effects. The inclusion of all patients was similar to other studies of complete cohorts of patients with TB (4, 13, 14), and the outcome definitions were also comparable to other series (4, 6, 13, 14). This should enhance the comparability and generalizability of results. Estimation of the incidence of side effects in person-time units took into account the differences in duration of treatment with each drug. Person-time estimates are important when considering the advantages or disadvantages of shorter regimens, such as 2 months of RIF and PZA for latent TB infection.
Nevertheless, there were several important limitations, inherent in retrospective studies. For example, abnormal baseline liver transaminases were noted in 38 patients. Although many had a history of alcohol abuse, or positive hepatitis B or C serology, no explanation was found for a substantial number of patients. HIV status was missing for a large number of patients, for three reasons. Physicians did not perform this test if they believed there was low probability of HIV infection, such as for many older persons. Many immigration applicants refused, because of concerns that a positive result could jeopardize their chances of acceptance. Third, HIV test results were not filed in the records but sent directly to the treating physicians. This meant results were much less accessible, particularly if negative. Positive results were unlikely to have been missed because such results would have important clinical repercussions including further evaluation, referral to HIV specialists in our institution, antiretroviral therapy, and specific follow-up.
The small number with serious side effects provided limited power to detect significant associations with a number of characteristics such as HIV infection, which was found in few patients. In addition, because HIV infection was a risk factor for serious adverse reactions, incidence may be higher in populations with higher HIV seroprevalence. Another limitation is that there were few patients with extrapulmonary disease in this study, thus limiting inferences to this group of patients who might tolerate therapy differently. Finally, this study was restricted to adults. Serious side effects can occur in young children, and may be associated with different risk factors (16), although PZA is still the most common offender (16).
Serious adverse events had to be divided between more than 1 possible drug in 10 of the 46 (22%) major adverse events. Six patients had sufficiently severe hepatotoxicity that their treating physicians did not rechallenge them with either INH or PZA. As a result, the incidence of PZA- or INH-induced hepatitis could have been over- or underestimated. Single-drug hepatitis was most commonly due to PZA, but attribution was divided equally between INH and PZA for the six patients, meaning PZA-associated hepatitis may have been underestimated. If the hepatitis in all six patients had been attributed to INH, instead of divided with PZA, the incidence of hepatotoxicity due to INH and PZA would have been 0.3/100 and 0.3/100 persons-months, respectively, and of all side effects, 0.6/100 and 1.3/100 person-months, respectively. Even using this most conservative estimate, PZA was still associated with the highest incidence of all serious side effects, and equal to INH in terms of risk of hepatotoxicity.
The most striking finding was the common occurrence of major adverse reactions due to PZA. This is in contrast to three large trials in which patients randomized to receive PZA had a similar occurrence of adverse events as patients who did not (10–12). However, our results are in agreement with those of four series with a total of 3,112 patients treated in specialized centers in Britain (13), Denmark (14), Germany (4), and Argentina (22). In these studies, PZA was the most common causative drug for all side effects (4, 13, 14, 22), hepatitis (4, 14), or rash (13). Incidence of PZA-associated adverse events in the present study was similar to that reported in series where the majority were outpatients (13, 14) but was lower than in series of hospitalized (4) or retreatment (22) patients. The higher incidence in these patients may reflect the greater likelihood of side effects among sicker patients (14). This was also suggested in this study, by the somewhat higher incidence among passively diagnosed patients who had more extensive disease.
The incidence of PZA-associated side effects in this and other unselected cohorts was higher than in randomized trials. Was this because patients in randomized trials are more carefully selected, or was it due to poor management in the cohort studies? Poor management seems unlikely as all five cohorts were reported from centers with considerable experience. In fact, of 3,542 patients reviewed in the 5 reports, only 1 died from a serious adverse reaction (4, 13, 14, 22).
The contrast between the incidence of PZA-associated side effects in randomized trials (10–12) and cohorts of all patients (the present work and see references 4, 13, 14, and 22) is remarkably similar to the difference in reported rates of hepatoxicity with the 2-month RIF and PZA regimen. In three randomized clinical trials the safety and tolerability of 2 months of RIF and PZA were excellent, and better than INH (23–25). Yet following more widespread use, serious and even fatal hepatotoxicity was reported (17). In more recent studies, the incidence of drug-induced hepatitis was 7.7% (18) or 9.4% (19) in patients taking 2 months of RIF and PZA and 3.5% (26) or 18% (27) in patients taking levofloxacin (Levoquin) and PZA for 6 months.
This experience is remarkably similar to the INH story. During randomized, placebo-controlled trials involving more than 100,000 persons, only 0.1% of those receiving INH developed hepatitis (28). However, shortly after the widespread introduction of INH into routine practice in the United States, serious hepatitis and deaths were reported (1, 29, 30), with incidence greater than 1% (i.e., 10 times higher than in randomized trials). Reported incidence of serious hepatotoxicity is much lower 25 years later (5). However, these observed differences in occurrence of serious side effects, between randomized trials and routine clinical use, reinforce the need for careful surveillance after the introduction of new drugs, or regimens, into routine clinical practice (31).
A number of other findings in the study are worthy of comment. Female sex was consistently associated with increased occurrence of adverse events even after adjusting for potential confounding factors in multivariate analysis. Female sex has been associated with development of drug-induced hepatitis in two studies (14, 32) but not in two others (1, 4) and with the development of rash in a single study (13). These contradictory findings, although inconclusive, suggest that closer monitoring of females during antituberculosis therapy is warranted. Asian-born patients had higher risk of adverse events, particularly due to PZA. This interesting finding warrants further investigation, particularly given the finding of an association of HLA haplotype with hepatotoxicity (33). However, this finding should be interpreted with particular caution as it results from post-hoc groupings in the analysis, and may have been due to chance, or features unique to our patient population.
At doses of 15 mg/kg, EMB-associated visual toxicity, known to be dose related (8), occurred in only one patient, and no other serious side effects were identified, as reported in three other series (4, 13, 22). In this study no instance of hepatitis was attributed to RIF, and all 12 patients with drug-induced hepatitis had reversion of transaminases to normal while still receiving treatment with RIF-containing regimens. In two other series, RIF was also the least likely to cause hepatoxicity (4, 14). Of elderly Chinese silicotic males treated with RIF alone, none developed any alteration of hepatic function (34). Taken together, these findings suggest that RIF does not commonly cause drug-induced hepatitis, although it might synergistically increase the risk of INH-induced hepatitis (7)—a possibility that could not be explored with this retrospective design.
In conclusion, serious adverse reactions to antituberculosis drugs were common, and resulted in substantial health care services utilization, as well as prolongation of therapy. Better tolerated therapy should be part of the TB research agenda. Female, HIV-infected, older, and Asian-born patients had greater risk. Incidence of PZA-induced hepatotoxicity and rash was significantly higher than for the other first-line anti-TB drugs. On the basis of these findings, the drug most likely responsible for the occurrence of hepatitis or rash during therapy for active TB is PZA.
| 1. | Kopanoff DE, Snider DE, Caras GJ. Isoniazid-related hepatitis. Am Rev Respir Dis 1978;117:991–1001. |
| 2. | Franks AL, Binkin NJ, Snider DE, Rokaw WM, Becker S. Isoniazid hepatitis among pregnant and postpartum hispanic patients. Public Health Rep 1989;104:151–155. |
| 3. | Durand F, Bernuau J, Pessayre D, Samuel D, Belaiche J, Degott C, Bismuth H, Belghiti J, Erlinger S, Rueff B, et al. Deleterious influence of pyrazinamide on the outcome of patients with fulminant or subfulminant liver failure during antituberculous treatment including isoniazid. Hepatology 1995;21:929–932. |
| 4. | Schaberg T, Rebhan K, Lode H. Risk factors for side-effects of isoniazid, rifampin and pyrazinamide in patients hospitalized for pulmonary tuberculosis. Eur Respir J 1996;9:2026–2030. |
| 5. | Nolan CM, Goldberg SV, Buskin SE. Hepatoxicity associated with isoniazid preventive therapy: a 7-year survey from a public health tuberculosis clinic. JAMA 1999;281:1014–1018. |
| 6. | Van den Brande P, Steenbergen WV, Verdoot G, Demedts M. Aging and hepatoxicity of isoniazid and rifampin in pulmonary tuberculosis. Am J Respir Crit Care Med 1995;152:1705–1708. |
| 7. | Steele MA, Burk RF, DesPrez RM. Toxic hepatitis with isoniazid and rifampin. Chest 1991;99:465–471. |
| 8. | Leibold JE. The ocular toxicity of ethambutol and its relation to dose. Ann N Y Acad Sci 1966;135:904–909. |
| 9. | American Thoracic Society. Treatment of tuberculosis and tuberculosis infection in adults and children. Am Rev Respir Dis 1994;149:1359–1374. |
| 10. | Combs DL, O'Brien RJ, Geiter LJ. USPHS tuberculosis short course chemotherapy trial 21: effectiveness, toxicity, and acceptability. Ann Intern Med 1990;112:397–406. |
| 11. | British Thoracic Association. A controlled trial of 6-months isoniazid and rifampin therapy for pulmonary tuberculosis: first report: results during drug therapy. Br J Dis Chest 1981;75:141–153. |
| 12. | Zierski M, Bek E. Side-effects of drug regimens used in short-course chemotherapy for pulmonary tuberculosis: a controlled clinical study. Tubercle 1980;61:41–49. |
| 13. | Ormerod LP, Horsfield N. Frequency and type of reactions to antituberculosis drugs: observations in routine treatment. Tuber Lung Dis 1996;77:37–42. |
| 14. | Dossing M, Wilcke TR, Askgaard DS, Nyboo B. Liver injury during antituberculosis treatment: an 11-year study. Tuber Lung Dis 1996;77:335–340. |
| 15. | Danan G, Pessayre D, Larrey D, Benhamou JP. Pyrazinamide fulminant hepatitis: an old hepatotoxin strikes again. Lancet 1981;7:1056–1057. |
| 16. | Ohkawa K, Hashiguchi M, Ohno K, Kiuchi C, Takahashi S, Kondo S, Echizen H, Ogata H. Risk factors for antituberculous chemotherapy-induced hepatotpxicity in Japanese pediatric patients. Clin Pharmacol Ther 2002;72:220–226. |
| 17. | American Thoracic Society, Centers for Disease Control and Prevention. Fatal and severe liver injuries associated with rifampin and pyrazinamide for latent tuberculosis infection, and revisions in the American Thoracic Society/CDC recommendations. MMWR 2001;50:733–735. |
| 18. | Jasmer RM, Saukkonen JJ, Blumberg HM, Daley CL, Bernardo J, Vittinghoff E, King MD, Kawamura LM, Hopewell PC. Short-Course Rifampin and Pyrazinamide for Tuberculosis Infection (SCRIPT) Study Investigators. Short-course rifampin and pyrazinamide compared with isoniazid for latent tuberculosis infection: a multicenter clinical trial. Ann Intern Med 2002;137:640–647. |
| 19. | Lee AM, Mennone JZ, Jones RC, Paul WS. Risk factors for hepatotoxicity associated with rifampin and pyrazinamide for the treatment of latent tuberculosis infection: experience from three public health tuberculosis clinics. Int J Tuberc Lung Dis 2002;6:995–1000. |
| 20. | Dawson-Saunders B, Trapp RG. Basic and clinical biostatistics. Norwalk: Appleton & Lange; 1990. |
| 21. | Cox DR. Regression models and life-tables [with discussion]. J. R. Stat. Soc. B 1972; 34:187–220. |
| 22. | Gonzalez Montaner LJ, Dambrosi A, Manassero M, Dambrosi V, Dambrosi ML. Adverse effects of antituberculosis drugs causing changes in treatment. Tubercle 1982;63:291–294. |
| 23. | Mwinga A, Hosp M, Godfrey-Faussett P, Quigley M, Mwaba P, Mugala BN, Nyirenda O, Luo N, Pobee J, Elliott AM, McAdam KP, Porter JD. Twice weekly tuberculosis preventive therapy in HIV infection in Zambia. AIDS 1998;12:2447–2457. |
| 24. | Halsey NA, Coberly JS, Desormeaux J, Losikoff P, Atkinson J, Moulton LH, Contave M, Johnson M, Davis H, Geiter L, Johnson E, Huebner R, Boulos R, Chaisson RE. Randomized trial of isoniazid versus rifampin and pyrazinamide for prevention of tuberculosis in HIV-1 infection. Lancet 1998;351:786–792. |
| 25. | Gordin F, Chaisson RE, Matts JP, Miller C, de Lourdes Garcia M, Hafner R, Valdespino JL, Coberly J, Schechter M, Klukowicz AJ, Barry MA, O'Brien RJ. Rifampin and pyrazinamide vs isoniazid for prevention of tuberculosis in HIV-infected persons. JAMA 2000;283:1445–1450. |
| 26. | Lou HX, Shullo MA, McKaveney TP. Limited tolerability of levofloxacin and pyrazinamide for multidrug-resistant tuberculosis prophylaxis in a solid organ transplant population. Pharmacotherapy 2002;22:701–704. |
| 27. | Papastavros T, Dolovich LR, Holbrook A, Whitehead L, Loeb M. Adverse events associated with pyrazinamide and levofloxacin in the treatment of latent multi-drug-resistant tuberculosis. CMAJ 2002;167:131–136. |
| 28. | Ferebee SH. Controlled chemoprophylaxis trials in tuberculosis. Adv Tuberc Res 1969;17:28–106. |
| 29. | Garibaldi RA, Drustin RE, Ferebee SH, Gregg MB. Isoniazid-associated hepatitis. Am Rev Respir Dis 1972;106:357–365. |
| 30. | Israel HL. Isoniazid-associated hepatitis: reconsideration of the indication for administration of isoniazid. Gastroenterology 1975;69:539–542. |
| 31. | Burman WJ, Reves RR. Hepatotoxicity from rifampin plus pyrazinamide: lessons for policymakers and messages for care providers. Am J Respir Crit Care Med 2001;164:1112–1113. |
| 32. | Teleman MD, Chee CBE, Earnest A, Wang YT. Hepatotoxicity of tuberculosis chemotherapy under general programme conditions in Singapore. IUATLD. Int. J. Tuberc. Lung Dis. 2002;6:699–705. |
| 33. | Sharma SK, Balamurugan A, Saha PK, Pandey RM, Mehra NK. Evaluation of clinical and immunogenetic risk factors for the development of hepatotoxicity during antituberculosis treatment. Am J Respir Crit Care Med 2002;166:916–919. |
| 34. | Hong Kong Chest Service/Tuberculosis Research Centre, MBMRC. A double-blind placebo-controlled clinical trial of three antituberculosis chemoprophylaxis regimens in patients with silicosis in Hong Kong. Am Rev Respir Dis 1992;145:36–41. |
