Although patients with end-stage renal disease receiving maintenance hemodialysis are at high risk for tuberculosis, the optimal treatment regimen for latent tuberculosis infection (LTBI) in this group has scarcely been studied for predictors of completion rate and adverse drug events (ADE). We prospectively enrolled dialysis patients for LTBI intervention from three medical centers in Taiwan.
KEYWORDS: dialysis, drug-related side effects and adverse reactions, interferon gamma release tests, latent tuberculosis
ABSTRACT
Although patients with end-stage renal disease receiving maintenance hemodialysis are at high risk for tuberculosis, the optimal treatment regimen for latent tuberculosis infection (LTBI) in this group has scarcely been studied for predictors of completion rate and adverse drug events (ADE). We prospectively enrolled dialysis patients for LTBI intervention from three medical centers in Taiwan. LTBI treatments were 3 months of weekly rifapentine plus isoniazid (3HP) and 9 months of daily isoniazid (9H). Completion rate, ADE, and reasons for treatment termination were recorded. Factors associated with treatment termination and ADE were analyzed using multivariate logistic regression. In all, 91 treatment courses (41 9H and 50 3HP) were surveyed. The completion rates were 61% for 9H and 82% for 3HP (P = 0.046). Use of 9H and development of ADE with a grade of ≥2 (≥grade 2 ADE) were associated with treatment termination. Hypersensitivity occurred in 29.2% of subjects in the 3HP group and 10.8% in the 9H group (P = 0.035) and independently correlated with 3HP regimen, diabetes mellitus (DM), and peritoneal dialysis (PD). Similarly, the independent predictors of ≥grade 2 ADE were use of 3HP regimen, presence of DM, and use of PD, whereas ≥grade 3 ADE were associated with eosinophil counts of >700/mm3 after 2 weeks of LTBI treatment even after adjustment for age and gender. In conclusion, for patients on dialysis, 3HP showed a higher rate of completion but also a higher rate of ≥grade 2 ADE than 9H. In addition, DM and PD were risk factors for ≥grade 2 ADE. Eosinophilia after 2-week treatment might be an alert for severe ADE.
INTRODUCTION
Tuberculosis (TB) is a major issue in public health worldwide. An estimated one-quarter of the world’s population have Mycobacterium tuberculosis infections (1). Although the incidence of TB has decreased in the past decade (2), around 10.0 million TB cases and 1.2 million TB deaths were recorded in 2018 (3). Treatment for latent tuberculosis infection (LTBI) is an important key step in the To END TB post-2015 strategy (4, 5). In addition to TB contacts, high-risk populations needing LTBI screening include patients with end-stage renal disease (ESRD) on dialysis, those with immunodeficiency, and those using biological agents (4).
Notably, the incidence of TB is 8 to 25 times higher in the dialysis population (6–8) than in the general population. Because of atypical symptoms in patients with dialysis, delayed diagnosis of TB and higher mortality have been reported (9, 10). Therefore, LTBI detection and treatment in this group are important for prevention of active TB (11).
The 9-month daily isoniazid (9H) regimen is considered too lengthy to complete (12). In contrast, the 3-month weekly high-dose rifapentine-plus-isoniazid (3HP) regimen, which has lower hepatotoxicity and a shorter duration than the 9H course, has been widely accepted for LTBI treatment in the general population (12). However, a high incidence of flu-like symptoms has been reported for health care workers with LTBI (13). Patients with renal failure have recently been reported as being at high risk for drug adverse events (ADE) of LTBI treatment (14) even though isoniazid and rifamycins are metabolized in the liver. A recent report on hemodialysis patients showed high rates of ADE (69%) during the 3HP course but lacked head-to-head comparison to other regimens (15). In the dialysis population, a particularly fragile group, the most concerning aspect of LTBI treatment for both patients and physicians is medication-related ADE. Serious ADE may not be acceptable for treating this subclinical condition. Therefore, before further implementing LTBI treatment in patients on maintenance dialysis, the safety profile, completion rate, and relevant predictors must be studied in this population by comparing the 3HP and 9H regimens.
(This study was presented in the tuberculosis oral presentation session of the European Respiratory Society International Congress 2020.)
RESULTS
Enrollment and demographics.
From September 2014 to April 2020, we enrolled 91 patients who received LTBI treatment courses, including 41 receiving 9H and 50 receiving 3HP (Fig. 1) (for screened numbers, please see point 3 in the supplemental material). Due to ADE, three patients switched from the 9H group to the 3HP group, while 2 switched from the 3HP group to the 9H group, with 1 failure. Demographics of the population are listed in Table 1. The mean age was 56.6 years, and the majority (69.2%) were male. Diabetes mellitus (DM) was a prevalent comorbidity (27.5%), and peritoneal dialysis (PD) accounted for 11%. There were no significant differences between the two groups.
FIG 1.
CONSORT diagram for the two treatment regimens. Two patients switched from 3 months of weekly rifapentine plus isoniazid (3HP) to 9 months of daily isoniazid (9H), one due to an adverse event and the other due to drug-drug interaction resulting in unstable immunosuppressant drug concentration; three patients from the 9H group shifted to the 3HP group due to adverse events. The switched cases were counted as treatment discontinuation, and subsequent courses were not included in the analysis. Abbreviations: ESRD, end-stage renal disease; FEMH, Far Eastern Memorial Hospital; IGRA, interferon gamma release assay; LTBI, latent tuberculosis infection; NTUH, National Taiwan University Hospital; TVGH, Taipei Veterans General Hospital.
TABLE 1.
Baseline characteristicsa
| Characteristic | Value for: |
P valueb | ||
|---|---|---|---|---|
| Total (n = 91) | 9H (n = 41) | 3HP (n = 50) | ||
| Age (yrs), mean ± SD | 56.6 ± 13.0 | 54.8 ± 13.3 | 58.0 ± 12.7 | 0.250 |
| Age group (%) | 0.424 | |||
| 18–40 yrs | 12 (13.2) | 7 (17.1) | 5 (10.0) | |
| 41–64 yrs | 54 (59.3) | 25 (61.0) | 29 (58.0) | |
| ≥65 yrs | 25 (27.5) | 9 (22.0) | 16 (32.0) | |
| Male, no. (%) | 63 (69.2) | 27 (65.9) | 36 (72.0) | 0.527 |
| Active smoker, no. (%) | 19 (21.1) | 6 (15.0) | 13 (26.0) | 0.204 |
| Body mass index, mean ± SD | 23.5 ± 2.5 | 22.9 ± 3.7 | 24.1 ± 3.5 | 0.117 |
| Comorbidities, no. (%) | ||||
| Diabetes mellitus | 25 (27.5) | 14 (34.1) | 11 (22.0) | 0.197 |
| Viral hepatitisc | 10 (11.0) | 5 (12.2) | 5 (10.0) | 0.750 |
| Autoimmune disease | 4 (4.4) | 2 (4.9) | 2 (4.0) | 0.999 |
| Malignancy | 1 (1.1) | 1 (2.4) | 0 (0.0) | 0.451 |
| Peritoneal dialysis | 10 (11.0) | 4 (9.8) | 6 (12.0) | 0.999 |
| Risk for tuberculosis, no. (%) | ||||
| Immunosuppressive therapy | 7 (7.7) | 5 (12.2) | 2 (4.0) | 0.237 |
| Contact history | 1 (1.1) | 1 (2.4) | 0 (0.0) | 0.451 |
| BCG scar presence | 86 (94.5) | 39 (95.1) | 47 (94.0) | 0.999 |
| Eosinophil count, baseline (/μl) | 346 ± 304 | 348 ± 361 | 345 ± 266 | 0.968 |
| QFT-GIT titer | 2.52 ± 2.48 | 2.31 ± 2.2.1 | 2.70 ± 2.69 | 0.468 |
Abbreviations: BCG, bacillus Calmette-Guérin vaccination; QFT-GIT, QuantiFERON-TB Gold In-tube; 3HP, 3 months of weekly rifapentine plus isoniazid; 9H, 9 months of daily isoniazid.
P values were compared between 9H and 3HP regimens using chi-squared test or Fisher’s exact test for categorical variables and Student’s t test or Mann-Whitney U test for continuous variables, where appropriate.
Viral hepatitis included infection with hepatitis B and C viruses.
Completion rate and discontinuation reasons.
The completion rate was significantly higher in the 3HP group than in the 9H group (82% versus 61%; P = 0.046). Nearly 24.4% of the 9H group did not complete the LTBI treatment due to patient refusal or withdrawal (Table 2), compared to 4% in the 3HP group (P = 0.176). Terminations due to ADE and other reasons were similar in the two groups (14.6 versus 14.0%). Six patients (4 in the 9H and 2 in the 3HP group) who dropped without receiving the first dose of a regimen were excluded from further ADE analysis. For those with termination, malaise (33%) was the most common symptom, with no intergroup difference, whereas gastrointestinal symptoms (poor appetite and nausea/vomiting) were reported exclusively in the 3HP group (P = 0.120) (see Table S1 in the supplemental material). For ADE within the initial 2 weeks, patients who eventually terminated LTBI treatment had higher proportions of weakness (P = 0.050) and hypersensitivity syndrome (P = 0.084) (Table S2).
TABLE 2.
Treatment completion and adverse drug events in two regimen groups
| Variable | Value for groupa |
P valueb | |
|---|---|---|---|
| 9H (n = 41) | 3HP (n = 50) | ||
| Treatment completed, no. (%) | 25 (61.0) | 41 (82.0) | 0.046 |
| Treatment not completed reasons, no. (%) | 0.176 | ||
| Participant refusal | 10 (24.4) | 2 (4.0) | |
| Termination due to ADE or other reasonsc | 6 (14.6) | 7 (14.0) | |
| ADE, no. (%)d | 9H (n = 37) | 3HP (n = 48) | |
| Hypersensitivity | 4 (10.8) | 14 (29.2) | 0.040 |
| Flu-like syndrome | 3 (8.1) | 7 (14.6) | 0.502 |
| Gastrointestinal symptoms | 6 (16.2) | 14 (29.2) | 0.163 |
| Hepatotoxicity | 0 (0) | 0 (0) | 1.000 |
| Hospitalization, attributed to LTBI treatment | 0 (0) | 3 (6.3) | 0.249 |
| Reported maximal grade of ADEd | 9H (n = 37) | 3HP (n = 48) | 0.001 |
| Grade 1 | 12 (32.4) | 17 (35.4) | |
| Grade 2 | 2 (5.4) | 16 (33.3) | |
| Grade 3 | 3 (8.1) | 5 (10.4) | |
| Grade 4 | 0 | 1 (2.1) | |
| Grade 2 or more | 5 (13.5) | 22 (45.8) | 0.002 |
| Grade 3 or more | 3 (8.1) | 6 (12.5) | 0.725 |
Abbreviations: ADE, adverse drug events; LTBI, latent tuberculosis infection.
P values were compared between 3HP and 9H groups using chi-squared test or Fisher’s exact test for categorical variables.
3HP group: one patient’s serum level of immunosuppressant was too unstable, and the index case of another patient was later reported to be multidrug-resistant tuberculosis.
The number for ADE comparison excluded the participants who withdrew before receiving the first dose of the regimen.
Logistic multivariate regression model (Table 3) indicated that both 3HP (adjusted odds ratio [aOR], 0.17; 95% confidence interval [CI], 0.04 to 0.69; P = 0.013) and ADE with a grade of ≥2 (≥grade 2 ADE) (aOR, 6.67; 95% CI, 1.66 to 26.80; P = 0.008) were independently associated with treatment termination.
TABLE 3.
Logistic regression models for treatment terminationa
| Variable | Crude OR | 95% CI | P value | Adjusted OR | 95% CI | P value |
|---|---|---|---|---|---|---|
| 3HP vs 9H | 0.39 | 0.15–0.99 | 0.049 | 0.17 | 0.04–0.69 | 0.013 |
| Age over 60 | 0.58 | 0.23–1.49 | 0.259 | |||
| Male | 1.30 | 0.47–3.56 | 0.616 | |||
| Active smoking | 1.60 | 0.55–4.66 | 0.392 | |||
| Diabetes mellitus | 2.08 | 0.78–5.54 | 0.141 | |||
| Peritoneal dialysis | 1.79 | 0.46–6.94 | 0.401 | |||
| ≥Grade 2 ADE | 2.82 | 1.00–7.96 | 0.050 | 6.67 | 1.66–26.80 | 0.008 |
| Eosinophil countb | ||||||
| <349/μl | Reference | |||||
| 350–699/μl | >100 | 0.999 | ||||
| ≥700/μl | >100 | 0.999 |
Only variables with a P value of <0.1 in univariate regression are included in the multivariate model. Abbreviations: CI, confidence interval; OR, odds ratio.
Laboratory test obtained at 2 weeks after initiating LTBI treatment. Eosinophil count cutoff was based on 2 times the normal upper limit value.
Adverse drug events during treatment.
The most commonly reported ADE in both groups was malaise (24.3% versus 54.2%; P = 0.006). Chills (0 versus 10.5%; P = 0.043), fever (0 versus 14.6%; P = 0.017), dizziness (10.8% versus 33.4%; P = 0.015), and myalgia (2.7% versus 16.7%; P = 0.038) were more commonly reported in the 3HP group. Hypersensitivity was more common in the 3HP group (P = 0.040) (Table 2), while flu-like syndrome and gastrointestinal symptoms were similar in both groups.
Three subjects using 3HP had severe ADE leading to hospitalization. Two patients had hypertensive crisis; of these, one patient was taking valsartan plus amlodipine and the other was taking bisoprolol for hypertension. Another subject had suspected aspiration pneumonia related to severe malaise. Before the 3HP regimens, they did not have experiences of the same presentation. Overall, more patients in the 3HP group experienced symptoms of ≥grade 2 ADE (45.8% versus 13.5%; P = 0.002), but symptoms of ≥grade 3 ADE presented at similar rates in the two groups (12.5% versus 8.1%; P = 0.725) (Table 2).
Neither severe hematological ADE nor clinically relevant hepatitis was found in serial laboratory follow-ups. Slight but significant reductions of white blood cell and platelet counts and increases in eosinophil counts, aspartate transaminase, and total bilirubin were found in the 3HP group, while a slight increase in aspartate transaminase was observed in the 9H group (Fig. S1).
Adverse drug events by treatment regimens and their associated factors.
The logistic regression model for hypersensitivity (Table 4) showed that the 3HP regimen (crude OR, 3.40; 95% CI, 1.01 to 11.39; P = 0.048), comorbidity with DM (crude OR 3.32 [1.10 to 10.08]; P = 0.034), and PD (crude OR 4.77 [1.20 to 18.89]; P = 0.026) were associated with hypersensitivity in univariate analysis. In the multivariate model after adjustment for age and gender, underlying 3HP (aOR, 5.21 [1.31 to 20.76]; P = 0.019), DM (aOR, 6.66 [1.78 to 24.90]; P = 0.005), and PD (aOR, 8.59 [1.80 to 41.07]; P = 0.007) were independent risk factors.
TABLE 4.
Logistic regression models for association of adverse drug eventsa
| Outcome and variable | Crude OR | 95% CI | P value | Adjusted OR | 95% CI | P value |
|---|---|---|---|---|---|---|
| Hypersensitivity | ||||||
| 3HP vs 9H | 3.40 | 1.01–11.39 | 0.048 | 5.21 | 1.31–20.76 | 0.019 |
| Age over 60 | 1.05 | 0.37–2.99 | 0.930 | |||
| Male | 0.91 | 0.30–2.76 | 0.872 | |||
| Active smoking | 1.57 | 0.47–5.19 | 0.461 | |||
| Diabetes mellitus | 3.32 | 1.10–10.08 | 0.034 | 6.66 | 1.78–24.90 | 0.005 |
| Peritoneal dialysis | 4.77 | 1.20–18.89 | 0.026 | 8.59 | 1.80–41.07 | 0.007 |
| Eosinophil countb | ||||||
| <349/μl | Reference | |||||
| 350–699/μl | 1.00 | 0.31–3.25 | 0.999 | |||
| ≥700/μl | 0.91 | 0.19–6.49 | 0.907 | |||
| ≥Grade 2 ADE | ||||||
| 3HP (vs 9H) | 5.42 | 1.8–16.27 | 0.003 | 9.77 | 2.55–37.49 | 0.001 |
| Age over 60 | 0.54 | 0.21–1.39 | 0.199 | |||
| Male | 0.90 | 0.34–2.39 | 0.832 | |||
| Active smoking | 2.67 | 0.91–7.78 | 0.073 | |||
| Diabetes mellitus | 3.30 | 1.18–9.21 | 0.023 | 7.73 | 2.06–29.06 | 0.002 |
| Peritoneal dialysis | 3.86 | 0.99–15.06 | 0.052 | 7.21 | 1.45–35.98 | 0.016 |
| Eosinophil countb | ||||||
| <349/μl | Reference | |||||
| 350–699/μl | 0.74 | 0.25–2.20 | 0.584 | |||
| ≥700/μl | 1.20 | 0.25–5.82 | 0.821 | |||
| ≥Grade 3 ADE | ||||||
| 3HP (vs 9H) | 1.62 | 0.38–6.96 | 0.517 | |||
| Age over 60 | 0.62 | 0.14–2.65 | 0.517 | |||
| Male | 0.54 | 0.13–2.21 | 0.393 | |||
| Active smoking | 2.00 | 0.45–8.94 | 0.364 | |||
| Diabetes mellitus | 2.78 | 0.67–11.50 | 0.159 | |||
| Peritoneal dialysis | 2.43 | 0.43–13.75 | 0.316 | |||
| Eosinophil countb | ||||||
| <349/μl | Reference | Reference | ||||
| 350–699/μl | 1.54 | 0.20–11.69 | 0.675 | 1.54 | 0.20–11.69 | 0.675 |
| ≥700/μl | 11.00 | 1.48–83.53 | 0.019 | 11.00 | 1.48–83.53 | 0.019 |
Only variables with a P value of <0.1 in univariate regression are included in the multivariate model.
Laboratory test obtained since 2 weeks after initiating LTBI treatment. Eosinophil count cutoff was based on 2 times the normal upper limit value.
High-grade ADE and their associated factors are also listed in Table 4. 3HP and DM were significantly associated with patients experiencing ≥grade 2 ADE, and PD had borderline significance in the univariate analysis. The multivariate logistic model showed independent associations of 3HP (aOR, 9.77 [2.55 to 37.49]; P = 0.001), DM (aOR, 7.73 [2.06 to 29.06]; P = 0.002), and PD (aOR, 7.21 [1.45 to 35.98]; P = 0.016) with occurrence of ≥grade 2 ADE.
Development of ≥grade 3 ADE was associated with a higher eosinophil count (≥700/μl, 2 times the normal upper limit) at 2 weeks posttreatment (42.9% versus 7.6%; P = 0.033 by Fisher’s exact test [Table S3]). In logistic regression, an eosinophil count of ≥700/μl was associated with ≥grade 3 ADE with a crude OR of 11.00 (1.48 to 83.58; P = 0.019 [Table 4]) compared with an eosinophil count of <350/μl, and the aOR (13.92 [1.52 to 127.41]; P = 0.020) remained statistically significant after adjustment for age and gender. For the time course, we compared the eosinophil counts between patients with and without ≥grade 3 ADE at baseline and 2 weeks and 4 weeks after LTBI treatment commencement (Fig. 2). The eosinophil counts were similar at baseline but higher in those with ≥grade 3 ADE at 2 weeks (534/μl versus 311/μl; P = 0.023) and 4 weeks (526/μl versus 337/μl; P = 0.058 by Mann-Whitney U test).
FIG 2.
Eosinophil count at baseline and at 2 weeks and 4 weeks after commencement of treatment for latent tuberculosis infection in dialysis patients. We divided patients by presence of ≥grade 3 adverse drug events (ADE) and compared eosinophil counts at the indicated time by Mann-Whitney U test. The eosinophil counts are shown as means and standard errors. * and #, P values of 0.023 and 0.058, respectively.
DISCUSSION
The present study compared LTBI treatment regimens in a population receiving maintenance dialysis, a high-priority group for LTBI intervention. Despite the higher rate of ADE in the 3HP group, its completion rate was significantly higher than that of the 9H group. In multivariate analysis, treatment termination was significantly associated with the 9H regimen and ≥grade 2 ADE. Hypersensitivity and ≥grade 2 ADE were more common for 3HP, DM, and PD. Moreover, it was found that a posttreatment eosinophil count of >700/μl could be a predictor for ≥grade 3 ADE.
The completion rate of the 3HP regimen was strikingly higher than that of the 9H regimen in this study, as in previous reports in the general population (12, 16). Another similarity was that termination of 9H was most often due to consent withdrawal or unwillingness (12, 16), possibly due to its daily lengthy daily course. For treatment termination, ≥grade 2 ADE was an independent factor in addition to 9H regimen. This indicates that significant ADE can influence the willingness for LTBI treatment. Due to the subclinical nature of LTBI, any obvious ADE occurrence would be stressful for both subjects and practitioners. Although the 3HP regimen led to a higher incidence of ≥grade 2 ADE in dialysis patients than did the 9H regimen, incidences of ≥grade 3 ADE were not insignificantly different between the 3HP and 9H groups, indicating that dangerous ADE exist in both regimens. Therefore, in treating LTBI in the dialysis population, the 3HP regimen should be suggested first with surveillance for ADE. If 3HP is contraindicated, the 9H regimen should be recommended with adherence monitoring.
In our study, relatively high percentages of dialysis patients developed ≥grade 2 ADE in both the 9H (13.5%) and 3HP (45.8%) groups, compared with 3.8% and 15.2% in the 9H and 3HP groups of TB contacts, respectively, reported by Sun et al. (16). The percentages of ≥grade 3 ADE (8.1% for 9H and 12.5% for 3HP) were higher than those in the PREVENT trial (6.5% for 9H and 5.7% for 3HP) (12). Hypersensitivity syndrome, a sign of systemic drug reaction reported for the 3HP regimen in previous studies (12, 17), was recorded for 29.2% of our 3HP group, compared with 1.9% in the PREVENT trial (12, 17). Although isoniazid and rifapentine are mainly metabolized in the liver, the uremic toxin might decrease cytochrome P450 enzymes and the efficiency of drug transport (18, 19), reducing the metabolism of the LTBI regimen. In addition, rifapentine may have drug-drug interactions and might reduce the effects of some medications for the comorbidities among dialysis patients, such as hypertension (20). Such interactions may explain why two of the 3HP users in this study had hypertensive crisis leading to hospitalization due to reduced serum concentrations of amlodipine and bisoprolol.
Risk factors for developing hypersensitivity and ≥grade 2 ADE in the dialysis population are unknown, and PD and DM were identified as independent risk factors along with the 3HP regimen. Severe drug reaction to the 3HP regimen was reportedly related to concomitant nonstudy medication (17). Polypharmacy was prevalent in our population, which had underlying diseases such as DM, and it might explain the higher rate of high-grade ADE. In addition, allergy is reportedly a common comorbidity of DM (21), possibly due to Th2 axis shift. Furthermore, DM patients might have complicated neuropathy and be susceptible to malaise, weakness, and hypotension. The metabolism of numerous antidiabetic agents might be affected by the 3HP regimen, and rifapentine might compete with these agents for plasma protein binding sites (22). PD, which preserves better immune status than does hemodialysis, might have a higher response to LTBI treatment (23). Acute rejection after renal transplantation was associated with pretransplantation PD (24). These clinical features could be used for ADE prioritization but will require further validation.
Eosinophilia was an independent factor for severe ADE, irrespective of LTBI regimen, possibly due to immunologically mediated drug hypersensitivity caused by both isoniazid and rifapentine (17, 25). Notably, eosinophils significantly increased after LTBI treatment in patient with ≥grade 3 ADE (Fig. S2). The two-time upper normal limit of the eosinophil count might be a useful predictor; it had a crude OR of 11 and an aOR of 8.44 after adjustment for age and gender. If validated, this phenomenon could alert clinicians to patients with high risk of developing severe ADE after 2 weeks of LTBI treatment. Further study is required to investigate the role of eosinophils in the ADE of LTBI treatment in dialysis patients.
Our study had some limitations. First, the choice of LTBI treatment was not randomized. Second, the 3HP group received medications through directly observed preventive therapy (DOPT), and some patients in the 9H group did not. We did not record the details of medication usage during the LTBI treatment course, which could have affected the side effects. In addition, the ADE might have resulted from dialysis, comorbidities, or some events other than the LTBI drug regimen; the causes could not be easily discriminated and may have led to overestimation of ADE occurrence. The sample size was small, so some statistical models might have overfitted or underestimated some factors with borderline significance. Last, this study was conducted in Taiwan, so the results may not be generalizable to other populations.
This study found a higher incidence of ADE during LTBI treatment in patients receiving maintenance dialysis. The 3HP users had a higher completion rate than that of the 9H users, but 3HP, together with DM and PD, was independent for hypersensitivity and ≥grade 2 ADE. An eosinophil count of >700/μl was correlated with ≥grade 3 ADE regardless of the LTBI regimen. In further LTBI interventions in dialysis, 3HP could be first prescribed along with ADE surveillance.
MATERIALS AND METHODS
Patient enrollment.
In this prospective cohort study, we enrolled ESRD patients receiving maintenance dialysis with positive results of LTBI from September 2014 to March 2020. This study was conducted at three medical centers in northern Taiwan, including National Taiwan University Hospital (NTUH) and its Jin-Shan branch, Taipei Veterans General Hospital (TVGH), and Far Eastern Memorial Hospital (FEMH), under approval of the respective institutional review boards (no. 201309056RINC, 201609062RIND, 201709038RINB, 108067-F, and 2017-12-004B). Diagnosis of LTBI was defined as positive interferon gamma release assay (IGRA) using the QuantiFERON-TB Gold In-tube (QFT-GIT) test. Exclusion criteria for the study included pregnancy, known infection with human immunodeficiency virus, and evidence of active tuberculosis infection.
Study design.
No patients had contraindication for 3HP or 9H, and all received explanations of the pros and cons of the two regimens. Each patient could choose his or her preferred regimen, so the present study had no special regimen assignment (for the dosages, please see the supplemental material). All participants receiving 3HP or 9H at FEMH took medication through directly observed preventive therapy (DOPT), while the remaining participants on 9H took medication without DOPT.
Complete blood counts, aminotransferases, and bilirubin levels were followed up at baseline and at 2 weeks, 4 weeks, 8 weeks, and 12 weeks during LTBI treatment. Reasons for discontinuation were recorded in case of termination.
ADE.
During LTBI treatment, the participants, the study nurse, and the physician could report any treatment-related adverse drug events (ADE), defined as additional symptoms not present in previous daily activities or dialysis processes. ADE were graded according to common terminology criteria for adverse events (CTCAE). Hypersensitivity and flu-like syndromes were defined according to a modified previous definition (17). Although the two syndromes overlapped, we still kept the two categories for clinical purposes. Hypersensitivity syndrome was defined as the presentation of either hypotension or at least three of the following symptoms: nausea, vomiting, shortness of breath, weakness, sweats, chills, fatigue, headache, fever, aches, dizziness, and flushing. Flu-like syndrome was defined as having two of the three following symptom clusters: (i) fever or chills, (ii) fatigue, myalgia, or weakness, and (iii) syncope, dizziness, or flushing. Gastrointestinal symptoms included gastrointestinal upset, nausea, vomiting, and diarrhea. Hepatotoxicity was defined as serum alanine transaminase (ALT) at >3-fold the upper limit of the normal range (ULN) with symptoms of nausea, vomiting, abdominal pain, jaundice, or unexplained fatigue or asymptomatic ALT at >5-fold the ULN (26).
Data collection.
Demographics, including age, comorbidity, current use of immunosuppressive medication, and BCG (bacillus Calmette-Guérin) scar were reviewed by the study nurse. Current smokers were defined as those who had smoked >100 cigarettes, with the last time of smoking within 1 month prior to the study (27).
Statistical analysis.
Demographics were compared between the 3HP and 9H groups using Student’s t test or the Mann-Whitney U test for continuous variables and the chi-squared or Fisher’s exact test for categorical variables. Serial treatment laboratory results were compared by paired t test. The Mann-Whitney U test was used to compare mitogen responses. Logistic regression (for high-grade ADE) was used to adjust for variates with a P value of <0.1 in univariate analysis. All statistics were performed in Stata R14 (StataCorp) and GraphPad Prism 8.00 (GraphPad) software.
Supplementary Material
ACKNOWLEDGMENTS
We acknowledge the assistance in case management provided by Xiu-Mei Hsu at the NTUH Jin-shan branch and Yu-Fen Liu at Far Eastern Memorial Hospital. We thank the staff of the Eighth Core Lab of the Department of Medical Research of National Taiwan University Hospital for their technical support. We express our gratitude to the staff of the National Taiwan University Hospital–Statistical Consulting Unit (NTUH-SCU) for statistical consultation and analyses.
Chin-Chung Shu and Ping-Huai Wang were responsible for the conception, design, and execution of the study and for the manuscript writing; Shu-Yung Lin contributed to case enrollment, data analysis, and manuscript writing; Chin-Hao Chang advised on statistical analysis; Jia-Yih Feng, Chih-Yuan Lee, Yi-Chih Lin, Yu-Hsiang Chou, and Kuan-Yin Lin assisted in case enrollment; and Yu-Feng Wei, Jann-Yuan Wang, and Chong-Jen Yu provided consultation on the entire study.
We declare no conflict of interest.
This study was supported by a Ministry of Health and Welfare and research grant from Far Eastern Memorial Hospital (no. FEMH-2020-C-027).
The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Footnotes
Supplemental material is available online only.
REFERENCES
- 1.Cohen A, Mathiasen VD, Schon T, Wejse C. 2019. The global prevalence of latent tuberculosis: a systematic review and meta-analysis. Eur Respir J 54:1900655. doi: 10.1183/13993003.00655-2019. [DOI] [PubMed] [Google Scholar]
- 2.Wejse C. 2015. Tuberculosis elimination in the post Millennium Development Goals era. Int J Infect Dis 32:152–155. doi: 10.1016/j.ijid.2014.11.020. [DOI] [PubMed] [Google Scholar]
- 3.World Health Organization. 2019. Global tuberculosis report 2019. World Health Organization, Geneva, Switzerland. [Google Scholar]
- 4.Getahun H, Matteelli A, Abubakar I, Aziz MA, Baddeley A, Barreira D, Den Boon S, Borroto Gutierrez SM, Bruchfeld J, Burhan E, Cavalcante S, Cedillos R, Chaisson R, Chee CB, Chesire L, Corbett E, Dara M, Denholm J, de Vries G, Falzon D, Ford N, Gale-Rowe M, Gilpin C, Girardi E, Go UY, Govindasamy D, D Grant A, Grzemska M, Harris R, Horsburgh CR, Jr, Ismayilov A, Jaramillo E, Kik S, Kranzer K, Lienhardt C, LoBue P, Lonnroth K, Marks G, Menzies D, Migliori GB, Mosca D, Mukadi YD, Mwinga A, Nelson L, Nishikiori N, Oordt-Speets A, Rangaka MX, Reis A, Rotz L, Sandgren A, et al. 2015. Management of latent Mycobacterium tuberculosis infection: WHO guidelines for low tuberculosis burden countries. Eur Respir J 46:1563–1576. doi: 10.1183/13993003.01245-2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lonnroth K, Migliori GB, Abubakar I, D’Ambrosio L, de Vries G, Diel R, Douglas P, Falzon D, Gaudreau MA, Goletti D, Gonzalez Ochoa ER, LoBue P, Matteelli A, Njoo H, Solovic I, Story A, Tayeb T, van der Werf MJ, Weil D, Zellweger JP, Abdel Aziz M, Al Lawati MR, Aliberti S, Arrazola de Onate W, Barreira D, Bhatia V, Blasi F, Bloom A, Bruchfeld J, Castelli F, Centis R, Chemtob D, Cirillo DM, Colorado A, Dadu A, Dahle UR, De Paoli L, Dias HM, Duarte R, Fattorini L, Gaga M, Getahun H, Glaziou P, Goguadze L, Del Granado M, Haas W, Jarvinen A, Kwon GY, Mosca D, Nahid P, et al. 2015. Towards tuberculosis elimination: an action framework for low-incidence countries. Eur Respir J 45:928–952. doi: 10.1183/09031936.00214014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Chen DY, Shen GH, Hsieh TY, Hsieh CW, Lan JL. 2008. Effectiveness of the combination of a whole-blood interferon-gamma assay and the tuberculin skin test in detecting latent tuberculosis infection in rheumatoid arthritis patients receiving adalimumab therapy. Arthritis Rheum 59:800–806. doi: 10.1002/art.23705. [DOI] [PubMed] [Google Scholar]
- 7.Lundin AP, Adler AJ, Berlyne GM, Friedman EA. 1979. Tuberculosis in patients undergoing maintenance hemodialysis. Am J Med 67:597–602. doi: 10.1016/0002-9343(79)90240-7. [DOI] [PubMed] [Google Scholar]
- 8.Smirnoff M, Patt C, Seckler B, Adler JJ. 1998. Tuberculin and anergy skin testing of patients receiving long-term hemodialysis. Chest 113:25–27. doi: 10.1378/chest.113.1.25. [DOI] [PubMed] [Google Scholar]
- 9.Venkata RKC, Kumar S, Krishna RP, Kumar SB, Padmanabhan S, Kumar S. 2007. Tuberculosis in chronic kidney disease. Clin Nephrol 67:217–220. doi: 10.5414/cnp67217. [DOI] [PubMed] [Google Scholar]
- 10.Fang HC, Lee PT, Chen CL, Wu MJ, Chou KJ, Chung HM. 2004. Tuberculosis in patients with end-stage renal disease. Int J Tuber Lung Dis 8:92–97. [PubMed] [Google Scholar]
- 11.Rose DN. 2000. Benefits of screening for latent Mycobacterium tuberculosis infection. Arch Intern Med 160:1513–1521. doi: 10.1001/archinte.160.10.1513. [DOI] [PubMed] [Google Scholar]
- 12.Sterling TR, Villarino ME, Borisov AS, Shang N, Gordin F, Bliven-Sizemore E, Hackman J, Hamilton CD, Menzies D, Kerrigan A, Weis SE, Weiner M, Wing D, Conde MB, Bozeman L, Horsburgh CR, Jr, Chaisson RE, TB Trials Consortium PREVENT TB Study Team. 2011. Three months of rifapentine and isoniazid for latent tuberculosis infection. N Engl J Med 365:2155–2166. doi: 10.1056/NEJMoa1104875. [DOI] [PubMed] [Google Scholar]
- 13.Jo KW, Kim JS, Kwon HS, Park YE, Kim JY, Hong MJ, Shim TS. 2019. Adverse event and treatment completion rates of a 12-dose weekly isoniazid and rifapentine course for South Korean healthcare workers. Respir Med 158:42–48. doi: 10.1016/j.rmed.2019.10.005. [DOI] [PubMed] [Google Scholar]
- 14.Feng JY, Huang WC, Lin SM, Wang TY, Lee SS, Shu CC, Pan SW, Chen CY, Lin CB, Wei YF, Tung CL, Li CP, Su WJ. 2020. Safety and treatment completion of latent tuberculosis infection treatment in elderly population—a prospective observational study in Taiwan. Int J Infect Dis 96:550–557. doi: 10.1016/j.ijid.2020.05.009. [DOI] [PubMed] [Google Scholar]
- 15.Lin SY, Chiu YW, Lu PL, Hwang SJ, Chen TC, Hsieh MH, Chen YH. 2019. Three months of rifapentine and isoniazid for latent tuberculosis infection in hemodialysis patients: high rates of adverse events. J Microbiol Immunol Infect 52:158–162. doi: 10.1016/j.jmii.2018.05.003. [DOI] [PubMed] [Google Scholar]
- 16.Sun HY, Huang YW, Huang WC, Chang LY, Chan PC, Chuang YC, Ruan SY, Wang JY, Wang JT. 2018. Twelve-dose weekly rifapentine plus isoniazid for latent tuberculosis infection: a multicentre randomised controlled trial in Taiwan. Tuberculosis (Edinb) 111:121–126. doi: 10.1016/j.tube.2018.05.013. [DOI] [PubMed] [Google Scholar]
- 17.Sterling TR, Moro RN, Borisov AS, Phillips E, Shepherd G, Adkinson NF, Weis S, Ho C, Villarino ME, Tuberculosis Trials Consortium. 2015. Flu-like and other systemic drug reactions among persons receiving weekly rifapentine plus isoniazid or daily isoniazid for treatment of latent tuberculosis infection in the PREVENT Tuberculosis Study. Clin Infect Dis 61:527–535. doi: 10.1093/cid/civ323. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Yeung CK, Shen DD, Thummel KE, Himmelfarb J. 2014. Effects of chronic kidney disease and uremia on hepatic drug metabolism and transport. Kidney Int 85:522–528. doi: 10.1038/ki.2013.399. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Evers R, Piquette-Miller M, Polli JW, Russel FGM, Sprowl JA, Tohyama K, Ware JA, de Wildt SN, Xie W, Brouwer KLR, International Transporter C. 2018. Disease-associated changes in drug transporters may impact the pharmacokinetics and/or toxicity of drugs: a white paper from the International Transporter Consortium. Clin Pharmacol Ther 104:900–915. doi: 10.1002/cpt.1115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.de Kanter R, Sidharta PN, Delahaye S, Gnerre C, Segrestaa J, Buchmann S, Kohl C, Treiber A. 2016. Physiologically-based pharmacokinetic modeling of macitentan: prediction of drug-drug interactions. Clin Pharmacokinet 55:369–380. doi: 10.1007/s40262-015-0322-y. [DOI] [PubMed] [Google Scholar]
- 21.Dales R, Chen Y, Lin M, Karsh J. 2005. The association between allergy and diabetes in the Canadian population: implications for the Th1-Th2 hypothesis. Eur J Epidemiol 20:713–717. doi: 10.1007/s10654-005-7920-1. [DOI] [PubMed] [Google Scholar]
- 22.Zheng C, Hu X, Zhao L, Hu M, Gao F. 2017. Clinical and pharmacological hallmarks of rifapentine’s use in diabetes patients with active and latent tuberculosis: do we know enough? Drug Des Devel Ther 11:2957–2968. doi: 10.2147/DDDT.S146506. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Schollmeyer P, Bozkurt F. 1988. The immune status of the uremic patient: hemodialysis vs. CAPD. Clin Nephrol 30(Suppl 1):S37–S40. [PubMed] [Google Scholar]
- 24.Ducloux D, Legendre M, Bamoulid J, Rebibou JM, Saas P, Courivaud C, Crepin T. 2018. ESRD-associated immune phenotype depends on dialysis modality and iron status: clinical implications. Immun Ageing 15:16. doi: 10.1186/s12979-018-0121-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Rubira N, Baltasar MA, Marti E. 1999. Hypersensitivity syndrome from isoniazid. Allergy 54:1011–1012. doi: 10.1034/j.1398-9995.1999.00161.x. [DOI] [PubMed] [Google Scholar]
- 26.Saukkonen JJ, Cohn DL, Jasmer RM, Schenker S, Jereb JA, Nolan CM, Peloquin CA, Gordin FM, Nunes D, Strader DB, Bernardo J, Venkataramanan R, Sterling TR, ATS (American Thoracic Society) Hepatotoxicity of Antituberculosis Therapy Subcommittee. 2006. An official ATS statement: hepatotoxicity of antituberculosis therapy. Am J Respir Crit Care Med 174:935–952. doi: 10.1164/rccm.200510-1666ST. [DOI] [PubMed] [Google Scholar]
- 27.Lin HH, Ezzati M, Chang HY, Murray M. 2009. Association between tobacco smoking and active tuberculosis in Taiwan: prospective cohort study. Am J Respir Crit Care Med 180:475–480. doi: 10.1164/rccm.200904-0549OC. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.