Plasma Concentrations of Isoniazid and Rifampin Are Decreased in Adult Pulmonary Tuberculosis Patients with Diabetes Mellitus

Abstract

Plasma isoniazid and rifampin concentrations, but not pyrazinamide and ethambutol concentrations, were decreased by about 50% (P < 0.05) in diabetic pulmonary tuberculosis patients. The prevalences of subnormal plasma isoniazid, rifampin, pyrazinamide, and ethambutol concentrations were 49% or 100% (P < 0.01), 66% or 100% (P < 0.05), 30% or 50% (P = 0.198), and 32% or 21% (P = 0.742) in nondiabetic or diabetic tuberculosis patients, respectively. These data show that plasma concentrations of isoniazid and rifampin were greatly reduced in diabetic tuberculosis patients.

TEXT

Diabetes mellitus (DM) is among the well-known major risk factors for tuberculosis (TB) (1–5), and the presence of DM affects the outcome of drug treatment negatively, reduces the cure rate, and enhances the risk of relapse and emergence of drug resistance (6–10). A recent study (11) has shown that DM is associated with slow response, and circulating levels of TB drugs (e.g., rifampin) are likely to be below the expected therapeutic ranges among diabetic TB patients. The circulating concentration of rifampin was also reported to decline in some (12) but not all (13, 14) studies in diabetic TB patients. Available data, however, are inconclusive with regard to whether DM affects serum therapeutic levels of anti-TB drugs.

Hence, the present study was designed to determine plasma concentrations of isoniazid, rifampin, pyrazinamide, and ethambutol, the first line anti-TB drugs, in adult Turkish pulmonary tuberculosis patients with DM during the intensive phase of the short-course therapy under conditions of directly observed therapy (DOT).

A total of 70 adult patients with newly diagnosed active pulmonary TB were enrolled in the study. Patients were divided into two groups. The first group consisted of 14 TB patients with type 2 DM who were receiving DM therapy with good glucose control (glycosylated hemoglobin level < 6.5%) at the time of recruitment, and they continued to use their antidiabetes drugs during the study. The second group consisted of 56 nondiabetic TB patients who were screened for DM, and patients with suspicious test results (fasting blood glucose > 100 mg/dl and/or hemoglobin A1c [HbA1c] > 6%) were excluded. Diagnosis of pulmonary TB was based on clinical symptoms, chest radiological examination, sputum microscopy, and culture. Patients with a comorbid disease (except for DM) as well as those using concomitant medications (except for antidiabetics) and having abnormal renal and/or hepatic function were excluded.

Written informed consent was obtained from all patients before the commencement of the study. The study protocol and the contents of the written consent form were approved by the Institutional Ethics Committee (Acibadem University Ethical Committee, Istanbul, Turkey).

All patients were administered orally identical drugs from a national manufacturer (Koçak Farma, Istanbul, Turkey) under conditions of DOT. Patients received daily administration of 300 mg of isoniazid, 600 mg of rifampin, 1,500 mg of pyrazinamide, and 1,000 mg of ethambutol.

Blood samples were collected 2 h after simultaneous drug ingestion on days 14 and 30 after the initiation of treatment regimen. Plasma isoniazid, rifampin, pyrazinamide, and ethambutol concentrations were determined by high-performance liquid chromatography (15, 16).

Data were given as means ± standard errors of the means (SEM). Correlations between age, body weight, body mass index, and plasma concentrations of drugs were determined by the Pearson moment correlation procedure.

Mean age, body weight, body mass index, drug doses per kilogram of body weight, blood glucose, and HbA1c were significantly higher in diabetic TB patients than in the nondiabetic group (Table 1). No difference was observed between any other patient characteristics (Table 1).

Table 1.

Characteristics of patients

Patient characteristic	Values for TB patient group		P
	Nondiabetic (n = 56)	Diabetic (n = 14)
Age (year)	39.0 ± 2.2	56.5 ± 3.5	<0.001
Body wt (kg)	56.7 ± 1.3	68.7 ± 3.4	<0.001
Height (cm)	167.7 ± 1.2	166.2 ± 2.4	0.578
Body mass index (kg/m2)	20.1 ± 0.4	25.1 ± 1.6	<0.001
No. (%) of males	34 (61)	12 (86)	0.116
No. (%) of females	22 (39)	2 (14)	0.313
No. (%) of smokers	29 (52)	11(79)	0.168
Drug dose (mg/kg of body wt)
Isoniazid	5.5 ± 0.1	4.5 ± 0.2	<0.001
Rifampin	11.0 ± 0.3	9.0 ± 0.4	<0.001
Pyrazinamide	27.4 ± 0.6	22.5 ± 1.1	<0.001
Ethambutol	18.3 ± 0.4	15.0 ± 0.7	<0.001
Laboratory data
Total protein (g/liter)	7.4 ± 0.2	7.2 ± 0.2	0.631
Albumin (g/liter)	3.3 ± 0.1	3.1 ± 0.1	0.337
Hemoglobin (g/liter)	12.6 ± 0.2	13.3 ± 0.7	0.189
Aspartate aminotransferase (U/liter)	20.8 ± 1.6	16.7 ± 2.5	0.238
Alanine aminotransferase (U/liter)	18.4 ± 1.5	14.5 ± 2.7	0.240
Creatinine (mg/dl)	0.78 ± 0.02	0.83 ± 0.04	0.267
Blood urea nitrogen (mg/dl)	11.7 ± 0.6	13.7 ± 1. 2	0.141
Fasting blood glucose (mg/dl)	85.8 ± 1.1	95.5 ± 2.1	<0.001
HbA1c (%)	3.8 ± 0.1	6.1 ± 0.1	<0.001
No. (%) sputum smear+	37 (66)	13 (93)/	0.098
No. (%) culture+	48 (86)	14 (100)	0.302
No. (%) with radiological lung lesions
Unilateral	40 (71)	8 (57)	0.344
Bilateral	16 (29)	6 (43)	0.617
Cavity	15 (27)	4 (29)	0.840

Plasma isoniazid and rifampin concentrations in diabetic TB patients were 50% of the values for the nondiabetic TB patients (Table 2). Plasma pyrazinamide and ethambutol concentrations were similar in the two groups (Table 2).

Table 2.

Plasma concentrations of antituberculosis drugs at 14 or 30 days after initiation of the short-course drug treatment regimen

Serum drug	Plasma concn (μg/ml) for TB patient group on indicated day
	Nondiabetic		Diabetic
	14	30	14	30
Isoniazid	3.2 ± 0.2	2.9 ± 0.2	1.5 ± 0.2a	1.2 ± 0.2a
Rifampin	5.1 ± 0.5	5.4 ± 0.5	2.9 ± 0.2a	3.2 ± 0.5a
Pyrazinamide	23.4 ± 2.0	27.9 ± 1.6	25.9 ± 1.4	23.4 ± 1.9
Ethambutol	2.81 ± 0.30	3.6 ± 0.3	3.70 ± 0.64	4.3 ± 0.7

^aValue is significantly (P < 0.05) lower than the values observed in nondiabetic TB patients on the corresponding days.

Negative correlations were found between body weight and plasma concentrations of isoniazid (r = −0.337; P = 0.005) and rifampin (r = −0.302; P = 0.012).

On day 30, plasma isoniazid, rifampin, pyrazinamide, and ethambutol concentrations in 27 (49%) or 14 (100%) (P < 0.01), 37 (66%) or 14 (100%) (P < 0.05), 17 (30%) or 7 (50%) (P = 0.192), and 18 (32%) or 3 (21%) (P = 0.742) nondiabetic or diabetic TB patients were below normal (isoniazid, >3 μg/ml, rifampin, >8 μg/ml, pyrazinamide, >20 μg/ml, and ethambutol, >2 μg/ml) at 2 h (17), respectively.

These data show that DM is associated with a significant decline in plasma isoniazid and rifampin concentrations but not pyrazinamide or ethambutol concentrations. Furthermore, achieved plasma isoniazid and rifampin concentrations at 2 h were below normal in all diabetic TB patients.

A significant decline in the serum rifampin concentration in TB patients with DM was also reported in some (12) but not all (13, 14) previous studies. The decline in plasma isoniazid and rifampin concentrations could have resulted from a decrease in gastrointestinal absorption of the ingested isoniazid and rifampin and/or from an increase in the volume distribution (18) due to the greater body weight of diabetic TB patients. The latter interpretation is supported by the observation that plasma isoniazid and rifampin concentrations correlated negatively with body weight in TB patients, as has been suggested previously (12). The absence of a decline in plasma pyrazinamide and ethambutol concentrations in diabetic TB patients suggests that the impact of DM on plasma pyrazinamide and ethambutol concentrations is different from that on isoniazid and rifampin, as is consistent with a recent report (13).

A recent retrospective study among pulmonary TB patients with slow response to treatment indicates that DM is a major risk factor for having low serum rifampin concentrations at 2 h (11). In accordance with the results of that study, we observed that the prevalences of subnormal isoniazid and rifampin concentrations were 100% or 49% and 100% or 66% in diabetic or nondiabetic TB patients, respectively. Similarly, Nijland et al. (12) reported that the prevalences of low rifampin concentrations are 94% in diabetic and 53% in nondiabetic TB patients.

On the other hand, the present study suffered from a number of limitations. First, the sample size of the diabetic TB patient group was small. Second, plasma concentrations were assessed in blood samples collected 2 h after administration of drugs. A second blood collection at 6 h could have had an additional value in distinguishing patients with delayed absorption. Third, we did not assess the relationships between plasma drug concentrations and clinical outcomes.

In conclusion, our findings show that serum concentrations of isoniazid and rifampin were reduced strongly in pulmonary TB patients with DM. The achieved plasma isoniazid and rifampin concentrations in all diabetic TB patients were below normal concentrations. Isoniazid and rifampin have excellent bactericidal activity and are essential components of first-line TB pharmacotherapy (19, 20). Low plasma concentrations of isoniazid and rifampin may be involved, in part, in known adverse outcomes of the pharmacotherapy for pulmonary tuberculosis in diabetic patients. Optimizing isoniazid and rifampin doses using therapeutic drug monitoring may be necessary in diabetic pulmonary TB patients (11, 17, 21).