[Leukopenia due to anti-tuberculous chemotherapy including rifampicin and isoniazid].

OBJECTIVE

To examine the incidence rate by age and gender of leukopenia caused by chemotherapy including rifampicin (RFP) and isoniazid (INH), and to study the relationships between the leukopenia and the hepatic side effect or other haematological disorders such as thrombocytopenia.

METHODS

Out of the tuberculosis patients who were admitted to our hospital in 1987-88, 1991-92, and 1996-2000, 1,525 patients (1,153 men, 372 women) were chosen for our study who had the white blood cell counts (WBC) in the peripheral blood more than 3,000/mm3 before chemotherapy, and underwent haematologic examination at least twice within 3 months after starting chemotherapy.

METHODS

The definition of leukopenia was as follows: 1) WBC became less than 3,000/mm3 during chemotherapy for patients with pre-treatment WBC more than 4,000/mm3, or 2) WBC decreased more than 1,000/mm3 in patients with pre-treatment WBC between 3,000 and 4,000/mm3. The incidence rates of leukopenia by age, gender, and regimens of chemotherapy were calculated. The case-control study was done between the control and the leukopenia groups excluding patients suffered from agranulocytosis to clarify the hematological and biochemical characteristics of the leukopenia group. The control patients were chosen in the following way. For each patient with leukopenia, a patient with the same admission year, same gender, same regimen of chemotherapy, and the nearest age was chosen as a control patient. The changes in counts of white blood cell, granulocyte, and platelet, in hemoglobin concentration, and in hepatic enzyme levels before and during chemotherapy were compared between the leukopenia and the control groups. Thrombocytopenia was defined as platelet count less than 15 x 10(4)/mm3 and hepatic dysfunction as either asparate aminotransferase (AST) higher than 31 IU/l or alanine aminotransferase (ALT) higher than 34 IU/l.

RESULTS

(1) Incidence rate of leukopenia The leukopenia appeared in 36 patients (14 men, 22 women), two (one man, one woman) of whom showed agranulocytosis. The incidence rate was 1.2% (14/1,153) for men and 5.9% (22/372) for women. The incidence rate of women was higher than that of men in the age groups between 20 to 79 y.o., but no difference was seen in the age groups elder than 80 y.o. There were no differences in the incidence rate among groups treated with HRE (E: ethambutol), HRS (S: streptomycin), and HREZ (Z: pyrazinamide). The chemotherapy was continued in 30 patients after the appearance of leukopenia, and the natural recovery from leukopenia was seen in 19 patients, while the leukopenic state lasted during the chemotherapy in the remaining 11 patients. In two patients who exhibited agranulocytosis all drugs were discontinued. In the remaining 4 patients one or more drugs were discontinued. (2) Case-control study between leukopenia (N = 34) and the control (N = 34) groups There were no differences in age, sputum culture positivity on admission, degree of roentgenographic extent of the disease, ratio of cavity formation, and quantity of daily doses between the two groups. There was also no difference between the days until leukopenia appeared after starting chemotherapy (47.6 +/- 29.5 days) in the leukopenia group, and the days until WBC count became minimum within 3 months after starting chemotherapy (41.7 +/- 21.0 days) in the control group. The negativity of tuberculin skin testing was higher in the leukopenia group [7/14 (50%)] than in the control group [1/10 (10%)], however, the difference was statistically not significant due to rather small size of cases. Before the starting chemotherapy, the counts of WBC (7,230 +/- 1,530 vs 5,500 +/- 1,510/mm3, p < 0.001), neutrophil (5,230 +/- 1,450 vs 4,320 +/- 1,620/mm3, p < 0.05), lymphocyte (1,440 +/- 830 vs 830 +/- 440/mm3, p < 0.001) and platelet (34.9 +/- 12.2 vs 24.1 +/- 6.4 x 10(4)/mm3, p < 0.001) in the peripheral blood and the globulin level (3.71 +/- 0.61 vs 3.35 +/- 0.61 g/dl, p < 0.05) in the serum were significantly higher in the control group than in the leukopenia group. The decrements in the counts of WBC and granulocyte during chemotherapy were larger in the leukopenia group than in the control group (delta WBC: 2,880 +/- 1,530 vs 1,910 +/- 1,520/mm3, and delta Neut: 2,840 +/- 1,510 vs 1,820 +/- 1,380/mm3, p = 0.01, respectively), but the counts of lymphocyte were similar in both groups. The platelet counts also decreased in both groups, but to the mid-normal level in the control group, and to the lowest normal level in the leukopenia group, in which 15 out of 34 patients (44%) showed thrombocytopenia. The levels in the serum of hepatic enzymes such as AST, ALT, and gamma-GTP (gamma-glutamyl aminotransferase) increased during chemotherapy in the leukopenia group, while decreased in the control group, and the facts indicate that in the former not only bone marrow cells but also hepatic cells were impaired by anti-tuberculosis drugs.

CONCLUSIONS

Leukopenia may occur in the course of treatment with anti-tuberculosis drugs, but it is not necessary to stop the chemotherapy immediately, because the WBC count recovers spontaneously or remains under stable leukopenic state during chemotherapy in most cases. But when leukopenia appears, the peripheral blood counts must be checked cautiously, and the chemotherapy should be stopped if the WBC count progressively decreases. The patients who showed leukopenia due to anti-tuberculosis drugs may have had weaker natural and acquired (cell-mediated) immunologic response to tuberculosis infection, and more vulnerable bone marrow cells and hepatic cells to anti-tuberculosis drugs than the control.