å¡ ORIGINAL
ARTICLE
å¡
A Randomized Trial of Reduced Doses of Azidothymidine in Japanese Patients with Human Immunodeficiency Virus Type 1 Infection Satoshi Kimura, Shinichi Oka, Takuo Toyoshima, Yoshihiro Hirabayashi, Yoshimi Kikuchi, Keiji Mitamura and Kaoru Shimada Adverse reactions to the standard dose (1,200 mg/day-1,500 mg/day) of azidothymidine (AZT) are serious. An in vitro pharmacokinetic study of intracellular AZT-5'-triphosphate suggested the feasibility of a clinical trial with reduced doses of AZT. A randomized trial with reduced doses of AZT (group A; 400mg/day, n= 15, group B; 800mg/day, n= 13), was conducted enrolling 28 patients with human immunodeficiency virus infection. The effective rate of AZT on CD4+ lymphocyte counts was similar for both groups, but the duration of the effect of AZT was significantly longer in group A (p < 0.05). In group B, adverse reactions were more frequently observed (p < 0.01), and AZT was withdrawn or the dose was reduced more frequently (p < 0.05). These results suggest that AZT at a dose of 400 mg/day is less toxic, and is (Internal Medicine 871-876,treatment. 1992) more beneficial for31: long-term Key words: azidothymidine-5'-monophosphate, azidothymidine-5'-diphosphate, azidothymidine-5 '-triphosphate , pharmacokinetic study
Introduction form AZT-5'-triphosphate in the cells (1). However, studies on AZT-5'-triphosphate have been very limited Azidothymidine (AZT, 3'-azido-2',3'-dideoxythybecause it has been measured only by using radioisotopes midine) is ahave potent inhibitor of the replication of human (5, 6). We recently developed a new sensitive non immunodeficiency virus typeintracellular 1 (HIV-1) in vitro (1). isotopic method to measure concentrations When AZT at a daily dose of 1,500mg was administered of phosphorylated forms of AZT (7), and found that an to patients with the acquired immunodeficiency synAZT concentration of higher than 2 ^M did not increase drome (AIDS) or AIDS-related complex, it prolonged intracellular concentration AZT-5 '-triphosphate (7). In the present study,the wefrequency firstofexamined the pharmaco survival, decreased of opportunistic infections, and kinetics oftransiently intracellular increased AZT-5'-triphosphate, CD4+ lymphocyte and then counts (2). Although the clinical clinical efficacytrial of AZT is obvious, conducted a randomized in which the safety AZT treatment has been frequently associated with and efficacy of daily doses of 400mg and 800mg of AZT severe adverse reactions including anemia, leukopenia, were compared. study was(3). started in earliest September nausea, myalgia, The and insomnia In the 1988, and the entry to the study was closed in March Japanese trial with daily doses of 600mg-1,200mg, the 1990. The results suggested that a daily dose 400mg doses had to be reduced or discontinued in 9 ofof12 cases is as effective as, lessreactions toxic than, more beneficial for because of adverse (4).and Considering the differlong-term treatment than a daily dose of 800mg. ence in body weight between American and Japanese Subjects and patients, it is Methods conceivable that Japanese patients can be treated with lower doses of AZT than the standard daily Measurement of Intracellular AZT-5'-Triphosphate dose in USA (1,200mg-1,500mg). AZT is thought to exert its anti-HIV effects when it is and Other Metabolites'. For in vitrophosphorylated analysis, MT-4 to cells,
From the Department of Infectious Diseases and Applied Immunology, the Institute of Medical Science, University of Tokyo, Tokyo Received for publication October 17, 1991; Accepted for publication April 9, 1992 Reprint requests should be addressed to Dr. Satoshi Kimura, the Department of Infectious Diseases and Applied Immunology, the Institute o Medical Science, The University of Tokyo, 4-6- 1 Shirokanedai, Minato-ku, Tokyo 108, Japan Internal Medicine Vol. 31, No. 7 (July 1992) 871
Kimura et al a cloned cell line established from ATL cells (8), and Molt-4 cells (9), a T cell line, were incubated under constant stirring at 37°C in RPMI 1640 medium contain ing 10% fetal calf serum. The concentration of AZT in the medium was continuously changed, using computer controlled pumps, so that the concentrations of AZT mimicked those in the sera of subjects who were orally administered of 200mg of AZT. At the designated time points, the cells were harvested, washed and extracted with 60% acetonitrile. The extracts were treated as described previously (7), dried, and finally dissolved The amounts of AZT-5'-monophosphate, AZT-5' with distilled water as equivalent to 107 cells/50/i. diphosphate and AZT-5'-triphosphate in the extracts were measured using HPLC under the conditions de veloped recently in our laboratory (7). Patient Population: Japanese patients with HIV-1 infection at least 15 years of age were eligible to enter the study if they met the following criteria; a hemoglobin value above 9.0 g/dl, a granulocyte count above 1,000/^1, a platelet count above 7 x 104//i, serum creatinine con centration lower than 2.0 mg/dl; no serious liver dysfunc tion, and no active secondary infection. HIV-1 infection was documented by a licensed particle-agglutination method (Fujirebio Inc., Tokyo) and confirmed by Informed consent for participation in the Lab., studyTokyo). was Western blot analysis (Nippon Bio-Rad obtained. Between September 1988, and March 1990, 31 patients were enrolled in the study, and randomly assigned to group A or group B. Three patients were, however, lost to follow-up shortly after the enrollment and were excluded from evaluation. Therefore, the
data analysis was possible for 28 patients (group A; 15 patients, group B; 13 patients). They consisted of 22 male hemophiliacs, three homosexual men and three heterosexually-transmitted men. The background of these patients is summarized in Table 1. No significant differences were noted between the two groups in their pretreatment Study Design:characteristics. The patients assigned to group A re ceived AZT at a daily dose of 400mg orally in four divided doses, and those assigned to group B received AZT at a daily dose of 800mg orally in four divided doses. The patients were requested to continue receiving their assigned treatments for at least 12 months. How ever, the dosages were reduced by half or discontinued if moderate to serious adverse reactions appeared. The patients were seen at least every three months, and clinical and laboratory findings were checked. Routine laboratory testing included complete blood cell counts with differentials, counting of lymphocyte subsets (CD4+ , CD8+) and evaluation of renal and liver functions. Serum HIV-1 p24 antigen levels were measured byLymphocyte ELISA Stratification of the Effect ofAZT on CD4+ (Abbott Lab., Abbott Park, IL, USA). Counts' . The effects of AZT on CD4+ lymphocyte counts
were arbitrarily classified into three grades after modi fication and combination of criteria previously described (10, ll); grade 1, an increase in CD4+ lymphocyte count by 10% or more of base-line count with minimal increase by lOcells/^1, grade 2, an increase by 30% or more with minimal increase by 30cells//il, grade 3, an increase by Criteria Adverse Reactions on Hemoglobin Concen 50% or for more with minimal increase by SOcells/^l. trations and Neutrophil Counts: A decrease in hemoglobin concentration of greater than 2g/dl, and a decrease in Table 1. Characteristics of Patients at Entry According to Dose of neutrophil count of greater than 30% of the base-line AZT Received No. of patients* value were taken as adverse reactions of AZT, if they Statistical were observed Analysis: after For the start the analysis of treatment. of CD4+ lym Group A Group B phocyte counts and the CD4+/CD8+ ratios, both 15 13 Total Wilcoxon's rank-sum test and Student's t test were used. For other comparisons between the two groups, both Age Fisher's exact test and Chi-squared test (after Yates' 15years - 24years 463 correction) were used. Greater p values were taken for 25years - 34years 35 years or more the interpretation of the data. A p value of less than 0.05 Risk factors for HIV-1 transmission was considered significant. Hemophilia 10 1 Heterosexual men Homosexual men CDC classification Group II Group III Group IV-C CD4+ lymphocyte count 99/^1 or less 100/^1 - 199/^1 200/^1 - 499//il
Distribution 872
was not statistically
2
different.
Results Pharmacokinetics of Intracellular AZ T-5 '' -triphosphate Figure 1A shows the average serum concentration of AZT when 200mg of AZT was ingested at time zero. When the concentration of AZT in the culture medium was changed exactly as shown in Fig. 1A, the concentra tion of AZT-5'-monophosphate in MT-4 cells followed the change of the concentration of AZT (Fig. IB). On the other hand, intracellular concentrations of AZT-5'Internal
Medicine
Vol. 31, No. 7 (July
1992)
Reduced Doses of AZT for HIV-1 Infection Among patients who could not complete the study with the full assigned doses, a total of three patients died; two patients with AIDS in group A and one also with AIDS in group B. One of the two patients in group A was a hemophiliac, and had a history of cytomegalovirus (CMV) infection, and suddenly died of cerebellar bleeding four months after the reduction of AZT from 400mg/day to 200mg/day, although he was doing quite well until the bleeding. Thus, his death was not due to progression of HIV-1 infection. The other patient in group A died of malignant lymphoma 10 months after withdrawal of AZT treatment. One case in group B had a history of CMV retinitis, and died of systemic CMV infection nine months after withdrawal of AZT treat ment in another hospital. Therefore, in the latter two cases, their death may not be due to the failure of AZT, but occurred several months after withdrawal of AZT. Changes in CD4+ Lymphocyte Counts and CD4+ICD8+ Effects of two different doses of AZT on immu nological markers, CD4+ lymphocyte counts and the CD4+/CD8+ ratios, were evaluated only for data ob tained during the period in which the full assigned doses of AZT were administered in order to avoid the influence of dose modification. The immunological effects of different doses of AZT were first compared by assessing the effects on individual Fig. 1. Pharmacokinetics of AZT and intracellular AZT-5'-mono patients. The effects on CD4+ lymphocyte counts were phosphate, AZT-5'-diphosphate and AZT-5'-triphosphate. MT-4 cells stratified into three grades as defined in "Subjects and were cultured in RPMI 1640 containing 10% fetal calf serum. Concen As shown in Table 2, both doses of AZT tration of AZT in the culture medium was simulated to match thatMethods". in were similarly effective, and there was no difference in the serum attained when 200mg of AZT was ingested (A). At design ated times, cells were harvested and AZT-5'-monophosphate (B),In theorder ratestoofassess effectiveness between the two groups. the effects of different doses of AZT-MP: AZT-5 '-diphosphate AZT-5'-monophosphate, and AZT-5'-triphosphate AZT-DP: AZT-5'-diphosphate, (C) were measured. AZT on CD4+ lymphocyte counts as a mass, the rates AZT-TP: AZT-5'-triphosphate, UD: undetectable. of changes in CD4+ lymphocyte counts over the basal diphosphate and AZT-5'-triphosphate showed plateau until 4 hours and 6 hours, respectively, after the simulated ingestion ofAZT (Fig. 1C). Similar results were obtained with Molt-4 cells, indicating that the intracellular con centration of AZT-5'-triphosphate may well be main tained at least for 6 hours. Clinical Effects of AZT In group A, seven patients could continue to take the full assigned dose (400mg/day) of AZT as long as 12 months, while only three could do so in group B. Among those who could take the full assigned doses at least for 12 months, none showed changes in clinical stages except for an asymptomatic carrier in group B, who developed persistent generalized lymphadenopathy 1 1 months after the start of AZT treatment. Neither new opportunistic infection nor recurrence of opportunistic infection was Weight gain of 2kg or more wasduring observed and observed among these patients the in 12three months. two cases of group A and group B, respectively. Internal
Medicine
Vol. 31, No. 7 (July
1992)
count in each patient at each time point were calculated, and the course of the rates over time was compared (Fig. 2). This parameter was used instead of the absolute number of CD4+ lymphocyte count in order to minimize the influence of withdrawal from the study of patients with relatively low or high CD4+ lymphocyte counts. As shown in Fig. 2, the rates of changes in CD4+ lymphocyte o .CD4+ of p atien ts (% ) Table 2. Comparison of AZT Effects Non Lymphocyte Counts
E ffect o n C D 4+ lym p ho cyte cou nt G rad e 1 G rade 2 G rad e 3
T otaln o .o f evalu ated p atien ts*
G ro up A
G ro up B
7 (50) 6 (43) 3 (21)
5 (56) 4 (44) 2 (22)
14 (100 )
'(100)
* Patients who were treated for more than three months. Difference between the two groups was not significant. 873
Kimura et al Table 4. Adverse Reactions which Led to Dose Reduction or Withdrawal of AZT No. of patients (%) with
adverse reactions
GroupA
12
Anorexia/Nausea Anemia Leukopenia CNS-related Total
GroupB
(69) (31) (23) (23)
3 (20) 4 (27) 2 (13) 1(7) 7 (47)*
re ductio n/withdrawal
13 (100)
GroupA
3 (20) 2 (13) 1(7) 0 6 (40)*
GroupB
8 (62) 2 (15) 2 (15) 3 (23) 12 (92)
Fig. 2. Rates of changes of CD4+ lymphocyte counts. Ratios of * p
ARTICLE
å¡
A Randomized Trial of Reduced Doses of Azidothymidine in Japanese Patients with Human Immunodeficiency Virus Type 1 Infection Satoshi Kimura, Shinichi Oka, Takuo Toyoshima, Yoshihiro Hirabayashi, Yoshimi Kikuchi, Keiji Mitamura and Kaoru Shimada Adverse reactions to the standard dose (1,200 mg/day-1,500 mg/day) of azidothymidine (AZT) are serious. An in vitro pharmacokinetic study of intracellular AZT-5'-triphosphate suggested the feasibility of a clinical trial with reduced doses of AZT. A randomized trial with reduced doses of AZT (group A; 400mg/day, n= 15, group B; 800mg/day, n= 13), was conducted enrolling 28 patients with human immunodeficiency virus infection. The effective rate of AZT on CD4+ lymphocyte counts was similar for both groups, but the duration of the effect of AZT was significantly longer in group A (p < 0.05). In group B, adverse reactions were more frequently observed (p < 0.01), and AZT was withdrawn or the dose was reduced more frequently (p < 0.05). These results suggest that AZT at a dose of 400 mg/day is less toxic, and is (Internal Medicine 871-876,treatment. 1992) more beneficial for31: long-term Key words: azidothymidine-5'-monophosphate, azidothymidine-5'-diphosphate, azidothymidine-5 '-triphosphate , pharmacokinetic study
Introduction form AZT-5'-triphosphate in the cells (1). However, studies on AZT-5'-triphosphate have been very limited Azidothymidine (AZT, 3'-azido-2',3'-dideoxythybecause it has been measured only by using radioisotopes midine) is ahave potent inhibitor of the replication of human (5, 6). We recently developed a new sensitive non immunodeficiency virus typeintracellular 1 (HIV-1) in vitro (1). isotopic method to measure concentrations When AZT at a daily dose of 1,500mg was administered of phosphorylated forms of AZT (7), and found that an to patients with the acquired immunodeficiency synAZT concentration of higher than 2 ^M did not increase drome (AIDS) or AIDS-related complex, it prolonged intracellular concentration AZT-5 '-triphosphate (7). In the present study,the wefrequency firstofexamined the pharmaco survival, decreased of opportunistic infections, and kinetics oftransiently intracellular increased AZT-5'-triphosphate, CD4+ lymphocyte and then counts (2). Although the clinical clinical efficacytrial of AZT is obvious, conducted a randomized in which the safety AZT treatment has been frequently associated with and efficacy of daily doses of 400mg and 800mg of AZT severe adverse reactions including anemia, leukopenia, were compared. study was(3). started in earliest September nausea, myalgia, The and insomnia In the 1988, and the entry to the study was closed in March Japanese trial with daily doses of 600mg-1,200mg, the 1990. The results suggested that a daily dose 400mg doses had to be reduced or discontinued in 9 ofof12 cases is as effective as, lessreactions toxic than, more beneficial for because of adverse (4).and Considering the differlong-term treatment than a daily dose of 800mg. ence in body weight between American and Japanese Subjects and patients, it is Methods conceivable that Japanese patients can be treated with lower doses of AZT than the standard daily Measurement of Intracellular AZT-5'-Triphosphate dose in USA (1,200mg-1,500mg). AZT is thought to exert its anti-HIV effects when it is and Other Metabolites'. For in vitrophosphorylated analysis, MT-4 to cells,
From the Department of Infectious Diseases and Applied Immunology, the Institute of Medical Science, University of Tokyo, Tokyo Received for publication October 17, 1991; Accepted for publication April 9, 1992 Reprint requests should be addressed to Dr. Satoshi Kimura, the Department of Infectious Diseases and Applied Immunology, the Institute o Medical Science, The University of Tokyo, 4-6- 1 Shirokanedai, Minato-ku, Tokyo 108, Japan Internal Medicine Vol. 31, No. 7 (July 1992) 871
Kimura et al a cloned cell line established from ATL cells (8), and Molt-4 cells (9), a T cell line, were incubated under constant stirring at 37°C in RPMI 1640 medium contain ing 10% fetal calf serum. The concentration of AZT in the medium was continuously changed, using computer controlled pumps, so that the concentrations of AZT mimicked those in the sera of subjects who were orally administered of 200mg of AZT. At the designated time points, the cells were harvested, washed and extracted with 60% acetonitrile. The extracts were treated as described previously (7), dried, and finally dissolved The amounts of AZT-5'-monophosphate, AZT-5' with distilled water as equivalent to 107 cells/50/i. diphosphate and AZT-5'-triphosphate in the extracts were measured using HPLC under the conditions de veloped recently in our laboratory (7). Patient Population: Japanese patients with HIV-1 infection at least 15 years of age were eligible to enter the study if they met the following criteria; a hemoglobin value above 9.0 g/dl, a granulocyte count above 1,000/^1, a platelet count above 7 x 104//i, serum creatinine con centration lower than 2.0 mg/dl; no serious liver dysfunc tion, and no active secondary infection. HIV-1 infection was documented by a licensed particle-agglutination method (Fujirebio Inc., Tokyo) and confirmed by Informed consent for participation in the Lab., studyTokyo). was Western blot analysis (Nippon Bio-Rad obtained. Between September 1988, and March 1990, 31 patients were enrolled in the study, and randomly assigned to group A or group B. Three patients were, however, lost to follow-up shortly after the enrollment and were excluded from evaluation. Therefore, the
data analysis was possible for 28 patients (group A; 15 patients, group B; 13 patients). They consisted of 22 male hemophiliacs, three homosexual men and three heterosexually-transmitted men. The background of these patients is summarized in Table 1. No significant differences were noted between the two groups in their pretreatment Study Design:characteristics. The patients assigned to group A re ceived AZT at a daily dose of 400mg orally in four divided doses, and those assigned to group B received AZT at a daily dose of 800mg orally in four divided doses. The patients were requested to continue receiving their assigned treatments for at least 12 months. How ever, the dosages were reduced by half or discontinued if moderate to serious adverse reactions appeared. The patients were seen at least every three months, and clinical and laboratory findings were checked. Routine laboratory testing included complete blood cell counts with differentials, counting of lymphocyte subsets (CD4+ , CD8+) and evaluation of renal and liver functions. Serum HIV-1 p24 antigen levels were measured byLymphocyte ELISA Stratification of the Effect ofAZT on CD4+ (Abbott Lab., Abbott Park, IL, USA). Counts' . The effects of AZT on CD4+ lymphocyte counts
were arbitrarily classified into three grades after modi fication and combination of criteria previously described (10, ll); grade 1, an increase in CD4+ lymphocyte count by 10% or more of base-line count with minimal increase by lOcells/^1, grade 2, an increase by 30% or more with minimal increase by 30cells//il, grade 3, an increase by Criteria Adverse Reactions on Hemoglobin Concen 50% or for more with minimal increase by SOcells/^l. trations and Neutrophil Counts: A decrease in hemoglobin concentration of greater than 2g/dl, and a decrease in Table 1. Characteristics of Patients at Entry According to Dose of neutrophil count of greater than 30% of the base-line AZT Received No. of patients* value were taken as adverse reactions of AZT, if they Statistical were observed Analysis: after For the start the analysis of treatment. of CD4+ lym Group A Group B phocyte counts and the CD4+/CD8+ ratios, both 15 13 Total Wilcoxon's rank-sum test and Student's t test were used. For other comparisons between the two groups, both Age Fisher's exact test and Chi-squared test (after Yates' 15years - 24years 463 correction) were used. Greater p values were taken for 25years - 34years 35 years or more the interpretation of the data. A p value of less than 0.05 Risk factors for HIV-1 transmission was considered significant. Hemophilia 10 1 Heterosexual men Homosexual men CDC classification Group II Group III Group IV-C CD4+ lymphocyte count 99/^1 or less 100/^1 - 199/^1 200/^1 - 499//il
Distribution 872
was not statistically
2
different.
Results Pharmacokinetics of Intracellular AZ T-5 '' -triphosphate Figure 1A shows the average serum concentration of AZT when 200mg of AZT was ingested at time zero. When the concentration of AZT in the culture medium was changed exactly as shown in Fig. 1A, the concentra tion of AZT-5'-monophosphate in MT-4 cells followed the change of the concentration of AZT (Fig. IB). On the other hand, intracellular concentrations of AZT-5'Internal
Medicine
Vol. 31, No. 7 (July
1992)
Reduced Doses of AZT for HIV-1 Infection Among patients who could not complete the study with the full assigned doses, a total of three patients died; two patients with AIDS in group A and one also with AIDS in group B. One of the two patients in group A was a hemophiliac, and had a history of cytomegalovirus (CMV) infection, and suddenly died of cerebellar bleeding four months after the reduction of AZT from 400mg/day to 200mg/day, although he was doing quite well until the bleeding. Thus, his death was not due to progression of HIV-1 infection. The other patient in group A died of malignant lymphoma 10 months after withdrawal of AZT treatment. One case in group B had a history of CMV retinitis, and died of systemic CMV infection nine months after withdrawal of AZT treat ment in another hospital. Therefore, in the latter two cases, their death may not be due to the failure of AZT, but occurred several months after withdrawal of AZT. Changes in CD4+ Lymphocyte Counts and CD4+ICD8+ Effects of two different doses of AZT on immu nological markers, CD4+ lymphocyte counts and the CD4+/CD8+ ratios, were evaluated only for data ob tained during the period in which the full assigned doses of AZT were administered in order to avoid the influence of dose modification. The immunological effects of different doses of AZT were first compared by assessing the effects on individual Fig. 1. Pharmacokinetics of AZT and intracellular AZT-5'-mono patients. The effects on CD4+ lymphocyte counts were phosphate, AZT-5'-diphosphate and AZT-5'-triphosphate. MT-4 cells stratified into three grades as defined in "Subjects and were cultured in RPMI 1640 containing 10% fetal calf serum. Concen As shown in Table 2, both doses of AZT tration of AZT in the culture medium was simulated to match thatMethods". in were similarly effective, and there was no difference in the serum attained when 200mg of AZT was ingested (A). At design ated times, cells were harvested and AZT-5'-monophosphate (B),In theorder ratestoofassess effectiveness between the two groups. the effects of different doses of AZT-MP: AZT-5 '-diphosphate AZT-5'-monophosphate, and AZT-5'-triphosphate AZT-DP: AZT-5'-diphosphate, (C) were measured. AZT on CD4+ lymphocyte counts as a mass, the rates AZT-TP: AZT-5'-triphosphate, UD: undetectable. of changes in CD4+ lymphocyte counts over the basal diphosphate and AZT-5'-triphosphate showed plateau until 4 hours and 6 hours, respectively, after the simulated ingestion ofAZT (Fig. 1C). Similar results were obtained with Molt-4 cells, indicating that the intracellular con centration of AZT-5'-triphosphate may well be main tained at least for 6 hours. Clinical Effects of AZT In group A, seven patients could continue to take the full assigned dose (400mg/day) of AZT as long as 12 months, while only three could do so in group B. Among those who could take the full assigned doses at least for 12 months, none showed changes in clinical stages except for an asymptomatic carrier in group B, who developed persistent generalized lymphadenopathy 1 1 months after the start of AZT treatment. Neither new opportunistic infection nor recurrence of opportunistic infection was Weight gain of 2kg or more wasduring observed and observed among these patients the in 12three months. two cases of group A and group B, respectively. Internal
Medicine
Vol. 31, No. 7 (July
1992)
count in each patient at each time point were calculated, and the course of the rates over time was compared (Fig. 2). This parameter was used instead of the absolute number of CD4+ lymphocyte count in order to minimize the influence of withdrawal from the study of patients with relatively low or high CD4+ lymphocyte counts. As shown in Fig. 2, the rates of changes in CD4+ lymphocyte o .CD4+ of p atien ts (% ) Table 2. Comparison of AZT Effects Non Lymphocyte Counts
E ffect o n C D 4+ lym p ho cyte cou nt G rad e 1 G rade 2 G rad e 3
T otaln o .o f evalu ated p atien ts*
G ro up A
G ro up B
7 (50) 6 (43) 3 (21)
5 (56) 4 (44) 2 (22)
14 (100 )
'(100)
* Patients who were treated for more than three months. Difference between the two groups was not significant. 873
Kimura et al Table 4. Adverse Reactions which Led to Dose Reduction or Withdrawal of AZT No. of patients (%) with
adverse reactions
GroupA
12
Anorexia/Nausea Anemia Leukopenia CNS-related Total
GroupB
(69) (31) (23) (23)
3 (20) 4 (27) 2 (13) 1(7) 7 (47)*
re ductio n/withdrawal
13 (100)
GroupA
3 (20) 2 (13) 1(7) 0 6 (40)*
GroupB
8 (62) 2 (15) 2 (15) 3 (23) 12 (92)
Fig. 2. Rates of changes of CD4+ lymphocyte counts. Ratios of * p
