Stability of Diltiazem and Its Metabolites in Human Blood Samples JEAN-LOUIS BONNEFOUS**, ROSELYNEBOULIEU**~, AND CHRISTIANE L A
H ~
Received June 18, 1990, from the *Institut des Sciences Pharmaceutiques et Biologiques, Laboratoire de Pharmacie Clinique, 8 avenue Rockefeller, 69373 L on Cedex 08, France, the *H6pitaI Cardiovasculaire et Pneumol ique, Service Pharmaceutique-BP Lyon Montchat, 69394 Lyon Cedex d3, France, and the Blnstitut des Sciences Pharmaceutrques et Bi?og,ques, Laboratoire de Mathbmatiques et Informatique, 8 avenue Rockefeller, 69373 Lyon Cedex 08, France. Accepted for publication June 1, 1991. Abstract OThe stability of diltiazem and its metabolites in blood samples from patients under chronic diltiazem therapy was investigated. When whole blood was kept for 1 h at room temperature between sampling and centrifugation, the concentration of Ndemethyldiltiazem (MA) decreased significantly, with an average loss of 24%. Under the same conditions, an average loss of 14% of diltiazem occurred, whereas the concentrations of the metabolites deacetyldiltiazem and Ndemethyldeacetyldiltiazem did not change significantly.No significant decrease in MA and diltiazem concentrationswas observed when whole blood was stored for 1 h in an ice bath. In spiked plasma samples kept at room temperature, only MA was unstable, with an average loss of 13% after 4 h. The present study shows the importance of observing rigorous conditions for the transport and treatment of blood samples. To achieve accurate determination of diltiazem and related compounds, the blood must be centrifuged immediately after collection or kept on Ice for up to 1 h. The plasma samples must be immediatelyfrozen at -80 "C and can be stored for up to 5 weeks before analysis. Using these rigorous conditions, we observed that MA is the main metabolite of diltiazem in plasma from patients under chronic oral diltiazem therapy.
Diltiazem is a calcium channel blocker that is widely used in the treatment of variant angina, hypertension, and supraventricular tachyanythmias.1.2 Diltiazem is extensively metabolized by deacetylation, N-demethylation, O-demethylation, and conjugation.3 Several authors1,4.6 have reported that deacetyldiltiazem (MUwas the main metabolite found in plasma after oral administration of diltiazem, whereas others6.7 have recently produced some evidence suggesting that Ndemethyldiltiazem (MA)may be the predominant metabolite in plasma. In a previous study involving HPLC analysis of diltiazem and its metabolites,s we found that 50% of the parent drug was converted to MA whereas only 12% of the parent drug was converted to M1 in plasma samples from patients under chronic diltiazem therapy. Furthermore, we observed a significant decrease in MA levels in blood samples lef€ a t room temperature before analysis. Significant decreases in MA concentrations in plasma samples kept for 4 days at 4 "C have recently been observed by Dub6 and co-workers9J0 in stability studies of diltiazem and its metabolites in spiked plasma. In the present study, we did a more detailed investigation of the stability of diltiazem and its metabolites in blood and plasma samples. This study allowed us to establish the conditions for sample collection, treatment, and storage necessary for the accurate determination of diltiazem and its metabolites in plasma.
Experimental Section Reagents-Diltiazem, M1, MA, N-demethyldeacetyldiltiazem ( M 2 ) , and propionyldeacetyldiltiazem (used as internal standard) were generously supplied by the Clinical Research Department, Laboratoires #Etudes et de Recherches Synthblabs, Paris, France. The drug supplies were evaluated for identity and purity by UV spectroecopy. CXZ2-3549/92/04CW-0341$02.50/0 0 1992, American Pharmaceutical Association
Acetonitrile (Uvasol) and ammonium dihydrogen phosphate and orthophosphoric acid (both of "Suprapur" grade) were purchased from Merck (Nogent-sur-Marne, France). Triethylamine (analytical grade) was obtained from Fluka (Buchs, Switzerland). Chromatographic Instrumentation and C o n d i t i o l g T h e apparatus consisted of a model 510 HPLC pump equipped with a model 481 variable-wavelength absorbance detector and a model 712 WISP sample processor (all from Waters, Saint Quentin-Yvelines, France). Chromatographic separation was achieved with a Nucleosil C18 column (15 cm x 4.6-mm inner diameter), with a 3-pm particle size (Interchim, Montlucon, France). Aprecolumn (1.5 cm x 4.6-mm inner diameter) packed with Nucleosil C18 ( 5 - l ~ particle n size) was used as a guard column. The mobile phase consisted of acetonitri1e:O.l M ammonium dihydrogen phosphate buffer (37:63, v/v), containing 0.08% triethylamine. The pH of the final solution was adjusted to 5.9 with orthophosphoric acid. The flow rate was 1 mumin, and the detection was performed at 237 nm. Sample Preparation-Before chromatographic analysis, diltiazem and its metabolites were extracted from plasma by a solid-phase extraction procedure that we have described previously.11The main steps of the extraction procedure are the following. The solid-phase extraction columns (100 mg of sorbent, Syva-Biombrieux) were activated with 2 x 3 mL of acetonitrile followed by 2 x 3 mL of 0.1 M ammonium dihydmgen phosphate buffer. Plasma samples (1 mL) diluted with 1 mL of acetonitri1e:O.l M ammonium dihydrogen phosphate (37:63, v/v) were then passed through the columns. The columns were washed with 2 x 1mL of acetonitri1e:water (20:80, v/v) followed by 2 x 500 pL of acetonitri1e:water (40:60, v/v). The compounds of interest were then eluted with 500 pL of acetonitri1e:O.l M ammonium dihydrogen phosphate (80:20, v/v) containing 0.06% triethylamine (pH of the final solution, 6.8). The eluates were evaporated to dryness under nitrogen at 40-45 "C. The residue was reconstituted with 100 pL of mobile phase, and an aliquot of 60 pL was injected onto the HPLC column. Procedure €or Stability S t u d i e d t a b i l i t y in Blood Samples from Patienh-The study was carried out with blood samples from eight patients under chronic oral diltiazem therapy (60 mg, twice daily). Blood samples were drawn just prior to the morning oral dose of 60 mg of diltiazem (trough level). Blood (2 x 10 mL) was collected by venipuncture in a heparinized tube. One blood sample was immediately placed on ice, and the other was kept at room temperature (20 "C). For each sample, 4-mL aliquots were withdrawn immediately (to)and 1h (t,) after blood collection. At these times, blood samples were centrifuged without delay at low temperature (4 "0. Plasma was decanted and frozen at -80 "C, and then analyzed within 3 weeks of blood collection. Stability in Spiked Plasma Samples-The stability of diltiazem and its metabolites in plasma stored at mom temperature or in an ice bath was investigated. Diltiazem and its metabolites, MA, M1, and M2, were added to pooled human bank plasma at a concentration of 100 ng/mL. Separate 8-mL aliquota of plasma were stored at mom temperature and in an ice bath. At the following intervals, 2 mL of plasma was withdrawn for analysis: 0,1, and 4 h. Simultaneously, the effect of storage a t -20 and at -80 "C on the concentration of the compounds of interest was assessed. Four spiked plasma samples were divided into 2-mL aliquots and stored either at -20 or at -80 "C and analyzed within 1, 2, 3, and 5 weeks. This study was replicated four times. Statistical A n a l y s i g T h e nonparametric Mann-Whitney test was used for statistical comparisons at a level of significance of p < 0.05. Journal of Pharmaceutical Sciences / 341 VOl. 81, No. 4, April 1992
Statistical analysis of the study of stability in patient blood samples was performed with the t test for paired data and the nonparametric Wilcoxon paired rank test, with p < 0.05 considered significant (p value was determined with a one-tailed test).
A
Results and Discussion The results of the assay on the stability of the compounds in blood samples are given in Table I, and the typical changes in the concentrations of the compounds are illustrated in Figure 1.When a blood sample from a treated patient was left for 1 h a t room temperature between sampling and centrifugation, the MA concentration decreased significantly from 37.92 11.1 to 30.02 11.7ng/mL (p 5 0.001 for ttest, p 5 0.004 for Wilcoxon T test), for an average loss of 24% of the original concentration. Under the same conditions, a significant decrease in diltiazem concentration from 62.9 2 18.0 to 53.9 f 16.4ng/mL was observed (p I 0.001for t test and p 5 0.004 for Wilcoxon T test), for an average loss of 14.3% of the original concentration. The concentrations of the other measured metabolites, M1 and M2,did not change significantly. When whole blood was stored for 1 h in an ice bath before centrifugation, no significant decrease in MA and diltiazem concentrations was observed. Under these conditions, an average of 98% of the original concentration of the drug and 99% of the original concentration of the mqjor metabolite (MA)were found. Therefore, significant variations in the concentrations of the compounds of interest, especially diltiazem and MA, occurred in whole blood left for only 1 h a t room temperature between sampling and centrifugation. Slight signs of degradation of diltiazem and MA in patient blood samples kept for 4 h at room temperature have been observed by Caill6 et al.10; 89 f 7.5and 88 f 8% of the initial concentration of diltiazem and MA, respectively, were found. However, the compounds were stable in whole blood left for 2 h at room temperature. To explain the mechanism of degradation of diltiazem and MA, Caill6 et a1.10 suggested a hydrolysis of diltiazem and MA to M1 and M2,respectively, in accordance with the metabolic pathway of diltiazem in humans proposed by Sugiharas (Scheme I). However, in our study, the significant decrease in MA levels wae not related to a significant increase in M 2 levels. No degradation of diltiazem, its metabolites, and the internal standard was observed during the analytical procedures (extraction and HPLC analysis). Hence, our results suggest that the degradation of diltiazem and MA cannot be explained only by metabolic transformations of diltiazem and MA to M1 and M2,respectively, as suggested by Caill6 et a1.10 Other mechanisms involving metabolic or chemical hydrolysis may be occurring. As shown in Table 11, only MA was unstable, with an average loss of 13% in plasma samples kept for 4 h at room
,
I
0
2
.
#
I
4 6 8 minutes
I
I
1 0 1 2
0
2
4
6 8 1 minutes
0
1
2
Figure l-chromatogram of plasma from a patient under chronic dittiazem therapy: (A) blood sample was immediately centrifuged after collection;(6)blood sample was kept 1 h at room temperature between sampling and centrifugation. Key to peaks: (1) M2; (2) M1; (3) MA; (4) diltiazem; (5) propionyldeacetyldiltiazem (internal standard).
temperature, and no significant degradation occurring after 1 h. When plasma samples were stored for 4 h in an ice bath, instead of being left at room temperature, MA levels did not decrease. Storage of diltiazem and its metabolites at -20 and at -80 "C did not result in significant changes in concentrations over a 3-week period of storage (Table 11). However, a significant decrease in MA levels occurred, with an average loss of 22%, after 5 weeks of storage at -20 "C. Under the same conditions, a significant decrease in diltiazem levels was alao observed (p < 0.011,whereas the concentrations of M1 and M 2 did not change significantly. No decrease in MA and diltiazem levels was observed in spiked plasma samples stored a t -80 "C. This investigation illustrates the importance of applying rigorous conditions for the transport and treatment of blood samples. To achieve accurate and reproducible results with samples from patients treated with diltiazem, the blood must be placed on ice for transport and centrifuged without delay a t low temperature, and the plasma must be immediately
Table CStablltty of Dlitlazem (D) and its Metabolite9 In Blood Samples from Patkntr under Chronic Oral Dlitlazem Therapy
Compound
n
Storage Temperature,
Concentration in Plasma (mean f SD), ng/mL
t Test for
Paired Data
Wilcoxon T Test for Paired Data
.-
"C
D MA M1 M2
D MA M1 M2
20 20 20 20 0 0 0 0
b' 62.9 f 18.0 37.9 2 11.1 17.4 t 8.9 14.6 t 1.7 62.9 t 18.0 37.9 5 11.1 17.4 2 8.9 14.6 t 1.7
4 53.9 2 16.46 30.0 t 11.76 15.6 f 8.96 15.4 f 2.0b 61.6 f 16.44 37.4 f 11.16 16.1 f 8.84 14.9 f 2.1"
t 9.0 6.64 1.25 1.71 0.74 0.58 1.16 0.73
P
H ~
Received June 18, 1990, from the *Institut des Sciences Pharmaceutiques et Biologiques, Laboratoire de Pharmacie Clinique, 8 avenue Rockefeller, 69373 L on Cedex 08, France, the *H6pitaI Cardiovasculaire et Pneumol ique, Service Pharmaceutique-BP Lyon Montchat, 69394 Lyon Cedex d3, France, and the Blnstitut des Sciences Pharmaceutrques et Bi?og,ques, Laboratoire de Mathbmatiques et Informatique, 8 avenue Rockefeller, 69373 Lyon Cedex 08, France. Accepted for publication June 1, 1991. Abstract OThe stability of diltiazem and its metabolites in blood samples from patients under chronic diltiazem therapy was investigated. When whole blood was kept for 1 h at room temperature between sampling and centrifugation, the concentration of Ndemethyldiltiazem (MA) decreased significantly, with an average loss of 24%. Under the same conditions, an average loss of 14% of diltiazem occurred, whereas the concentrations of the metabolites deacetyldiltiazem and Ndemethyldeacetyldiltiazem did not change significantly.No significant decrease in MA and diltiazem concentrationswas observed when whole blood was stored for 1 h in an ice bath. In spiked plasma samples kept at room temperature, only MA was unstable, with an average loss of 13% after 4 h. The present study shows the importance of observing rigorous conditions for the transport and treatment of blood samples. To achieve accurate determination of diltiazem and related compounds, the blood must be centrifuged immediately after collection or kept on Ice for up to 1 h. The plasma samples must be immediatelyfrozen at -80 "C and can be stored for up to 5 weeks before analysis. Using these rigorous conditions, we observed that MA is the main metabolite of diltiazem in plasma from patients under chronic oral diltiazem therapy.
Diltiazem is a calcium channel blocker that is widely used in the treatment of variant angina, hypertension, and supraventricular tachyanythmias.1.2 Diltiazem is extensively metabolized by deacetylation, N-demethylation, O-demethylation, and conjugation.3 Several authors1,4.6 have reported that deacetyldiltiazem (MUwas the main metabolite found in plasma after oral administration of diltiazem, whereas others6.7 have recently produced some evidence suggesting that Ndemethyldiltiazem (MA)may be the predominant metabolite in plasma. In a previous study involving HPLC analysis of diltiazem and its metabolites,s we found that 50% of the parent drug was converted to MA whereas only 12% of the parent drug was converted to M1 in plasma samples from patients under chronic diltiazem therapy. Furthermore, we observed a significant decrease in MA levels in blood samples lef€ a t room temperature before analysis. Significant decreases in MA concentrations in plasma samples kept for 4 days at 4 "C have recently been observed by Dub6 and co-workers9J0 in stability studies of diltiazem and its metabolites in spiked plasma. In the present study, we did a more detailed investigation of the stability of diltiazem and its metabolites in blood and plasma samples. This study allowed us to establish the conditions for sample collection, treatment, and storage necessary for the accurate determination of diltiazem and its metabolites in plasma.
Experimental Section Reagents-Diltiazem, M1, MA, N-demethyldeacetyldiltiazem ( M 2 ) , and propionyldeacetyldiltiazem (used as internal standard) were generously supplied by the Clinical Research Department, Laboratoires #Etudes et de Recherches Synthblabs, Paris, France. The drug supplies were evaluated for identity and purity by UV spectroecopy. CXZ2-3549/92/04CW-0341$02.50/0 0 1992, American Pharmaceutical Association
Acetonitrile (Uvasol) and ammonium dihydrogen phosphate and orthophosphoric acid (both of "Suprapur" grade) were purchased from Merck (Nogent-sur-Marne, France). Triethylamine (analytical grade) was obtained from Fluka (Buchs, Switzerland). Chromatographic Instrumentation and C o n d i t i o l g T h e apparatus consisted of a model 510 HPLC pump equipped with a model 481 variable-wavelength absorbance detector and a model 712 WISP sample processor (all from Waters, Saint Quentin-Yvelines, France). Chromatographic separation was achieved with a Nucleosil C18 column (15 cm x 4.6-mm inner diameter), with a 3-pm particle size (Interchim, Montlucon, France). Aprecolumn (1.5 cm x 4.6-mm inner diameter) packed with Nucleosil C18 ( 5 - l ~ particle n size) was used as a guard column. The mobile phase consisted of acetonitri1e:O.l M ammonium dihydrogen phosphate buffer (37:63, v/v), containing 0.08% triethylamine. The pH of the final solution was adjusted to 5.9 with orthophosphoric acid. The flow rate was 1 mumin, and the detection was performed at 237 nm. Sample Preparation-Before chromatographic analysis, diltiazem and its metabolites were extracted from plasma by a solid-phase extraction procedure that we have described previously.11The main steps of the extraction procedure are the following. The solid-phase extraction columns (100 mg of sorbent, Syva-Biombrieux) were activated with 2 x 3 mL of acetonitrile followed by 2 x 3 mL of 0.1 M ammonium dihydmgen phosphate buffer. Plasma samples (1 mL) diluted with 1 mL of acetonitri1e:O.l M ammonium dihydrogen phosphate (37:63, v/v) were then passed through the columns. The columns were washed with 2 x 1mL of acetonitri1e:water (20:80, v/v) followed by 2 x 500 pL of acetonitri1e:water (40:60, v/v). The compounds of interest were then eluted with 500 pL of acetonitri1e:O.l M ammonium dihydrogen phosphate (80:20, v/v) containing 0.06% triethylamine (pH of the final solution, 6.8). The eluates were evaporated to dryness under nitrogen at 40-45 "C. The residue was reconstituted with 100 pL of mobile phase, and an aliquot of 60 pL was injected onto the HPLC column. Procedure €or Stability S t u d i e d t a b i l i t y in Blood Samples from Patienh-The study was carried out with blood samples from eight patients under chronic oral diltiazem therapy (60 mg, twice daily). Blood samples were drawn just prior to the morning oral dose of 60 mg of diltiazem (trough level). Blood (2 x 10 mL) was collected by venipuncture in a heparinized tube. One blood sample was immediately placed on ice, and the other was kept at room temperature (20 "C). For each sample, 4-mL aliquots were withdrawn immediately (to)and 1h (t,) after blood collection. At these times, blood samples were centrifuged without delay at low temperature (4 "0. Plasma was decanted and frozen at -80 "C, and then analyzed within 3 weeks of blood collection. Stability in Spiked Plasma Samples-The stability of diltiazem and its metabolites in plasma stored at mom temperature or in an ice bath was investigated. Diltiazem and its metabolites, MA, M1, and M2, were added to pooled human bank plasma at a concentration of 100 ng/mL. Separate 8-mL aliquota of plasma were stored at mom temperature and in an ice bath. At the following intervals, 2 mL of plasma was withdrawn for analysis: 0,1, and 4 h. Simultaneously, the effect of storage a t -20 and at -80 "C on the concentration of the compounds of interest was assessed. Four spiked plasma samples were divided into 2-mL aliquots and stored either at -20 or at -80 "C and analyzed within 1, 2, 3, and 5 weeks. This study was replicated four times. Statistical A n a l y s i g T h e nonparametric Mann-Whitney test was used for statistical comparisons at a level of significance of p < 0.05. Journal of Pharmaceutical Sciences / 341 VOl. 81, No. 4, April 1992
Statistical analysis of the study of stability in patient blood samples was performed with the t test for paired data and the nonparametric Wilcoxon paired rank test, with p < 0.05 considered significant (p value was determined with a one-tailed test).
A
Results and Discussion The results of the assay on the stability of the compounds in blood samples are given in Table I, and the typical changes in the concentrations of the compounds are illustrated in Figure 1.When a blood sample from a treated patient was left for 1 h a t room temperature between sampling and centrifugation, the MA concentration decreased significantly from 37.92 11.1 to 30.02 11.7ng/mL (p 5 0.001 for ttest, p 5 0.004 for Wilcoxon T test), for an average loss of 24% of the original concentration. Under the same conditions, a significant decrease in diltiazem concentration from 62.9 2 18.0 to 53.9 f 16.4ng/mL was observed (p I 0.001for t test and p 5 0.004 for Wilcoxon T test), for an average loss of 14.3% of the original concentration. The concentrations of the other measured metabolites, M1 and M2,did not change significantly. When whole blood was stored for 1 h in an ice bath before centrifugation, no significant decrease in MA and diltiazem concentrations was observed. Under these conditions, an average of 98% of the original concentration of the drug and 99% of the original concentration of the mqjor metabolite (MA)were found. Therefore, significant variations in the concentrations of the compounds of interest, especially diltiazem and MA, occurred in whole blood left for only 1 h a t room temperature between sampling and centrifugation. Slight signs of degradation of diltiazem and MA in patient blood samples kept for 4 h at room temperature have been observed by Caill6 et al.10; 89 f 7.5and 88 f 8% of the initial concentration of diltiazem and MA, respectively, were found. However, the compounds were stable in whole blood left for 2 h at room temperature. To explain the mechanism of degradation of diltiazem and MA, Caill6 et a1.10 suggested a hydrolysis of diltiazem and MA to M1 and M2,respectively, in accordance with the metabolic pathway of diltiazem in humans proposed by Sugiharas (Scheme I). However, in our study, the significant decrease in MA levels wae not related to a significant increase in M 2 levels. No degradation of diltiazem, its metabolites, and the internal standard was observed during the analytical procedures (extraction and HPLC analysis). Hence, our results suggest that the degradation of diltiazem and MA cannot be explained only by metabolic transformations of diltiazem and MA to M1 and M2,respectively, as suggested by Caill6 et a1.10 Other mechanisms involving metabolic or chemical hydrolysis may be occurring. As shown in Table 11, only MA was unstable, with an average loss of 13% in plasma samples kept for 4 h at room
,
I
0
2
.
#
I
4 6 8 minutes
I
I
1 0 1 2
0
2
4
6 8 1 minutes
0
1
2
Figure l-chromatogram of plasma from a patient under chronic dittiazem therapy: (A) blood sample was immediately centrifuged after collection;(6)blood sample was kept 1 h at room temperature between sampling and centrifugation. Key to peaks: (1) M2; (2) M1; (3) MA; (4) diltiazem; (5) propionyldeacetyldiltiazem (internal standard).
temperature, and no significant degradation occurring after 1 h. When plasma samples were stored for 4 h in an ice bath, instead of being left at room temperature, MA levels did not decrease. Storage of diltiazem and its metabolites at -20 and at -80 "C did not result in significant changes in concentrations over a 3-week period of storage (Table 11). However, a significant decrease in MA levels occurred, with an average loss of 22%, after 5 weeks of storage at -20 "C. Under the same conditions, a significant decrease in diltiazem levels was alao observed (p < 0.011,whereas the concentrations of M1 and M 2 did not change significantly. No decrease in MA and diltiazem levels was observed in spiked plasma samples stored a t -80 "C. This investigation illustrates the importance of applying rigorous conditions for the transport and treatment of blood samples. To achieve accurate and reproducible results with samples from patients treated with diltiazem, the blood must be placed on ice for transport and centrifuged without delay a t low temperature, and the plasma must be immediately
Table CStablltty of Dlitlazem (D) and its Metabolite9 In Blood Samples from Patkntr under Chronic Oral Dlitlazem Therapy
Compound
n
Storage Temperature,
Concentration in Plasma (mean f SD), ng/mL
t Test for
Paired Data
Wilcoxon T Test for Paired Data
.-
"C
D MA M1 M2
D MA M1 M2
20 20 20 20 0 0 0 0
b' 62.9 f 18.0 37.9 2 11.1 17.4 t 8.9 14.6 t 1.7 62.9 t 18.0 37.9 5 11.1 17.4 2 8.9 14.6 t 1.7
4 53.9 2 16.46 30.0 t 11.76 15.6 f 8.96 15.4 f 2.0b 61.6 f 16.44 37.4 f 11.16 16.1 f 8.84 14.9 f 2.1"
t 9.0 6.64 1.25 1.71 0.74 0.58 1.16 0.73
P
