Stereoselective in vitro degradation pattern of mivacurium in human plasma{dagger}

J. Laurin1, F. Donati2 and F. Varin*,1

1 Faculté de Pharmacie, Université de Montréal C.P. 6128, Succ. Centre-Ville, Montréal (QC), Canada H3C 3J7. 2 Département d’Anesthésie, Faculté de Médecine, Université de Montréal C.P. 6128, Succ. Centre-Ville, Montréal (QC), Canada H3C 3J7 france.varin@umontreal.ca


{dagger}This research was supported by the Canadian Institutes of Health Research (MA-10274). A CIHR-Rx&D studentship was awarded to Julie Laurin.

Accepted for publication: July 10, 2002


    Abstract
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Background. Mivacurium is a mixture of three isomers, two of which are rapidly broken down in vivo by plasma cholinesterases. This study investigates the stereospecificity of mivacurium in vitro degradation to determine if it accounts for its in vivo behaviour.

Methods. The in vitro rate of degradation of each isomer of mivacurium and the in vitro rate of formation of their primary (monoesters and alcohols) and secondary (alcohols) metabolites were examined using human plasma from six healthy volunteers. The in vitro rate of degradation of the monoester metabolites was also assessed. All these determinations were made using a stereospecific high-performance liquid chromatography assay.

Results. The in vitro rate of disappearance of the two active isomers of mivacurium was very rapid, with mean values for the trans trans and cis trans isomers of 0.803 and 0.921 min–1 respectively. These values are twofold faster than published in vivo data. The in vitro rate of disappearance was much slower for the cis cis isomer, with a mean value of 0.0106 min–1. The cis trans isomer was converted exclusively to cis monoester and trans alcohol, while only metabolites in the trans and cis configuration were found for the trans trans and cis cis isomers respectively. Mean in vitro rates of disappearance for the trans and cis monoester were 0.00750 and 0.000633 min–1 respectively.

Conclusions. The in vitro rates of hydrolysis of the active isomers of mivacurium confirm that plasma cholinesterases play a major role in their in vivo degradation, but that in vivo elimination is slowed by extravascular distribution. Mivacurium hydrolysis is stereoselective, the ester group in the trans configuration being more accessible to enzymatic attack. This stereoselective pattern, along with the relatively slow breakdown of the cis cis isomer, sheds light on the in vivo disposition of the cis alcohol metabolite.

Br J Anaesth 2002; 89: 832–8

Keywords: blood, plasma; metabolism, mivacurium; metabolism, stereoselective; neuromuscular block, mivacurium


    Introduction
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Mivacurium chloride consists of a mixture of three stereoisomers. The two most active are the trans trans and cis trans isomers (57 and 37% w/w respectively), which are equipotent; the cis cis isomer (6% w/w) has only one-tenth of the activity of the others in cats and monkeys.1 Mivacurium isomers are believed to be hydrolysed mostly by plasma cholinesterases to give primary (monoester and alcohol) and secondary (alcohol) metabolites (Fig. 1).2 The in vivo elimination half-life in humans is approximately 2 min for the trans trans and cis trans isomers and 30 min for the cis cis isomer.3 The maximum plasma concentration of the metabolites is reached within 35 s of injection, with an average elimination half-life of 90 min, except for the cis alcohol, which is minimally and transiently detected.



View larger version (16K):
[in this window]
[in a new window]
 
Fig 1 Putative degradation pathway of a mivacurium isomer. Reproduced from Savarese et al., Anesthesiology 1998; 68: 723–33, with permission of the publisher.

 
Cook and colleagues4 were the first to determine an in vitro degradation rate for the mixture of mivacurium but their assay was not stereoselective. Referring to a personal communication, Lien and colleagues5 reported an in vitro half-life for the individual isomers (0.83, 1.30 and 276 min for the trans trans, cis trans and cis cis isomers respectively). However, it is unclear whether the isomers were incubated individually or as a mixture. Wiesner and colleagues6 have recently reported shorter in vitro half-lives of 0.71 and 0.61 min for the trans trans and cis trans isomers respectively after incubation of the mixture. These studies did not determine the concentrations of metabolites and therefore the actual pathway of their metabolite formation remains to be elucidated.

The present study was designed to elucidate the individual degradation pathways of each isomer of mivacurium and its monoester metabolites in human plasma, and represents the first attempt to propose an overall pattern for the in vivo disposition of mivacurium.


    Methods
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Chemicals
The three isomers of mivacurium chloride, the monoester and quaternary alcohol metabolites, and laudanosine were provided by Glaxo Wellcome (Research Triangle Park, NC, USA). All analytes were supplied individually with the exception of the monoester metabolite standard, which was supplied as a mixture of the cis and trans isomers. The relative proportions of the cis and trans monoester isomers in the mixture were reported by the supplier to be approximately 26.3 and 66.3% respectively. All organic solvents were HPLC (high-performance liquid chromatography) grade (Anachemia, Montreal, Canada). Echothiophate iodide, a plasma cholinesterase inhibitor, was supplied by Wyeth-Ayerst Laboratories (Montreal, Canada).

Plasma collection
Blood was obtained from six healthy fasting volunteers after they had given written informed consent. The study protocol was approved by the Hotel-Dieu Hospital Ethics Committee (Montreal, Canada). Among the volunteers, there were three men and three women and their age varied between 23 and 40 yr. Approximately 100 ml of blood was obtained in a heparinized syringe. After collection, the blood was centrifuged at 1650 g for 15 min and the plasma was separated and kept on an ice-water bath until incubation.

Incubation
The three isomers of mivacurium and the mixture of monoester metabolites were incubated separately at 37°C in fresh plasma obtained from every volunteer. The in vitro study was performed within 1 h of blood collection to ensure integrity of the medium, especially with respect to enzyme activity. For each analyte, a starting solution (10 ml) was prepared by adding an appropriate amount of corresponding analyte to each preincubated individual plasma. Final concentrations of 4500 ng ml–1 (4.37 µmol litre–1) for trans trans mivacurium, 2500 ng ml–1 (2.43 µmol litre–1) for cis trans mivacurium, 500 ng ml–1 (0.49 µmol litre–1) for cis cis mivacurium and 3000 ng ml–1 (5.15 µmol litre–1) for monoester metabolites were prepared. These final concentrations were chosen to mimic the maximum plasma concentrations observed in patients after an intubating dose of the commercial mixture of the three isomers.3 According to the relative proportions of the trans and cis monoester in the mixture, the monoester metabolite starting solution corresponded to 1989 ng ml–1 (3.41 µmol litre–1) and 789 ng ml–1 (1.35 µmol litre–1) of trans and cis monoester respectively. Each starting solution was separated in nine aliquots of 1 ml, to which 20 µl of echothiophate iodide 0.04 M was added at predetermined times to stop the reaction and to prevent further degradation of the compound. The time points for stopping the reaction after addition of the analyte were 0, 1, 2, 3, 4, 5, 6, 7 and 8 min for the trans trans and cis trans mivacurium metabolites; 0, 15, 30, 45, 60, 75, 90, 105 and 120 min for the cis cis mivacurium metabolites; and 0, 60, 120, 180, 240, 300, 360 and 420 min for the monoester metabolites. After adding the inhibitor, the aliquots were frozen using dry ice and then stored at –70°C until HPLC analysis.

Sample preparation and chromatographic apparatus and conditions
Mivacurium isomers and their metabolites were determined by HPLC using a method similar to that published for the determination of cisatracurium and its metabolites in human urine.7 Bond Elut phenyl solid-phase extraction cartridges (Varian, Harbor City, CA, USA) were conditioned with acetonitrile 1 ml and 5 mM sulphuric acid 1 ml. The plasma sample (0.75 ml) and internal standard (75 µl of laudanosine 1 µg ml–1) were combined in the reservoir and then aspirated through the sorbent. A vacuum of 50–80 kPa was applied to the manifold of the Vac-Elut chamber (Analytichem International, Harbor City, CA, USA) throughout the extraction procedure. The cartridges were washed sequentially with 5 mM sulphuric acid 0.75 ml and a mixture of methanol:water (50:50, 0.75 ml). Analytes were eluted with 80 mM sodium sulphate 2x300 µl in 5 mM sulphuric acid:acetonitrile (40:60). The eluents were then reduced to half of their volume by evaporation using a Speed-Vac concentrator (Model SC210A; Savant Instruments, Farmingdale, NY, USA). Aliquots were injected directly into the HPLC system according to the following procedure.

The chromatographic system consisted of a Constametric 4100 pump (LDC Analytical, Riviera Beach, FL, USA) which was programmed to deliver 14 mM sodium sulphate in 0.5 mM sulphuric acid:acetonitrile (40:60) for 5 min followed by 70 mM sodium sulphate in 0.5 mM sulphuric acid:acetonitrile (40:60) for 6 min, returning to the initial condition to re-equilibrate for 3 min. The flow rate was 2.0 ml min–1 and the mobile phase was not allowed to recirculate. Samples were injected using an auto-injector (SIL-9A; Shimadzu, Kyoto, Japan). Separation of mivacurium isomers and their metabolites was performed on a 5 µm Spherisorb strong cation exchanger column (150x4.6 mm internal diameter; Phenomenex, Torrance, CA, USA). Peaks were detected with a Hewlett Packard 1046 A fluorescence detector (Hewlett Packard, Waldbroom, Germany) at excitation and emission wavelengths of 202 and 320 nm respectively. A Shimadzu C-R3A integrator was used.

The method proved to be sensitive, with a lower limit of quantitation of 4.88 ng ml–1 (0.0047 µmol litre–1) for each isomer of mivacurium and for the quaternary alcohol metabolites, and 10.36 ng ml–1 (0.0232 µmol litre–1) and 4.11 ng ml–1 (0.0071 µmol litre–1) for the trans and cis monoesters respectively. The assay proved to be linear up to 5000 ng ml–1 (4.86 µmol litre–1) for the isomers of mivacurium, 5000 ng ml–1 (11.20 µmol litre–1) for the quaternary alcohol metabolites, 2652 ng ml–1 (4.55 µmol litre–1) for the trans monoester and 1052 ng ml–1(1.81 µmol litre–1) for the cis monoester. The coefficient of determination (r2) was always higher than 0.9930 for each analyte. The method proved to be reproducible for each analyte, with a coefficient of variation always less than 20% over the linear range.

Data analysis
The rate of in vitro disappearance (kin vitro) of each analyte from human plasma was determined by fitting the plasma data to a non-compartmental model with bolus input using a non-linear least-squares computer program (WinNonlin® software; Pharsight Scientific Consulting, Palo Alto, CA, USA) that applied linear regression to all data points. The in vitro elimination half-life (T1/2 in vitro) was calculated using 0.693 kin vitro–1. The mass balance (number of moles disappeared vs number of moles formed) was examined in all subjects.


    Results
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
After incubation in human plasma, trans trans mivacurium disappeared very rapidly, with a mean kin vitro value of 0.803 min–1 (Table 1). This isomer was completely and equally (on a molar basis) converted to trans monoester and trans alcohol metabolites (Fig. 2A).


View this table:
[in this window]
[in a new window]
 
Table 1 Comparison of in vitro and in vivo parameters. The kin vitro values are mean (SD). The T1/2 in vivo values are taken from reference 3
 


View larger version (29K):
[in this window]
[in a new window]
 
Fig 2 In vitro degradation in human plasma after separate incubation of the (A) trans trans, (B) cis trans and (C) cis cis isomers of mivacurium and (D) a mixture of the cis and trans monoesters. The left panels illustrate the mass balance between the degradation of parent compound and the formation of metabolites on an ordinal scale. The right panels illustrate the exponential decline of the plasma concentration–time profile for the parent compound.

 
The disappearance of cis trans mivacurium from human plasma was also very rapid, with a mean kin vitro value of 0.921 min–1 (Table 1). Moreover, the hydrolysis reaction was almost complete at 4 min, when about 98% of the cis trans mivacurium was degraded. Figure 2B shows that the plasma concentrations of cis trans mivacurium and its corresponding metabolites did not change significantly after 4 min of incubation. The individual kin vitro values were therefore calculated between 0 and 4 min. Our results also show that cis trans mivacurium is almost exclusively converted to cis monoester and trans alcohol. Although we were able to quantify some cis alcohol after the incubation of the cis trans isomer, less than 1% of the cis trans isomer was converted into cis alcohol after 8 min. This finding is consistent with the 2% fraction of the cis trans isomer that was not hydrolysed after 4 min. This was considered to be negligible and within the error of the assay, and the data are therefore not presented in the tables and figures.

After incubation in human plasma, the cis cis isomer of mivacurium disappeared at a much slower rate than that of the other two isomers. After 120 min of incubation, the hydrolytic reaction was incomplete, with 30% of the initial cis cis mivacurium plasma concentration remaining. Only metabolites in the cis configuration resulted from hydrolysis of this isomer. Hence, this isomer was equally (on a molar basis) converted to cis monoester and cis alcohol metabolites (Fig. 2C). The resulting mean kin vitro value was 0.0106 min–1 (Table 1).

After a 7-h incubation, the trans monoester metabolite was completely converted to trans alcohol metabolite, whereas only 20% of the cis monoester metabolite was converted to cis alcohol metabolite (Fig. 2D). The resulting mean kin vitro value was 0.00750 min–1 for the trans monoester metabolite and 0.000633 min–1 for the cis monoester metabolite (Table 1). The mean terminal elimination half-life (T1/2 in vivo) obtained in eight healthy patients undergoing elective surgery2 is also presented in Table 1 for each isomer of mivacurium and for the monoester metabolites.

As demonstrated in Figure 2, the mass balance (number of moles disappeared vs number of moles formed) was good, the values varying from 86 to 101%. The slight variation was probably the result of the error inherent in the analytical method (±15%).

In view of these results, a pattern is proposed for the stereoselective degradation of mivacurium (Fig. 3).



View larger version (17K):
[in this window]
[in a new window]
 
Fig 3 Stereoselective degradation pattern of mivacurium. Note the very slow breakdown of cis monoester to cis alcohol.

 

    Discussion
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Mivacurium is an ester type of neuromuscular blocking agent and therefore undergoes enzymatic breakdown by plasma cholinesterases. Because mivacurium contains two ester groups, enzymatic breakdown can occur at either one of these sites. Thus, the hydrolysis of one mole of mivacurium will result in the formation of one mole of monoester metabolite and one mole of alcohol metabolite. Theoretically, when mivacurium is in the trans trans or cis cis geometrical configuration, the metabolites will be in either the trans or the cis configuration respectively if there is no interconversion in vivo. Our results support the absence of in vivo interconversion. On the other hand, when mivacurium is in the cis trans configuration, metabolites could be in the cis or trans configuration. More specifically, if the ester group in the trans configuration is attacked, it will lead to the formation of trans alcohol and cis monoester. Therefore, the hydrolysis of cis trans mivacurium could theoretically result in the formation of the trans alcohol and cis monoester metabolites and/or the cis alcohol and trans monoester metabolites. Our results show clearly that cis trans mivacurium is converted exclusively to trans alcohol and cis monoester metabolites, suggesting that the ester group of the isomer in the trans configuration is more labile.

There is no clear explanation for the flattening of the curve for cis trans degradation in plasma. However, we are sure that this finding is not an artefact. There are two good reasons why analytical sensitivity can be ruled out. First, the levels are well above the lower limit of quantitation (LLOQ; at least four times the LLOQ). Secondly, the flattening of the cis trans concentration curve was consistent in all patients but never observed for the trans trans isomer. After 4 min, the measured plasma concentration of the cis trans isomer averaged 50 ng ml–1 (0.0486 µmol litre–1), which represents approximately 2% of the initial concentration. One possible explanation for the flattening of the curve would be that the rate of ester hydrolysis is different for the cis and trans positions. Hydrolysis at the trans position would be complete at 4 min and the flattening of the curve could be attributed to slower hydrolysis at the cis position.

After the hydrolysis of mivacurium isomers into monoester and alcohol metabolites, the monoester metabolites are in turn degraded into alcohol metabolites. However, as the kin vitro values for the monoesters are at least 10-fold slower than those for the isomers, the relative importance of this secondary metabolic pathway to the formation of the alcohol metabolite is thought to be negligible.

Our results also show that the cis monoester metabolite does not undergo any significant in vitro hydrolysis in comparison with the trans monoester metabolite. This finding supports the results obtained for the isomers, i.e. that the ester group in the trans configuration is more susceptible to attack by the enzyme. The presence of greater steric hindrance when the ester group is in the cis configuration may explain the greater resistance to enzymatic degradation.

For drugs undergoing extensive hydrolysis in human plasma (non-organ-based elimination), the in vitro rate of degradation (kin vitro) is often used as a substitute in pharmacokinetic models assuming peripheral elimina tion.810 In addition, direct comparison of in vitro and in vivo elimination half-lives provides information about the relative contribution of organ-based elimination (e.g. by the kidneys and liver) to the overall body elimination. Our results indicate that, for the two most active isomers (trans trans and cis trans), the T1/2 in vitro is 2-fold faster than the T1/2 in vivo, confirming that hydrolysis by plasma cholinesterases plays a major role in the degradation of these two isomers. The possibility that this apparent discrepancy resulted from the fact that two different populations of subjects were studied cannot be ruled out. However, Wiesner and colleagues6 recently reported in vitro half-lives similar to those obtained in this study. Another explanation for the slower in vivo elimination could be that extravascular distribution of these two isomers delays their overall elimination from the body. The presence of plasma cholinesterases in the cerebrospinal fluid has been demonstrated in animals,11 although their activity may differ markedly between tissues and plasma. Despite the fact that the two active isomers of mivacurium (trans trans and cis trans) are eliminated rapidly by plasma cholinesterases, a contribution of the peripheral compartment to their elimination cannot be excluded. Recent data have confirmed the presence of an arterial–venous gradient of approximately 30% across muscle tissue in anaesthetized patients.12

In contrast, the T1/2 in vivo for the cis cis isomer is approximately 2-fold faster than the T1/2 in vitro. This finding is an indication that the overall elimination of this isomer is not solely the result of its hydrolysis by plasma cholinesterases. Indeed, the contribution of renal function to the overall elimination of this isomer has been demonstrated by Head-Rapson and colleagues.13 Moreover, Lien and colleagues5 have shown that the clearance of this isomer is not related to plasma cholinesterase activity.

For the monoester metabolites, the in vivo and in vitro values were similar for the trans monoester, indicating a minor contribution of organ-based elimination to total body disposition. Conversely, the in vitro half-life of the cis monoester was considerably slower than that observed in vivo. However, this value should be interpreted cautiously as the duration of incubation did not allow accurate determination of the half-life. Nonetheless, our findings suggest that the contribution of enzymatic hydrolysis to the overall in vivo elimination of the cis monoester metabolite is negligible.

When examining the in vivo formation and elimination of the alcohol and monoester metabolites after an i.v. bolus of the commercial mixture of mivacurium in human patients, Lacroix and colleagues3 noted that the cis alcohol metabolite was detected minimally and transiently. An initial mixing peak was observed at approximately 30 s for both the cis and the trans alcohol metabolite. This was attributed to the alcohols (known degradation or synthesis by-products of the isomers) that are already contained in the injectable solution. Of importance, the levels of the cis alcohol were not quantifiable (below the LLOQ) after only 8 min, while levels almost 10-fold higher allowed characterization of the kinetics of the trans alcohol for up to 250 min. As the later portion of the plasma concentration–time curve was more compatible with in vivo formation of the metabolites, no explanation can be provided for this finding. Our in vitro results now shed some light on the in vivo disposition of the cis alcohol metabolite. First, less than 2% of the cis alcohol is formed after in vitro degradation of the cis trans isomer. Secondly, the in vitro rate of formation of cis alcohol from the cis monoester is questionable. Finally, as the cis cis isomer represents only 6% of the injectable mixture of mivacurium, its contribution to the in vivo formation of the cis alcohol is certainly minimal.

In conclusion, the in vitro rates of hydrolysis of the active isomers confirm that plasma cholinesterases play a major role in their in vivo degradation, but also suggest that their in vivo elimination is slowed by extravascular distribution. Mivacurium hydrolysis is stereoselective, the trans configuration being more amenable to ester hydrolysis than the cis configuration. This stereoselective pattern, along with the relatively slow breakdown of the cis cis isomer, explains the in vivo disposition of the cis alcohol metabolite. Hydrolysis is facilitated by the presence of another esterified moiety at the opposite end of the molecule, the hydrolysis of the cis cis and trans trans isomers being considerably faster than that of their respective monoesters. From the results of this in vitro study we propose an overall stereoselective pattern for the degradation of mivacurium that allows the contribution of organ-based elimination to be dissociated from that of non-organ-based elimination.


    Acknowledgements
 
The authors would like to thank Louisa Cale and Johanne Couture for their technical assistance.


    References
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
1 Belmont MR, Beemer G, Bownes P, Russo J, Wisowaty J, Savarese JJ. Comparative pharmacology of mivacurium and isomers in rhesus monkeys [abstract]. Anesth Analg 1993; 76: S18[ISI]

2 Savarese JJ, Hassan HA, Salvatore JB, et al. The clinical neuromuscular pharmacology of mivacurium chloride (BW B1090U). Anesthesiology 1988; 68: 723–32[ISI][Medline]

3 Lacroix M, Donati F, Varin F. Pharmacokinetics of mivacurium isomers and their metabolites in healthy volunteers after intravenous bolus administration. Anesthesiology 1997; 86: 322–30[ISI][Medline]

4 Cook DR, Stiller RL, Weakly JN, Chakravorti S, Brandom BW, Welch RM. In vitro metabolism of mivacurium chloride (BW B1090U) and succinylcholine. Anesth Analg 1989; 69: 452–6

5 Lien CA, Schmith VDE, Embree PB, Belmont MR, Wargin WA, Savarese JJ. The pharmacokinetics and pharmacodynamics of the stereoisomers of mivacurium in patients receiving nitrous oxide/opioid/barbiturate anesthesia. Anesthesiology 1994; 80: 1296–306[ISI][Medline]

6 Wiesner G, Gruber M, Keyl C, Schneider A, Drescher J, Hobbhahn J. In vitro effects of fluoride on pseudocholinesterase activity and the metabolism of the cis-trans and trans-trans isomers of mivacurium. Anesthesiology 2001; 95: 806–7[ISI][Medline]

7 Bryant BJ, James CD Jr, Cook DR, Harrelson JC. High performance liquid chromatography assay for cisatracurium and its metabolites in human urine. J Liquid Chromatogr Relat Technol 1997; 20: 2041–51

8 Fisher DM, Canfell PC, Fahey MR, et al. Elimination of atracurium in humans: Contribution of Hofmann elimination and ester hydrolysis versus organ-based elimination. Anesthesiology 1986; 65: 6–12[ISI][Medline]

9 Tran TV, Fiset P, Varin F. Pharmacokinetics and pharmacodynamics of cisatracurium after a short infusion in patients under propofol anesthesia. Anesth Analg 1998; 87: 1158–63[Abstract]

10 Bergeron L, Bevan DR, Berrill A, Kahwaji R, Varin F. Concentration–effect relationship of cisatracurium at three different dose levels in the anesthetized patient. Anesthesiology 2001; 95: 314–23[ISI][Medline]

11 Ummenhofer WC, Brown SM, Bernards CM. Acetyl cholinesterase and butyrylcholinesterase are expressed in the spinal meninges of monkeys and pigs. Anesthesiology 1998; 88: 1259–65[ISI][Medline]

12 Ezzine S, Donati F, Varin F. Mivacurium arteriovenous gradient during steady state infusion in anesthetized patients. Anesthesiology 2002; 97: 622–9[ISI][Medline]

13 Head-Rapson AG, Devlin JC, Parker CJR, Hunter JM. Pharmacokinetics and pharmacodynamics of the three isomers of mivacurium in health, in end-stage renal failure and in patients with impaired renal function. Br J Anaesth 1995; 75: 31–6[Abstract/Free Full Text]





This Article
Abstract
Full Text (PDF)
E-Letters: Submit a response to the article
Alert me when this article is cited
Alert me when E-letters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (1)
Disclaimer
Request Permissions
Google Scholar
Articles by Laurin, J.
Articles by Varin, F.
PubMed
PubMed Citation
Articles by Laurin, J.
Articles by Varin, F.