- Email: [email protected]
MRI findings in epileptic patients on vigabatrin for more than 5 years
Seizure 1992; 1: 163-l 65
MRI findings in epileptic more than 5 years LEONARDO
patients
on vigabatrin
for
COCITO, MAURA MAFFINI & CARLO LOEB
Clinica Neurologica del/‘tJniversitA di Genova, ha/y Correspondence
to Prof. Carlo Loeb, Clinica Neurologica dell’Universith di Genova, Via A. De Toni, 5, 16132 Genova, Italy
Although vigabatrin is a promising new antiepileptic drug, its safety has been challenged by the report of dosedependent central nervous system myelin vacuolation in some preclinical animal studies. Since it has been shown that vacuolation is associated with specific magnetic resonance imaging (MRI) findings in rats and dogs, MRI of the brain was performed in 11 patients with complex partial seizures who had been receiving vigabatrin for 64-78 months (mean 74.0 ? 5.0 sd) as additional treatment for epilepsy, with a cumulative exposure ranging 4200 to 9360 g. In no case did MRI show white matter changes similar to the pathological findings of microvacuolation observed in animals. These results would appear to confirm that current doses of vigabatrin do not cause myelin vacuolation in humans, even for treatment periods of longer than 5 years. Key words: antiepileptic imaging; vigabatrin.
drugs; brain microvacuolation;
INTRODUCTION Vigabatrin is a specific irreversible inhibitor of y-aminobutyric acid (GABA) transaminase, which enhances inhibitory neurotransmission by increasing concentrations of GABA in the brain and whose efficacy as add-on therapy in refractory epilepsy has been well-established by several short- and long-term trials’. Major concerns about the safety of vigabatrin were raised by reports of the occurrence of dosedependent central nervous system microvacuolation caused by intramyelinic edema in preclinical toxicology studies in mice, rats and dogs2-? Indirect evidence that vacuolation does not occur in humans has been provided by the lack of abnormalities of somatosensory evoked potentials (SEPs) in patients receiving vigabatrin5v6, whereas histopathological vacuolation in dogs is associated with increased central latency of SEPs7. More direct information comes from occasional post-mortem studies of patients treated with vigabatrin, in whom no central nervous system microvacuolation was observed4s8*‘. Quite recently, magnetic resonance imaging (MRI) was reported to detect vigabatrin-induced lesions both in rats” 1059- 1311192/030163+03
$06.00/O
clinical trials; long-term safety; magnetic resonance
and dogs’i. Since it is conceivable that similar information might also be provided in humans, MRI of the brain was performed in patients who had been receiving vigabatrin for longer than 5 years.
PATIENTS AND METHODS Eleven patients (6 females, 5 males, mean age 41.3 years + 14.1sd) with complex partial seizures were studied, who had been receiving vigabatrin 2-4 g/day (mean daily dose 51.0 mgl kg + 11.2sd) for 64-78 months (mean 74 + 5 sd) as additional treatment for epilepsy. The main clinical and pharmacological features of individual patients are reported in Table 1. All the patients had been entered in a l-year single-blind trial, according to inclusion and exclusion criteria reported in detail elsewhere12, and continued the treatment after the trial endpoint because of sustained benefit. The duration of epilepsy ranged from 12 to 50 years (mean 24.3 + 11.2 sd) and the aetiologies were post-traumatic in one patient, perinatal in two patients and unknown in eight patients. Secondarily generalized seizures occurred in five @ 1992 Bailli&e Tindall
L. cocito et i?/.
164 Table 1: Clinical and pharmacological Patient no.
Sex Age (years)
Aetiology
1 2 3 4 5 6 I 8 9 10 11
F M F F F M M F M F M
Unknown Unknown Unknown Unknown Unknown Perinatal Unknown Unknown Perinatal Unknown Posttraumatic
Mean sd
31 56 50 41 30 35 34 61 24 28 64
features of the patients CT scan
Epilepsy duration (years)
Secondary generalization
Treatment duration (months)
Normal Abnormal Abnormal Normal Normal Abnormal Abnormal Abnormal Normal Abnormal Normal
19 50 30 35 23 22 13 26 12 13 24
+ + + + +
41.3 14.1
VGB = vigabatrin,
24.3 11.2 CBZ = carbamazepine,
PB = phenobarbital,
Cumulative exposure to VGB (g)
Concomitant antiepileptic drugs
76 78 76 76 76 78 18 70 64 66 76
Daily VGB dose hglkg) 55 41 62 14 47 49 52 33 60 58 31
6840 9360 7980 9120 6840 9360 7020 4200 5760 7425 5700
CBZ, CBZ CBZ CBZ CBZ CBZ CBZ, CBZ, CBZ, CBZ, CBZ,
74.0 5.0
51.0 12.7
7236.8 1651.6
PHT = phenytoin,
PB PHT PB PRI PB
PRI = primidone.
onal slices were obtained as well; images after paramagnetic contrast ment with Cd-DTPA were obtained patients. MR images were evaluated enced neuroradiologists who had involved in the trial.
patients. CT (computed tomography) scan performed before starting vigabatrin was normal in five patients and showed minor abnormalities in six. The cumulative exposure to vigabatrin since the onset of treatment ranged from 4200 to 9360g (mean 7237 g t- 1652 sd). All patients were concomitantly receiving carbamazepine (700-1200 mg/day), four were additionally taking phenobarbital (125-150 mg/ day), one phenytoin (250mg/day) and one primidone (500 mg/day). MRI studies were performed with a 0.5 T superconducting MR unit (Esatom 5000, Esaote Biomedica, Italy), using a spin-echo technique. In all patients Tl-weighted and T2weighted axial slices were obtained (slice thickness 8mm). In nine patients 5-mm cor-
PB
additional enhancein eight by experinot been
RESULTS
The MRI findings are reported in detail in Table 2. Brain MRI was normal in three patients and showed some degree of abnormality in the other eight patients. Abnormalities detected by MRI were consistent with the available CT information in five patients and revealed unsuspected lesions in three (see
Table 2: Summary of the MRI findings Gd-DTPA
Global judgment
PVH
1 2 3
Sex/ age F/31 Ml56 F/50
+ +
Normal Abnormal Abnormal
-
4
F/41
+
Abnormal
-
F/30 Ml35 Ml34 F/61
+ + + +
Normal Abnormal Abnormal Abnormal
M/24
+
Patient no.
10 11
F/28 Ml64
VD
CA
WMA
Other findings
-
-
-
-
Moderate Mild
-
-
-
-
-
-
Moderate -
-
Normal
-
-
-
-
Abnormal Abnormal
Minimal
-
-
-
None None Right medial temporal T2-hyperintense area (gliosis) Small left cerebellar (cortical) T%-hyperintense contrast-enhanced area (probable meningioma) None None Left periventricular calcified lesion Bilateral parietal-occipital porencephalic cysts None (extracerebral: right maxillary sinusitis) Slight ventricular asymmetry None
PVH = periventricular hyperintensity areas, VD = ventricular abnormalities (increased T2-decreased Tl).
dilation,
CA = cortical atrophy, WMA = white matter
MRI in patients on long-term
vigabatrin
Table 2). No areas of increased T2 and decreased Tl signals in the white matter, possibly related to microvacuolation as described in animal studies”,‘i, were observed in any patient.
165
We are indebted to the Institute of Radiology of the University of Genova (Prof. G. Cittadini) for performing the MRI investigations.
REFERENCES DISCUSSION * Brain MRI in these 11 epileptic patients on long-term vigabatrin treatment often revealed abnormalities, possibly related to the cause of seizures. However, no areas of abnormal signal in the white matter were detected. MRI studies in rats” and dogs l1 have indicated that vigabatrin-induced microvacuolation gives rise to specific MRI findings. Since there is no reason to expect that the occurrence of this abnormality would not be detected in humans, it can be argued that no evidence of microvacuolation was found in our patients. No reports on MRI findings in patients on long-term vigabatrin have been published to date, and therefore our results need confirmation in larger groups of patients. Nevertheless, the lack of white matter MRI abnormalities after receiving therapeutic doses of vigabatrin for longer than 5 years (with a cumulative exposure of several thousands grams) suggests that microvacuolation is unlikely to occur in humans. These findings are in keeping with both the results of post-mortem studies4,“’ and the indirect evidence provided by electrophysiological investigations5’6, and represent further support for the safety of vigabatrin, even for long-term treatment.
ACKNOWLEDGMENTS This work was supported in part by CNR Centro di Studio per la Neurofisiologia Cerebrale di Genova.
1. Ring, H.A. and Reynolds, E.H. Vigabatrin. In: Recent Advances in Epilepsy, Volume 5 (Eds T.A. Pedley and B.S. Meldrum). Edinburgh, Churchill Livingstone, 1992: pp. 177-195. 2. Butler, W.H., Ford, G.P. and Newbeme, J.W. A study of the effects of vigabatrin on the central nervous system and retina of Sprague Dawley and Listerhooded rats. Toxicological Pathology 1987; 15: 143148. 3. Graham, D. Neuropathology of vigabatrin. British Journal of Clinical Pharmacology 1989; 27 (Suppl. 1): 43s-45s. 4. Hauw, J-J., Trottier, S., Boutry, J-M. et al. The neuropathology of vigabatrin. British Journal of Clinical Practice 1988; Suppl. 61: 10-13. 5. Liegeois-Chauvel, C., Marquis, P., Gisselbrecht, D. et al. Effects of long-term vigabatrin on somatosensoryevoked potentials in epileptic patients. Epilepsia 1989; 30 (Suppl. 3): S23-S25. 6. Cosi, V., Callieco, R., Galimberti, CA. et al. Effects of vigabatrin on evoked potentials in epileptic patients. British Journal of Clinical Pharmacology 1989; 27 (Suppl. 1): 61S-68s. 7. Arezzo, J.C., Schroeder, C.E., Litwak, M.S. et al. Effects of vigabatrin on evoked potentials in dogs. British Journul of Clinical Pharmacology 1989; 27 (Suppl. 1): 53S-60s. 8. Pederson, B., Hojgaard, K. and Dam, M. Vigabatrin: no microvacuoles in human brain. Epilepsy Research 1987; 1: 74-76. 9. Butler, W.H. The neuropathology of vigabatrin. Epilepsia 1989; 30 (Suppl. 3): S15-S17. 10. Jackson, J.D., Williams, S.F., van Bruggen, N. et al. Vigabatrin-induced cerebellar and cortical lesions are demonstrated by quantitative magnetic resonance imaging. Epilepsia 1991; 32 (Suppl. 1): 13. 11. Sussman, N.M., Weiss, K.L., Schroeder, C.E. et al. Vigabatrin: effect on in vivo and ex vivo magnetic resonance imaging of dog brains. Epilepsia 1991; 32 (Suppl. 1): 13. 12. Cocito, L., Maffini, M., Perfumo, P. et al. Vigabatrin in complex partial seizures: a long-term study. Epilepsy Research 1989; 3: 160-166.
MRI findings in epileptic more than 5 years LEONARDO
patients
on vigabatrin
for
COCITO, MAURA MAFFINI & CARLO LOEB
Clinica Neurologica del/‘tJniversitA di Genova, ha/y Correspondence
to Prof. Carlo Loeb, Clinica Neurologica dell’Universith di Genova, Via A. De Toni, 5, 16132 Genova, Italy
Although vigabatrin is a promising new antiepileptic drug, its safety has been challenged by the report of dosedependent central nervous system myelin vacuolation in some preclinical animal studies. Since it has been shown that vacuolation is associated with specific magnetic resonance imaging (MRI) findings in rats and dogs, MRI of the brain was performed in 11 patients with complex partial seizures who had been receiving vigabatrin for 64-78 months (mean 74.0 ? 5.0 sd) as additional treatment for epilepsy, with a cumulative exposure ranging 4200 to 9360 g. In no case did MRI show white matter changes similar to the pathological findings of microvacuolation observed in animals. These results would appear to confirm that current doses of vigabatrin do not cause myelin vacuolation in humans, even for treatment periods of longer than 5 years. Key words: antiepileptic imaging; vigabatrin.
drugs; brain microvacuolation;
INTRODUCTION Vigabatrin is a specific irreversible inhibitor of y-aminobutyric acid (GABA) transaminase, which enhances inhibitory neurotransmission by increasing concentrations of GABA in the brain and whose efficacy as add-on therapy in refractory epilepsy has been well-established by several short- and long-term trials’. Major concerns about the safety of vigabatrin were raised by reports of the occurrence of dosedependent central nervous system microvacuolation caused by intramyelinic edema in preclinical toxicology studies in mice, rats and dogs2-? Indirect evidence that vacuolation does not occur in humans has been provided by the lack of abnormalities of somatosensory evoked potentials (SEPs) in patients receiving vigabatrin5v6, whereas histopathological vacuolation in dogs is associated with increased central latency of SEPs7. More direct information comes from occasional post-mortem studies of patients treated with vigabatrin, in whom no central nervous system microvacuolation was observed4s8*‘. Quite recently, magnetic resonance imaging (MRI) was reported to detect vigabatrin-induced lesions both in rats” 1059- 1311192/030163+03
$06.00/O
clinical trials; long-term safety; magnetic resonance
and dogs’i. Since it is conceivable that similar information might also be provided in humans, MRI of the brain was performed in patients who had been receiving vigabatrin for longer than 5 years.
PATIENTS AND METHODS Eleven patients (6 females, 5 males, mean age 41.3 years + 14.1sd) with complex partial seizures were studied, who had been receiving vigabatrin 2-4 g/day (mean daily dose 51.0 mgl kg + 11.2sd) for 64-78 months (mean 74 + 5 sd) as additional treatment for epilepsy. The main clinical and pharmacological features of individual patients are reported in Table 1. All the patients had been entered in a l-year single-blind trial, according to inclusion and exclusion criteria reported in detail elsewhere12, and continued the treatment after the trial endpoint because of sustained benefit. The duration of epilepsy ranged from 12 to 50 years (mean 24.3 + 11.2 sd) and the aetiologies were post-traumatic in one patient, perinatal in two patients and unknown in eight patients. Secondarily generalized seizures occurred in five @ 1992 Bailli&e Tindall
L. cocito et i?/.
164 Table 1: Clinical and pharmacological Patient no.
Sex Age (years)
Aetiology
1 2 3 4 5 6 I 8 9 10 11
F M F F F M M F M F M
Unknown Unknown Unknown Unknown Unknown Perinatal Unknown Unknown Perinatal Unknown Posttraumatic
Mean sd
31 56 50 41 30 35 34 61 24 28 64
features of the patients CT scan
Epilepsy duration (years)
Secondary generalization
Treatment duration (months)
Normal Abnormal Abnormal Normal Normal Abnormal Abnormal Abnormal Normal Abnormal Normal
19 50 30 35 23 22 13 26 12 13 24
+ + + + +
41.3 14.1
VGB = vigabatrin,
24.3 11.2 CBZ = carbamazepine,
PB = phenobarbital,
Cumulative exposure to VGB (g)
Concomitant antiepileptic drugs
76 78 76 76 76 78 18 70 64 66 76
Daily VGB dose hglkg) 55 41 62 14 47 49 52 33 60 58 31
6840 9360 7980 9120 6840 9360 7020 4200 5760 7425 5700
CBZ, CBZ CBZ CBZ CBZ CBZ CBZ, CBZ, CBZ, CBZ, CBZ,
74.0 5.0
51.0 12.7
7236.8 1651.6
PHT = phenytoin,
PB PHT PB PRI PB
PRI = primidone.
onal slices were obtained as well; images after paramagnetic contrast ment with Cd-DTPA were obtained patients. MR images were evaluated enced neuroradiologists who had involved in the trial.
patients. CT (computed tomography) scan performed before starting vigabatrin was normal in five patients and showed minor abnormalities in six. The cumulative exposure to vigabatrin since the onset of treatment ranged from 4200 to 9360g (mean 7237 g t- 1652 sd). All patients were concomitantly receiving carbamazepine (700-1200 mg/day), four were additionally taking phenobarbital (125-150 mg/ day), one phenytoin (250mg/day) and one primidone (500 mg/day). MRI studies were performed with a 0.5 T superconducting MR unit (Esatom 5000, Esaote Biomedica, Italy), using a spin-echo technique. In all patients Tl-weighted and T2weighted axial slices were obtained (slice thickness 8mm). In nine patients 5-mm cor-
PB
additional enhancein eight by experinot been
RESULTS
The MRI findings are reported in detail in Table 2. Brain MRI was normal in three patients and showed some degree of abnormality in the other eight patients. Abnormalities detected by MRI were consistent with the available CT information in five patients and revealed unsuspected lesions in three (see
Table 2: Summary of the MRI findings Gd-DTPA
Global judgment
PVH
1 2 3
Sex/ age F/31 Ml56 F/50
+ +
Normal Abnormal Abnormal
-
4
F/41
+
Abnormal
-
F/30 Ml35 Ml34 F/61
+ + + +
Normal Abnormal Abnormal Abnormal
M/24
+
Patient no.
10 11
F/28 Ml64
VD
CA
WMA
Other findings
-
-
-
-
Moderate Mild
-
-
-
-
-
-
Moderate -
-
Normal
-
-
-
-
Abnormal Abnormal
Minimal
-
-
-
None None Right medial temporal T2-hyperintense area (gliosis) Small left cerebellar (cortical) T%-hyperintense contrast-enhanced area (probable meningioma) None None Left periventricular calcified lesion Bilateral parietal-occipital porencephalic cysts None (extracerebral: right maxillary sinusitis) Slight ventricular asymmetry None
PVH = periventricular hyperintensity areas, VD = ventricular abnormalities (increased T2-decreased Tl).
dilation,
CA = cortical atrophy, WMA = white matter
MRI in patients on long-term
vigabatrin
Table 2). No areas of increased T2 and decreased Tl signals in the white matter, possibly related to microvacuolation as described in animal studies”,‘i, were observed in any patient.
165
We are indebted to the Institute of Radiology of the University of Genova (Prof. G. Cittadini) for performing the MRI investigations.
REFERENCES DISCUSSION * Brain MRI in these 11 epileptic patients on long-term vigabatrin treatment often revealed abnormalities, possibly related to the cause of seizures. However, no areas of abnormal signal in the white matter were detected. MRI studies in rats” and dogs l1 have indicated that vigabatrin-induced microvacuolation gives rise to specific MRI findings. Since there is no reason to expect that the occurrence of this abnormality would not be detected in humans, it can be argued that no evidence of microvacuolation was found in our patients. No reports on MRI findings in patients on long-term vigabatrin have been published to date, and therefore our results need confirmation in larger groups of patients. Nevertheless, the lack of white matter MRI abnormalities after receiving therapeutic doses of vigabatrin for longer than 5 years (with a cumulative exposure of several thousands grams) suggests that microvacuolation is unlikely to occur in humans. These findings are in keeping with both the results of post-mortem studies4,“’ and the indirect evidence provided by electrophysiological investigations5’6, and represent further support for the safety of vigabatrin, even for long-term treatment.
ACKNOWLEDGMENTS This work was supported in part by CNR Centro di Studio per la Neurofisiologia Cerebrale di Genova.
1. Ring, H.A. and Reynolds, E.H. Vigabatrin. In: Recent Advances in Epilepsy, Volume 5 (Eds T.A. Pedley and B.S. Meldrum). Edinburgh, Churchill Livingstone, 1992: pp. 177-195. 2. Butler, W.H., Ford, G.P. and Newbeme, J.W. A study of the effects of vigabatrin on the central nervous system and retina of Sprague Dawley and Listerhooded rats. Toxicological Pathology 1987; 15: 143148. 3. Graham, D. Neuropathology of vigabatrin. British Journal of Clinical Pharmacology 1989; 27 (Suppl. 1): 43s-45s. 4. Hauw, J-J., Trottier, S., Boutry, J-M. et al. The neuropathology of vigabatrin. British Journal of Clinical Practice 1988; Suppl. 61: 10-13. 5. Liegeois-Chauvel, C., Marquis, P., Gisselbrecht, D. et al. Effects of long-term vigabatrin on somatosensoryevoked potentials in epileptic patients. Epilepsia 1989; 30 (Suppl. 3): S23-S25. 6. Cosi, V., Callieco, R., Galimberti, CA. et al. Effects of vigabatrin on evoked potentials in epileptic patients. British Journal of Clinical Pharmacology 1989; 27 (Suppl. 1): 61S-68s. 7. Arezzo, J.C., Schroeder, C.E., Litwak, M.S. et al. Effects of vigabatrin on evoked potentials in dogs. British Journul of Clinical Pharmacology 1989; 27 (Suppl. 1): 53S-60s. 8. Pederson, B., Hojgaard, K. and Dam, M. Vigabatrin: no microvacuoles in human brain. Epilepsy Research 1987; 1: 74-76. 9. Butler, W.H. The neuropathology of vigabatrin. Epilepsia 1989; 30 (Suppl. 3): S15-S17. 10. Jackson, J.D., Williams, S.F., van Bruggen, N. et al. Vigabatrin-induced cerebellar and cortical lesions are demonstrated by quantitative magnetic resonance imaging. Epilepsia 1991; 32 (Suppl. 1): 13. 11. Sussman, N.M., Weiss, K.L., Schroeder, C.E. et al. Vigabatrin: effect on in vivo and ex vivo magnetic resonance imaging of dog brains. Epilepsia 1991; 32 (Suppl. 1): 13. 12. Cocito, L., Maffini, M., Perfumo, P. et al. Vigabatrin in complex partial seizures: a long-term study. Epilepsy Research 1989; 3: 160-166.