CME
`
`New onset geriatric epilepsy
`A randomized study of gabapentin, lamotrigine, and
`carbamazepine
`
`A.J. Rowan, MD; R.E. Ramsay, MD; J.F. Collins, ScD; F. Pryor, MPH; K.D. Boardman, RPh;
`B.M. Uthman, MD; M. Spitz, MD; T. Frederick, MD; A. Towne, MD; G.S. Carter, MD, PhD; W. Marks, MD;
`J. Felicetta, MD; M.L. Tomyanovich, MD; and the VA Cooperative Study 428 Group
`
`Abstract—Objective: To determine the relative tolerability and efficacy of two newer antiepileptic drugs, lamotrigine
`(LTG) and gabapentin (GBP), as compared to carbamazepine (CBZ) in older patients with epilepsy. Methods: This was an
`18-center, randomized, double-blind, double dummy, parallel study of 593 elderly subjects with newly diagnosed seizures.
`Patients were randomly assigned to one of three treatment groups: GBP 1,500 mg/day, LTG 150 mg/day, CBZ 600 mg/day.
`The primary outcome measure was retention in trial for 12 months. Results: Mean age was 72 years. The most common
`etiology was cerebral infarction. Patients had multiple medical conditions and took an average of seven comedications.
`Mean plasma levels at 6 weeks were as follows: GBP 8.67 ⫾ 4.83 ␮g/mL, LTG 2.87 ⫾ 1.60 ␮g/mL, CBZ 6.79 ⫾ 2.92 ␮g/mL.
`They remained stable throughout the trial. Early terminations: LTG 44.2%, GBP 51%, CBZ 64.5% (p ⫽ 0.0002). Signifi-
`cant paired comparisons: LTG vs CBZ: p ⬍ 0.0001; GBP vs CBZ: p ⫽ 0.008. Terminations for adverse events: LTG 12.1%,
`GBP 21.6%, CBZ 31% (p ⫽ 0.001). Significant paired comparisons: LTG vs CBZ: p ⬍ 0.0001; LTG vs GBP: p ⫽ 0.015.
`There were no significant differences in seizure free rate at 12 months. Conclusions: The main limiting factor in patient
`retention was adverse drug reactions. Patients taking lamotrigine (LTG) or gabapentin (GBP) did better than those taking
`carbamazepine. Seizure control was similar among groups. LTG and GBP should be considered as initial therapy for older
`patients with newly diagnosed seizures.
`NEUROLOGY 2005;64:1868–1873
`
`than LTG in the treatment of older patients with
`new onset seizures.10 These considerations led us to
`design a clinical trial of both GBP and LTG in el-
`derly patients with newly diagnosed epileptic sei-
`zures, using CBZ, widely considered to be a drug of
`choice for partial onset seizures, as a comparator.
`
`Epidemiologic data indicate that the incidence of ep-
`ilepsy increases markedly after age 60, exceeding
`that of any other age group, including children, by
`several fold.1,2 Factors complicating the treatment of
`seizures in older age groups include concurrent med-
`ical diseases, polytherapy, changes in pharmacoki-
`netics (pK), and altered CNS pharmacodynamics.3-7
`Many antiepileptic drugs (AEDs) pose problems in
`the aged due to limited tolerability. With the intro-
`duction of gabapentin (GBP) in 1993 and lamotrigine
`(LTG) shortly thereafter, both appeared to have fa-
`vorable pK and side effect profiles that might offer
`treatment benefits.8,9 One controlled study found
`that carbamazepine (CBZ) was less well tolerated
`
`Additional material related to this article can be found on the Neurology
`Web site. Go to www.neurology.org and scroll down the Table of Con-
`tents for the June 14 issue to find the title link for this article.
`
`Methods. The study commenced in 1998 at 18 Veterans Affairs
`Medical Centers. The design, similar to that of two previous VA
`studies, was modified for the elderly population.11,12 Approval was
`obtained from the central VA Human Rights Committee and all
`local institutional review boards. All participants gave their in-
`formed consent. Initially, patients aged 65 and older with newly
`diagnosed seizures of any type were randomly assigned to blinded
`treatment with GBP, LTG, or CBZ. After the first year, age at
`entry was lowered to 60 to improve enrollment.
`Clinical evaluations were carried out at enrollment, biweekly
`to week 8, monthly to week 28, and bimonthly to week 52. Pa-
`tients remaining in the study for a second year were evaluated
`every 3 months. Patients continued on the assigned AED until the
`end of the trial, or until they exited the study for any reason.
`Editorial, see page 1834
`
`From VA Medical Center (Dr. Rowan), Bronx, NY; VA Medical Center (Dr. Ramsay, F. Pryor), Miami, FL; VA Medical Center (Dr. Collins), Perry Point, MD;
`VA Cooperative Studies Program (K.D. Boardman), Albuquerque, NM; VA Medical Center (Dr. Uthman), Gainesville, FL; VA Medical Center (Dr. Spitz),
`Denver, CO; VA Medical Center (Dr. Frederick), New Orleans, LA; VA Medical Center (Dr. Towne), Richmond, VA; VA Medical Center (Dr. Carter), Dallas, TX; VA
`Medical Center (Dr. Marks), San Francisco, CA; VA Medical Center (Dr. Felicetta), Phoenix, AZ; and VA Medical Center (Dr. Tomyanovich), Chicago, IL.
`Supported by the Department of Veterans Affairs, Cooperative Studies Program. GSK and Pfizer provided study medications, placebos, and drug plasma
`levels.
`R. Eugene Ramsay has served as a consultant and speaker for Pfizer and GSK. Basim M. Uthman has served as a consultant and speaker for and has
`received research grants from Pfizer, GSK, and Novartis. Mark Spitz has served as a speaker for GSK, Novartis, and Pfizer.
`Received September 15, 2004. Accepted in final form April 14, 2005.
`Address correspondence and reprint requests to Dr. A. James Rowan, Neurology Service (127), Bronx VA Medical Center, 130 West Kingsbridge Road, Bronx,
`NY 10468; e-mail: aj.rowan@med.va.gov or a.james.rowan@mssm.edu
`
`1868 Copyright © 2005 by AAN Enterprises, Inc.
`
`ARGENTUM Exhibit 1092
` Argentum Pharmaceuticals LLC v. Research Corporation Technologies, Inc.
`IPR2016-00204
`
`Page 00001
`
`

`
`Entry criteria. Eligible patients were newly diagnosed with
`epileptic seizures and were untreated, treated only acutely (⬍4
`weeks), or treated but with subtherapeutic levels. A minimum of
`one seizure during the 3 months preceding enrollment was re-
`quired. No restriction regarding concomitant diseases was im-
`posed excepting those conditions likely to lead to a life expectancy
`of less than 12 months, progressive neurologic disease, or condi-
`tions that would significantly affect the response to treatment. All
`comedications were allowed save chronic AEDs. Additional exclu-
`sion criteria included those with severe psychiatric conditions,
`current alcoholism, illicit drug use, or a history of noncompliance.
`Study design. This was a randomized, double-blind, parallel
`trial comparing three monotherapy treatments: GBP (target dose:
`1,500 mg/day), LTG (target dose: 150 mg/day), and CBZ (target
`dose: 600 mg/day). Subjects received two dosage forms labeled
`alpha (tablets) and beta (capsules). Alpha was LTG 25 mg or its
`matching placebo given twice daily; beta was GBP 300 mg, over-
`encapsulated CBZ 200 mg tablets, or placebo given three times
`daily. All patients were randomized to receive one active and one
`placebo formulation with equal numbers entered into each treat-
`ment arm. GBP was started at 300 mg/day and increased by 300
`mg/day every 3 days to the target of 1,500 mg/day. LTG was
`titrated at 25 mg/day for 2 weeks, 50 mg/day for 2 weeks, 100
`mg/day for 1 week, followed by 150 mg/day. CBZ was titrated by
`200 mg every 2 weeks to 600 mg/day. It is emphasized that target
`doses were estimates of effective, well-tolerated doses in this pop-
`ulation. They were not intended to be fixed throughout the study.
`Incremental increases above target were allowed at any time if
`seizure control was inadequate. Similarly, incremental decreases
`were allowed if the patient experienced toxicity. AEDs being taken
`at enrollment were tapered to zero during titration of study drug.
`At enrollment 239 patients (40.3%) were taking enzyme-inducing
`AEDs, 219 (36.9%) of whom were taking phenytoin. A total of 340
`(57.3%) were taking no AEDs. If a patient experienced seizures
`during titration, short-term treatment with a benzodiazepine was
`permitted. Seizures were individually recorded by date, time of
`occurrence, and type: simple partial (SPS), complex partial (CPS),
`generalized tonic-clonic (GTC), GTC and partial, and mixed par-
`tial. Incremental increases above target dose to toxicity were allowed
`if seizure control was inadequate. The blind was maintained by hav-
`ing patients simultaneously increase or decrease both capsules
`and tablets (one active, one placebo). If toxicity coexisted with
`inadequate seizure control, the patient exited the study. Compli-
`ance was monitored with pill counts at each visit.
`Randomization was done separately for each site using varying
`block sizes. To randomize a patient, the nurse coordinator/site
`investigator telephoned the study’s Data Coordinating Center
`(DCC) where a staff member assigned a nonconsecutive, site-
`specific patient number from a computer generated randomization
`list. This patient number corresponded to a patient drug kit in the
`site’s pharmacy. A prescription for the specified drug kit was then
`prepared for the patient, who usually took the first dose of study
`medication on the day of randomization.
`The primary outcome measure was retention in the trial for 12
`months, a measure of both efficacy and tolerability. Decisions
`concerning retention in the study rested on the clinical judgment
`of the principal investigator in concert with a patient-investigator
`discussion. Those retained for 12 months were considered success-
`ful completers.
`Secondary endpoints included seizure freedom at 12 months,
`time to first seizure, and drug toxicity.
`Statistical analysis. The trial’s proposed original sample size
`of 720 patients was based on being able to detect a 15-percentage
`point difference among the treatment groups on the primary out-
`come measure of retention at 12 months assuming an estimate of
`65% retention for the standard drug, CBZ. This estimate was
`based on two previous multicenter trials of AEDs.11,12 Power of
`0.90 and a two-sided test were assumed. A significance level of
`0.0167 was used to account for the three possible treatment com-
`parisons (CBZ vs GBP, CBZ vs LTG, GBP vs LTG).
`␹2 analysis was used for the overall comparison of the three
`treatment groups for the primary outcome measure. Fisher’s exact
`tests were used to analyze the paired comparisons. Three patients
`who were terminated solely because a participating center was
`closed are not included in the retention analyses, but are included
`where appropriate in all other analyses. ␹2 techniques were used
`to compare discrete baseline and secondary outcome measures,
`
`Figure 1. Flow diagram of patients’ progress through the
`study.
`
`while analysis of variance techniques were used to compare
`continuous baseline and secondary outcome measures. Time to
`seizures and time to early termination were analyzed using
`Kaplan-Meier curves and log rank statistics. Paired comparisons
`were only performed if the overall test was p ⱕ 0.05 and were
`considered significant only if p ⱕ 0.0167. p Values between 0.0167
`and 0.05 were considered trends.
`
`Results. Patient demographics. Of 1,358 patients
`screened for possible enrollment, 593 met inclusion criteria
`(figure 1). Enrollment continued from January 1998 to
`April 2002. All patients were eligible to remain in the trial
`for 12 months with the option of continuing an additional
`12 months. Final follow-up visits occurred in April 2003.
`Patients were randomly and equally assigned to one of
`three treatment groups: GBP (n ⫽ 195), LTG (n ⫽ 200),
`and CBZ (n ⫽ 198).
`The main reasons for exclusion (not mutually exclusive)
`were under minimum age (n ⫽ 97), no seizure during pre-
`ceding 3 months (n ⫽ 187), satisfied with current treat-
`ment (n ⫽ 147), unstable medical condition (n ⫽ 162),
`questionable compliance (n ⫽ 131), unwilling to enter
`study (n ⫽ 131), and unable to give consent (n ⫽ 143).
`The most common seizure type was CPS (251/581;
`43.2%), followed by GTC, SPS, GTC and partial, and mixed
`partial (see table E-1 on the Neurology Web site at www.
`neurology.org). There were no significant differences
`across treatment groups. Similarly, primary etiology did
`not differ significantly across groups, the most common of
`which was cerebral infarction (177/592; 29.9%), followed by
`arteriosclerosis (93/592; 15.7%) and head trauma (42/592;
`7.1%, see table E-1). Unknown causes accounted for 24.0%
`(142/592) with all other causes at less than 2% each.
`The majority of patients had vascular disease as evi-
`denced by concurrent medical problems such as hyperten-
`sion (391/593; 65.9%), stroke (302/593; 50.9%), and cardiac
`
`June (1 of 2) 2005 NEUROLOGY 64 1869
`
`Page 00002
`
`

`
`Figure 2. Percentage of patients remaining in the trial
`over time (52 weeks).
`
`Figure 3. Percentage of patients remaining in the trial
`over time (6-week titration time).
`
`disease (286/593; 48.2%). At baseline, mild cognitive im-
`pairment was found in 35.0% (207/592), and neurologic
`findings included gait disturbances (312/593; 52.6%), ab-
`normal sensory examination (183/593; 30.9%), memory
`problems (153/593; 25.8%), abnormal station (140/593;
`23.6%), diminished motor power (132/593; 22.3%), and ab-
`normal coordination (86/593; 14.5%).
`Study drug doses and serum levels. Total daily doses
`of the study drugs approached target doses. At 6 weeks
`mean dosages were GBP 1,424 ⫾ 285 mg/day, CBZ 558 ⫾
`144 mg/day, and LTG 131 ⫾ 34 mg/day. At 52 weeks, mean
`dosages were GBP 1,422 ⫾ 288 mg/day, CBZ 582 ⫾ 218
`mg/day, and LTG 152 ⫾ 33 mg/day.
`Mean serum levels at 6 weeks were GBP 8.67 ⫾ 4.83
`␮g/mL, CBZ 6.79 ⫾ 2.92 ␮g/mL (unbound 1.0 ⫾ 0.45), and
`LTG 2.87 ⫾ 1.60 ␮g/mL. At 52 weeks, mean serum levels
`were GBP 8.54 ⫾ 5.57 ␮g/mL, CBZ 6.48 ⫾ 3.72 ␮g/mL
`(unbound 0.81 ⫾ 0.43), and LTG 3.46 ⫾ 1.68 ␮g/mL.
`Dosage reductions for side effects occurred in 31.3%
`(171/547) while dosage increases above target for inade-
`quate seizure control occurred in 21.4% (117/547). Dosage
`increases above target occurred more often in patients re-
`ceiving LTG as compared to CBZ (27.1% [51/188] vs 14.0%
`[25/179], p ⫽ 0.002). Overall, medication compliance was
`89% without significant group differences.
`Outcome measures. Of 590 patients enrolled and not
`administratively terminated 276 (46.8%) completed 1 year
`in trial. The overall three-group comparison was signifi-
`cant (p ⫽ 0.00022). In paired-group comparisons, CBZ had
`more early terminators than either GBP (p ⫽ 0.008) or
`LTG (p ⬍ 0.0001). Reasons for early termination are listed
`in table E-2. Fewer LTG patients terminated for adverse
`reactions than either CBZ (p ⬍ 0.0001) or GBP (p ⫽ 0.015)
`patients. Relatively few patients exited the study due pri-
`marily or in part to uncontrolled seizures, with no differ-
`ences among treatment groups. Times to early termination
`before 12 months and before 2 months are shown in the
`Kaplan-Meier curves (figures 2 and 3).
`Between weeks 4 and 5 (see figure 3) the groups began
`to separate with better retention for LTG. When early
`terminations for adverse events occurring during the
`6-week titration phase were considered, there were fewer
`LTG patients terminating for adverse events (8/199; 4.0%)
`than for either CBZ (41/199; 20.8%; p ⬍ 0.0001) or GBP
`(29/194; 14.9%; p ⫽ 0.0002).
`1870 NEUROLOGY 64 June (1 of 2) 2005
`
`Early terminators had a mean age of 73.0 (SE ⫽ 0.42)
`while completers had a mean age of 71.5 (SE ⫽ 0.45) (p ⫽
`0.0193). With respect to serum levels of early terminators,
`the last available values before termination were LTG (n ⫽
`60) mean ⫽ 2.67 ␮g/mL, SD 2.29 ␮g/mL, 95% CI 2.08 to
`3.26 ␮g/mL; GBP (n ⫽ 69) mean ⫽ 10.14 ␮g/mL, SD 9.45
`␮g/mL, 95% CI 7.87 to 12.41 ␮g/mL; CBZ (n ⫽ 85) mean ⫽
`4.95 ␮g/mL, SD 3.44 ␮g/mL, 95% CI 4.21 to 5.69 ␮g/mL;
`free CBZ (n ⫽ 85) mean ⫽ 0.69 ␮g/mL, SD 0.52 ␮g/mL,
`95% CI 0.58 to 0.80 ␮g/mL.
`Efficacy. Seizure freedom, a secondary outcome mea-
`sure, was analyzed at 3, 6, and 12 months after start of
`treatment. Patients remaining in the study dropped to 402
`at 3 months (LTG 157, GBP135, CBZ 110), to 333 at 6
`months (LTG 132, GBP 113, CBZ 88), and 276 at 12
`months (LTG 111, GBP 95, CBZ 70). Of those remaining in
`the study for 3, 6, and 12 months, the seizure-free rates
`were 63.2% at 3 months (LTG 63.1%, GBP 62.2%, CBZ
`64.5%), 58.6% at 6 months (LTG 56.6%, GBP 56.6%, CBZ
`64.8%), and 53.3% at 12 months (LTG 51.4%, GBP 47.4%,
`CBZ 64.3%). There were no noteworthy group differences
`(overall p values: 0.93 at 3 months, 0.39 at 6 months, and
`0.09 at 12 months). When seizures occurring during the
`6-week titration phase were excluded, seizure-free rates
`increased to 80.1% at 3 months (LTG 80.3%, GBP 80.0%,
`CBZ 80.0%), to 70.6% at 6 months (LTG 68.2%, GBP
`71.7%, CBZ 72.7%), and to 63.4% at 12 months (LTG
`61.3%, GBP 60.0%, CBZ 71.4%). Again, there were no sig-
`nificant differences (overall p values: 1.00 at 3 months,
`0.73 at 6 months, and 0.27 at 12 months).
`We considered time to first, second, fifth, and tenth
`seizure during the first year. A total of 233 patients had at
`least one, 182 at least two, 101 at least five, and 54 at least
`10 seizures. The log rank statistics for overall group com-
`parisons were not different for any of these time to sei-
`zure(s) analyses (p ⫽ 0.18, 0.13, 0.74 and 0.95). Figure 4
`shows the graphic for time to first seizure. When seizures
`during the 6-week titration period were excluded, these
`analyses were still not different (p ⫽ 0.39, 0.19, 0.11, and
`0.34). Here, however, LTG patients tended to do worse for
`time to first and second seizure, and GBP patients tended
`to do better for time to fifth seizure than the other groups.
`Again, the results were not significant.
`A third measure of efficacy that was considered was
`seizure-free retention. These analyses included all patients
`
`Page 00003
`
`

`
`link between drug and cause of death. None of the deaths
`was determined to be clearly due to study drug. One pa-
`tient died 2 weeks after stopping study drug due to a
`probable hypersensitivity reaction that led to multiple sys-
`tem organ failure. This patient was in the CBZ arm and
`had received phenytoin for 1 week before enrollment, thus
`obscuring the proximate cause. Other causes of death
`ranged from cardiac and pulmonary disease to sepsis and
`cancer.
`
`Discussion. This multicenter clinical trial of sei-
`zures in an older population is the largest to date.
`When the protocol was designed, two newly approved
`AEDs (GBP and LTG) appeared to offer advantages
`over the standard AEDs, particularly with respect to
`their pharmacokinetic and side effect profiles.13
`There was a need to compare the new drugs with a
`standard AED. Phenytoin, valproic acid, and CBZ
`were considered, and CBZ was selected due to its
`worldwide acceptance as a treatment of choice for
`partial onset epilepsy.
`Hepatic and renal function decline with age; thus,
`lower total daily doses are usually suggested for
`older adults.14 Hence, selection of appropriate target
`doses and titration schedules were major issues and
`discussed in depth with our outside advisory commit-
`tee. We also consulted the then limited available lit-
`erature and the relevant pharmaceutical companies.
`The selected target doses were generally lower than
`what might be considered standard doses for younger
`adults, recognizing that the protocol contained built-in
`mechanisms for both decreasing and increasing the
`doses at any time as clinically indicated. Further, the
`titration schedules were slower than usually employed
`in clinical practice. We found that final dosages of the
`study drugs were similar to the target doses, and the
`range of serum levels for the three drugs remained low
`to moderate and relatively stable throughout the trial.
`The patients were newly diagnosed with epilepsy
`and treated with AED monotherapy, circumventing
`complications associated with add-on and cross-over
`trials. Concurrent medical diseases were allowed to
`ensure that the study would reflect medical realities
`of the elderly population. Because of the high recur-
`rence rate in the aged after a first seizure (66% to
`90%),15-19 the potential consequences of recurrent sei-
`zures, and the high incidence of risk factors such as
`cerebrovascular disease, we felt the occurrence of at
`least one seizure during the 3-month window preced-
`ing enrollment was a justified enrollment criterion.
`Complex partial seizures alone were the most com-
`mon seizure type (43.2%). Only 25.3% presented
`with GTCs alone, a lower proportion than reported
`in epidemiologic studies that predominantly include
`younger adults.20 These findings are likely due to a
`different predominate etiology for seizures in older
`patients—namely, vascular disease involving the an-
`terior and middle cerebral arteries. One would there-
`fore expect an increased occurrence of seizures that
`originate in the frontoparietal region.
`The high incidence of seizures in older patients
`June (1 of 2) 2005 NEUROLOGY 64 1871
`
`Figure 4. Percentage of patients remaining seizure-free
`over time (time to first seizure).
`
`at all rating periods and counted early terminators as if
`they had had seizures. As with the seizure-free rate analy-
`ses, this variable was analyzed at 3, 6, and 12 months, and
`was done using 1) all seizures and 2) excluding seizures
`occurring during the 6-week titration period. For all sei-
`zures the seizure-free retention rates were 43.1% at 3
`months (LTG 49.7%, GBP 43.3%, CBZ 36.0%), 33.1% at 6
`months (LTG 37.2%, GBP 33.0%, CBZ 28.9%), and 24.9%
`at 12 months (LTG 28.6%, GBP 23.2%, CBZ 22.8%). A
`significant difference was seen only at 3 months (overall p
`values: 0.02 at 3 months, 0.22 at 6 months, and 0.33 at 12
`months) with the LTG group doing significantly better
`than the CBZ group (p ⫽ 0.006). When seizures occurring
`during the 6-week titration period are excluded, LTG pa-
`tients (63.3%) again did better at 3 months (p ⫽ 0.001)
`than CBZ patients (44.7%) with GBP patients in between
`(55.7%). Similarly, at 6 months LTG patients (45.2%) also
`had better seizure-free retention (p ⫽ 0.009) than did CBZ
`patients (32.5%). The differences at 12 months were not
`significant (overall p value ⫽ 0.16).
`Adverse reactions. Table E-3 reports the systemic and
`neurotoxicities that occurred during the first 12 months for
`those patients having at least one submitted follow-up
`form. Significantly more patients on GBP had weight gain
`during the first 12 months than either those on CBZ (p ⫽
`0.002) or LTG (p ⫽ 0.001). More patients on GBP had large
`weight gain (⬎18 pounds) than those on CBZ (p ⫽ 0.005)
`or LTG (p ⫽ 0.014). Water retention was significantly
`greater with GBP than with CBZ (p ⫽ 0.004) or LTG (p ⫽
`0.02). More patients lost weight with LTG than with GBP
`(p ⫽ 0.002), but the proportion of patients who gained
`(47.5%) or lost weight (36.1%) while on LTG was similar.
`Hypersensitivity (rash of any degree) occurred more fre-
`quently with CBZ than with LTG (p ⫽ 0.007). Of seven
`patients hospitalized for hypersensitivity reaction, six
`were in the CBZ group and one was treated with LTG.
`Hyponatremia (sodium less than 130 mg %) occurred more
`frequently in CBZ than in GBP patients. There were no
`significant differences in other systemic toxicities. Consid-
`ering neurotoxicities, there were no differences among the
`treatment groups over 12 months. Severe neurotoxicities
`were reported in 43 patients (8.1%).
`Thirty-nine deaths occurred during the trial: 15 in the
`CBZ group, 11 GBP, and 8 LTG. There was no clustering of
`causes for death in any of the treatment arms to suggest a
`
`Page 00004
`
`

`
`has not been widely appreciated. We suspect that
`CPS may not be recognized due to their subtle or
`different clinical presentations. For example, CPS
`may present only with periods of confusion or staring
`for brief periods with little if any motor activity (au-
`tomatisms). Of the 25.3% with GTCs alone, none had
`evidence of primary generalized epilepsy, for exam-
`ple generalized spike-wave discharges in the EEG.
`Our primary outcome measure, retention in trial
`for 12 months, showed a highly significant difference
`with CBZ showing poorer retention than either GBP
`or LTG. The data suggest that this is not due to
`differences in efficacy. Methods used to evaluate effi-
`cacy included 1) percent of patients seizure-free for
`12 months, 2) time to first seizure, and 3) seizure-
`free retention rate. There were no significant differ-
`ences using methods 1, 2, or 3. Of the patients
`remaining in the study for 52 weeks, the seizure free
`rate was highest with CBZ. Methods 1 and 2, how-
`ever, favor a poorly tolerated drug, inasmuch as sei-
`zures could only be counted in patients who
`remained in the study, and patients who withdrew
`might also be the ones likely to have recurrent sei-
`zures. A higher seizure-free retention rate was found
`with LTG using an intent-to-treat analysis (method
`3), a method favoring a well-tolerated drug. In fact,
`at 3 months, there was a significantly better seizure
`free rate for LTG. This difference disappeared at 6
`and 12 months. (See Outcome measures, efficacy.)
`Differences in efficacy were less evident when sei-
`zures occurring during the 6-week titration phase
`were ignored. The primary factor accounting for pa-
`tients remaining in the trial, therefore, appears to be
`the incidence of adverse events and not poor seizure
`control.
`Although the mean age of early terminators was
`greater than that of completers (73 vs 71.5, p ⫽
`0.0193), we do not believe this difference materially
`affected the results of the trial. With respect to se-
`rum levels at or close to the time of termination, we
`found that the levels of LTG and CBZ were slightly
`less than those seen during the maintenance phase
`of the study (after 6 weeks). On the other hand, GBP
`at termination was above those levels obtained dur-
`ing the course of the trial. The lower termination
`levels of CBZ (mean 4.95 ␮g/mL) and free CBZ
`(mean 0.69 ␮g/mL) than those found during mainte-
`nance add support to the thesis that the choice of
`initial dose and titration schedule for this drug was
`conservative and thus not a sufficient reason for its
`intolerability. The opposite may be true for GBP
`where higher termination doses were found than
`those obtained during maintenance.
`Because our study population was made up pre-
`dominantly of men, it is possible that these results
`may not be generalizable to the population at large.
`It is noted, however, that seizure occurrence in
`younger women is influenced by hormonal fluctua-
`tions—not a factor in older patients. Thus, we postu-
`late that the results of this study should be broadly
`applicable.
`1872 NEUROLOGY 64 June (1 of 2) 2005
`
`The protocol allowed for increasing the dose be-
`yond target if seizures were not controlled and de-
`creasing the dose to reduce/eliminate side effects.
`During the first 12 months, dosage reductions varied
`from 30.7% for CBZ to 32.2% for GBP. During the
`titration phase, dosage reduction ranged from 11.7%
`for LTG to 16.2% for CBZ. Had the initial dose for
`any of the treatments been too high, a greater drop-
`out rate would have been expected for that drug
`during the first 2 weeks, recalling that the dose of
`CBZ was not changed during that time, and that its
`titration schedule was slower than usually employed
`in clinical practice. In fact, retention was similar in
`the three treatment arms up to the end of the third
`week (see figure 3), with differences becoming evi-
`dent during weeks 4 to 6. Thus, the dosing schedules
`did not significantly alter the outcome of the study.
`Would use of an extended release preparation of
`CBZ (ER-CBZ) have altered the results? Because of
`the longer CBZ half-life in the elderly, due largely to
`reduced hepatic metabolism, the peak-trough effect
`is potentially less prominent. This would mitigate a
`possible advantage of an ER-CBZ. Moreover, the rel-
`atively lower CBZ serum levels in this population
`compared to those usually sought in younger adults
`would also reduce the impact of an ER-CBZ. We
`therefore believe that using an ER-CBZ would not
`have led to a significant difference in our results.
`
`Appendix
`Chairmen’s Office: R. Eugene Ramsay, MD, Miami, FL, Study Co-Chair-
`man; A. James Rowan, MD, Bronx, NY, Study Co-Chairman; Flavia Pryor,
`RN, MPH, Miami, FL, National Study Coordinator. Cooperative Studies
`Program Coordinating Center (CSPCC), Perry Point, MD: Joseph F. Col-
`lins, ScD, Director; Susan Stinnett, Linda Linzy, Colleen Crigler, Chuan-
`Shue Lee, Pat Grubb, Beverly Calvert. Cooperative Studies Program
`Clinical Research Pharmacy Coordinating Center (CSPCRPCC), Albuquer-
`que, NM: Mike Sather, PhD, FASHP, Cindy Colling, RPh, MS, Kathy D.
`Boardman, RPh, Jenine Peterson, BS. Data Safety Monitoring Board: John
`Pellock, MD, Dan Berlowitz, MD, Carla Herman, MD, Steven Schachter,
`MD, James Willmore, MD, Nancy Temkin, PhD, Kerry Lee, PhD. Human
`Rights Committee, Cooperative Studies Program Coordinating Center,
`Perry Point, MD: Clint McSherry, PhD, Rev. James Jones, James Crothers,
`Anthony Harris, MD, Lettie Carr, Rose Kurz, Thaddeus Prout, MD, Adele
`M. Gilpin, PhD, JD, Alan Fix, MD, MS, Mary Zorzi. Participating centers
`(VA Cooperative Study #428 Group): Birmingham, AL—R. Edward Faught,
`MD, Diane Wilhite, Cheryl Hall; Boston, MA—Thomas Browne, MD, Bar-
`bara Dworetzky, MD, Menisha Thakore, MD, Sheila Savickis, RN; Bronx,
`NY—Maria Muxfeldt, MD, Martin Gluck, MD, PhD, Helene Price, MD,
`Linda Tuchman, LPN; Chicago, IL—Mary Lou Tomyanovich, MD, Cristina
`Orfei, MD, Rita Shapire, DO, Susan Winkler, PharmD; Dallas, TX—
`Gregory Carter, MD, PhD, Amy Choate, BS; Denver, CO—Mark Spitz, MD,
`Jacci Bainbridge, PharmD; Gainesville, FL—Bassim Uthman, MD, Brenda
`Smith, RN; Hines, IL—Meenal Mamdani, MD, Sudha Gupta, MD, Katar-
`zyna Olejniczak; Miami, FL—John DeToledo, MD, Juanita Johnson, BS;
`New Orleans, LA—Tim Frederick, MD, Kathryn LaRussa, RN; Oklahoma
`City, OK—Peggy Wisdom, MD, K.J. Oommen, MD, Terry Rogers-Neame,
`MD, Richard Dasheiff, MD, Neil Holland, MD, Kersi Bharucha, MD, Mar-
`sha DeWitt, RN; Phoenix, AZ—Richard Matthews, MD, Jaswant Sachdev,
`MD, Halina Roznowski; Pittsburgh, PA—Anne Van Cott, MD, Regina Fen-
`ton, MSN; Portland, OR—Martin D. Salinsky, MD, Debbie Johnstone, RN;
`Richmond, VA—Alan Towne, MD, Heather Shebelski, RN; San Diego, CA—
`Vincent Iragui, MD, Karen Wetzel, PA; San Francisco, CA—William Marks,
`MD, Elaine Lanier, RN; San Antonio, TX—Jose Cavazos, MD, Laura
`Moreno, RN.
`
`References
`1. Hauser WA, Annegers JF, Kurland LT. Incidence of epilepsy and un-
`provoked seizures in Rochester, Minnesota 1935–1984. Epilepsia 1993;
`34:453–458.
`2. Hauser WA. Epidemiology of seizures and epilepsy in the elderly. In:
`Rowan AJ, Ramsay RE, eds. Seizures and epilepsy in the elderly. New
`York: Butterworth-Heinemann, 1997;7–20.
`
`Page 00005
`
`

`
`3. Vestal RE, Cusack BJ. Pharmacology and aging. In: Schneider EL,
`Rowe JW, eds. Handbook of the biology of aging. San Diego: Academic
`Press, 1990;349–383.
`4. Rowe JW, Andres RA, Tobin JD, Norris AH, Shock NW. The effect of
`age on creatinine clearance in man: a cross sectional and longitudinal
`study. J Gerontol 1976;31:155–163.
`5. Woodhouse KW, Wynne HA. Age-related changes in liver size and he-
`patic blood flow. The influence of drug metabolism in the elderly. Clin
`Pharmacokinet 1988;15:287–294.
`6. Salzman C, Shader RI, Greenblatt DJ, Harmatz JS. Long versus short
`half-life benzodiazepines in the elderly. Kinetics and clinical effects of
`diazepam and oxazepam. Arch Gen Psychiatry 1983;40:293–297.
`7. Feely J, Coakley D. Altered pharmacodynamics in the elderly. Clin
`Geriat Med 1990;6:269–283.
`8. The US Gabapentin Study Group. The long-term safety and efficacy of
`gabapentin (Neurontin) as add-on therapy in drug-resistant partial ep-
`ilepsy. Epilepsy Res 1994;18:67–73.
`9. Goa KL, Ross SR, Chrisp P. Lamotrigine: a review of its pharmacologi-
`cal properties and clinical efficacy in epilepsy. Drugs 1993;46:152–176.
`10. Brodie MJ, Overstall PW, Giorgi L. Multicentre, double-blind, random-
`ized comparison between lamotrigine and carbamazepine in elderly
`patients with newly diagnosed epilepsy. The UK Lamotrigine Elderly
`Study Group. Epilepsy Res 1999;37:81–87.
`11. Mattson RH, Cramer JA, Collins JF, et al. Comparison of carbamaz-
`epine, phenobarbital, phenytoin, and primidone in partial and sec-
`ondarily generalized tonic-clonic seizures. N Engl J Med 1985;313:
`145–151.
`
`12. Mattson RH, Cramer JA, Collins JF, DVA Cooperative Study No. 264
`Group. A comparison of valproate with carbamazepine for the treat-
`ment of complex partial seizures and secondarily generalized tonic-
`clonic seizures in adults. N Engl J Med 1992;327:765–771.
`13. Ramsay RE, Pryor FM. Gabapentin: adverse effects. In: Levy RH, Matt-
`son RH, Meldrum BS, Perucca E, eds. Antiepileptic drugs. Philadel-
`phia: Lippincott Williams & Wilkins, 2002;354–359.
`14. Leppik IE, Cloyd JC. Epilepsy in the elderly. In: Levy RH, Mattson RH,
`Meldrum BS, Perucca E, eds. Antiepileptic drugs. Philadelphia: Lippin-
`cott Williams & Wilkins, 2002;149–158.
`15. So EL, Annegers JF, Hauser WA, O’Brien PC, Whisnant JP.
`Populatio