CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 70 NUMBER 7 JULY 2003 649

REDEEPAK LACHHWANI, MB, BS, MDDepartment of Neurology, The Cleveland Clinic

HANS LDERS, MD, PhDChairman, Department of Neurology, The Cleveland Clinic

In refractory temporal lobeepilepsy, consider surgery sooner

INTERPRETING KEY TRIALS

n ABSTRACTDespite evidence that surgery fortemporal lobe epilepsy is safe andeffective, physicians continue to view it asa last resort in people who do notrespond to medical therapy. A randomizedstudy in Canada has demonstratedsignificantly higher rates of freedom fromseizures in patients who underwentsurgery rather than medical therapy. If wewait too long to rule a patients epilepsymedically refractory, we increase thepatients risk of morbidity and death, andwe miss a window of opportunity toeliminate seizures and permit the patienta full and productive life.

PILEPSY SURGERY is perhaps the mostunderused therapy in all of medicine

today.More than 2 million people in the United

States have epilepsyas many women ashave breast cancer and as many men as havelung cancer. Of these 2 million people,400,000 to 600,000 have seizures that areresistant to drug therapy,1 even with the newantiepileptic medications, and as many as25% to 50% of those with drug-resistantepilepsy100,000 to 300,000 patientsmaybe candidates for epilepsy surgery (ie, theyhave a small, discrete, area of epileptogenicitythat can be safely removed).

Yet a 1990 survey found that only 1,500surgical procedures for epilepsy were per-formed in the United States that year.2 Thesituation is the same in other industrializedcountries and worse in developing coun-tries.3,4

Epilepsy affects people of all ages, races,and incomes and impairs their lives substan-tially.1,5,6 Uncontrolled seizures carry a signif-icant burden of disease and death.7,8

And epilepsy surgery is effective: the curerates at most centers are 70% to 90% inpatients with surgically remediable (but med-ically refractory) epilepsy syndromes.9

Why then do physicians and patients shyaway from surgery, even when drug therapy withmultiple drugs is unsuccessful?10 Surgery isexpensive and invasive and has an inherent riskof complications. More important, until recent-ly, there has been no randomized, controlledtrial comparing surgery with drug therapy.11

Such a study has been ethically difficult todesign: in view of the high cure rates withsurgery it would be unethical to deny a clear-ly superior treatment to a control group.

Now, however, such a randomized con-trolled trial has been performed and unequiv-ocally proves the superiority of surgery formedically refractory temporal lobe epilepsy.12

It shows that even though surgery carriesthe risk of complications, avoiding or delayingsurgery in patients with medically intractableepilepsy is also fraught with grave risks.

n EPILEPSY SURGERY STUDY DESIGN

Wiebe and colleagues in London, Ontario,were able to conduct an ethical randomizedtrial of surgery vs medical therapy (FIGURE 1) bytaking advantage of what some may consider aweakness of the Canadian health care system: along (1-year) waiting list for surgery.12

Study populationPatients age 16 years or older with temporallobe epilepsy were informed about the ratio-

E

Epilepsysurgery maybe the mostunderusedtherapy inmedicine

CREDITCME

on April 30, 2014. For personal use only. All other uses require permission.www.ccjm.orgDownloaded from

650 CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 70 NUMBER 7 JULY 2003

nales for medical and surgical therapy, as wellas about the purpose of the study. They werealso informed that randomization meant theyhad a 50% chance of being assigned to receivemedical treatment. Written informed consentwas obtained from those enrolled.

Patients in the medical group wereassigned to the usual 1-year waiting period, atthe end of which they were admitted for apresurgical evaluation and offered surgerywithin 4 weeks if they were eligible.

The patients in the surgical group under-went presurgical evaluation within 48 hours ofrandomization; then, if eligible, they under-went surgery within 4 weeks.

Inclusion and exclusion criteriaPatients were eligible only if they had hadseizures with strong temporal lobe semiologyfor more than 1 year that were refractory totwo or more anticonvulsant drugs.

Patients were not eligible for the study ifthey had acute brain lesions requiring urgent

surgery, progressive central nervous systemdisease, previous epilepsy surgery, or evidenceof extratemporal disease on electrophysiolog-ic or magnetic resonance imaging (MRI)studies.

Evaluation and treatmentBefore surgery, patients were admitted to anepilepsy monitoring unit, where their seizureswere characterized and their electroen-cephalographic (EEG) data were analyzed. Ifthe surface EEG data were inadequate tolocate the source of the seizure, invasive EEGmonitoring was carried out. Memory functionwas assessed by intracarotid amobarbital test-ing.

Those with seizures originating in onetemporal lobe and who had concordant MRIand neuropsychological data underwent ante-rior temporal lobe resection within 4 weeksafter randomization. They then received opti-mal medical therapy for 1 year.

Two epileptologists who were blinded to

EPILEPSY LACHHWANI AND LDERS

Surgical group (n=40) Medical group (n=40)

Presurgical evaluationand surgery within 4 weeks

Optimal medical therapyFollow-up months 3, 6, 9, 12

Primary outcome: free from seizures impairingawarenessSecondary outcome: free from all seizures,improved quality of life

Enrollmentand

randomization

FIGURE 1. Schematic design of the Canadian epilepsy surgery trial.12

on April 30, 2014. For personal use only. All other uses require permission.www.ccjm.orgDownloaded from

the patients identities reviewed the adequacyof medical therapy and the status of eachpatients seizures at each quarterly follow-upvisit by reviewing the written clinical infor-mation from the patients clinic visit. Aftersurgery, the antiepileptic drugs were contin-ued unchanged in therapeutic doses for atleast 9 months, even if the patient was seizure-free.

Patients in the medical group were placedon the 1-year waiting list for admission to theepilepsy monitoring unit, according to themedical institutions policies. Both treatmentgroups were seen every 3 months by epileptol-ogists who tailored each patients antiseizuredrug therapy as needed.

Outcomes measuredThe primary outcome was freedom fromseizures impairing awareness at 1 year. Thetrial was designed to include enough patients(40 in each treatment group) to detect anabsolute difference of 34% in this outcomebetween the groups, with 90% power at a two-sided significance level of 0.05%.

Quality of life after surgery was assessedwith the following: Liverpool seizure severity scale The Quality of Life in Epilepsy inventory General health questionnaire Center of Epidemiologic Studies depres-sion scale.

Employment status and school atten-dance were also measured.

Data were analyzed after accounting forany imbalances in demographic and clinicalcharacteristics of the patients that might haveaffected the statistical significance of the dif-ference between the two groups.

n RESULTS:A MARKED REDUCTION IN SEIZURES

Of 92 patients screened, 86 were eligible, and80 agreed to participate. Forty patients wereassigned to each treatment group. At baseline,the groups were well matched by demograph-ic and clinical characteristics.

In the surgical group, 6 patients (15%)needed invasive monitoring with subduralgrids, and 4 (10%) did not undergo surgery.One declined surgery, 2 were not found to be

good candidates at the completion of theirevaluation, and 1 did not have any seizuresduring the evaluations conducted in the hos-pital.

Twenty-four patients had operations onthe left side and 12 on the right side. Twopatients had a surgical procedure that differedslightly from the procedure described in theprotocol. One underwent a selective amyg-dalohippocampectomy due to concerns aboutspeech and memory in the dominant hemi-sphere,12 and 1 needed a more extensive tem-poral resection.

Of 40 patients treated surgically, 4 hadsurgical complications: 1 had a small thalamicinfarct resulting in a sensory deficit in the leftthigh, 1 had a wound infection, and 2 had adecline in memory that interfered with theiroccupations at 1 year. As was expected,because resection involved some fibers thatserve visual field function, 22 patients in thesurgical group had an asymptomatic superiorsubquadrantic visual field deficit. This deficitwas less than the size of a quadrant and, sincepatients compensated for it very well, it wasnot considered clinically important.

Rates of depression were 18% to 20% inboth treatment groups.

In the medical group. One patient in themedical group died suddenly of an unknowncause 7.5 months into the study. No patientswere lost to follow-up, and no patients crossedover from the medical group to the surgicalgroup. All patients in the medical group need-ed an adjustment or a change in theirantiepileptic drugs vs only 22% in the surgicalgroup.

Number needed to treat. In the surgicalgroup, 58% were free of seizures impairingawareness at 1 year vs 8% in the medicalgroup (P < .001); 38% in the surgical groupwere free of all seizures including auras vs only3% in the medical group. Therefore, twopatients need to undergo surgery to render oneadditional patient free of seizures impairingawareness at the end of 1 year, and threepatients need to undergo surgery to render oneadditional patient free of all seizures at 1year.12 The severity of persisting seizures wassimilar in both groups.

Quality of life improved in both groups,but at 1 year it was statistically significantly

CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 70 NUMBER 7 JULY 2003 651

Three patientsneed to betreatedto render oneadditionalpatient seizure-free at 1 year

on April 30, 2014. For personal use only. All other uses require permission.www.ccjm.orgDownloaded from

652 CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 70 NUMBER 7 JULY 2003

better in the surgical group than in the med-ical group. More patients in the surgical groupwere employed or attending school at 1 year,but the difference did not reach statistical sig-nificance.

n TAKE-HOME POINTSFROM THE CANADIAN STUDY

Wiebe et al12 showed that at 1 year, patientswho underwent surgery for temporal lobeepilepsy were more likely to be free of

seizures than those who underwent medicaltherapy (FIGURE 2, FIGURE 3), and their quality oflife was better.

Shortcomings of the studySeizure-free rate in surgical patients was

comparatively low. Compared with reports of70% of surgical patients free of seizures,13,14the rate of 58% reported by Wiebe et al mayseem low; however, the surgery group includedseveral patients with complicated epilepsy (sixneeded invasive monitoring studies), whichmay explain the lower rate of seizure-free out-come. The lower seizure-free rate might alsobe due to the fact that four of those random-ized to the surgical group did not, in fact,undergo surgery.

Follow-up was limited to 1 year, which isshort for demonstrating the beneficial effectsof successful surgery on quality of life andsocial functioning.15 Even so, the significantlyhigher scores on quantitative measure of thequality of life at the end of 1 year, and thetrend toward higher rates of employment andschool attendance, are meaningful.

Why the Canadian study is important:The no-surgery approach is also riskyThe Canadian randomized, controlled trial isa milestone in several ways: It confirms the effectiveness of surgery in

0 128 10642

100

80

60

40

20

0

Surgical group (n=40)

Surgical group (n=40)

Medical group (n=40)

Medical group (n=40)

58%

8%

38%

3%0 128 106

Months

42

100

80

60

40

20

0

Perc

enta

ge

wit

ho

ut

seiz

ure

sim

pai

rin

g a

war

enes

sPe

rcen

tag

e w

ith

ou

tan

y se

izu

res

Seizure free

8

58

Seizure free(including auras)

70

50

60

40

30

10

20

0

Perc

enta

ge

of

pat

ien

ts

3

38

MedicalSurgical

FIGURE 2. Kaplan-Meier event-free survival curvescomparing the cumulative percentages of patients inthe two groups who were free of seizures impairingawareness (top) and free of all seizures (bottom). Inboth analyses, more patients in the surgical groupwere free of seizures (P < .001).

FROM WIEBE S, BLUME WT, GIRVIN JP, ELIASZIW M. A RANDOMIZED, CONTROLLED TRIAL OFSURGERY FOR TEMPORAL-LOBE EPILEPSY. N ENGL J MED 2001; 345:311318.

FIGURE 3. Comparison of patients in thesurgical and medical groups who werefree of seizures at 1 year.

DATA FROM WIEBE S, BLUME WT, GIRVIN JP, ELIASZIW M. ARANDOMIZED, CONTROLLED TRIAL OF SURGERY FOR TEMPORAL-LOBE

EPILEPSY. N ENGL J MED 2001; 345:311318.

EPILEPSY LACHHWANI AND LDERS

Trial data: Surgery beats medical therapy for temporal lobe epilepsy

on April 30, 2014. For personal use only. All other uses require permission.www.ccjm.orgDownloaded from

appropriately selected patients with temporallobe epilepsy It indicates that a no-surgery approachin patients with medically intractable epilepsyis fraught with risks, the most serious onebeing the risk of death (sudden unexplaineddeath in epilepsy [SUDEP] or death due toother reasons), as seen with one of thepatients in the medical arm.

A narrow window of opportunity:Surgical evaluation is not a last resortThe study results should prompt physicians toconsider surgery not as a last resort, but as apreferred option in patients whose seizures areuntreatable with drugs.

There is mounting evidence that drug-resis-tance can be successfully predicted after only

one or two carefully chosen drugs have provenineffective.16,17 In a study by Kwan and Brodie,1647% of patients became free of seizures aftertreatment with the first antiepileptic drug; how-ever, of those who did not become free of seizureswith the first drug, only 11% became free ofseizures with medical management.

We have evidence to suggest that years ofuncontrolled epilepsy lead to cognitivedecline.18 Earlier surgical intervention toeliminate seizures would offer a window ofopportunity to allow patients a full and pro-ductive life, rather than years of ineffectivepharmacotherapy with its consequent toll oftime wasted and perhaps an opportunity lostforever. Therefore, consideration for referringa patient for surgical evaluation should bemade soon after failure of two drugs.

Opt for surgeryearly to takeadvantage ofnarrow windowof opportunity

n REFERENCES1. Hauser WA, Hesdorffer DC. Incidence and prevalence. In:

Hauser WA, Hesdorffer DC, editors. Epilepsy: Frequency,Causes, and Consequences. New York: Demos, 1990: 151.

2. Engel J Jr, Shewmon DA. Who should be considered asurgical candidate? In: Engel J Jr, editor. SurgicalTreatment of the Epilepsies. 2nd ed. New York: RavenPress, 1993: 2334.

3. Wieser HG, Silfvenius H. Overview: epilepsy surgery indeveloping countries. Epilepsia 2000; 41(suppl 4):S3S9.

4. Murray CJ, Lopez AD, Jamison DT. The global burden ofdisease in 1990: summary results, sensitivity analysis andfuture directions. Bull World Health Organ 1994;72:495509.

5. Wiebe S, Bellhouse DR, Fallahay C, Eliasziw M. Burden ofepilepsy: the Ontario Health Survey. Can J Neurol Sci1999; 26:263270.

6. Kale R. Bringing epilepsy out of the shadows. BMJ 1997;315:23.

7. Sperling MR, Feldman H, Kinman J, Liporace JD,OConnor MJ. Seizure control and mortality in epilepsy.Ann Neurol 1999; 46:4550.

8. Pilcher WH, Roberts DW. Complications of epilepsysurgery. In: Engel J Jr, editor. Surgical Treatment of theEpilepsies. 2nd ed. New York: Raven Press, 1993: 565581.

9. Engel J Jr. Surgery for seizures. N Engl J Med 1996;334:647652.

10. Engel J Jr. The timing of surgical intervention for mesialtemporal lobe epilepsy: a plan for a randomized clinicaltrial. Arch Neurol 1999; 56:13381341.

11. Engel J Jr, Van Ness PC. Outcome with respect to epilepticseizures. In: Engel J Jr, editor. Surgical Treatment of theEpilepsies. 2nd ed. New York: Raven Press, 1993: 609621.

12. Wiebe S, Blume WT, Girvin JP, Eliasziw M. A randomized,controlled trial of surgery for temporal-lobe epilepsy. NEngl J Med 2001; 345:311318.

13. Sperling MR, OConnor MJ, Saykin AJ, Plummer C.Temporal lobectomy for refractory epilepsy. JAMA 1996;276:470475.

14. Foldvary N, Nashold B, Mascha E, et al. Seizure outcomeafter temporal lobectomy for temporal lobe epilepsy: aKaplan-Meier survival analysis. Neurology 2000;54:630634.

15. Wass CT, Rajala MM, Hughes JM, et al. Long-term follow-up of patients treated surgically for medically intractableepilepsy: results in 291 patients treated at Mayo ClinicRochester between July 1972 and March 1985. Mayo ClinProc 1996; 71:11051113.

16. Kwan P, Brodie MJ. Effectiveness of first antiepilepticdrug. Epilepsia 2001; 42:12551260.

17. Kwan P, Brodie MJ. Early identification of refractoryepilepsy. N Engl J Med 2000; 342:314319.

18. Jokeit H, Ebner A. Long-term effects of refractory tempo-ral lobe epilepsy on cognitive abilities: a cross sectionalstudy. J Neurol Neurosurg Psychiatry 1999; 67:4450.

ADDRESS: Deepak Lachhwani, MD, Department of Neurology,S51, The Cleveland Clinic Foundation, 9500 Euclid Avenue,Cleveland, OH 44195.

on April 30, 2014. For personal use only. All other uses require permission.www.ccjm.orgDownloaded from