gene and long-term cancer risk. They propose that the surgical decision on whether to retain the rectum should depend on phenotype in the patient and the family, the mutation site, and patient preference. Although we agree with this individualised approach we emphasise the heterogeneous nature of the phenotype with mutations 3' of codon 1250,1–3 and that expert advise should be sought on the interpretation of APC-mutation data. Notwithstanding the other factors in decision making, perhaps the most important factor is the true level of morbidity associated with colectomy with ileorectal anastomosis and restorative proctocolectomy (pouch). The former is a much simpler procedure with few complications. Where pouch surgery is highly centralised, such as it is in France, adverse effects arise rarely.4 However, in most countries where centres are dispersed, distressing side-effects of pouch surgery such as faecal incontinence and impotence, can be unacceptably high, bearing in mind that recipients are usually very young and symptom-free at presentation. Another factor in professional assessment of the appropriate intervention is the growing interest in long-term chemoprevention. If the inprogress randomised trials of nonsteroidal anti-inflammatory agents, or future attempts at topical gene therapy are successful, the prognosis for the retained rectum in patients undergoing ileorectal anastomosis is likely to be improved greatly. *D Gareth R Evans, James Hill, Tracey Dudding, John Burn, Eamonn R Maher *Department of Medical Genetics, St Mary’s Hospital, Manchester M13 0JH, UK; Department of Surgery, Manchester Royal Infirmary, Manchester; Department of Human Genetics, University of Newcastle, Newcastle upon Tyne; and University Department of Human Genetics, Birmingham Maternity Hospital, Birmingham 1
Scott R, Froggatt NJ, Trembath RC, Evans DGR, Maher ER. Familial infiltrative fibromatosis (Desmoid tumours) caused by a recurrent 3' APC mutation. Hum Melec Genet 1996; 5: 1921–24. Scott RJ, van der Luijt RB, Spycher M, et al. Novel germline APC mutation in a large familial adenomatous polyposis kindred displaying variable phenotypes. Gut 1995; 36: 731–36. Eccles DM, van der Luijt R, Breukel C, et al. Hereditary Desmoid disease due to a frameshift mutation at codon 1924 of the APC gene. Am J Hum Genet 1996; 59: 1193–201. Kartheuser AH, Parc R, Penna CP, et al. Ileal pouch-anal anastomosis as the first choice operation in patients with familial adenomatous polyposis: a ten-year experience. Surgery 1996; 119: 615–23.
Hippocampal damage caused by seizures in temporal lobe epilepsy Sir—Tuuli Salmenperä and colleagues (Jan 3, p 35)1 suggest that hippocampal damage is caused by seizures in temporal lobe epilepsy (TLE), and that, therefore, good seizure control may be neuroprotective. We wholeheartedly agree that every effort should be made to control seizures, but wish to raise some points about the methodology. How were the controls chosen and matched? In the investigators’ reference to hippocampal volume measurement, it is stated that for statistical calculations, measurements of hippocampal volume are “normalised” by division by brain area measured in a coronal plane “at the level of the anterior commissure”. The choice of plane is unspecified (rotation of a coronal plane through this landmark, can make the plane axial!), which could affect the correction made. Published cases of rapid volume loss after status epilepticus and subsequent chronic seizures exist;2 there are cases with longstanding apparently typical TLE without any loss of hippocampal volume. Hippocampal-volume loss is a single change, as seen on magneticresonance imaging, but may be heterogeneous in aetiology and effect. Without the support of histology (were any patients offered surgical treatment?), cross-sectional analysis may obscure pathological subtleties that need exploration. Thus, in their own group, the patients with the longest duration of TLE and “damage” were more likely to have begun their seizures at age 5 years or younger. It could be argued that these patients do not have the variety of TLE most commonly seen in adults with refractory partial seizures, in which group habitual seizures frequently follow a quiescent period after an initial injury.3 The cross-sectional design of the study does not exclude the possibility that those patients with the longest duration of epilepsy had the most severe loss of hippocampal volume before the onset of habitual epilepsy and had the most refractory epilepsy. Only a longitudinal study can settle this issue. One possibility, not entirely heretical,4 that cannot be excluded by the data provided is that chronic exposure to antiepileptic drugs may cause hippocampal damage. Did the patients with frequent seizures have a significantly different drug profile from those with “rare” seizures? Longitudinal studies of outcome show that most patients with partial seizures are no less
likely to remain intractable than any other group of patients.5 If Salmenperä and colleagues’ findings are correct, they may apply only to a select group of patients who tend to accumulate in tertiary referral centres. We believe that their findings are not supported by the methods and results and should not be generally applied to patients with partial seizures. *S M Sisodiya, S L Free, J W A Sander *Department of Neurology, Radcliffe Infirmary, Oxford OX2 6HE, UK; and Epilepsy Research Group, University Department of Clinical Neurology, National Hospital for Neurology and Neurosurgery, London; and National Society for Epilepsy, Chalfont St Peter 1
3 4 5
Salmenperä T, Kälviäinen R, Partanen K, Pitkänen A. Hippocampal damage caused by seizures in temporal lobe epilepsy. Lancet 1998; 351: 35. Weishmmann UC, Woermann FG, Lemieux L, et al. Development of hippocampal atrophy: a serial magnetic resonance imaging study in a patient who developed epilepsy after generalised status epilepticus. Epilepsia 1997; 38: 1238–41. Engel J Jr. Update on surgical treatment of epilepsies. Neurology 1993; 43: 1612–17. Loiseau P. Do antiepileptic drugs exacerbate seizures? Epilepsia 1998; 39: 2–5. Manford M, Hart YM, Sande JWA, Shorvon SD. National General Practice Study of Epilepsy (NGPSE): partial seizure patterns in a general population. Neurology 1992; 42: 1911–17.
Author’s reply Sir—Sanjay Sisodiya and colleagues raise several methodological concerns. We regret that due to the limited length, we were not able to include all the methodological or clinical details in the report. We measured the volume of the hippocampus with magnetic resonance imaging (MRI) in 25 controls (13 men, 12 women, mean age 33 [SD 12] years) and 153 patients with TLE (78 men, 75 women, mean age 36 [SD 13] years). There were no between-group differences in age and sex distribution. All controls were interviewed to exclude those with neurological diseases. 15 of 48 TLE patients with a history of over 21 years of frequent seizures have been treated surgically. Significant loss of hippocampal volume was detected in 13 of these patients on preoperative MRI. Finally, our clinic is the only unit in the district to diagnose and treat patients with epilepsy. Therefore, our selection criteria does not consist of only the most severe cases as in tertiary referral centres, but also includes newly or recently diagnosed patients. The method used to correct hippocampal volumes for individual variance in head size has been described in detail previously.1,2 “The coronal plane” was a plane oriented at right
THE LANCET • Vol 351 • April 11, 1998
angle to the long axis of the hippocampus. The brain area used for normalisation of the hippocampal volumes was outlined in a coronal section that was imaged at the level where the anterior commissure appears for the first time. Is the hippocampal damage present before the onset of TLE? Our current findings do not exclude the possibility that hippocampal damage may exist before the onset of TLE in some patients. When the analysis is done in a large group of patients, however the number of recurrent seizures is a major factor associated with the appearance and severity of hippocampal damage. We found that in patients with newly diagnosed cryptogenic TLE, the mean volume of the hippocampus was normal. Moreover, there was a significant reduction in the hippocampal volume in patients with chronic drug-refractory cryptogenic TLE compared with controls, and the severity of damage correlated with the lifetime number of seizures.2 We agree with Sisodiya and coworkers that the cross-sectional design of the study does not allow us to determine the progression rate of seizure-induced damage on individual basis. Also, some “varieties of TLE” associated with hippocampal damage may remain undetected. For those purposes. we would need a follow-up study. A follow-up of TLE patients with the same MRI protocol for up to 20 years certainly forms a major challenge for epileptologists. All patients included in the study (except those with newly diagnosed epilepsy) had antiepileptic medication at the time of imaging. The exposure of patients to various antiepileptic drugs varies with each individual, and therefore, contribution of drugs to hippocampal damage is difficult to assess. Our aim was to find out factors that are associated with the hippocampal volume reduction in MRI in a large population of patients with TLE. The data obtained cannot be extrapolated to other types of epilepsy without further studies. Moreover, our populationbased results do not question the value of MRI follow-up of individual patients when hippocampal damage, its progression, or both need to be assessed. *Tuuli Salmenperä, Reetta Kälviäinen, Kaarina Partanen, Asla Pitkänen Departments of *Neurology and Radiology, Kuopio University Hospital, PO Box 1777 Kuopio, Finland; and A I Virtanen Institute, University of Kuopio, Kuopio 1
Soininen HS, Partanen K, Pitkänen A, et al.
THE LANCET • Vol 351 • April 11, 1998
Volumetric MRI analysis of the amygdala and the hippocampus in subjects with ageassociated memory impairment. Neurology 1994; 44: 1660–68. Kälviäinen R, Salmenperä T, Partanen K, et al. Recurrent seizures may cause progressive hippocampal damage in epilepsy. Neurology (in press).
Nevirapine and rashes S IR —Kelly Warren and colleagues (Feb 21, p 567)1 describe a patient who developed Stevens-Johnson syndrome after treatment with nevirapine for HIV-1 infection and presented data from a literature survey. Several points must be made to provide both balance and critical missing information to this letter. Severe rash and Stevens-Johnson syndrome are recognised complications of nevirapine and are described in the labelling with a prominent warning. To help reduce the risk, Boehringer Ingelheim, with international experts, has developed guidelines for rash management. These guidelines have been widely disseminated and are in the prescribing information. The great majority of rashes occur within the first 4–6 weeks of treatment. There are data to support the importance of using a 2week lead-in dose (200 mg once daily) to reduce the risk of rash of any severity. Furthermore, the nevirapine dose should not be escalated in the face of any rash during the lead-in period until the rash has resolved. Nevirapine should be discontinued in patients with severe rash or any rash accompanied by constitutional symptoms. These patients should never be rechallenged. One of the drug-related deaths cited by Warren et al was in a patient whose dose escalated while experiencing a rash during the lead-in period. The data described in Warren and colleagues’ letter do reflect the entire nevirapine safety database available to the US Food and Drug Administration. Boehringer Ingelheim has maintained regular communication with the FDA and has been updating nevirapine labelling based on the following safety information. In an analysis of the four major controlled development trials, the frequency of rash of any severity was 35% in 350 nevirapine-treated patients compared with 19% in 308 controls. In these same patients, 6·6% of nevirapine recipients experienced severe rashes compared with 1·3% of controls. To evaluate more fully the risk of StevensJohnson syndrome or toxic epidermal necrolysis in all nevirapine trials for our 1997 European licensing submission approved in February, 1998, we analysed over 2800 nevirapine recipients; 0·3% of patients
experienced Stevens-Johnson syndrome or toxic epidermal necrolysis, well below the rates presented in Warren’s letter. Since the US launch in August, 1996, with an estimated 40 000 patients now treated with nevirapine, there is no suggestion that the risk of these events has increased, though we acknolwedge the limitations of the data on drug exposure and the reality of under-reporting of adverse events in general. In assessing treatment options for AIDS, the balance between safety and efficacy is different from that for treatments for more benign diseases. Nevirapine is generally well tolerated, with a low incidence of side-effects other than rash. The substantial additional benefit provided by nevirapine in combination with other antiretrovirals has been described in several trials.2–5 Given the major challenges facing HIV-1-infected individuals and their physicians, it would be unfortunate to limit treatment options unduly for patients because inadequate information was available upon which to make a treatment decision. While Warren et al highlight the most significant adverse event associated with nevirapine we hope that our response will allow for a more accurate understanding of the risk of severe rashes and the recommended management guidelines, which can then be used to determine an appropriate risk/benefit assessment by the treating physician. *Andreas Barner, Maureen Myers Boehringer Ingelheim Pharmaceuticals Inc, PO Box 368, Ridgefield, CT 06877, USA
Warren KJ, Boxwell DE, Kim NY, et al. Nevirapine-associated Stevens-Johnson syndrome. Lancet 1998; 351: 567. D’Aquila RT, Hughes MD, Johnson VA, et al. Nevirapine, zidovudine and didanosine compared with zidovudine and didanosine in patients with HIV-1 infection: a randomized, double-blind, placebocontrolled trial. Ann Intern Med 1996; 124: 1019–30. Montaner JSG, Reiss P, Cooper D, et al. A randomized double-blind trial comparing combinations of nevirapine, didanosine and zidovudine for HIV infected patients: the INCAS trial. JAMA (in press). Vella S, Floridia M, Tomino C, et al. A triple combination of reverse transcriptase inhibitors (2 NRTI+1 NNRTI) induces pronounced and sustained effects on RNA and CD4 in antiretroviral naïve patients with very advanced disease (trial ISS 047). Proceedings of 37th Interscience Conference on Antimicrobial Agents and Chemotherapy (Toronto, 1997). Washington: American Society for Microbiology, 1997: abstr LB-7. Luzuriaga K, Bryson Y, Krogstad P, et al. Combination treatment with zidovudine and nevirapine in infants with human immunodeficiency virus type I infection. N Engl J Med 1997; 336: 1343–49.