524 Maters arising prolong the lives of those suffering, thereby Most of these problems are not taken into 6 Poser C, Paty D, Scheinberg L, et al. New J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.55.6.524-b on 1 June 1992. Downloaded from providing the necessity for self-action by the consideration in Sibley's article. diagnostic criteria for multiple sclerosis. Ann patient. It is germane to point out that the "scien- Neurol 1983;13:227-31. Perhaps when we advise patients and tific" evidence of epidemiology and biostatis- families who face difficult decisions, we tics overlooks the fact that such studies can at Sibley et al reply: would be better to heed the available evi- best only be estimates within a cohort and dence about who will do well and who will do cannot be generalised; epidemiological stud- We age- and sex-matched MS patients with badly rather than withholding recommenda- ies are most useful in providing aetiological healthy controls solely to establish the degree tions because unduly cautious statisticians clues but cannot be used to deny possible of accident-proneness of MS patients. It choose to neglect the downside effects of causal relationships.2 As was pointed out by probably needs to be clearly restated that the their theoretical arguments. Neurologists are Schoenberg, "Statistical significance (or lack patients were used as their own controls in an especially favourable position to under- thereof) does not equal biological when they were not "at risk". stand the misery that accompanies chronic, significance". Physical trauma had no positive influence overwhelming brain damage. This knowledge It is difficult to understand, on the basis of on either exacerbation rate or progression of places them in a position to understand the what is already known about the pathogen- MS in our group of MS patients as a whole, humane needs in such cases and advise esis of MS, how trauma such as laceration of or in various subgroups divided according to families accordingly. the hand, contusion of the right finger, a root sex, age, degree of disability, or duration of FRED PLUM canal procedure, or a thorn in the foot can disease-all of these data could not Department ofNeurology, be and Neuroscience, possibly lead to the appearance or recurrence presented in our paper in the interests of The New York Hospital-Cornell Medical Center, of symptoms of MS. Sibley et al dismiss the brevity. We have not analysed separately 525 East 68th Street, alteration of the blood-brain barrier in the based on sites of residence or socioeconomic New York, NY 10021, USA genesis of the MS lesion; in doing so they factors. disregard the numerous recent publications I am sure that Dr Poser is aware that we 1 Bates D. Defining prognosis in medical coma. which have amply confirmed this fact, have no way ofidentifying MS patients whose J Neurol, Neurosurg Psychiatry 1991;54: including statements by such authorities as disease has not 569-7 1. yet been recognised. His 2 Levy DE, Bates D, Caronna JJ, et al. Prognosis McDonald and Barnes' who commented statement that some of them will not have in non-traumatic coma. Ann Int Med that "A consistent, very early event is a symptoms until subjected to trauma, makes 198 1;94:293-301. breakdown of the blood-brain barrier which causative assumptions that our study data do is largely repaired over weeks, leading to not support. Bates replies: marked changes in the size of acute lesions" With regard to his comments concerning I am grateful to Professor Plum for emphasis- and by Compston' who wrote, "Blood-brain the validity of statistical methods, we believe ing the important exclusions in the Inter- barrier damage is necessary early in the that it is common scientific practice to national Study on Prognosis in Non-trau- development of focal demyelination". generalise on the basis of data obtained in matic Coma and appreciate his comments One cannot expect that all, or even most representative samples of a group. The cal- relating to the use of unfavourable early signs. MS patients will have exacerbations follow- culation of annual exacerbation rates was for I did not intend to suggest that clinical or ing trauma, since it is a common experience the convenience of the reader; the chi-square laboratory indicators of poor outcome were for MS patients to have repeated viral infec- calculations were done on actual exacerba- not to be used as part of the process to make tions without changes in the course of their tion numbers during periods at risk and not decisions in patient management, and am disease. at risk for trauma, not on the annual rates. very conscious of what Professor Plum refers The relationship between trauma to the Our study tested the hypothesis, advanced to as the "downside statistical feature" of head, neck or back that I have supported, is by others, that trauma, including peripheral prolonging an insentient life and the con- based upon clinical, neuropathological, and trauma, could cause exacerbations ofMS. We sequent burden on the family. My purpose experimental evidence which were reviewed are happy to see that Dr Poser agrees that was rather to emphasise that predictors of in 1986' and based on the obligatory step of peripheral trauma, at least, can be dismissed poor outcome should not be used as the sole the alteration of the blood-brain barrier for as a cause. He continues to believe, however, factor in making decisions about life support, the formation of the MS lesion. This has now that CNS trauma that causes a breach in the though they should be used in discussions been amply documented. blood-brain barrier can cause new lesions in with relatives of the patient and with our Additional problems seemed to have MS. The anecdotal case reports that form the colleagues in helping to arrive at an appro- plagued this study: for reasons which are not clinical basis for his belief are selected from priate clinical decision. explained, symptoms alone are not recog- large numbers of cases where there is no http://jnnp.bmj.com/ nised as exacerbations. Nevertheless, diplo- temporal relationship between trauma and pa, paresthesiae and vertigo, classic symp- exacerbation and cannot be accepted as valid toms of MS are quite often unaccompanied evidence that trauma is a risk factor in MS. by objective changes in the neurological We agree that if trauma were a causative A prospective study of physical trauma examination. It is curious that Sibley et al factor that all traumas would not be followed and multiple sclerosis used 48 hours' duration for an exacerbation by excacerbations, since many new lesions of as opposed to the more generally accepted 24 MS are symptomatic. However, a study such I read with great interest the article by Sibley hours.6 Finally, it is difficult to draw any as ours should have shown a higher propor- et al.' Because of my own involvement in this conclusions from data given on the basis of tion of excacerbations when at risk for on September 24, 2021 by guest. Protected copyright. particular problem, I wish to make the mean annual exacerbation rates for the sim- trauma in 170 patients over a five year period following comments. ple reason that exacerbations in MS are never since the frequency of trauma was high. There can be no more difficult task than to uniformly distributed; periods of complete The Multiple Sclerosis NMR Group at conduct epidemiological surveys of multiple quiescence lasting several years not uncom- Queen Square has found that in some sclerosis (MS) because of its unpredictability monly occur after a series ofseveral exacerba- patients gadolinium leakage may precede and symptomatic variability. It is therefore tions within a single year. other evidence of new lesion formation; they unsuitable for measurements with the yard- CHARLES M POSER attribute this to inflammation, not to trau- sticks currently at hand. Indeed, to design an Department ofNeurology, ma.' Certainly barrier breakdown due to the Harvard Medical School, epidemiologically valid study, it would be Boston, MA, USA perivenular inflammatory lesions is an early necessary to match MS patients (not just event, but we believe that the cause of this healthy controls) for age, sex, ethnic origin, 1 Sibley WA, Bamford CR, Clark K, Smith MS, inflammatory response is unknown and our duration and geographic sites of residence, Laguna JF. A Prospective study of physical data suggest that trauma is not a factor in living and socio-economic environment, trauma and multiple sclerosis. Y Neurol Neu- new lesion formation. duration and of type of rosurg Psychiatry 1991;54:584-9. severity disease, 2 Schoenberg B. Neurological epidemiology. New Our decision to define exacerbations on clinical course, stage of activity, and number York: Raven Press, 1978;1 1, 43. the basis of objective neurological changes and location oflesions using modern imaging 3 McDonald I, Barnes D. Lessons from magnetic only are in keeping with the Schumacher techniques such as MRI. It would be neces- resonance imaging in multiple sclerosis. Committee criteria for Trends Neurosci 1989;12:376-9. definition of a sary to consider not only patients who have 4 Compston A. Immunological aspects of multi- relapse.2 The 48 hour minimum duration of already been diagnosed as having MS, but ple sclerosis. In: MatthewsW (ed) McAlpine's new symptoms was an arbitrary departure also those who have been symptomatic but Multiple Sclerosis, 2nd ed. Edinburgh: Chur- from the Schumacher criteria to help ensure not yet recognised as well as all those chill-livingstone, 1991; 330. patients 5 Poser C. The pathogenesis of multiple sclerosis: that fluctuations in symptoms due to fatigue whose symptoms of underlying MS will not a critical reappraisal.