The Legacy of Patient H.M. for Neuroscience
Total Page：16
File Type：pdf, Size：1020Kb
Neuron NeuroView The Legacy of Patient H.M. for Neuroscience Larry R. Squire1,2,* 1Veterans Affairs Healthcare System, San Diego, CA 92161, USA 2Departments of Psychiatry, Neurosciences, and Psychology, University of California, San Diego, La Jolla, CA 92093, USA *Correspondence: [email protected] DOI 10.1016/j.neuron.2008.12.023 H.M. is probably the best known single patient in the history of neuroscience. His severe memory impairment, which resulted from experimental neurosurgery to control seizures, was the subject of study for five decades until his death in December 2008. Work with H.M. established fundamental principles about how memory functions are organized in the brain. In 1952, Brenda Milner was completing of anticonvulsant medication, that he The findings from H.M. established the her doctoral research at McGill University could not work or lead a normal life. Sco- fundamental principle that memory is under the direction of Donald Hebb. At ville offered H.M. an experimental proce- a distinct cerebral function, separable about this time, she encountered two dure that he had carried out previously in from other perceptual and cognitive abili- patients (P.B. and F.C.) who had become psychotic patients, and the surgery was ties, and identified the medial aspect of severely amnesic following unilateral then performed with the approval of the the temporal lobe as important for removal of the medial structures of the patient and his family. memory. The implication was that the left temporal lobe for the treatment of When Milner first visited H.M., she saw brain has to some extent separated its epileptic seizures (Penfield and Milner, that the epilepsy was now controlled but perceptual and intellectual functions 1958). This unfortunate outcome was that his memory impairment was even from its capacity to lay down in memory entirely unexpected, and it was proposed more severe than in Penfield’s two the records that ordinarily result from that in each case there had been a preex- patients, P.B. and F.C. What she engaging in perceptual and intellectual istent, but unsuspected, atrophic lesion in observed was someone who forgot daily work. the medial temporal lobe of the opposite events nearly as fast as they occurred, hemisphere. In that way, the unilateral apparently in the absence of any general The Medial Temporal Lobe Memory surgery would have resulted in a bilateral intellectual loss or perceptual disorder. System lesion, an idea that was confirmed at He underestimated his own age, apolo- The early paper is sometimes cited incor- autopsy some years later for patient P.B. gized for forgetting the names of persons rectly as evidence that the hippocampus After the two cases were presented at to whom he had just been introduced, and is important for memory, but this partic- the 1955 meeting of the American Neuro- described his state as ‘‘like waking from ular point could not of course be estab- logical Association, Wilder Penfield (the a dream . every day is alone in itself.’’ lished from a lesion that, by the surgeon’s neurosurgeon in both cases) received (Milner et al., 1968, p. 217). description, included the hippocampus, a call from William Scoville, a neurosur- The first observations of H.M., and the amygdala, and the adjacent parahippo- geon in Hartford, Connecticut. Scoville results of formal testing, were reported campal gyrus. As Milner subsequently told Penfield that he had seen a similar a few years later (Scoville and Milner, wrote, ‘‘Despite the use of the word memory impairment in one of his own 1957). This publication became one of ‘hippocampal’ in the titles of my papers patients (H.M.) in whom he had carried the most cited papers in neuroscience with Scoville and Penfield, I have never out a bilateral medial temporal lobe resec- (nearly 2500 citations) and is still cited claimed that the memory loss was solely tion in an attempt to control epileptic with high frequency. H.M. continued to attributable to the hippocampal lesions’’ seizures. As a result of this conversation, be studied for five decades, principally (Milner, 1998). Indeed, the original paper Brenda Milner was invited to travel to by Brenda Milner, her former student ends, quite appropriately, with the state- Hartford to study H.M. Suzanne Corkin, and their colleagues ment: H.M. had been knocked down by (Corkin, 1984, 2002; Milner et al., 1968). a bicycle at the age of 7, began to have He died on December 2, 2008, at the It is concluded that the anterior minor seizures at age 10, and had major age of 82. It can be said that the early hippocampus and hippocampal seizures after age 16. (The age of the descriptions of H.M. inaugurated the gyrus, either separately or together, bicycle accident is given as 9 in some modern era of memory research. Before are critically concerned in the reports; for clarification see Corkin, H.M., due particularly to the influence of retention of current experience. It 1984.) He worked for a time on an Karl Lashley, memory functions were is not known whether the amygdala assembly line but, finally, in 1953 at the thought to be widely distributed in the plays any part in this mechanism, age of 27 he had become so incapaci- cortex and to be integrated with since the hippocampal complex tated by his seizures, despite high doses intellectual and perceptual functions. has not been removed alone, but 6 Neuron 61, January 15, 2009 ª2009 Elsevier Inc. Neuron NeuroView always together with uncus and At this juncture, several points became information for a period of time after it amygdala. (Scoville and Milner, clear. First, H.M.’s lesion was less was presented. Thus, he could carry on 1957, p. 21). extensive than described originally by the a conversation, and he exhibited an intact surgeon in that it extended a little more digit span (i.e., the ability to repeat back The findings from H.M. were initially than 5 cm caudally from the temporal a string of six or seven digits). Indeed, met with some resistance, especially pole (not 8 cm). As a result the posterior information remained available so long because of the difficulty for many years parahippocampal gyrus was largely as it could be actively maintained by of demonstrating anything resembling spared (specifically, the parahippocampal rehearsal. For example, H.M. could retain his impairment in the experimental animal. cortex or what in the monkey is termed a three-digit number for as long as 15 min Efforts to establish an animal model in fact area TH TF). Second, the reason that by continuous rehearsal, organizing the began almost immediately when Scoville H.M.’s memory impairment was so severe digits according to an elaborate himself came to Montreal and did the was that the bilateral damage included the mnemonic scheme. Yet when his atten- same surgery in monkeys that he had parahippocampal gyrus (anteriorly) and tion was diverted to a new topic, he forgot done with H.M. But these monkeys and was not restricted to the hippocampus. the whole event. In contrast, when the others with medial temporal lesions Damage limited to the hippocampus material was not easy to rehearse (in the seemed able to learn tasks that H.M. causes significant memory impairment case of nonverbal stimuli like faces or could not learn. Only much later did it but considerably less impairment than in designs), information slipped away in become understood that apparently H.M. Third, memory impairment more less than a minute. These findings sup- similar tasks can be learned in different severe than H.M.’s could now be under- ported a fundamental distinction between ways by humans and monkeys. For stood, as when the damage includes the immediate memory and long-term example, the visual discrimination task, structures damaged in H.M. but also memory (what William James termed which is learned gradually by the monkey extends far enough posteriorly to involve primary memory and secondary memory). over hundreds of trials, proved to involve the parahippocampal cortex (patients Primary memory [immediate memory] what one would now call habit learning. E.P. and G.P.; Kirwan et al., 2008). In the monkey, this kind of learning In the early years, the anatomy of the .comes to us as belonging to the depends on the basal ganglia, not the medial temporal lobe was poorly under- rearward portion of the present medial temporal lobe. Eventually, tasks stood, and terms like hippocampal zone space of time, and not to the were developed for the monkey that and hippocampal complex were often genuine past (James, 1890, p. 647). were exquisitely sensitive to medial used to identify the area of damage. With temporal lobe lesions (for example, the the elucidation of the boundaries and Secondary memory [long-term memory] is one-trial, delayed nonmatching to sample connectivity of the structures adjacent to quite different. task), and an animal model of human the hippocampus and the discovery that memory impairment thereby became these structures are important for An object which has been recol- available (Mishkin, 1978). memory, vague terms like hippocampal lected . is one which has been Cumulative work with the animal model complex became unnecessary (though absent from consciousness alto- over the next decade, together with one can still find them in contemporary gether, and now revives anew. It is neuroanatomical studies, succeeded in writing). It is now possible to achieve care- brought back, recalled, fished up, identifying the anatomical components ful descriptions based on anatomical so to speak, from a reservoir in of what is now termed the medial measurement and modern terminology. which, with countless other temporal lobe memory system (Squire H.M.