This Article Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation

Guest Editorial: The Autopsy in Gerontological Research

A Retrospective

^a Division of Gerontology, Department of Medicine, St. Louis University School of Medicine, and Aging and Development Program, Department of Psychology, Washington University, St. Louis, Missouri

Herman T. Blumenthal, 6203 Washington, St. Louis, MO 63130.

AS Hill and Anderson ⁽¹⁾ have recorded, almost two and a half centuries have elapsed since Morgagni published his observations on some 700 autopsies. It heralded a systematic correlation between the clinical manifestations of a disease and the anatomical findings. The era of discovery that followed led to the establishment of pathology as an independent discipline by Rokitansky and to the elucidation of discrepancies between premortem diagnosis and pathological manifestations.

Hill and Anderson cite studies showing that as recently as 1984 in about one third of cases at 33 hospitals, some associated with medical schools, the primary clinical diagnoses were not substantiated at autopsy; and in about one third of these, "if the correct diagnosis had been made and appropriate therapy instituted, the outcome would have been expected to be different." In a report also of 1980s vintage, Nemetz and colleagues ⁽²⁾ cite 33 studies in which the rate of discrepancies between the clinical diagnosis and autopsy findings ranged between 0.3% and 55.5%. One reason for this wide discrepancy is that some studies dealt with the totality of cases coming to autopsy with a wide range of diseases, while others focused on single disease categories such as cancer.

Hill and Anderson also list 16 ways in which the autopsy makes a significant contribution to medical practice and health policy. Among those other than determining cause of death and discrepancies between clinical and pathological diagnoses—some relevant to current areas of research—are (i) discovery of new diseases, (ii) evaluation of new surgical techniques, (iii) evaluation of the efficacy of new drugs, (iv) contributing to medical and epidemiological research, (v) evaluation of new prostheses in cardiac and orthopedic research, (vi) evaluation of the effects of treatments in the developing area of genetic engineering, and (vii) provision of essential tissues for research.

But today, the autopsy is an endangered species, on its way to becoming extinct. Following are some reasons for this state of affairs: (i) The elimination of the 20% rate for hospital accreditation, (ii) there is no provision for reimbursing the costs of postmortem examinations, and (iii) the view that scanning techniques serve to replace the autopsy.

The information provided by scanning techniques are comparable to the gross examination of an organ at autopsy, but they do not yield the information provided by conventional and electron microscopy. If only these imaging techniques had been used back when the neuritic (amyloid) plaques (NPs), the neurofibrillary tangles (NFTs), or the fibrillar-inclusion Lewy bodies (LBs) of major neurodegenerative diseases were first described, we would have had no knowledge of the existence of these lesions. Moreover, the use of imaging techniques have led to overestimates of disease prevalence ⁽³⁾.

The effects of these limitations on the performance of necropsies are palpable. In the United States, the necropsy rate has declined from 50% in 1945 to less than 10% at present. With regard to matters of particular interest to gerontologists, the lowest necropsy rates are on elders, yet this age group constitutes the largest proportion of deaths ⁽⁴⁾. Necropsies on patients dying in nursing homes are a particularly valuable source of information, but they are almost nonexistent.

Given this state of affairs, a review of what necropsies have contributed in the past to our knowledge of matters of interest to gerontologists are in order; and that is the principal focus of this article.

	An Historical Perspective

Top An Historical Perspective The Future of the... Conclusion References

 
As a result of the decline in the autopsy rate in recent years, the best sources of information relevant to gerontologists are from autopsy studies carried out in earlier periods when the rates were considerably higher.

Cause of Death
A 1982 study by Kohn ⁽⁵⁾ based on 200 aged subjects was primarily aimed at resolving the dichotomy that separates biological aging from the age-related diseases by determining the causes of death in subjects of advanced age. He found no cause of death in 30% of the cases, and in an attempt to bridge the dichotomy, he proposed that senescence be regarded as a cause of death and accepted as a disease. This proposal has failed to satisfy the dichotomists as evidenced by the fact that the debate continues ⁽⁶⁾.

Furthermore, there are flaws in Kohn's study that bring into question the 30% level of cases with no cause of death. This study does not include data on autopsies in younger subjects in which no cause of death was found. It is not certain, therefore, whether Kohn's findings are characteristic only of autopsies in the aged or reflect a more general limitation of the necropsy. Also lacking in Kohn's study, and in most others, including those on centenarians noted below, is a routine analysis for the presence of amyloid, particularly of the heart and brain. As discussed later, there are studies on the prevalence of amyloidosis that bring us much closer to resolving the aging-disease dichotomy. Nor did his study address the prevalence of comorbid lesions, also of importance to gerontologists, and also discussed below.

A more recent 1999 autopsy study ⁽⁷⁾ on 114 subjects ranging in age between 97 and 106 years provides data that are at variance with Kohn's report. This study found no cause of death in only 5% of cases, with vascular and respiratory diseases accounting for 70% and cancer for 6% of deaths. However, here, too, there were no analyses for amyloid, and the prevalence of comorbidities was also not determined. Nevetheless, the authors concluded that "senescence is not a cause of death."

Other studies carried out between 1951 and 1987 ^{(8)(9)(10)(11)(12)(13)(14)(15)}, some in European countries where an autopsy was carried out in virtually all deaths, provide the most reliable data. Those covering the years between 1951 and 1975 contain an abundance of data on octa- and nonagenarians. Taken together, they also show that the most common causes of death were vascular disease of the heart and brain, and cancer, with prevalence rates of about 35%–50% for coronary heart disease and stroke and about 25%–30% for cancer. And they provide data suggesting that the rates for these diseases may be lower in those of more advanced age than in subjects in their sixth and seventh decades. However, they provide no data on the prevalence of Alzheimer's disease.

Autopsy data on centenarians, most relevant to the cause of death at the end of the human life span, are, of course, sparse. One report ⁽¹³⁾ contains findings in two centenarians, one of whom had a carcinoma of a bronchus. Two other studies ^(14)(15) (a total of 55 cases) showed a prevalence of vascular disease of the heart and brain, and of malignancies, that are at variance with the studies in octa- and nonagenarians in that they did not show a decline in the prevalence of these diseases. Cardiac and cerebral atherosclerosis of a degree to be considered clinically significant were almost universal, with thromboembolic phenomena present in 52%. Cancers sufficiently widespread to be considered a cause of death were present in more than 40%. Thus, the data on mortality-causing malignancies in centenarians suggest a significant increase in prevalence over the 20%–30% in octa- and nonagenarians.

It is notable that these many older studies contain no information on the prevalence of amyloid and, relatedly (because the neuritic plaques contain amyloid), of Alzheimer's disease. This appears to reflect the aforementioned fact that procedures for the detection of amyloid were not routinely carried out. However, Glenner ⁽¹⁶⁾ has noted the high prevalence of amyloid deposits in the brain in biological aging and in Alzheimer's disease, in the islets of Langerhans in biological aging and diabetes, and in the heart, concluding that in subjects older than 65 the prevalence of amyloidosis exceeds that of coronary heart disease, cancer, and other aging-associated diseases. The few autopsy studies that directly address the prevalence of amyloid deposits in the heart in subjects of advanced age ^{(17)(18)(19)(20)} place its prevalence at 80% to 100%. Gray and colleagues ⁽²¹⁾ have concluded that myocardial fibrillation due to amyloid deposits in the heart may be the leading cause of death in the industrialized world. It would appear, therefore, that amyloidosis is a greatly underappreciated disorder of biological aging and of the age-related diseases and that it merits serious consideration in the resolution of the aging-disease dichotomy.

Currently, the clinical diagnosis of Alzheimer's disease in patients with dementia, implying the presence of many NPs and NFTs, enjoys overwhelming popularity, with multi-infarct dementia a distant second and dementia with LBs an even more distant third. However, autopsy studies indicate a more complex array of lesions. There are a number of brain lesions other than NPs and NFTs that are present in greater number in patients with dementia than in age-matched controls ⁽²²⁾. But, perhaps more to the point, cerebral amyloid angiopathy (CAA) is present in 70% of nondemented subjects over age 80, in 90% of demented subjects with NPs and NFTs, and in cases of dementia with only CAA ⁽²³⁾. There are, in addition, cases of dementia attributable to diffuse LB disease, as well as LBs in the dementia in Parkinson's disease.

The prevalence of clinical dementia may be as high as 50% in those older than 85 ⁽²⁴⁾, but autopsy studies indicate that to assume that all these cases can be cured if their brains can be cleared of NPs and NFTs may be an oversimplification. There are many paths to the same destination.

Comorbidities and Their Relevance to Causation
The realization that the aged are particularly prone to several concomitant diseases is hardly a new one. Zeman ⁽²⁵⁾ in his 1962 scholarly history of studies on the pathology of old age cites Engel's comment on this subject as follows:

The majority of these people right up to their last illness suffered not only from the predispositions that advanced age brings with it, but also from some certain definite more or less important other diseases; indeed some persons are a veritable storehouse of the most varied morbid processes. If one counts all these diseases of single organs which may either bring about deaths in longer or shorter time, or which may have dangerous sequelae, without considering all the many minor alterations, we get the following interesting facts: the 326 women suffered from 565, the 220 men from 373 important or dangerous sicknesses, or in other words—in each 150 sick men or women, about 70 are dangerously sick from two diseases of different organs. (pp. 132–133)

A mid-20th century study by Howell and Piggott ⁽⁹⁾ reports a progressive increase with age in comorbid lesions (Table 1 ), and in an analysis of 100 octagenarians, Howell ⁽¹³⁾ recorded fewer than 10 nonfatal lesions in 68 patients, but between 10 and 17 such lesions in the remaining 32.

Stoddard ⁽²⁷⁾ has proposed a unique way of looking at comorbid diseases. He observes that we presently regard a patient's clinical manifestations as indicative of "one and only one disease." But as biomedical data have multiplied, it has become evident that there are crossover similarities among cases in different disease categories. He therefore proposes a new way of classifying diseases—on the basis of "biopathological populations, or sets." He defines a "set" as follows: "A set is a collection of elements or objects having some common property that is selected to define the set and so to identify its member elements... " (p. 232).

This definition implies a causal commonality for the diseases of a set. It also provides a unique way of looking at the connection between biological aging and the age-related diseases. The frequency with which several pathologies are present in an aged individual, most frequently neoplasms and vascular disease of the heart and brain (add Alzheimer's disease), is in agreement with this concept. The concept of "sets," therefore, is in accord with the view that stochastic biological aging phenomena may provide the mechanisms that generate comorbidities.

As Olshansky and colleagues ⁽²⁸⁾ have commented, occlusive vascular disease of the heart and brain, and cancer, now account for about three fourths of deaths in those older than 65, and as deaths become more concentrated at older ages, the competition among causes of death becomes more pronounced. Thus, there are increases in opportunities for comorbidities. Not included here are the comorbidities in dementia patients that show prevalence rates for heart disease, stroke, diabetes, and cancer comparable to those in age-matched nondemented subjects ⁽²⁹⁾. Other examples of sets are the common associations of diabetes with hypertension and vascular disease and the multiplicity of brain lesions associated with dementia as noted above.

Likely suspects in this scenario are oxidative stress and the accumulation of advanced glycosylation end (AGE) products, which have been implicated in both biological aging and age-dependent disorders such as cardiovascular disease, cancer, and neurodegenerative diseases ^(30)(31).

Vital Statistics
While death certificates, in the absence of autopsy verification, offer reliable data as to age-at-death and mortality rate by gender, as the foregoing discussion on discrepancies between clinical and autopsy findings indicate, they lack precision as to cause of death. Despite this unreliability, death certificates, nevertheless, also support a high rate of comorbidities. More than one cause of death was reported in 74% of all death certificates, with 33% recording two causes, 25% three causes, and 16% four or more causes ⁽³²⁾. Yet much planning to provide for the future needs of a growing population of the aged remains dependent on data provided by death certificates lacking autopsy confirmation.

The Autopsy as an Epidemiological Tool
While autopsy data may have the potential for providing more accurate epidemiological information than clinical data, their validity is contingent on demonstrating an absence of demographic bias ^(33)(34). Demographic bias was found not to be present at an autopsy rate of 56%. However, given the current low autopsy rate, meeting this criterion would rarely be obtainable. Epidemiological studies on the prevalence of the age-related diseases based on necropsy observations are therefore meager.

	The Future of the Autopsy in the Genomic Age

Top An Historical Perspective The Future of the... Conclusion References

 
The techniques applied to tissues in the study of pathological lesions have evolved over several centuries. First, sections were stained simply to better visualize cells. Later, special stains were introduced with which to identify particular lesions, as for example congo red to distinguish amyloid from hyaline, followed by histochemical stains to identify particular chemical entities such as the periodic acid–Schiff stain with which to identify polysaccharides. Then came fluorescence and other tags in combination with entities such as antibodies with which to identify proteins and components of the immune system present in or on cells. Along the way came electron microscopy, which provided the visualization of cellular components not visible by conventional light microscopy, followed by a separate histochemistry adapted for application to electron microscopy.

But the autopsy has not become obsolete, nor have the tools for analyzing the nature of lesions been exhausted. We are now entering the era of the genomic revolution, and with it a new journal titled The Journal of Molecular Diagnostics sponsored by The American Society of Investigative Pathology. The new investigative tools provide the capacity to identify genes in both their normal and mutated forms in both fixed archival as well as fresh tissues. While applied largely to biopsy specimens in which the information may have immediate diagnostic and therapeutic implications, the application of these techniques to autopsy specimens have the potential to provide information useful to research on both biological aging and the age-related diseases. And stimulated by the need for tissues with which to carry out research on Alzheimer's disease, there is a resurgence of attempts to increase the number of autopsies, particularly on the aged ⁽³⁵⁾.

With remarkable foresight, pathologist Emanuel Farber, in a 1987 editorial ⁽³⁶⁾, stated the following:

... with enough studies of genes, messenger RNA, proteins, receptors, etc., everything will fall into place. Stating the problem somewhat simplistically, all we have to do to solve any complex jigsaw problem is to collect all the pieces, shake them up, and let them fall at random, and somehow they will find their appropriate mates and solve the problem. (p. 533)

In principle, that is what microarray or DNA chip technology can accomplish. An array consists of thousands of DNA spots, each corresponding to a gene. When an RNA from a tissue is processed and applied to the array, the spots corresponding to active genes light up ⁽³⁷⁾. Protein arrays are already on the horizon. This technology can reveal an individual's propensity to a host of disorders including the likelihood of acquiring Alzheimer's disease, diabetes, and specific cancers. In the context of implementing information on autopsied tissues, it provides the capacity to compare the genetic makeup of primary and metastatic tumors. And with respect to the primary focus of this article, the microarray technique applied to autopsied tissues from subjects of a wide range of ages can provide information as to aging-related changes in the genetic makeup of a variety of organ systems and their relevance to age-associated diseases.

	Conclusion

Top An Historical Perspective The Future of the... Conclusion References

 
As noted early in this article, the autopsy has the potential to provide useful information relevant to a variety of medically related activities. In the main, however, the focus has been on information of particular importance to gerontologists. Verifying the causes of death in old age has been important in addressing the connection between biological and pathological aging. In this regard, there are studies on amyloidosis, although few, that suggest that this lesion may be close to universal in advanced old age. The increasing prevalence of comorbidities in relation to age has been discussed with regard to commonalities in the causation of different age-related diseases. The autopsy also has the potential to provide more reliable epidemiological data, but the present low necropsy rate precludes this possibility.

With the advent of new technologies for the analysis of the genetic makeup of normal and diseased tissues, there are new opportunities for studying biological aging in human autopsy material and the relevance of information so derived to the genesis of the age-related diseases. While archival tissues can be used for such studies, their use is limited, and there is, therefore, a need for the resurgence of the autopsy. Demographic studies indicate a marked future growth in the population of the aged, and autopsy studies can provide important information as to future needs in the provision of medical care in the aged.

Received February 14, 2002

Accepted February 18, 2002

	References

Top An Historical Perspective The Future of the... Conclusion References

 

1. Hill RB, Anderson RE, 1989. The evolving purposes of the autopsy: twenty-first century values from an eighteenth century procedure. Perspect Biol Med 32:223-233. [Medline]
2. Nemetz PN, Ludwig J, Kurland LT, 1987. Review article: assessing the autopsy. Am J Pathol 128:362-379. [Abstract]
3. Black WX, Welch HG, 1993. Advances in diagnostic imaging and overestimation of disease prevalence and the benefits of therapy. N Engl J Med 328:1237-1243. [Free Full Text]
4. Research after death. Lancet. 344:19941517-1518. [Medline]
5. Kohn RR, 1982. Cause of death in very old people. JAMA 247:2793-2797. [Abstract/Free Full Text]
6. Hayflick L, 2000. The future of ageing. Nature 408:267-269. [Medline]
7. Bordin PD, Col PG, Puzzo P, et al. 1999. Causes of death and clinical diagnostic errors in extreme aged hospitalized people: a retrospective clinical-necropsy survey. J Gerontol Med Sci 54A:M554-M559. [Abstract]
8. Israel RA, Rosenberg HM, Curtin LR, 1986. Analytic potential for multiple cause-of-death data. Am J Epidemiol 124:151-179.
9. Howell TG, Piggott AP, 1951(b) Parts III and IV. Findings in earlier and later seventies. Geriatrics. 1952;7:137–142. (c) Part V. Findings in the later sixties. Geriatrics. 1953;8:215–218. (d) New malignant growths in old age. Geriatrics. 1958;13:176–179.. Morbid anatomy in old age. (a) Parts I and II. Pathological findings in the 9th and 10th decades. Geriatrics 6:85-95. [Medline]
10. Monroe RT. Diseases of Old Age. Cambridge, MA: Harvard University Press; 1951.
11. Mulligan RM, 1960. Geriatric pathology. AMA Arch Pathol 69:9-42.
12. McKeown F. Pathology of the Aged. London: Butterworth; 1965.
13. Howell TH. Old Age. London: HR Lewis and Co. Ltd; 1975.
14. Ishii T, Sternby NH, 1978II. Urogenital and digestive systems. J Am Geriatr Soc. 1978;26:391–396. III. Osseous system, malignant lesions, and causes of death. J Am Geriatr Soc. 1978;26:529–533.. Pathology of centenarians. I. The cardiovascular system and lungs. J Am Geriatr Soc 26:108-115. [Medline]
15. Klatt EC, Meyer PR, 1987. Geriatric autopsy pathology in centenarians. Arch Pathol Lab Med 111:367-369. [Medline]
16. Glenner GG, 1983. Editorial: Alzheimer's disease. The commonest form of amyloidosis. Arch Path Lab Med 107:281-282.
17. Pitkanen P, Westermarl P, Cornwell GG, III 1984. Senile systemic amyloidosis. Am J Pathol 117:391-399. [Abstract]
18. Cornwell GG, III Westermark P, Murdock W, Pitkanen P, 1982. A third distinctive type of age-related cardiovascular amyloidosis. Am J Pathol 108:135-139. [Abstract]
19. Cooper JM, 1974. Selective staining of amyloid as a function of amyloid composition and structure. Lab Invest 31:232-238. [Medline]
20. Kawamamura S, Takahashi M, Ishuhara T, Uchino E, 1995. Incidence and distribution of isolated atrial amyloid. Histologic and immunochemical analyses of 100 aging hearts. Pathol Int 45:335-342. [Medline]
21. Gray RA, Pertsov AM, Julife J, 1998. Spatial and temporal organization during cardiac fibrillation. Nature 391:75-82.
22. Von Dras DD, Blumenthal HT, 1992. Dementia of the aged: disease or atypical-accelerated aging? Biological and psychological perspectives. J Am Geriatr Soc 40:285-294. [Medline]
23. Vinters HF, 1987. Cerebral amyloid angiopathy: a critical review. Stroke 18:311-324. [Free Full Text]
24. Evans DA, Funkenstein HH, Albert MS, et al. 1989. Prevalence of Alzheimer's in a community. Higher than previously reported. JAMA 263:2335-2340.
25. Zeman FD, 1962. Pathologic anatomy of old age. Arch Pathol 73:126-143. [Medline]
26. Ludwig SC, 1981. Senescence: pathology facing medicine's ultimate issue. Arch Path Lab Med 105:445-451.
27. Stoddard LD, 1980. Toward a new human pathology. 1. Biopathological populations or sets: a substitute for the old pathology's diseases. Hum Pathol 11:228-239. [Medline]
28. Olshansky SJ, Carnes BA, Cassel CK, 1993. The aging of the human species. Sci Am 268:46-52. [Medline]
29. Thorpe J, Widman LP, Wallin A, et al. 1994. Co-morbidity of other chronic age-dependent diseases in dementia. Aging Clin Exp Res 6:159-166.
30. Ames BN, Shigenaga MK, Hagen TM, 1993. Oxidants, antioxidants and the degenerative diseases of aging. Proc Natl Acad Sci USA. 90:6925-7911.
31. Gerstein HC, Yusuf S, 1996. Dysglycemia and risk of cardiovascular disease. Lancet 347:949-950. [Medline]
32. Carter JR, 1985. The problematic death certificate. N Engl J Med 313:1285-1286. [Medline]
33. McFarlane MJ, Feinstein AR, Wells CK, et al. 1987. The epidemiologic necropsy. JAMA. 258:331-338. [Abstract/Free Full Text]
34. McFarlane MJ, 1987. Absence of demographic bias in selection of necropsy. Arch Pathol Lab Med 111:67-70. [Medline]
35. Schmitt FA, Wetherly MMC, Wekstein DR, et al. 2001. Brain donation in normal aging. Procedures, motivations, and donor characteristics from the biologically resilient adults in neurological studies (BRAiNS) project. Gerontologist 41:716-721. [Abstract/Free Full Text]
36. Farber E, 1987. Molecular biology and genetics, and our understanding of disease. Hum Pathol 18:533-534. [Medline]
37. Friend SH, Stoughton RB, 2002. The magic of microarrays. Sci Am 276:44-53.

This article has been cited by other articles:

				H. T. Blumenthal The Aging-Disease Dichotomy: True or False? J. Gerontol. A Biol. Sci. Med. Sci., February 1, 2003; 58(2): M138 - 145. [Full Text] [PDF]

				J.-P. Michel, S. Pautex, D. Zekry, G. Zulian, and G. Gold End-of-Life Care of Persons With Dementia J. Gerontol. A Biol. Sci. Med. Sci., October 1, 2002; 57(10): M640 - 644. [Abstract] [Full Text] [PDF]

This Article Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation