Alzheimer Lesions in the Autopsied Brains of People 30 to 50 Years of Age

Abstract

Objective

To test the hypothesis that asymptomatic Alzheimer disease lesions may appear before 50 years of age.

Background

Alzheimer disease has an asymptomatic stage during which people are cognitively intact despite having substantial pathologic changes in the brain. While this asymptomatic stage is common in older people, how early in life it may develop has been unknown.

Methods

We microscopically examined the postmortem brains of 154 people aged 30-39 years (n = 59) and 40-50 years (n = 95) for specific Alzheimer lesions: beta-amyloid plaques, neurofibrillary tangles, and tau-positive neurites. We genotyped DNA samples for the apolipoprotein E gene (APOE).

Results

We found beta-amyloid lesions in 13 brains, all of them from people aged 40 to 49 with no history of dementia. These plaques were of the diffuse type only and appeared throughout the neocortex. Among these 13 brains, 5 had very subtle tau lesions in the entorhinal cortex and/or hippocampus. All individuals with beta-amyloid deposits carried 1 or 2 APOE4 alleles. Among the individuals aged 40 to 50 with genotype APOE3/4, 10 (36%) had beta-amyloid deposits but 18 (64%) had none.

Conclusions

Our study demonstrates that beta-amyloid deposits in the cerebral cortex appear as early as 40 years of age in APOE4 carriers, suggesting that these lesions may constitute a very early stage of Alzheimer disease. Future preventive and therapeutic measures for this disease may have to be stratified by risk factors like APOE genotype and target people in their 40s or even earlier.

INTRODUCTION

Alzheimer disease (AD), an age-associated neurodegenerative disease, is the most common cause of dementia in older people. Clinicopathologic studies indicate that by the time patients develop clinical signs of AD, the pathologic hallmarks of beta-amyloid (Aβ) plaques and neurofibrillary tangles in the brain are already well established. These observations have led to the widely accepted concept that the cognitive deficits of AD are preceded by a prolonged asymptomatic (Driscoll and Troncoso, 2011; Driscoll et al, 2006) or preclinical (Reiman et al, 1996; Sperling et al, 2011) period during which the pathologic changes progress slowly and silently over many years. This asymptomatic period is of great interest because it may be a critical time when disease-modifying therapies might be most effective.

The notion of asymptomatic AD emerged from autopsies, first in cross-sectional and then in longitudinal studies (Bennett et al, 2006; Crystal et al, 1988; Davis et al, 1999; Dickson et al, 1992; Miller et al, 1984; Morris and Price, 2001; Morris et al, 2001; Price et al, 2009; Sparks et al, 1993; Tomlinson et al, 1968; Troncoso et al, 1996). These studies demonstrated that neuritic beta-amyloid (Aβ) plaques and neurofibrillary tangles are common and frequently meet histopathologic criteria for AD in older people who have documented normal cognition before death (Bennett et al, 2006; Hulette et al, 1998; Katzman et al, 1988; Knopman et al, 2003; Morris and Price, 2001; Price and Morris, 1999; Price et al, 2009; Schmitt et al, 2000; Troncoso et al, 1998).

More recently, neuroimaging and cerebrospinal fluid (CSF) biomarker studies have also substantiated the concept of asymptomatic AD (Jack et al, 2013). In vivo imaging of brain Aβ by positron emission tomography (PET) scanning (Klunk et al, 2004) has revealed Aβ deposits in the brains of asymptomatic people not only in old age (Aizenstein et al, 2008; Klunk et al, 2004; Mintun et al, 2006; Resnick et al, 2010; Villemagne et al, 2008), but also in middle age (Morris et al, 2010). In asymptomatic carriers of dominant AD mutations, PET with “Pittsburgh compound B” (PiB), a ligand for Aβ, has demonstrated Aβ deposits as early as 10 to 15 years before the expected age of symptom onset (Bateman et al, 2012).

Furthermore, an abnormally low concentration of Aβ₄₂ in CSF—a consistent marker of AD brain pathology (Fagan et al, 2014; Fortea et al, 2014; Gustafson et al, 2007; Li et al, 2007; Seppala et al, 2011; Shaw et al, 2009)—is seen in some normal middle-aged individuals with PiB-negative PET scans. This finding suggests that measurement of Aβ₄₂ in CSF may be the earliest biomarker of AD (Morris et al, 2010).

Morris et al (2010) reported that in 45- to 59-year-olds, the apolipoprotein gene (APOE) genotype correlates with CSF levels of Aβ, but not tau. The study also suggested that about 10% to 20% of individuals in this age range might have evidence of AD pathology, depending on their APOE4 status (Corder et al, 1993; Morris et al, 2010; Reiman et al, 2009).

In the present study, we tested the hypothesis of a possible extended asymptomatic phase of AD by examining a large series of 154 postmortem brains of people who had died between the ages of 30 and 50 years. We studied the brains for AD lesions, specifically, Aβ plaques, neurofibrillary tangles, and tau-positive neurites. We also determined the people's APOE genotypes. Because we did not find Aβ in the 30- to 39-year-olds (n = 59), we focused our analysis on the 40- to 50-year-olds (n = 95).

A better understanding of the preclinical stages of AD will give us further insight into the pathogenesis of the disease and guide the timing of potential therapeutic interventions.

METHODS

The neuropathologic studies that we report here are part of a comprehensive program established by the Lieber Institute for Brain Development in Baltimore, Maryland.

We report a total of 154 autopsies performed between September 2012 and June 2014. These autopsies were not consecutive. Our series included a high proportion of subjects with a history of a psychiatric or behavioral abnormality. Lieber Institute staff acquired detailed information about the subjects during life.

All of the autopsies were performed in Baltimore at the Office of the Chief Medical Examiner of the State of Maryland, following protocols authorized by the Institutional Review Board of the Johns Hopkins University School of Medicine and the State of Maryland. Brain examinations and dissections were performed at the Division of Neuropathology of the Johns Hopkins University School of Medicine.

Autopsy Subjects

Table 1 shows the age, sex, and racial distribution of the 154 people whose brains we studied.

TABLE 1.

Age, Sex, and Race of 154 Autopsied Subjects

	Age (Years)		Total (n = 154)
	30-39 (n = 59)	40-50 (n = 95)
Sex
Men	36	60	96
Women	23	35	58
Race
White	41	66	107
African American	14	28	42
Other	4	1	5

Staff at the Lieber Institute obtained a comprehensive medical history when consent for brain donation for research was granted. This information included general medical history, history of dementia and psychiatric disorders, medication history, history of alcohol consumption, use and abuse of prescribed analgesics and opiates, and use of illegal drugs. We also obtained information on history of head trauma and participation in contact sports.

The Office of the Chief Medical Examiner provided the cause and manner of death, based on a complete autopsy and toxicology examination.

Neuropathology

Brains were examined externally, weighed, and then cut into coronal slabs. Tissue blocks for histology were dissected, and other tissues were frozen. Tissue blocks for microscopic examination came from the superior frontal gyrus, middle and superior temporal gyri, inferior parietal cortex, occipital cortex, amygdala, hippocampus and entorhinal cortex, midbrain, pons, medulla, and cerebellum. Tissues were fixed in 10% buffered formaldehyde, processed, embedded in paraffin, and cut at 10-μm thickness. All sections were stained with hematoxylin and eosin. Selected sections were silver-stained by the Hirano method (Yamamoto and Hirano, 1986) and immunostained with antibodies against ubiquitin (Polyclonal Rabbit Anti-Ubiquitin, Dako North America, Carpinteria, California), phosphorylated monoclonal mouse anti-tau (PHF-1) (a gift of Peter Davies, PhD, Departments of Pathology and Neuroscience, Albert Einstein College of Medicine, Bronx, New York), and mouse monoclonal human against Aβ protein (clone: 6E10, Signet Laboratories Inc, Dedham, Massachusetts).

In brains with Aβ lesions, we performed single and double immunostains with antibodies for astrocytes (glial fibrillary acidic protein, DAKO North America, Carpinteria, California) and microglia (Iba1, Abcam, Cambridge, Massachusetts) to examine potential inflammatory reactions to the Aβ deposits.

All microscopic preparations were viewed by conventional light microscopy.

Genotyping

We genotyped all 154 brains for APOE by the method of Hixson and Vernier (1990).

RESULTS

Aβ Lesions

Among the 154 autopsied subjects, we found Aβ lesions in the brains of 5 women and 8 men, all of them 40 to 49 years old. In all 13 brains, the lesions were diffuse Aβ plaques scattered throughout the neocortex (Figure 1 and Table 2). Five of the brains showed Aβ lesions limited to the neocortex and not seen in the entorhinal cortex and hippocampus. In 8 brains, the Aβ plaques extended into the entorhinal cortex and transentorhinal cortices, and 4 of these 8 also showed plaques in the hippocampus (Table 2).

FIGURE 1.

Photomicrographs of neocortex from one of the autopsied subjects. The sections are immunostained with 6E10 antibody, revealing beta amyloid (Aβ) deposits and plaques. Panel A: Rare plaques. Panel B: Frequent plaques. Panel C: High magnification of a diffuse plaque shrouding several neurons. Panel D: Aβ deposit in the wall of a blood vessel.

TABLE 2.

Demographic Information, Neuropathologic Findings, and Apolipoprotein E (APOE) Genotypes of the 13 Autopsied Subjects Aged 40 to 50 Years Who Had Beta Amyloid (Aβ) Lesions

Subject	Age, Sex	Race	Location and Severity of Lesions											APOE Genotype
			Aβ*									Tau*
			Hippocampus	Transentorhinal Cortex	Entorhinal Cortex	Inferior Temporal Gyrus	Frontal Cortex	Superior & Middle Temporal Gyri	Inferior Parietal Lobule	Occipital Cortex	Blood Vessels	Mesial Temporal Lobe	Neocortex
1	47, W	AA	–	+	+	+	+	+	+	+	++	–	–	E4/E4
2	43, M	W	+	na	++	na	++	+++	+++	+	+	–	–	E4/E4
3	49, M	W	+	+	+	+++	+++	+++	+++	+++	+	–	–	E4/E4
4	46, W	W	+	+	+	++	+++	+++	+++	+++	–	–	–	E3/E4
5	43, M	W	na	na	na	na	++	+++	++	–	–	–	–	E3/E4
6	47, M	W	–	–	–	+	+	+	+	+	–	–	–	E3/E4
7	44, M	AA	–	+++	++	+++	++	+++	++	++	–	+	–	E3/E4
8	40, M	W	–	–	–	++	+	+	+	+	–	–	–	E3/E4
9	45, W	W	–	+++	+++	na	++	+	+	+	+	+	–	E3/E4
10	45, W	W	–	++	–	++	+	+++	+	++	–	–	–	E3/E4
11	48, W	W	–	+	+	+++	++	++	++	+	+	+	–	E3/E4
12	44, M	AA	+	+	+	+++	++	+++	+++	+	–	+	–	E3/E4
13	47, M	W	–	–	–	+	+	–	+	+	–	+	–	E3/E4

^*Number of lesions: – = none. + = rare. ++ = moderate. +++ = frequent.

W = woman. M = man. AA = African American. W = white. Aβ = beta amyloid protein. na = not available. Mesial Temporal Lobe = hippocampus and/or entorhinal cortex.

The severity of Aβ deposits was fairly uniform throughout the temporal, frontal, and parietal lobes, but in some brains it appeared lower, in the occipital regions. The pattern of Aβ deposits was also somewhat variable, ranging from very diffuse blotches of immunoreactivity to more condensed plaques (Figure 1).

We found vascular Aβ in 5 brains. While all 3 brains with APOE genotype 4/4 had vascular Aβ, only 2 of 10 brains with APOE3/4 had vascular Aβ (Table 2). We did not notice a correlation between vascular Aβ and severity of parenchymal Aβ deposits.

None of the 13 brains with Aβ deposits showed evidence of astroglial or microglial reactions.

Coexistence of Aβ and Tau Lesions

Among the 13 brains with Aβ plaques, 5 had rare and subtle tau lesions limited to the entorhinal cortex and/or hippocampus, with no involvement of the neocortex (Table 2). In 1 brain, we noted rare neurofibrillary tangles in both regions; in the other 4, the tau lesions were confined to the entorhinal cortex and consisted of punctate structures in the neuronal perikaryon and subtle neuropil threads similar to lesions in stages 1a and 1b of Braak et al (2011) (Figure 2). We saw no tau lesions in the neocortex. The remaining 8 brains bearing Aβ plaques were free of tau abnormalities in the cerebral cortex.

FIGURE 2.

Photomicrographs of hippocampus from one of the autopsied subjects. The sections are immunostained with phospho-tau antibody. Panel A: Pyramidal neuron with immunoreactive organelles in the perikaryon and proximal dendrites. Panel B: Neuropil threads in the entorhinal cortex.

APOE4 Genotypes

As shown in Table 2, all 13 individuals whose brain had Aβ deposits were 40 to 49 years of age and all carried 1 or 2 APOE4 alleles.

For our total series of 154 brains, Table 3 shows the distribution of APOE alleles by age group and race. We found no significant differences in the APOE allele frequency between the subjects 30 to 39 years old and those 40 to 50 years old, or between whites and African Americans.

TABLE 3.

Apolipoprotein E (APOE) Gene Alleles in Autopsied Subjects Aged 30-39 versus 40-50 Years

Race	# of Alleles	APOE Allele
		E2	E3	E4
59 subjects aged 30-39 APOE alleles: 59 × 2 = 118
White	82	9 (10.9%)	59 (71.9%)	14 (17%)
African American	28	3 (10.7%)	20 (71.4%)	5 (17.9%)
Other	8	0	8 (100%)	0
Total	118	12 (10.17%)	87 (73.7%)	19 (16.1%)
95 subjects aged 40-50 APOE alleles: 95 × 2 = 190
White	132	13 (9.8%)	93 (70.4%)	26 (19.7%)
African American	56	7 (12.5%)	38 (67.8%)	11 (19.6%)
Other	2	0	2 (100%)	0
Total	190	20 (10.5%)	133 (70%)	37 (19.5%)

Table 4 shows the APOE genotype distribution in the 95 subjects aged 40 to 50, with and without Aβ. All 3 brains with APOE4/4 had Aβ deposits. Of the brains with APOE3/4, 10 (36%) had Aβ deposits, but 18 (64%) did not. The brains with other genotypes, including APOE 2/4, were free of Aβ.

TABLE 4.

Apoprotein E (APOE) Genotypes in 95 Autopsied Subjects Aged 40-50 Years, by Beta Amyloid (Aβ) Status

APOE Genotypes	n	Aβ Status
		Positive	Negative
2/2	2	0 (0%)	2 (100%)
2/3	13	0 (0%)	13 (100%)
2/4	3	0 (0%)	3 (100%)
3/3	46	0 (0%)	46 (100%)
3/4	28	10 (36%)	18 (64%)
4/4	3	3 (100%)	0 (0%)
Total	95	13 (13.5%)	83 (86.4%)

Table 5 lists the cause and manner of death for each of the 13 subjects with Aβ deposits. As outlined in Table 6, we did not observe any significant differences in the manner of death between the subjects with and those without Aβ deposits.

TABLE 5.

Cause and Manner of Death in 13 Autopsied Subjects with Aβ Deposits

Subject #	Cause of Death	Manner of Death
1	Bupropion intoxication	Suicide
2	Narcotic intoxication	Undetermined
3	Multiple injuries	Accident
4	Arteriosclerotic cardiovascular disease and narcotic intoxication	Undetermined
5	Fentanyl and chlordiazepoxide intoxication	Undetermined
6	Drowning, alcohol and mixed drug intoxication	Undetermined
7	Drowning	Accident
8	Alcohol and narcotic intoxication	Undetermined
9	Cardiac arrhythmia	Natural
10	Arteriosclerotic cardiovascular disease	Natural
11	Arteriosclerotic cardiovascular disease	Natural
12	Intraoral gunshot wound	Suicide
13	Megacolon	Natural

TABLE 6.

Manner of Death and Beta Amyloid (Aβ) Status of 95 Autopsied Subjects Aged 40-50 Years

Manner of Death	Aβ Status
	Negative (n = 82)	Positive (n = 13)
Natural	40 (48.8%)	4 (30.8%)
Accident	13 (15.9%)	2 (15.3%)
Suicide	9 (10.9%)	2 (15.3%)
Undetermined	20 (24.4%)	5 (38.4%)

Table 7 compares the medical histories for the 40- to 50-year olds by their Aβ status. The histories were similar except for more heart disease in the Aβ-positive group.

TABLE 7.

Medical History of 95 Autopsied Subjects Aged 40-50 Years, by Beta Amyloid (Aβ) Status

History	Aβ Status
	Negative (n = 82)	Positive (n = 13)
Schizophrenia	5%	8%
Affective disorder	42%	31%
Alcohol abuse	23%	23%
Drug abuse	31%	23%
Use of narcotic analgesics	2%	8%
Traumatic brain injury	1%	0%
Post-traumatic stress disorder	6%	0%
Suicide attempts	4%	8%
Heart disease	4%	23%

Some subjects had > 1 psychiatric disorder, traumatic brain injury, or type of cardiovascular disease.

DISCUSSION

Our major finding in this study was Aβ deposits in 13 of 95 (13.7%) autopsied brains from non-demented individuals between 40 and 50 years of age. The frequency of Aβ lesions in this age range was 13.3% in men, 14.3% in women, 15.5% in whites, and 10.4 % in African Americans. We did not find Aβ deposits in any individual younger than 40.

This last observation is consistent with a study by Braak et al (2011) reporting that Aβ deposits started to appear at approximately age 40 and affected 4% of individuals between 40 and 50. Previous studies from our laboratory have also shown that Aβ lesions started to appear in the early 40s, and their frequency was 7.8% in 40- to 54-year-olds, without differences between the sexes or races (Sandberg et al, 2001).

Therefore, it appears that the beginning of the fifth decade is the inflection point for development of Aβ deposits and senile plaques, which may be advanced by genetic factors such as Down syndrome (Burger and Vogel, 1973; Cork, 1990; Mann and Esiri, 1989) or amyloid precursor protein and presenilin mutations (Rogaeva, 2002).

The Aβ lesions that we saw in this study were diffuse plaques, some perineuronal, and not associated with obvious neuritic or inflammatory components. These plaques were similar to ones that we had reported in the preclinical stages of AD, associated with endosomal and lysosomal abnormalities (Troncoso et al, 1998). We found the preclinical plaques throughout the neocortex, although less frequently in occipital regions.

Because in the present study we found these lesions only in the subjects aged 40 or older, the lesions likely represent the very early stages of AD pathology. Our belief is reinforced by finding that each of the affected individuals had at least 1 APOE4 allele.

Notably, only 5 of the 13 brains with Aβ deposits had any tau pathology in the cerebral cortex. The tau lesions were confined to the mesial temporal lobe. None were in the neocortex, where Aβ deposition was most severe. This strongly suggests that Aβ deposition is the earliest event in AD, preceding the development of local tau lesions and neurofibrillary tangles.

Our observations support the amyloid hypothesis of the pathogenesis of AD, which proposes that abnormal Aβ processing, deposition, and clearing are the primordial events in AD, coming before inflammatory changes, tau lesions, and neuronal degeneration (Hardy, 2009; Hardy and Higgins, 1992; Selkoe, 1991).

Few studies have been done of AD biomarkers in normal young people. We compared our findings to those of Morris et al's (2010) clinical biomarkers study in cognitively normal aging, and we found both differences and similarities. We detected Aβ deposits in 13.7% of our subjects 40 to 50 years of age, but Morris et al did not find any Aβ deposits on PiB PET scans of subjects aged 45 to 49. This discrepancy suggests that PiB PET scans may not be sensitive enough to detect the very early stages of Aβ deposition.

Our results concur with the finding by Morris et al that 18.2% of their subjects had low CSF Aβ42 concentrations. Both studies support the notion that this biomarker is sensitive to very early Aβ deposition. Studies of CSF biomarkers in autosomal-dominant AD have associated reduced CSF Aβ42 with Aβ plaques as well as with elevated CSF tau in asymptomatic mutation carriers; both markers can be found 10 to 20 years before the estimated age of onset (Fagan et al, 2014).

The APOE4 allele is the major known risk factor for the development of late-onset sporadic AD (Corder et al, 1993). This effect may be mediated by the increased Aβ deposition in the brain and by changes in synaptic repair and plasticity (Verghese et al, 2011). Furthermore, individuals carrying the APOE4 allele deposit Aβ in the brain in a dose-dependent manner during the preclinical stages of the disease (Morris et al, 2010; Reiman et al, 2009).

In our sample, all subjects with Aβ deposits in the brain carried 1 or 2 APOE4 alleles, reaffirming the importance of APOE4 for Aβ deposition (Corder et al, 1993). All 3 of our subjects who had the E4/E4 genotype had Aβ deposits. Although our number of APOE4-homozygous subjects was small, the finding that 100% of them had Aβ deposits versus only 36% of subjects with APOE3/4 is congruent with the concept of the gene dose effect of APOE (Morris et al, 2010; Reiman et al, 2009).

Based on our cross-sectional observation, we cannot predict whether these individuals, had they lived longer, would have developed cognitive decline or dementia. However, considering the epidemiologic data on the APOE4 carriers, some probably would have. Importantly, many of the subjects bearing a single APOE4 allele did not show Aβ deposits, suggesting that other genetic (Guerreiro et al, 2012, 2013) and environmental (Wirth et al, 2014) factors can modulate the risk effect of APOE for Aβ deposition and development of AD.

We also note that the frequency of the E4 allele in our African American subjects aged 40 to 50 was 19.6%, consistent with the 19.5% frequency reported in a previous study (Riudavets et al, 2006).

A major limitation of our study was the nature of the sample. People coming to autopsy in a medical examiner's office do not necessarily represent the population at large, and, in particular, may include a disproportionate number of men and women with a history of substance abuse and/or mental health disorders. This limitation, however, is balanced by the opportunity to examine a large number of brains in a short period of time. Furthermore, the composition of our sample is valuable as it includes a large proportion of African Americans, who have not been well represented in autopsy studies of AD.

In conclusion, our study demonstrates that APOE4 carriers as young as 40 years of age may have asymptomatic or preclinical AD characterized by Aβ deposits in the cerebral cortex. Our finding highlights the need to target future AD preventive and therapeutic measures at young people, perhaps triaged on the basis of genetic risk, and the need to develop sensitive biomarkers of AD that can be applied to people in their 40s or even younger.

IN MEMORIAM.

This work is dedicated to Dr Oscar S.M. Marin, my mentor, friend, and fellow countryman.

I went to medical school in Santiago, Chile, at the Universidad Católica, where Oscar Marin had been Professor of Neurology from 1961-1965. I arrived in 1967. During medical school, although I was supposed to become a surgeon, I fell in love with neurology and devoted as much time as possible to getting experience in neurology and neuroscience.

Only years later did I recognize how much Dr Marin had shaped both the clinical and academic environment in which I had developed my interest. He had either created or heavily influenced the hospital neurology services in which I learned, and he had taught my professors. Dr Jaime Court, my professor of neurology and mentor, had been a student of Dr Marin and coauthored a 1965 paper with him on necrotizing arteries.

After I graduated in 1973, I wished to enter a neurology training program in Santiago. But 1973 was also the year of the military coup in Chile. Things got quite difficult, and my wife, Gloria, and I decided to leave the country.

I wrote to many institutions outside Chile before finding help from a friend of mine, Dr Fernando Miranda, who was doing a neurology residency at the University of Maryland in Baltimore [see Dr Miranda's article in this issue of the journal]. Dr Miranda helped me obtain an internship in internal medicine at the South Baltimore General (now Harborview) Hospital. This internship was a prerequisite for applying for a neurology training program.

We arrived in Baltimore in the last week of June 1974—Gloria, who was pregnant with our future son Felipe, and I. On July 1, I began working at South Baltimore General Hospital.

After settling for a few weeks, I decide to visit Dr Marin, who was then on the faculty at the Johns Hopkins University School of Medicine in Baltimore. It was probably late July when I drove to the then Baltimore City Hospitals to meet Dr Marin. I brought a letter of reference from Dr Juan Allamand, a renowned Chilean surgeon who knew him.

I explained my situation to Dr Marin and asked him for advice about seeking training in neurology. He was very warm and receptive. He gave me his perspective on the current neurology training landscape and recommended 3 or 4 good East Coast neurology programs where I would be competitive. One of his recommendations was Hahnemann University Hospital in Philadelphia, where I eventually did my neurology residency under the mentorship of Dr Elliott Mancall.

In addition to giving professional advice, Dr Marin invited Gloria and me to visit him at his home. That was when we met the women of the Marin family, Señora Clara, Clarita, and Andrea (Oscar Jr was away). We saw each other socially several times. Then the Marins asked us if we would house-sit for them in December and January, while they stayed at their apartment in St Denis, Paris. To make a long story short, we were living at the Marins’ house when Felipe was born in early January of 1975.

That July, we moved to Philadelphia for my neurology residency. There I became very interested in neuropathology while working for 6 months with Dr David Dunn at the Medical College of Pennsylvania. In 1978, at the end of my clinical training, Dr Marin helped me get a neuropathology fellowship at Johns Hopkins. He was one of my sponsors, along with Drs Donald Price and Guy McKhann. Shortly after I began my neuropathology training, Dr Marin asked me to start doing brain cuttings and neuropathology conferences with him at Baltimore City Hospitals. During my second year, I also began attending as a clinical neurologist at Baltimore City Hospitals.

After Dr Marin left for Portland, Oregon, in 1979, I visited him several times. I have stayed on at Johns Hopkins.

In retrospect, I appreciate how much Dr Marin influenced my career even before I met him, because he was largely responsible for the thriving neurology environment that so excited me in medical school in Chile. Then, as I moved to the US, he gave me good advice and strongly supported my professional development. Dr Marin was a thoughtful mentor who guided my career wisely and at the same time was a wonderful friend.

— Juan C. Troncoso, MD

Glossary

AD

Alzheimer disease

Aβ

beta amyloid

APOE

apolipoprotein gene

CSF

cerebrospinal fluid

PET

positron emission tomography

PiB

Pittsburgh compound B