Notable Histologic Findings in a “Normal” Cohort: The National Institutes of Health Genotype-Tissue Expression (GTEx) Project Open Access

An average of 26 tissue types were collected from each donor at a variety of biospecimen source sites (BSSs), which included several medical centers and organ procurement organizations (OPOs) in the United States. Although inclusion and exclusion criteria for GTEx and details relating to consent, tissue procurement, and preservation have been reported previously,^2,3,6  the donor criteria applied may be study specific because of innate differences in aim and scope. Briefly, research authorization was obtained from a member of each donor’s family.^3,6,7  In the present study, any of the following resulted in donor exclusion: an antemortem transfusion within 48 hours of death, a history of metastatic malignancy, exposure to HIV/AIDS or the hepatitis C virus within the 5 years preceding death, repeated positive reactive screening tests for HIV-1 or HIV-2 antibodies, active sepsis, or multiple documented or confirmed pathologic conditions.

Eligible GTEx donors for the present study were between 21 and 70 years of age at the time of death and had a body mass index (BMI) between 18.5 and 35. A minimum of 5 essential target tissues (skeletal muscle, tibial nerve, tibial artery, skin, and adipose tissue) were required from each consented donor. In total, 40 different tissue types were collected for histologic evaluation. The number of aliquots collected per donor varied, secondary to the sex of the decedent, prior explantation to organ recipients, injury (traumatic death) or antemortem loss of an organ, or other proscriptions from the donor or authorizing next-of-kin. A list of all target donor tissues is provided in the Supplemental Table (see supplemental digital content at https://meridian.allenpress.com/aplm in the March 2025 table of contents). Demographic and clinical data collection was completed according to the GTEx Clinical Collection Case Report Form.⁸ 

GTEx prosectors were experienced staff at their respective BSS who underwent training and followed detailed guidance outlined in GTEx Standard Operating Procedures (SOPs),⁹  which specified the preferential collection of normal-appearing tissue and instructions to record any abnormalities observed. Additionally, BSSs provided and received ongoing feedback from project pathologists. Eligible GTEx donors were designated as either brain (n = 419) or nonbrain (n = 541) donors. For nonbrain donors, the first tissue was placed in PAXgene fixative (Qiagen, Germantown, Maryland) within 8 hours of documented cardiac cessation or recorded time of death. For brain donors, all tissues were collected and placed into PAXgene fixative within 24 hours of cardiac cessation. Aliquots were taken from tissue that appeared grossly normal.⁹  If no normal tissue was apparent, the organ was sampled, and abnormalities were recorded. Aliquots were trimmed to 10 × 10 mm with a thickness 4 mm or less to allow for uniform embedding in a tissue cassette. Tissues normally surrounded by adipose or other connective tissue (eg, arteries, nerve, adrenal) were trimmed prior to embedding to ensure tissue purity. Samples from mucosa-lined hollow viscera (stomach, colon, ileum) were collected as 10 × 10 mm aliquots and then trisected prior to paraffin embedding to facilitate “on-edge” histologic evaluation.

All nonbrain tissues were preserved in PAXgene fixation solution for 6 to 24 hours before transfer to PAXgene stabilizer (Qiagen) and paraffin embedding. For brain tissue, aliquots were collected from the cerebellum (cerebellar hemisphere) and cortex (frontal pole region of the frontal cortex) and preserved in PAXgene fixative as described for nonbrain tissue, whereas the remaining brain was shipped on dry ice for the expert dissection of 8 to 11 subregions and preservation by snap-freezing for molecular study; GTEx project pathologists did not conduct histologic review of the brain subregions beyond the 2 regions specified. All tissues were preserved according to project-specific SOPs.⁹  The PAXgene Tissue Fixation System was chosen because of the excellent preservation of both histomorphology and nucleic acids and to obviate the need for freezing and shipment of biospecimens using dry shippers from remote BSSs.¹⁰ 

One hematoxylin-eosin–stained slide from each PAXgene-fixed, paraffin-embedded tissue aliquot was prepared. A digital whole-slide image (at ×20 magnification) was generated using an Aperio Scanscope (Leica Biosystems). All images were reviewed by centralized project pathologists to (1) verify the correct target tissue, (2) assess and grade autolysis, (3) observe and record pathologic findings, and (4) verify the size and number of tissue pieces. Histologic assessments were reviewed and recorded prior to molecular analysis. Instances of diseased, severely autolyzed, or incorrectly collected (ie, the wrong target tissue) tissue aliquots were noted in electronic pathology reports to facilitate exclusion and prevent potential contamination of normal genetic and epigenetic data sets.

A Fisher exact test was used to compare the rate of occurrence of a microcystic expansion of the mid to deep gastric antral glands among GTEx donors who received mechanical ventilation prior to organ procurement to those that did not. One-tailed binomial probability was used to determine directionality. Statistical significance was set at P < .001. Statistical analyses were conducted using R software (version 4.0.2).¹¹ 

Although eligible GTEx donors represented a range of demographics (age, sex, race), most of the 960 eligible GTEx donors in the present study were white (814 donors, 85%) and male (644 donors, 67%; Figure 1; Table 1). The proportion of male to female donors was similar across donor age groups (male, 57%–70%; female, 30%–41%). Eligible GTEx donors ranged in age from 21 to 70 years; however, 654 of the 960 donors (68%) were between the fifth and seventh decade at time of death (Figure 1).

Many GTEx cases, particularly those from older individuals, had acute or chronic changes in tissues that, although not diseased, were not strictly normal. The abnormalities and chronic lesions that were more frequently observed among older GTEx donors are summarized in Table 2. Some of these abnormalities were changes induced by damage that occurs as a result of normal use and/or aging (eg, loss of aortic medial elasticity), whereas others were due to age-related hormonal alterations (eg, reduced spermatogenesis and endometrial atrophy). Other findings included histologic changes that likely developed over time from an underlying condition such as hypertension or diabetes. As reviewed in a companion paper,⁵  tissue samples that contained either chronic or acute lesions without extensive tissue loss were retained and considered acceptable for downstream molecular analysis, whereas those that displayed an extensive loss of normal tissue (eg, due to advanced cirrhosis or nephrosclerosis) were excluded. All abnormalities that were observed during histologic review of tissues from GTEx donors were also annotated in the pathologist’s report and have been made available along with other clinical and molecular data through dbGaP (dbGaP accession No. phs000424; https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs000424).

Despite the preferential sampling of normal-appearing tissue from GTEx donors, several notable incidental findings were observed by project pathologists during histologic review. These findings included features indicative of neoplastic, autoimmune, and genetic conditions and are discussed in detail below.

Incidental neoplastic findings were identified in 73 of the 960 postmortem GTEx cases (7.6%) and are described in Table 3. Of the 960 donors that underwent histologic evaluation, there were 55 cases that were either malignant or potentially malignant (5.7%), and the remaining 18 cases were benign (1.9%).

The vast majority of malignant and premalignant findings among GTEx donors were of ambiguous prognostic significance; for instance, the 4 thyroid papillary carcinomas observed did not exceed 2 to 3 mm in size, and the pancreatic intraepithelial neoplasia (PanIN) foci were all low-grade, in situ lesions (34 cases). Potentially significant tumors were observed in several donors and included diffuse non-Hodgkin lymphoma (1 case), posttransplantation lymphoproliferative disorder (1 case), and a prostate adenocarcinoma with a Gleason grade of 9 (1 case). Incidental and less aggressive carcinomas were present in other donors: prostatic adenocarcinoma (5 cases) and papillary renal carcinoma (3 cases).

Histologic evidence of Hashimoto thyroiditis that included diffuse interstitial lymphoid infiltration and fibrosis, atrophy, and/or scarring (Figure 2 shows one such case) was present in 85 of 898 donors from whom thyroid tissue was procured (9.5%). The sex-specific incidence was 13.1% in female donors (40 of 306) and 7.6% in male donors (45 of 592).

Thyroid neoplasms were also observed in 6 donors. Thyroid tissue samples from 4 donors contained subcentimeter papillary carcinomas. Samples from the remaining 2 donors contained neoplasms of uncertain malignant potential, as noted in Table 3.

Of the 931 donors from whom stomach samples were procured, 28 (3%; 21 male and 7 female donors) displayed a novel microcystic expansion of the mid to deep gastric antral glands, which was typically multifocal and unassociated with inflammation. However, 1 donor also exhibited evidence of patchy chronic active gastritis, with deep microabscesses present in other submitted aliquots/sections. The histologic appearance of the microcystic expansions observed in stomach sections of the 28 donors bears some similarity to the microcystic metaplasia seen in fundic gland polyps, but no other manifestations of polyps were observed in these GTEx donors, and the cystic change occurred in otherwise normal basal mucosa (Figure 3). At the time of death these donors were 21 to 70 years of age with a BMI of 23 to 33. Notably, 27 of the 28 affected donors (96%) received mechanical ventilation for 49 to 180 hours (2–8 days) prior to organ procurement, suggesting these abnormalities may be an agonal response to prolonged ventilation. The condition was observed in 27 of the 505 donors who received mechanical ventilation prior to death (5.4%) but in just 1 of the 430 donors who did not (0.2%), a difference that was significant (P < .001), affecting significantly fewer donors who did not receive mechanical ventilation relative to those who were on a ventilator preceding organ procurement (P < .001).

A total of 6 of the 960 donors (<0.1%) had incidental leiomyomata of the distal esophagus/gastroesophageal junction.

Macrovesicular steatosis was identified in 154 of 607 donors (25%) from whom liver specimens were collected. Incidence was similar among male (112 of 416 male donors; 27%) and female (42 of 191 female donors 22%) donors. The average BMI of affected donors was 28.1.

Other hepatic incidental findings included amyloidosis and cirrhosis. One case of amyloidosis was identified in a donor who also had manifestations of extensive amyloid deposits in the heart and kidneys (systemic amyloidosis). There were no lymphoid elements present in samples from this donor that would suggest lymphoma or plasmacytoma. Primary hepatic and “cardiac” cirrhosis were also observed at various degrees of intensity in numerous donors.

An incidental neuroendocrine tumor was identified in the pancreas tissue of 1 donor. Of the 861 donors from whom pancreas specimens were collected, 34 (4%) contained 1 or more foci of PanIN as defined by Hruban et al,¹²  all of which were low grade (PanIN-1).

Collection of high-quality kidney specimens proved challenging during the GTEx project. Histologic evaluation revealed moderate to advanced autolysis (score 2 or 3) in 438 of the 518 kidney samples collected from GTEx donors (84.6%); this corresponded to a slightly lower RNA integrity number than kidney specimens with no or mild autolysis (4.19 versus 5.35, respectively).⁵ 

Nodular glomerulosclerosis (Kimmelstiel-Wilson type) is a common syndrome that is present in many older patients. It is a characteristic of diabetic nephropathy and of arteriolar nephrosclerosis that is typically associated with hypertension. The severity of this common pathologic condition was highly variable among GTEx donors and was difficult to assess via histology alone, particularly when accompanied by advanced autolysis. Because most GTEx kidney specimens demonstrated moderate to advanced autolysis, the incidence could not be determined with confidence.

In terms of neoplastic lesions, incidental papillary renal neoplasms were identified in kidney specimens from 3 donors; advanced autolysis precluded more definitive assessment.

Myocardial infarcts (acute or remote) were identified in 122 of the 743 donors (16.4%) from whom cardiovascular specimens were procured. Affected donors were between 44 and 67 years of age and included both sexes (Table 2). In 9 of the 743 donors (1.2%), the infarction was either acute or subacute and likely related to the immediate cause of death; the remaining infarctions were remote.

Other cardiovascular lesions encountered among the 960 GTEx donors included singular cases of lymphocytic myocarditis (0.1%) and cardiac amyloidosis (0.1%). Lymphocytic myocarditis was present in a 59-year-old female donor with diffuse myocarditis and extensive related interstitial fibrosis. Cardiac amyloidosis was present in a 56-year-old male donor with diffuse amyloidosis that, in addition to the heart, also involved the liver, spleen, and kidney.

Monckeberg medial calcific sclerosis is a focal to ringlike calcification of the vascular media of small- to medium-sized vessels (Figure 4) without associated intimal thickening, although it may coexist with peripheral artery atherosclerotic disease. It is a well-known incidental finding in peripheral muscular arteries, particularly the leg, and more recently has been identified as a cause of nonneoplastic microcalcifications that are observed in mammograms. Of the 960 donors from whom tibial artery specimens were collected, 165 (17.2%) displayed evidence of tibial artery Monckeberg sclerosis; 43 of 316 female donors (13.6%) and 122 of 644 male donors (18.9%) were affected.

Gynecomastoid hyperplasia is a benign enlargement of the male breast that typically involves both stromal and ductal hyperplasia (Figure 5, A). Most cases are idiopathic in that no definite etiology can be determined. The histologic abnormalities reflect either an imbalance in the ratio of estrogen to androgen levels in the tissue or altered responsiveness to these hormones within the target tissue. Gynecomastoid hyperplasia was identified in the breast tissue of 178 of the 584 male GTEx donors in whom breast tissue was obtained (30.5%). Of the 178 donors with the condition, 42 (23.6%) were younger than 50 years, and 136 (76.4%) were 50 years or older.

Klinefelter syndrome was confirmed by both histologic findings and molecular analysis in 4 of 644 male GTEx donors (ages 38, 48, 66, and 68 years), reflecting an incidence of 1%. The testes of the affected donors demonstrate profound tubular atrophy and hyalinization with microaggregates of Leydig cell hyperplasia (Figure 5, B), as well as mammary tissue displaying varying degrees of gynecomastoid stromal and/or ductal hyperplasia (Figure 5, A). Breast tissue that was collected from 2 of the affected donors also displayed evidence of gynecomastoid hyperplasia (donor ages, 38 and 67 years); breast tissue was not collected from the 2 remaining donors with Klinefelter syndrome.

Six cases of prostate adenocarcinoma were identified among male GTEx donors. Five cases consisted of microfoci of Gleason sum score 6 to 7; and 1 case showed diffuse Gleason sum score 9 (high grade) carcinoma.

Incidental uterine leiomyomas occurred in 6 female donors. The uterine aliquot in GTEx donors was requested to include both endometrium and myometrium. In the vast majority of cases, the endometrium, if present, was rarely evaluable because of advanced autolysis. Given most donors were 50 to 70 years of age, those aliquots in which viable endometrium was present typically showed inactive, basalis-type endometrium.

Two cases of diffuse lymphoma were encountered among GTEx donors, both involving multiple organ sites. One donor was afflicted with a posttransplantation-associated lymphoma (posttransplantation lymphoproliferative disorder), and the other donor displayed a large cell phenotype identified by histology. Immunohistochemical or flow cytometric evaluation was not possible in either case because of the limitations of the collection process.

To better understand the relationship between genetic variation and tissue-specific gene expression in nondiseased tissue, the GTEx project accrued specimens and molecular data from nearly 30 000 samples reflecting up to 40 tissue types collected from more than 900 donors. Each tissue aliquot was histologically evaluated by a pathologist to ensure the data set was representative of physiologically normal tissue. Outcomes of histologic evaluation included incidental findings of age-related and/or unexpected lesions, as well as challenges in procurement for several tissue types that are addressed in a companion paper.⁵  A greater incidence than reported in the literature was observed for several pathologic conditions, which is notable given the targeted collection of grossly normal-appearing tissue for the GTEx project. Collectively, results of the histologic evaluation of GTEx tissue specimens support systematic assessment of tissue quality for large community-shared resources to safeguard the authenticity of the targeted sample set, be it normal or representative of a specific pathologic condition. Histologic assessment of GTEx aliquots was performed for quality assurance purposes to ensure the correct tissue was collected (and correctly identified prior to molecular analysis), that the size was sufficient for analysis, and to assess tissue quality (via autolysis score). Thus, assessment was for exclusionary purposes by design, to segregate “nonnormal tissues” regardless of etiology. Histologic evaluation of each tissue aliquot by a project pathologist was a substantial but justified investment. Although BSS prosectors deliberately avoided abnormal-appearing areas in target organs and tissues, collection of abnormal specimens was not precluded given prosectors were instructed to obtain samples even when no normal tissue was apparent.^1–3,13  Approximately 5% of the tissue aliquots collected for GTEx were excluded from subsequent molecular analysis after histologic assessment identified extensive tissue abnormalities (such as cirrhosis), excessive autolysis, or the inclusion of extraneous tissue in the sample.⁵  Histologic review also revealed the presence of abnormal but commonly encountered lesions in grossly normal-appearing tissues. These relatively minor lesions were associated with a variety of abnormalities and included age-related changes. Although the aim of the GTEx project was to facilitate the study of gene expression in normal tissues, the boundary between normal and abnormal can be less than clear cut. Is congested lung, liver, or spleen normal? Will the presence of large amounts of blood mask the tissue’s molecular signature? What extent of lymphoid infiltrate in an early case of lymphocytic thyroiditis would affect the molecular characteristics of a thyroid sample (Figure 2)? Do atherosclerotic plaques invalidate molecular data from a coronary artery or aortic sample? What extent of hepatic, pancreatic, cardiac, or renal fibrosis is acceptable without compromising downstream molecular studies? Because of a lack of knowledge about how such conditions influence “normal” gene expression, tissue specimens containing minor lesions and abnormalities were deemed acceptable for subsequent molecular analysis, and notes characterizing all abnormal deviations were recorded in the electronic pathology report and included in dbGaP genomic sequence data submissions. Similarly, relationships between gene expression and age-related changes in tissue remain largely unknown. By including tissue with age-influenced deviations from normal, the GTEx collection provides an opportunity to compare gene expression in individual tissues of the young versus the elderly, which could potentially shed light on age-dependent differences in gene expression that parallel age-related tissue changes (ie, normal “wear and tear”). Such studies might also indicate the stability of gene expression despite tissue changes.

Several alterations in tissue histologic features were more pronounced in older than younger donors, but all were commonly associated with advanced age. In general, most age-related tissue changes were anticipated and held limited prognostic impact. For example, the mean age of GTEx donors who displayed a loss of aorta elasticity was 60 years, which is in concert with an increase in arterial stiffness that occurs with age.^14,15  Most GTEx donors who displayed evidence of reduced spermatogenesis were 50 years or older (82%), which is consistent with a decline in daily sperm production after the age of 50 years.^16,17 

Unexpected malignant and premalignant neoplasias were observed in 5.7% of GTEx donors, which is on the lower end of the 5% to 8% range reported by other postmortem and autopsy studies.^18–20  Possible reasons for this modestly lower incidence of neoplasms among GTEx donors include the preferential collection of grossly normal-appearing tissue and the limited number and size of tissue aliquots (a single 10 × 10 mm sample per tissue) that were procured and evaluated. These strategies differ widely from the multiple blocks collected and submitted during a conventional autopsy. Because definitive tumor grading and staging (size, margins) were not possible for GTEx specimens given these constraints, all specimens with malignant neoplasms were excluded from molecular analysis to maintain the database as one of “normal” genetic and epigenetic results.

The potentially significant tumors observed in several donors, including diffuse non-Hodgkin lymphoma, posttransplantation lymphoproliferative disorder, and a prostate adenocarcinoma with a Gleason grade of 9, raised concerns about whether an organ or tissue from the same donor might be transplanted or stored in a tissue bank, and whether there might in theory be health implications for a potential recipient of that tissue. The GTEx project excluded donors with a known history of metastatic cancer or chemotherapy or radiation within 2 years; therefore, any potential malignancy would not be included in the donor’s medical history and an OPO or tissue bank might not be aware of any potential risks. GTEx addressed these medical and ethical concerns through a deliberative process described in Lockhart et al.²¹  Because GTEx staff were blinded to the identity of the donor and the potential uses of that donor’s organs and/or tissues, the process allowed for independent pathology review and potential information sharing back to the OPO for their review and possible action.

PanIN is a neoplastic lesion that is accepted as one of three morphologic forms that are precursor lesions to invasive pancreatic carcinoma.²²  The reported incidence of PanIN identified during autopsy varies widely (26%–86%) and is likely influenced by age, sex, grade, the size of the lesion, and the degree of sampling (reviewed in Longnecker and Suriawinata²³ ). PanIN was histologically confirmed in 4% of GTEx donors, although the incidence may be higher given that 19% of pancreata (165 of 861) were not histologically evaluable because of advanced autolysis or saponification and that GTEx pancreatic aliquots were solely obtained from the midpancreas (body). Reports also conflict as to whether PanIN primarily occurs in the head¹²  of the pancreas or whether PanINs, regardless of grade, can be found at similar frequencies in the head, body, and tail²⁴  of the pancreas. Based on observations in GTEx specimens, findings support a diffuse occurrence in the pancreas.

Several of the histologic changes observed among GTEx donors had a prevalence that differed between sexes, although these, too, carried limited prognostic implications. One such condition was Hashimoto thyroiditis. The overall frequency of Hashimoto thyroiditis in GTEx donors was 9.5%, which was substantially higher than the routinely reported frequency of 1% to 4% in the literature.^25,26  Generally, higher frequency among female donors relative to male donors is typical of autoimmune conditions, which was also observed among GTEx donors (13.1% of donors with Hashimoto thyroiditis were female and 7.6% were male). Clinical diagnosis of Hashimoto thyroiditis typically depends on serologic parameters that include markers of hypothyroidism (depressed T3 and T4) and the presence of antithyroid antibodies, but also evidence of lymphocytic infiltrates in the thyroid. A recent retrospective study that evaluated the frequency of Hashimoto thyroiditis by ultrasound-guided fine needle aspiration in 761 patients during a 3-year period reported a frequency of 13.4%, slightly higher than the GTEx data, with approximately 7.4% being euthyroid (subclinical) and 6% presenting with clinical hypothyroidism.²⁷  The frequency of Hashimoto thyroiditis in the GTEx cohort supports their observation and suggests that preclinical lymphocytic or autoimmune thyroiditis is more common than generally believed.

Gynecomastia is a sex-specific condition but one whose frequency is also influenced by patient age. Three classic peaks for gynecomastia are reported in males: at birth, secondary to maternal hormonal action on male breast tissues; at puberty; and in the fifth and later decades of life. The first two age categories are absent in GTEx donors. The frequency of the condition among GTEx donors who are 50 to 70 years of age (76%) is similar to that reported by others (72%; ages 50–69 years).²⁸  The overall frequency of gynecomastoid hyperplasia among GTEx male donors was 30.4%, which falls close to the 32% to 65% range reported in other studies.^29,30  A sex-specific condition that had a higher frequency among GTEx donors than reported in the literature was Klinefelter syndrome. The frequency of Klinefelter syndrome in the population has been variously reported as 0.1% to 0.2%,³¹  and a study in Australia has posited the prevalence to be approximately 0.22% in that country.³²  The GTEx data suggest a frequency of 1%, roughly 5 to 10 times greater than reported in the literature. The frequency of Klinefelter has long been suspected to be much higher than reported because of a lack of clinical suspicion, public health awareness, and testing. Although testicular biopsy is suggestive of the diagnosis, genetic analysis is definitive and confirmed our findings in these donors.

In stomach specimens, a novel microcystic expansion of the mid to deep gastric antral glands was observed in 3% of all GTEx donors and 6% of donors who received mechanical ventilation. Although the small sample size precludes the definitive identification of related variables, the lesion was observed in both males and females, in donors within a 50-year age range (21–70 years), and in donors who were of normal weight as well as those who were obese. Notably, all but 1 donor (27 of 28; 96%) received mechanical ventilation, which ranged in duration from 2 to 8 days, prior to death, and the rate of occurrence was significantly higher in GTEx donors who were ventilated compared with those who were not (5.4% versus 0.2%). Mechanical ventilation has been associated with complications of several organ systems, including the gastrointestinal tract.³³  However, elucidation of a potential association between mechanical ventilation and gastrointestinal abnormalities is hampered by presentation of a variety of clinical symptoms, the possibility that the gastrointestinal abnormality was a direct result of an underlying disease, an artefact of postmortem interval or tissue processing, and a lack of histologic details. One possibility is that the cystic change that occurred in normal basal mucosa of affected donors may have been an agonal response to prolonged mechanical ventilation.

Nonneoplastic changes with sex equivalence that were present in GTEx donors at a higher rate than previously reported included Monckeberg sclerosis and several progressive lesions that included primary hepatic and “cardiac” cirrhosis; macrovesicular steatosis occurred at the high end of the accepted frequency. Monckeberg sclerosis, characterized by calcification of the middle layer of small and/or medium-sized arteries, was observed in tibial artery specimens of 13.3% of female and 19% of male GTEx donors. Frequency among GTEx donors far exceeded the approximately 1% that has been reported in the literature.³⁴  Notably, the tibial artery is not typically evaluated during routine hospital or forensic autopsy, and few assessments of the frequency of Monckeberg sclerosis are available. Although the frequenceis of progressive conditions may be higher than those reported in the general population, such conditions may be relatively overrepresented in GTEx donors given that grossly abnormal or diseased livers are rejected from explantation. This could have raised the relative frequency of progressive disease among GTEx donors, which included organ donors.

Macrovesicular steatosis, a form of hepatic steatosis identifiable by displacement of the nucleus by a large fat droplet, was observed among liver specimens of 25% of GTEx donors. The frequency of macrovesicular steatosis among GTEx donors is on the high end of the range reported for this condition (10%–24%).³⁵  Possible etiologies for steatosis are numerous, including alcoholic and nonalcoholic steatohepatitis, diabetes, obesity, and environmental/toxic and drug-related injuries or reactions. Although we cannot speculate on many of the etiologies, the mean BMI of affected donors was 28.1, which falls within the overweight range.

The incidental findings noted for GTEx aliquots should be interpreted with caution because of the limitations of collection and the potential for bias. Organ and tissue aliquots were small in size and number; therefore, tissue samples would not be expected to capture all existing conditions (such as malignancies). Further, prosectors were trained to avoid grossly abnormal-appearing areas in tissues and organs, hence the extremely low number of uterine leiomyomas in uterine aliquots, for example. Collection of a limited number of small aliquots that grossly appeared normal is in contrast to the usual extensive sampling of abnormal tissue in cases of recognized lesions, particularly with respect to the neoplasms identified during analysis of GTEx aliquots. Further, the “true” incidence of lesions observed among GTEx donors may in fact be higher than reflected in the results presented here, potentially because of limited sampling (the number, size, and/or location of procured aliquots), or assessment due to postmortem autolysis in GTEx aliquots. Reliance on trained prosectors at individual BSSs for sampling may also be viewed as a limitation of the study and may have affected the frequency of the conditions reported, given the possibility that lesions that were unnoticeable to a prosector may have been observed by a highly trained pathologist. However, effects due to sampling errors were minimized through continuous communication between BSS prosectors and project pathologists during the collection phase of the project. The incidences reported here should be interpreted with caution given that GTEx donors do not represent a discrete population or a random sampling of individuals but are a collection of donors who met multiple study inclusion criteria, which may have introduced bias during donor selection and tissue sampling. For example, because most GTEx donors were in their fifth to seventh decades of life, it is plausible that the incidence of age-related abnormalities may be inflated. Nonetheless, the frequency of several conditions may prove informative, as might the novel microcytic change noted in gastric aliquots that may reflect an idiosyncratic reaction or response to prolonged ventilation time, which has not been previously described.

Nearly 30 000 tissue samples were collected from 980 donors during the GTEx project to establish a shared resource for the scientific community for the purpose of exploring potential relationships between genotype and tissue-specific gene expression in normal tissue. Histologic evaluation of each tissue specimen was essential to confirm the correct target was procured, to assess tissue quality, and to verify the tissue was free of significant alteration or lesion. Results of this in-depth review revealed several incidental findings that included a potential perimortem agonal tissue reaction in basal dilated gastric glands, and higher than expected incidences of conditions such as Klinefelter syndrome, Monckeberg medial sclerosis, and Hashimoto/autoimmune thyroiditis. Given the limitations associated with the project, which include the number and size of aliquots collected and the biased collection of grossly normal-appearing tissue, the incidental findings observed should be interpreted with caution. Nonetheless, the histologic findings from the GTEx project may serve to improve populational awareness of several conditions and present a unique opportunity for others to explore age- and sex-influenced conditions.

This resource would not have been possible without the generous donation of biospecimens and data by individuals and their families. We acknowledge the contributions of the many partners in the GTEx project, including the GTEx Consortium, Roswell Park Cancer Institute, National Disease Research Interchange and their network of contributing Organ Procurement Organizations, the Miami Brain Bank, the Van Andel Research Institute, the Broad Institute, Leidos Biomedical Research Inc, the NIH Common Fund, the National Human Genome Research Institute (NHGRI), the National Institute of Mental Health (NIMH), and the NCI. We thank Maria M. Tomaszewski, MD, for her contributions as a project pathologist and James Robb, MD, for his valuable contributions during protocol development. We also thank Brian Luke, PhD, for performing the statistical analysis. We gratefully acknowledge the leadership of Carolyn Compton, MD, PhD, in the early stages of this work. The GTEx project was supported by the Common Fund of the Office of the Director of the NIH, and by NCI, NHGRI, NIMH, the National Institute of Neurological Disorders and Stroke, the National Institute on Drug Abuse, and the National Heart, Lung, and Blood Institute.