Unexpected Autopsy Findings Arising From Postmortem Ocular Examination

Postmortem ocular examination can aid in identifying the cause of death, provide clues regarding the length and severity of multisystem disease processes, and reveal unexpected findings. Despite these potential benefits, eyes are not routinely included in postmortem examinations at many institutions. To our knowledge, no reports have addressed the prevalence of ocular findings in individuals who come to autopsy. In the present study, we detail the range of ophthalmic findings encountered at autopsy and address their clinical relevance.

We retrospectively reviewed postmortem ocular findings from unrestricted autopsies performed at our institution between 1995 and 1999. Our institutional autopsy authorization permit specifically states that the eyes may be removed as part of an unrestricted autopsy. This study was deemed exempt from Institutional Review Board purview, as Title 45, Code of Federal Regulations, Part 46 (Protection of Human Subjects1) defines a human subject as a living individual.

After formalin fixation, gross inspection of the enucleated globes was performed according to standard methods.2 Histologic assessment was made on a hematoxylin-eosin–stained central horizontal section of each globe. Additional tissue sections and other histochemical stains were evaluated in individual cases as necessary. Ocular findings were graded according to severity. Category I consisted of major findings that contributed to death, were likely to have required patient management, and/or may have important medical implications for close relatives. Minor findings that may have eventually required treatment were classified under category II, as were findings that would be reasonably expected after ocular surgery. Category III included findings considered to be incidental. Individual cases could be assigned more than one categorical designation, based on the number and severity of ocular findings.

Eyes were removed for pathologic evaluation in 277 unrestricted autopsies at our institution between 1995 and 1999, representing 20% of all postmortem examinations conducted during the 5-year period. Decedents ranged in age from less than 1 to 95 years. The median age was 62 years. The study population included 149 male and 128 female decedents. Major findings (category I) were identified in 32% of autopsies (Table 1). These findings ranged from age-related macular degeneration, a leading cause of irreversible blindness in the United States, to exceedingly rare ophthalmic disorders (Figures 1 through 10). We documented minor, or category II, findings in 62% of autopsies (Table 2). Within this category, uveitis was most commonly encountered. Thirty-four percent of cases contained ophthalmic findings that we considered incidental (category III), which are outlined in Table 3. The number and severity of ophthalmic findings was sufficient to warrant more than one categorical designation in 35% of autopsies. Only 14% of postmortem examinations revealed no pathologic ophthalmic diagnoses.

At many institutions, removal of the eyes is not a routine part of the postmortem examination. Pathologists are reluctant to remove the eyes at autopsy for a variety of reasons. Ophthalmic pathology is a field in which most pathologists have only limited experience. Some pathologists do not consider examination of eyes central to the clinical concerns to be addressed at autopsy. Additionally, there is a misconception that removal of the eyes causes cosmetic distortion of the face. When performed properly, the superior approach and the more commonly employed frontal method of postmortem enucleation are both nondistorting.

Autopsies play an essential role in establishing clinicopathologic correlations. In diseases of the retina and other ocular sites not readily amenable to tissue biopsy during life, postmortem examination is the only way to histologically confirm clinical impressions. In forensic investigation, examination of the eyes is critical in cases of apparent accidental death in which visual impairment may have been a contributing factor. Ophthalmic findings can also corroborate other evidence suggestive of intentional trauma, as in cases of shaken baby syndrome.3 

The most frequently encountered ophthalmic condition in this series, identified in 14% of autopsies, was diabetic retinopathy. The prevalence of diabetic retinopathy is high. Fifty percent of patients will develop retinopathy within 7 years after the onset of diabetes.4 Not surprisingly, nearly all patients in this series known to have diabetes mellitus exhibited some histologic evidence of diabetic eye disease.

Much of our understanding of age-related macular degeneration (ARMD), the principle cause of vision loss in older individuals in the United States,5 has been derived from postmortem ophthalmic studies. Prevention and treatment of this prevalent disorder remains a formidable challenge. Age-related macular degeneration was the second most common major finding in this series. The rate of ARMD in our series was somewhat less than the estimated prevalence of 9.4% in individuals aged 40 years or older in the United States.6 

The reported prevalence of glaucoma, a common cause of blindness in the United States, ranges from 1% to 6%.7,8 Histologic manifestations of glaucoma, which include cupping of the optic disc and optic nerve atrophy, usually require long-standing elevated intraocular pressure. Mild or recent-onset increased intraocular pressure may not be apparent histologically. For this reason, it is likely that in postmortem series such as ours, in which histologic manifestations of glaucoma were found in 2% of cases, the prevalence of glaucoma is likely underestimated.

The most commonly diagnosed intraocular neoplasm in the series was malignant melanoma of the uveal tract, a finding that is consistent with data from large eye tumor registries.9 One of the cases had been noted clinically and was histologically confirmed at autopsy. The second case was a clinically occult, small (5 mm in diameter) peripheral choroidal melanoma arising in a patient with Parkinson disease. The third case proved to be a metastasis from a skin melanoma. Secondary involvement of the eye by metastatic tumors has been shown to occur more frequently than primary intraocular melanoma.9 Visceral malignancies, in particular, those of breast and lung origin, have a propensity to metastasize to the eye. We detected metastatic carcinoma in 2 of our cases (1 breast, 1 pancreatic).

In addition to revealing common ophthalmic disorders, postmortem ocular examination uncovered a number of rare ophthalmic conditions. One case of retinitis pigmentosa was confirmed histologically at autopsy in a patient with a positive family history who had not undergone a full ophthalmologic evaluation while alive. The annual incidence of this hereditary condition in the United States is approximately 1 in 3500.10 There were also 2 cases of pigmentary retinopathy similar to retinitis pigmentosa. One was secondary to vitamin E deficiency resulting from celiac sprue, whereas the other was of uncertain etiology.

Another rare ophthalmic condition that was discovered at autopsy has important medical implications for close relatives. Bilateral nanophthalmos, or simple microphthalmos, is a condition in which the total axial length of the eye is reduced. Individuals with this heritable disorder have a high incidence of serious complications after ophthalmologic surgery.11,12 

Bilateral diffuse uveal melanocytic hyperplasia is an exceedingly rare paraneoplastic syndrome characterized by progressive vision loss associated with uveal thickening, retinal detachment, and rapid cataract formation. Since the initial description of this entity in 1966, fewer than 30 cases have been reported.13,14 We identified one case of bilateral diffuse uveal melanocytic hyperplasia. To our knowledge, this is the first case noted to have developed in association with histologically proven pancreatic adenocarcinoma.

It has long been recognized that a high percentage of autopsies yield important unexpected findings.15 Approximately 10% of autopsies produce unexpected findings significant enough to have changed patient management had they been recognized before death.2,15,16 Of the ophthalmic findings in our series that were severe enough to have impacted on patient management, a large proportion were not noted in the clinical records reviewed at the time of autopsy.

The detection of at least one pathologic ophthalmic finding that would have significantly impacted on clinical management and/or had important ramifications for close relatives in one third of cases in our series underscores the relevance of postmortem ocular examination. It is important that clinicians and families understand that postmortem study of eyes is potentially as beneficial as other parts of the autopsy. The clinicopathologic information gathered from postmortem ocular examination improves both ophthalmologic care of the living, as well as our understanding of ophthalmic disorders in general.