Selection Criteria Optimal for Recovery of Inner Ear Tissues from Deceased Organ Donors

Abstract

Objective:

To identify optimal conditions for recovering viable inner ear tissues from deceased organ donors.

Setting:

Tertiary recovery hospitals and Donor Network West Organ Recovery Center.

Interventions:

Recovering bilateral inner ear tissues and immunohistological analysis.

Main Outcome Measures:

Immunohistochemical analysis of utricles from human organ donors after brain death (DBD) or donors after cardiac death (DCD).

Results:

Vestibular tissues from 21 organ donors (39 ears) were recovered. Of these, 18 donors (33 utricles) were examined by immunofluorescence. The sensory epithelium was present in 7 utricles (two from DBD and five from DCD). Relative to DBD utricles, DCD organs more commonly displayed dense populations of hair cells and supporting cells. Relative to DBD, DCD had significantly shorter postmortem interval time to tissue recovery (less than 48 hours). Compared to donors with no sensory epithelium, donors with intact and viable sensory epithelium (both DCD and DBD) had significantly shorter lag time to resuscitation prior to hospital admission (6.4±9.2 versus 35.6±23.7 minutes, respectively) as well as a shorter time between pronouncement of death to organ recovery (22.6±30.4 versus 64.8 ±22.8 hours, respectively).

Conclusions:

Organ donors are a novel resource for bilateral inner ear organs. Selecting tissue donors within defined parameters can optimize the quality of recovered inner ear tissues, thereby facilitating future research investigating sensory and non-sensory cells.

INTRODUCTION:

Hearing loss and vestibular disorders are common disabilities worldwide¹. The current treatment of sensorineural hearing loss is limited to amplification devices or cochlear implants. On the other hand, vestibular dysfunction is managed by rehabilitation therapy, with the efficacy of vestibular implants being an active area of investigation.^2,3 Loss of inner ear hair cells is the most common underlying pathologic finding associated with auditory and vestibular disorders, yet we currently lack treatment options to reverse this pathology.

While research on the inner ear may reveal novel therapeutic options, many studies utilize non-mammalian vertebrates as model systems^4,5. Others use small animal models such as rodents, and less commonly non-human primates^.6,7 Moreover, protocols for recovering inner ear tissues from different animal species are well established, yet existing sources of human inner ear tissues are rather limited. The exact similarities and differences between inner ear tissues from various species compared to human tissues remain unclear. Current sources of human inner ear tissues are primarily from patients undergoing surgical resection for lateral skull base pathologies (e.g., meningiomas, vestibular schwannoma),^8–11 aborted fetuses,^12–14 post mortem cadavers,¹⁵ as well as from libraries of archived temporal bones.¹⁶ Patients with lateral skull base pathologies warranting surgical resection typically have hearing loss, with inner ear organs displaying hair cell loss.¹⁷ Fetal tissues are usually recovered during the first trimester, when the inner ear is maturing and may be considerably different from pediatric and adult inner ear. As the time to tissue recovery is expected to dictate their viability and thus quality, inner ear tissues from both aborted fetuses and deceased donor cadavers can be limiting. At present, we lack the ability to reliably recover fresh and non-diseased inner ear tissues.

The human inner ear is housed in the temporal bone, rendering it relatively inaccessible. This challenging access lengthens the time needed to recover inner ear tissues from humans, leading to rapid decay particularly of postmortem samples. Using a recently described surgical approach to access the labyrinth in human cadavers (Vaisbuch et al., submitted), we recovered bilateral vestibular tissues from human organ donors. With the goal of recovering healthy vestibular tissues for research on sensory and non-sensory cells, we have examined pre-recovery donor conditions associated with viable vestibular tissues.

MATERIALS AND METHODS

Demographics of Organ Donors Tissue Recovery

Between January 2018 and July 2020, a total of 21 organ donors were identified for this study through a not-for-profit organization, Donor Network West (DNW, San Ramon, CA, USA). Medical records were reviewed, and the following information was collected: gender, age, comorbidities, cause of death, time to resuscitation in the field (either time from witnessed cardiopulmonary arrest/asystole or last time the donor was known to be conscious/alive), radiological reports (both computer tomography and magnetic resonance imaging of the brain), post admission time to recovery of inner ear organs, as well as time from the pronouncement of death to organ recovery. Research protocol was approved by the DNW’s internal ethics committee (Research project: #STAN-17–200) and its medical advisory board, as well as by the Institutional Review Board at Stanford University (#50076).

Organ Donor Tissue Recovery

Organ donor tissue recovery was done through a retroauricular trans-canal, trans-otic, surgical approach and was conducted by two surgeons (Y.V. and K.A.A.), a more detailed methodology is described in Vaisbuch et al. (submitted). Inner ear tissues were recovered from donors at the same time as organ procurement for transplantation (either heart, lung, liver or kidney) by other surgical teams.

Two categories of donors were used: donors after brain death (DBD) and donors after cardiac death (DCD). As a group, DCDs maintain some brain reflexes but require cardiopulmonary support to maintain oxygenation and circulation. They typically serve as organ donors for liver and kidney transplantation. Immediately before organs were recovered, life-supporting therapy was withdrawn, and cardiorespiratory arrest occurred. The period after cardiorespiratory arrest is termed warm ischemia, when the donors’ systolic pressure dropped below 80 mm Hg and/or oxygen saturation dipped below 80%.^18,19 Following cessation of cardiac and respiratory function, the donor was observed for 5 minutes to assure that no cardiopulmonary auto-resuscitation has occurred, at which point the final declaration of death was pronounced.²⁰ At this point, the donor was brought into the operating room, again confirmed to lack active circulation, before organ recovery was initiated. While some donors’ hearts stopped within 5 minutes, others continued to maintain prolonged circulation time, even in a state of low oxygen. Sometimes the donors maintained circulation for longer than 90 minutes, in which case they were no longer suitable for organ transplantation due to concerns for poor organ viability, and recovery or organs had to be canceled.

For DBD cases, the donor was declared brain dead, and maintained on full life support until the procurement of organs. The interval between brain death status and organ recovery can span several days. DBD cases usually served as organ donors for cardiac, lung, liver and kidney transplantation. For these cases, the donor was brought into the operating room still intubated. The extraction of inner ear tissues from the first ear was performed while the donor was still under ventilatory support. Tissues from the second ear were recovered after administration of systemic anticoagulation, discontinuation of mechanical ventilation, organ perfusion with preservation solution and cessation of circulation.

Immunohistochemistry:

Previously established protocols were followed.²¹ Once the utricle was recovered, it was placed in chilled phosphate buffered solution (PBS) and microscopically evaluated. Tissue quality was assessed, and it was determined whether the sensory epithelium was grossly intact, damaged, or absent. This was done to provide immediate feedback to the surgeon to improve tissue recovery and technique for the second ear. Next, tissues were placed in 4% paraformaldehyde (in PBS, pH 7.4; Electron Microscopy Services, #15710, Hatfield, Pennsylvania, USA) in a 50 ml Eppendorf tube and transferred on ice to the laboratory. The tissue remained in fixative at 4°C for 20 hours. After that, tissues were rinsed 3 times with PBS and then permeabilized with PBS-0.1% TritonX-100 solution at room temperature (3×30 minutes each), then blocked with 5% donkey serum, 0.1% TritonX-100, 1% bovine serum albumin (BSA, Thermo Fisher Scientific, #BP1600–100), and 0.02% sodium azide (NaN₃, Sigma-Aldrich, #S2002–25G) in PBS at pH 7.4 for 1 hour at room temperature. Primary antibodies diluted in the same blocking solution were added overnight at 4°C. The next day, after washing with PBS-0.1% TritonX-100 (3×15 minutes each), tissues were incubated with secondary antibodies diluted in 0.1% TritonX-100, 0.1% BSA, and 0.02% NaN₃ solution in PBS at pH 7.4 for 2 hours at room temperature. After washing with PBS (3×15 minutes), tissues were mounted in antifade Fluorescence Mounting Medium (DAKO, Agilent, #S3023, Santa Clara, California, USA) and cover-slipped.

Antibodies against the following proteins were used: Myosin7a (1:1000, rabbit, #25–6790, Proteus Biosciences) and Sox2 (1:400, goat, #sc-17320, Santa Cruz Biotechnology or 1:200, goat, #AF2018, R&D). Secondary antibodies included Alexa Fluor donkey anti-goat 647 (1:250, #A21447, Thermo Fisher Scientific) and Alexa Fluor donkey anti-rabbit 546 (1:250, #A10040, Thermo Fisher Scientific). DAPI (1:10,000 from 5 mg/ml stock solution; #D1306, Thermo Fisher Scientific) was used for counterstaining.

Confocal Imaging and Image Analysis:

Whole-mount preparations of utricles were analyzed as described before.²¹ Briefly, tissues were imaged as Z-stacks on a Zeiss LSM700 or LSM880 confocal microscope and were captured with Zen Software (Carl Zeiss). Organs were imaged with the tiling (6% overlap; mode: bounding grid) and stitching function. Images were taken from 1–6 representative areas of the sensory epithelium and analyzed with ImageJ (NIH) and Photoshop CS6 (Adobe Systems). The numbers of hair cells and supporting cells were quantified from z-stack images and calculated as density per 10,000 μm². Density of hair cells and supporting cells per end organ was analyzed and compared.

Statistical analysis

For statistical analysis Student’s t-tests were conducted using Microsoft Office Excel (Microsoft) and Prism 8.4 Software (GraphPad). A chi-square test was used to test the association between two categorical variables. Chi-square was used when the number of subjects was five or more. p<0.05 was considered statistically significant.

RESULTS

Demographics of organ donors

We have successfully procured 39 utricles from 21 organ donors. Only one ear was recovered from three donors who all had craniotomies for intracranial bleeding that rendered the other ear inaccessible. The cohort consists of 12 males and 9 females with an average age of 48.6 (±20.6, ranged 2–79) years old (Table 1).

Table1.

Characteristics of organ donors

Organ donor number	Age (years)	Gender	Cause of death	Admission diagnosis	CT/MRI brain imaging findings	Time down in the field (min)	Post admission (hrs)	Post mortem (hrs)	Initial microscopic evaluation		Histology of Sensory Epithelium density (10,000μm2
									LEFT	RIGHT	LEFT	RIGHT
1	72	M	Brain Death	Stroke, ICH	Right SDH	>60	96	48	NP*	NP*	NP	NP
2	67	F	Brain Death	Stroke, ICH	Bilateral parietal infact	0	121.5	52	NP*	Good	NP	HC: 71.9 SC: 177.7
3	20	M	Brain Death	Asphyxia	Diffuse brain edema ABI	45	101	39.5	NP	NP	NP	NP
4	69	M	Brain Death	Blunt head Trauma	Right SDH Diffuse brain edema”	>60	116.2	81.2	NP*	NP*	NP	NP
5	62	M	Brain Death	Stroke, ICH	Left Hemorrhagic infarct IPH /IVH	0	71.5	59.6	NP	Good	NP	HC: 64.8 SC: 66.7
6	28	M	Brain Death	Asphyxia	Diffuse SAH Cerebral edema	30	94	74.9	NP	NP	NP	NP
7	38	F	Brain Death	Stroke, ICH	Diffuse SAH Cerebral edema	0	124	87	NP	NP	NP	NP
8	27	F	Brain Death	Asphyxia	Right large SDH	12	153	73	NP	NP	NP	NP
9	35	M	Cardiac Death	MI	None	12	88	0.3	Good	Good	HC: 93.9 SC: 188.0	HC: 98.6 SC: 215.8
10	37	M	Brain Death	Blunt head Trauma	SAH ABI!	49	163.3	75.7	NP	NA	NP	NA
11	73	F	Brain Death	Stroke	Right large IPH / IVH Uncal herniation”	>60	79.6	56.5	NP*	NP*	NP	NP
12	79	M	Brain Death	Asphyxia	Diffuse ABI	35	106.6	40.4	NP*	Good	NP	NP
13	2	M	Cardiac Death	Spinal muscular atrophy	Basal ganglia infarct	20	74.9	0.4	Good	Good	HC: 73.7 SC: 126.9	HC: 84.7 SC:130.3
14	33	M	Cardiac Death	MI	None	5	160.5	0.7	Good	Good	NA#	NA#
15	66	F	Brain Death	Blunt head Trauma	SAH / SDH”	0	153.6	117.9	Good	Good	NP	NP
16	70	F	Brain Death	Stroke, ICH	SAH / Cerebral edema Uncal herniation	30	92.8	54.7	NP	NP*	NP	NP
17	47	F	Brain Death	Asphyxia	Diffuse brain edema Uncal herniation	60	68.6	32	NP	NP	NA	NA
18	59	F	Cardiac Death	Stroke, ICH	Cerebellar infarct	0	117.9	0.6	Good	Good	NA#	NA#
19	43	F	Brain Death	Stroke, ICH	SAH / IPH / IVH	Unknown duration	103.1	50.7	NP	NP	NP	NP
20	48	M	Brain Death	Stroke, ICH	Intracerebral hemorrhage Herniation /Cerebral edema”	>60	72	53.9	Good	NA	NP	NA
21	46	M	Cardiac Death	Stroke, ICH	Right limited IPH / IVH Right Uncal herniation	0	160	0.7	Good	NA	HC: 86.8 SC: 189.2	NA
SE average (±SD)	42.4 (±25.9)					6.4 (±9.2)	103.2 (±37.4)	22.6 (±30.4)
no SE average (±SD)	52.8 (±21.1)					35.6 (±23.7)	107.7 (±25.5)	64.8 (±22.8)

MI: Myocardial infarction. ICH: Intracranial hemorrhage. SDH: Subdural hematoma

SAH: Subarachnoid hemorrhage ABI: Anoxic brain injury

IPH: intraparenchymal hemorrhage IVH: Intraventricular hematoma

“ Indicates evidence of hematoma or hemorrhage within the vestibular end organ when procured NP: Not present. Absent or unhealthy sensory epithelium was detected

NA: No tissue available

^#Samples were designated for single cell RNA-seq experiment, no histology data available

^*Samples with evidence of mechanical damage from tissue handling HC: hair cells

SC: supporting cells

The majority of patients (16) were DBDs, with admission diagnoses ranging from stroke with intracranial hemorrhage (n=8), asphyxia (n=5), and blunt head trauma (n=3). In the 5 DCDs, the admission diagnoses were stroke with intracranial hemorrhage (n=2), myocardial infarction (n=2), and sequalae of spinal muscular atrophy (n=1).

As part of the standard evaluation, all donors received CT and MRI imaging of their brains and the team was provided with the reports of the findings. DCDs had minimal or focal imaging abnormalities-small cerebellar infarct (n=1), basal ganglia infarct (n=1), right-sided intraparenchymal and intraventricular hemorrhage with unilateral uncal herniation (n=1), or no abnormalities reported (n=2) (Table 1). Almost all patients in the DBD cohort showed significant intracranial abnormalities, including large intracranial bleeds (either subarachnoid hemorrhage or subdural hematoma, n=11), diffuse cerebral edema (n=7), signs of anoxic brain injury (n=3), or bilateral uncal herniation (n=4).

Immunohistochemistry of organ donor utricles

Prior to immunohistochemistry, all 39 utricles collected from 21 donors were assessed under light microscopy. Of these, 9 had evidence of tissue damage likely caused by surgical dissection (Table 1). Fifteen organs were grossly normal and showed no or minimal tissue damage. The remaining 24 showed poor tissue integrity with no visible sensory epithelium as well as evidence of mechanical damage.

Next, 33 utricles were examined by immunohistochemistry, with the remaining 6 used for other experimental procedures (Table 1). To further assess the integrity of the sensory epithelium, we immunolabeled for the hair cell and supporting cell markers Myosin7a and Sox2, respectively (Figure 1). In 7 of 33 utricles, we detected sensory epithelia densely populated by hair cells and supporting cells (Figure 1A–B’’’’). No hair cells were detected in 13 utricles, while another 13 organs harbored a few Myosin7a⁺ hair cells (Figure 1C–D). The remaining utricles were devoid of both hair cells and supporting cells, suggesting that no sensory epithelium was present on many cases (Figure 1C–D, Table 1). Of note, of the 14 organs that appeared grossly intact immediately after recovery, only 7 were confirmed to contain sensory epithelia post tissue processing.

Figure 1: Immunohistochemistry of utricles from human organ donors.

A-D) Whole mount preparation of utricles were immunolabeled for the hair cell marker Myosin7a and the supporting cell markers Sox2. In utricles from organ donors #13 and #9, many hair cells and supporting cells were observed in (A and B). A’-B’’’) Representative confocal images of the sensory epithelium at high magnification. A’’’’-B’’’’) orthogonal views of A’ and B’. (C-D) High and low magnification images showing no or few hair cells and a variable number of supporting cells in utricles from organ donors #4 and #7. E-F) In seven utricles from five organ donors, hair cell density was found to be 79.8±13.4 per 10,000 μm² and supporting cell density 165.1±51.8 per 10,000 μm². Donor #9 is a cardiac death, from whom bilateral utricles showed the highest hair cell and supporting cell density (98.6 and 215.8 per 10,000 μm² respectively). Donor #5 is a brain death donor, from whom the utricle from the right ear had the lowest hair cell and supporting cell densities (64.8 and 66.7 per 10,000 μm² respectively). Each data point represents cell density from high magnification images (1–5 per utricle). Data shown as mean±S.D., scale bars: A-D) 200 μm, A’-D’) 20 μm.

Among the 7 utricles showing intact sensory epithelia, two were from DBD and 5 were from DCD procurement (Table 1). We quantified hair cells and supporting cells from high magnification images of the sensory epithelium and found their density to be 79.8±13.4 and 165.1±51.8 per 10,000 μm², respectively (Figure 1E–F). To further assess cell viability, we examined hair cells and supporting cells and found many pyknotic nuclei in a utricle where some hair cells were lost, indicating ongoing cell death (Figure 2A–D).

Figure 2: Pyknotic nuclei in utricular sensory epithelia from an organ donor suffering from brain injury.

A) Low magnification image of a whole mount preparation of utricle recovered from donor #15 who sustained blunt head trauma 117 hours prior to tissue procurement, showing the presence of nuclei and some hair cells. B-D) High magnification images taken from regions in (A). Nuclei were stained with DAPI showing many cells were pyknotic (asterisks). Scale bar: A) 500 μm, B-D) 20 μm.

Factors affecting donor tissue quality

We investigated data among organ donors to identify parameters that correlated with utricles that had intact viable sensory epithelia with dense populations of both hair cells and supporting cells. Only organs examined via immunohistochemistry were included in this analysis and those processed for other experiments were excluded.

There was no significant age difference between the donors with intact sensory epithelia (42.4±25.9 years old) versus those that did not (52.8±21.1 years old, p=0.39). Gender also did not play a role between these two groups (Chi-square, p=0.69).

On the other hand, donors with intact sensory epithelia had significantly shorter time between cardiopulmonary arrest and resuscitation prior to hospital admission than those who did not (6.4±9.2 versus 35.6±23.7 min, p=0.004). Donors with intact sensory epithelia also received organ procurement much sooner after pronouncement of death than those without (22.6±30.4 versus 64.8±22.8 hrs, p<0.0002). Lastly, utricles that consistently lacked intact sensory epithelia came from donors that were found unconscious or have asystole for a prolonged period (>60 min). Postadmission time was not found to affect the success rate of tissue recovery (Table 1).

In donors with preserved sensory epithelium, we noted that tissue procurement after the pronouncement of death began much earlier for the DCD than DBD group (0.5±0.2 hrs versus 55.8±5.4 hrs). Furthermore, 100.0% of DCD utricles (5 of 5) as compared with 7.1% of DBD (2 of 28) showed intact sensory epithelia on immunohistochemical analysis as a whole, (Chi-square, p<0.00001). Additionally, in the two DCD cases from whom bilateral utricles were procured, no noticeable difference in the hair cell and supporting cell counts was detected between the two sides. Together, these data suggest that shorter asystole time in the field, shorter time to tissue recovery, and donors from cardiac death are parameters optimal for the successful isolation of utricles.

DISCUSSION

Similar to neuronal tissues in the central nervous system, inner ear tissues critically depend on oxygen and other nutrients and undergo rapid cell death after donor’s’ demise. As such, shortening the time to procurement is expected to most reliably yield the most viable inner ear tissues. Obtaining fresh tissues from organ donors creates numerous opportunities in advancing inner ear research, such as morphological analysis of vestibular sensory and non-sensory cells, single cell transcriptomic analyses, and potentially a platform to derive human inner ear organoids which allow drug screening/testing. These areas of research require viable cells from non-diseased inner ear tissues that cannot be obtained from cadaveric temporal bones and inner ear tissues harvested in surgical patients. Using the extraction technique described by our group (Vaisbuch et al. submitted), we have identified essential parameters that should be considered for recovering inner ear tissues from organ donors. Our technique is simple, reproducible, and reasonably fast. Once past the learning curve, inner ear tissues can be recovered in less than 20 minutes.

Like other procedures, there is a learning curve for procuring inner ear tissues. We have noted that microscopic examination of the recovered tissues from the first ear provided useful, instant feedback for the surgeon, thereby improving on the extraction and tissue handling of the second ear. We believe that it is helpful for surgeons to get feedback particularly on the first 5–10 donors.

Although the second ear was typically recovered 20–30 min after the surgery on the first ear, we did not detect significant difference among tissue quality between the two ears. In fact, two utricles procured from a DBD case had intact sensory epithelia from the second ear but not from the first ear. Additionally, in the 2 DCD cases from whom bilateral utricles were recovered, no noticeable difference in the hair cell and supporting cell counts was detected between the two sides, suggesting that timely tissue procurement is feasible with limited ischemia time in appropriately selected patients.

Cardiac versus brain death

When discussing organ donors, it is important to understand the differences between DBD and DCD cases. In DCD cases, organs were retrieved after death was confirmed using circulatory criteria. When DCDs were brought to the operating room, their donor criteria were confirmed after withdrawal of life support (5–20 min), before inner organs were procured 18–42 min later. On the other hand, death in DBD patients was confirmed using neurological criteria prior to transport to the operating room. After they were pronounced brain dead, they received life support for days (32–118 hrs), before they were brought into the operating room when the organs were recovered for transplantation. Thus, the time interval is one main difference between the DBD and DCD groups.

Secondly, relative to DCDs, the DBD group had a significantly longer asystole time down in the field (6.4±9.2 versus 35.6±23.7 min, p=0.004), even though their post-admission durations were comparable (Table 1).

Thirdly, imaging studies of DBD cases usually showed large intracranial bleeds, asphyxia, or edema. Given that the inner ear contains perilymphatic fluid that communicates with cerebrospinal fluid and also shares vascular supply with the brain, it is likely that initial insults to the brain among DBD also negatively impacted the viability of inner ear tissues. Additionally, all patients with blunt head trauma had blood in the labyrinth when the vestibular end organs were recovered (Table 1), possibly causing asphyxia of the sensitive inner ear organs and leading to cell death. In support of this association, several reports in the literature have shown that inner ear hemorrhage can lead to sudden sensorineural hearing loss, tinnitus and vertigo.^22–24

By contrast, DCD cases typically had fewer brain abnormalities on imaging, while maintaining some brainstem reflexes. We postulate that these factors may contribute to higher qualities of inner ear tissues among the DCD cohort.

Viability of inner ear tissues

Although we only succeeded in recovering utricles with intact sensory epithelia from 5 patients, our study is the first to report such an approach and selection criteria among organ donors. The finding that not all tissues procured from organ donors were healthy and viable is highly significant, and points to the stark difference between recovering inner ear organs and other solid organs used for transplantation (e.g. liver, kidney). In our study, the average hair cell density of the sensory epithelium was 79.8±13.4 per 10,000 μm², which is with the range of densities previously reported on archived temporal bones (58–88 cells per 10,000 μm²).^25–27 Vestibular tissues with preserved sensory epithelia were more frequently recovered from DCD cases as compared to DBD cases for several reasons. First, DBD subjects demonstrated more severe intracranial injuries as stated above. Additionally, the DBD cohort had significantly longer time down or being resuscitated (in asystole) than DCD cases, which in turn could have led to longer anoxia to the central nervous system among DBD patients. Among the DBD cohort, intact sensory epithelia were successfully harvested from 2 donors (donors #2 and #5, Table 1), both with short time down in the field (0 min) despite long postmortem time to procurement (52 and 75 hours, respectively). Moreover, these donors’ imaging studies demonstrated minimal CNS injuries, suggesting that short time down in the field may have helped minimize injuries to the CNS and inner ear tissues as mentioned above.

It is a well-established that the brain is vulnerable to ischemia; interruption of blood flow to the region for 5 minutes or more causes loss of neurons and can ultimately lead to irreversible brain damage.²⁸ Furthermore, animal models demonstrate that labyrinthine artery occlusion can lead to hypoxia in the cochlea after only 1 minute; if circulation is not restored, auditory function decreases after 10 minutes and hair cell loss occurs after more than 15 minutes.^29,30 Hence, one may infer that conditions that lead to prolonged anoxia or hypoperfusion to the brain can similarly cause irreversible damage to the inner ear. Nonetheless, a recent study reports that it is possible to isolate cells with progenitor cell characteristics from the human cochlea even after more than 17 hours postmortem, albeit a lower viability than in comparison to fresh surgical specimens. Another paper reported the isolation of putative progenitors from the cochleae of vestibular schwannoma and DBD patients.³¹ However, whether hair cells and supporting cells or other specific cell types were preserved was unclear in both studies. Thus, it is possible that certain cell types or inner ear organs are differentially sensitive to hypoxia.

Additionally, we found that shorter time between death pronouncement and tissue recovery optimizes quality of the sensory epithelia. Since this duration is significantly shorter in DCD than DBD cases, it likely contributed to the higher quality of inner ear tissues procured from the former group. Along a similar vein, DBD cohort from whom inner ear tissues displayed sensory epithelia received tissue procured much sooner than those with suboptimal tissues (55 versus 63 hours, Table 1). Additionally, all utricles in the DCD group were procured within 0.76 hours from pronouncement of death. Finally, DBDs had significantly longer time down in the field than DCDs, and preserved inner ear tissues were only observed in the 2 DBDs that had no time down in the field. Based on these observations, we conclude that shorter postmortem time to recovery and time down in the field are significant parameters for isolation of high quality inner ear tissue.

While several studies have reported variable success in isolating cells from the inner ear of cadavers and donor patients,^15,31 preservation of the organ architecture, including the integrity of hair cells and supporting cells, have not been systematically examined. For example, the previous report on tissue recovery from DBDs had potential limitations.³¹ First of all, several methods to procure inner ear tissues described but not systematically examined whether one is more optimal than others. Second, factors among organ donors that may impact tissue recovery were not reported (e.g. including time from pronouncement of death, tissue quality, etc.). Third, no DCD patients were included in that study. Lastly, while they reported visualizing portions of the organ of Corti and stria vascularis under light microscopy, whether individual sensory and non-sensory cell types were present was not confirmed. In our study, we have found that only half of the organs deemed grossly normal under light microscopy were subsequently confirmed to contain hair cells and supporting cells, many of which appeared pyknotic. These results highlight the need to identify additional factors optimizing inner ear tissues recovery. We believe the current study should facilitate these efforts.

The limitations of our study include an initial learning curve, which probably has affected the quality of the recovered tissue from the initial donors (mostly DBDs). This effect could have potentially skewed data towards over-reporting poor tissue quality. Additionally, in this manuscript we only focused our analysis of the utricle, and that factors that help preserve cells residing in other inner ear sensory organs, particularly the cochlea, may differ. It is possible that cells in the cochlea are more sensitive to hypoxia than in the vestibular organs, and that conditions are different for various cell types such as hair cells, supporting cells, and potential progenitor cells, which provides opportunities for further studies.

In conclusion, we demonstrate that viable inner ear tissues can be procured from adult human organ donors. We further show only a subset of recovered utricles contain sensory epithelia with hair cells and supporting cells, with several pre-recovery donor conditions dictating the success rate. In our cohort, DCD cases represent the optimal group with the highest rates of preserved sensory epithelia, in part because of their shorter resuscitation time, less severe brain insults, and shorter time between pronouncement of death and tissue procurement.