EFFICACY OF HOPE: ANALYSIS OF ORGAN QUALITY AND AVAILABILITY AMONG DECEASED HIV-POSITIVE DONORS

Abstract

Background

Improved survival among people with HIV (PWH) has led to increased organ failure, necessitating transplantation. In 2013, the HIV Organ Policy Equity (HOPE) Act was passed, allowing PWH to donate organs to other PWH. No study has assessed organ quality and quantity among a national pool of PWH.

Methods

CFAR Network of Integrated Clinical Systems (CNICS), a multicenter study capturing data on PWH, was used to identify 6,504 deaths from 1999–2018. Exclusions included cause of death, chronic kidney disease, fibrosis-4 score≥3.25, and opportunistic infection at time of death. Donor quality was defined by HIV viremia and the Kidney Donor Profile Index (KDPI). The CDC Wonder database, which contains national death data, permitted estimation of deaths among PWH nationally from 1999–2018. Assuming CNICS was representative of PWH nationally, percentages of potential donors were applied to the CDC Wonder cohort.

Results

Within CNICS, there were 3,241 (65.9%) potential kidney donors and 3,536 (71.9%) potential liver donors from 1999–2018. Based on viremia and KDPI, 821 were lower-risk kidney donors (16.7%) and 1,206 (24.5%) were lower-risk liver donors. Within CDC Wonder, we identified 12,048 potential donors from 1999–2018. Extrapolating from CNICS to the national cohort suggested 396 kidney donors (792 kidneys) and 433 liver donors annually, with 100 kidney donors (200 kidneys) and 147 livers being lower-risk.

Conclusion

A substantial number of PWH meet donation criteria, a valuable source of organs for PWH in need of transplants. Our estimates suggest there may be more available organs from PWH than current transplant numbers indicate.

Graphical Abstract

Introduction

As a result of antiretroviral therapy (ART) introduced in the 1990’s, life expectancy at 20 years of age among people with HIV (PWH) has increased from 19.1 years in 1996 to 53.1 years in 2011¹. With increased longevity, PWH develop chronic and end-stage diseases of the kidneys and liver resulting in the need for organ transplants^2–4. Similar to HIV-negative counterparts, PWH are encumbered by the ongoing organ shortage but are particularly vulnerable as they are less likely to receive kidney or liver transplants than HIV-negative patients^5,6. Historically, PWH were forbidden to donate organs in the United States because of restrictions outlined in the National Organ Transplant Act⁷. However, in 2013 the HIV Organ Policy Equity (HOPE) Act was passed with the goal of mitigating the ongoing organ shortage by legalizing organ donations from PWH to other PWH with various end-stage diseases⁸.

Multiple studies have demonstrated the safety and efficacy of organ transplants between PWH^9–13. South Africa has led the global movement to transplant organs from PWH and has had high success rates. In South Africa, survival rates among recipients living with HIV who received donation transplant from a donor with HIV were comparable to survival rates in the United States among recipients living with HIV who received a donation from a donor with HIV. Within the South African study, five-year survival following kidney transplantation was 83.3%, while 78.7% patients had a functioning kidney transplant¹⁴. While the South African HIV epidemic presents a unique situation due to low ART resistance rates, which creates a comparatively favorable environment for transplanting organs from donors with HIV, results have been promising in the United States as well. In a prospective multicenter pilot study, Durand et. al found similar graft survival rates and recipient survival rates among recipients living with HIV who received a kidney from a donor with HIV compared to those who received a kidney from an HIV-negative donor¹⁴. However, while excellent outcomes have been achieved with donor organs from PWH¹⁴, concerns about donor quality related to an HIV infection remain, motivating further study of this potential donor pool. Among these concerns include the possibility of introducing a new strain of HIV from the donor, which could lead to superinfection and subsequently compromise the recipient’s treatment regimen, especially if this strain is ART resistant¹⁷. While superinfection was a concern early in the HOPE Act experience, the HOPE Act does not impose viral load as an eligibility criterion as it would limit the donor pool, particularly given anticipated difficulties in obtaining timely viral loads and/or antiretroviral resistance profiles^15,18. Thus, while HIV viremia was mentioned as a risk factor for superinfection, recent research has suggested that superinfection is rare and may not be a clinical concern¹⁸. Despite these excellent outcomes, the practice of transplantation of PWH with donor organs from PWH is contingent upon the number and quality of suitable donor organs available.

Previous studies have given varied estimates for the number of potential organs from donors with HIV made available through the HOPE Act. Using national data, Boyarsky et al. concluded that between 494 and 534 deceased donors were available for transplant annually, but did not exclude donors with comorbidities that preclude donation and did not examine viral load¹⁶. In contrast, Richterman et al. projected 192 kidneys and 247 livers would be available under the HOPE Act annually. However, this study was limited to deaths in Philadelphia, which may limit the generalizability of their study. Furthermore, several variables used in calculating donor risk indices were unavailable¹⁷. The discordance between these studies highlights current gaps in knowledge regarding the pool of potential donors with HIV, including important considerations of organ quality, donor availability, and HIV viral control. Therefore, we sought to leverage a national longitudinal cohort study of PWH to identify potential deceased donors across the United States and characterize both organ quality and viral control. We hypothesized the number of nationally available donor organs from PWH was greater than the estimates provided by Richterman et al. as they were restricted to PWH in Philadelphia but lower than the estimates provided by Boyarsky et al ^16,17 given exclusion of donors with comorbidities such as CKD/ESKD and fibrosis of the liver that preclude donation. Furthermore, we hypothesized that the quality of organs increased over the study period with a greater proportion of lower-risk potential donors existing in recent years given advancements in care of PWH.

Materials and Methods

Data Source

This study utilized the Center for AIDS Research (CFAR) Network of Integrated Clinical Systems (CNICS) cohort and the Centers for Disease Control (CDC) Wonder national death database. CNICS is a longitudinal study in which PWH are enrolled at eight geographically diverse academic medical centers across the United States (Fenway Health, University of North Carolina Chapel Hill, Johns Hopkins University, University of Alabama at Birmingham, Case Western Reserve University, University of California San Diego, University of California San Francisco, University of Washington). Clinical and demographic data, including medical comorbidities, medications and routine laboratory measurements, are collected at study visits and are reported to a central data repository on a periodic basis¹⁹. Comorbidities and diagnoses are ascertained via chart review at each site. The CDC Wonder database contains national data pulled from death certificates including underlying and multiple causes of death. This study was approved by the IRB at the University of Alabama at Birmingham (protocol number: 300001300).

Study Design

We performed a cross-sectional analysis of the CNICS database, which contains medical records for more than 34,000 PWH. Participants of CNICS are confirmed to have HIV and therefore no false-positives were included in this analysis. Between 1999 and 2018, 6,504 deaths occurred among PWH enrolled in CNICS care. All comorbidities and laboratory values were captured as close to the date of death as possible. These data were used in estimating the percentage of individuals who could have potentially served as deceased kidney and/or liver donors based on organ quality. From the 6,504 deaths recorded in CNICS care, exclusions were made based on cause of death (cancer or an AIDS-related death; ICD-10 Codes C00-D48 and B20-B24) that would preclude organ donation.

Comorbidities and Organ Quality

All diagnoses were coded in accordance with CNICS recommendations using medications and laboratory measurements captured as close to the date of death as possible. Diabetes was defined as an A1C ≥6.5% or use of a diabetes-specific medication. Hypertension was defined as a reported diagnosis of hypertension based on an elevated systolic and/or diastolic blood pressure or use of an antihypertensive medication. Coronary artery disease was defined as a reported diagnosis of coronary artery disease or myocardial infarction. Hepatitis B and C co-infections were determined by the presence of a positive antigen or antibody test result, and/or the presence of detectable viral DNA. Estimated glomerular filtration rate (eGFR) was calculated using the recently released race-less CKD-EPI eGFR equations²⁰. Chronic kidney disease (CKD) was defined as an eGFR of < 60 mL/min/1.73m² for more than 90 days without an intervening higher eGFR. Similarly, end-stage kidney disease (ESKD) was defined as the presence of an eGFR <15 mL/min/1.73m² for more than 90 days without an intervening higher eGFR. The following diagnoses were defined as opportunistic infections: candidiasis, cytomegalovirus (encephalitis, enterocolitis, esophagitis, pneumonitis, retinitis, polyradiculopathy), coccidiomycosis, cryptococcosis, cryptosporidiosis, herpes-simplex (bronchitis, pneumonitis, chronic ulcer, esophagitis), histoplasmosis, mycobacterium, isosporiasis, Kaposi’s sarcoma, cervical cancer, Non-Hodgkin’s lymphoma, pneumocystis, pneumonia, regressive multifocal leukoencephalopathy, salmonella, tuberculosis, and toxoplasmosis. The Kidney Donor Profile Index (KDPI) was calculated for all potential deceased donors in the CNICS cohort. KDPI incorporates demographics, comorbidities, and laboratory values to provide a summary statistic by which clinicians can quickly ascertain the relative quality of a donor kidney, and is used to allocate kidneys for transplant²¹. Potential deceased donors were then categorized by their KDPI into ranges of 0–20, 21–84, and 85 or greater. The Fibrosis-4 (FIB-4) score was used to assess liver quality based on the estimated amount of fibrosis²². FIB-4 incorporates ALT, AST, platelet count, and age. Potential deceased donors were then categorized by their FIB-4 scores into ranges of mild (<1.45), moderate (1.45–3.25), and severe (>3.25) fibrosis²³.

The percentages of potential deceased kidney and/or liver donors with HIV estimated in the CNICS cohort were then applied to the national CDC numbers. The CDC Wonder database was used to determine the number of deaths that occurred nationally among PWH during the study period. We assumed that the PWH identified in CNICS was a representative sample of the national population of PWH identified in the CDC Wonder database. Therefore, the deaths that occurred in CNICS care are a subset of the larger number of national deaths that occurred among PWH. This gave us a final estimate of the number of kidneys and livers available from potential deceased donors with HIV on a national scale.

Data Analyses

PWH were categorized by whether they were non-viremic (<200 copies/mL) or viremic (≥200 copies/mL) at the closest lab to death. The HOPE Act states it is anticipated that the risk of donor-derived superinfection is lower from deceased donors with undetectable viral loads at the time of death but did not preclude donation from donors with active viremia¹⁵. Therefore, viral load was not considered an exclusion criterion but was used to label a subset of our potential donors as lower-risk¹⁵. Potential lower-risk kidney donors are those with undetectable viral loads as well as a KDPI score below 85. In contrast, lower-risk liver donors are those with undetectable viral loads. Continuous characteristics were then compared across viral load status using the Wilcoxon ranked sum test. Categorical variables were compared using the Chi-squared test of association.

Sensitivity Analyses

The KDPI incorporates an inflation factor for kidneys that come from donors with an HCV infection²⁴. As a result, kidneys from donors with HCV are considered to have a higher risk of graft failure. However, 1-year patient and graft survival rates are similar between recipients who received a kidney from an HCV-positive donor and recipients who received a kidney from an HCV-negative donor²⁵. Due to the large percentage of our cohort with an HCV coinfection, this resulted in a substantial number of kidneys assigned a higher KDPI score indicative of poorer quality. Given comparable patient and graft survival rates in addition to the introduction of direct-acting antivirals, we recalculated KDPI, removing the HCV inflation factor, and compared these results to the results obtained using the official KDPI formula. Additionally, the recently released race-less 2021 CKD-EPI eGFR equation was used to calculate eGFR and subsequently define CKD²⁶. The CKD-EPI equation derived in 2009 has been widely used in clinical settings to estimate GFR and contains an inflation factor based on African American/Black race such that at any given serum creatinine level African American/Black PWH would have a higher eGFR than non-African American/Black PWH²⁶. Given the natural lag in adoption of the 2021 CKD-EPI eGFR equation, we performed the same analysis using the race-based CKD-EPI formula and compared these results to the results obtained from the race-less eGFR equation. Our timeframe includes years in the late 1990s and early 2000s during which ART was not widely available for use by the general population of PWH. Because of this, a higher number of deaths due to AIDS occurred towards the beginning of our study. We examined the percentages within a restricted timeframe and compared them with our main results.

A two-sided p-value of 0.05 was the threshold for statistical significance. All analyses were conducted using SAS 9.4 (Cary, NC).

Results

Overall Cohort

Among the 6,504 deaths that occurred in the CNICS cohort, 867 were excluded due to missing viral load measurement and an additional 716 were excluded due to a cancer or AIDS-related death (ICD-10 Codes C00-D48 and B20-B24). A cohort of 4,921 deceased PWH remained. The cohort was majority male (67.6%) and African American/Black (51.0%). A substantial portion were co-infected with HCV (37.7%) and had hypertension (29.5%). About half of the cohort had a history of smoking (42.3%) and a past opportunistic infection (47.0%) (Table 1). In total, 3,321 (67.4%) were viremic. Over the course of the twenty-year study period, the percentage of viremic PWH decreased (Figure 2). The non-viremic group was majority white (53.0%) and the viremic group was majority African American/Black (56.0%) (p<0.001). Diabetes, hypertension, coronary artery disease, and obesity were more prevalent in the non-viremic group. Current opportunistic infections were more prevalent in the viremic group (Table 1).

Table 1:

Eligible deaths within the CNICS cohort

	Cohort
	Total	Non-Viremic Viral Load	Viremic Viral Load	P-value
	N = 4,921	N = 1,600	N = 3,321
Demographics
Age (Years)	47 (41, 54)	51 (45, 58)	46 (39, 52)	<0.001
Male, Present Sex	3,327 (67.6)	1,116 (70.0)	2,211 (66.6)	<0.001
Transgender	43 (0.9)	7 (0.4)	36 (1.1)	0.03
Race				<0.001
African American/Black	2,509 (51.0)	650 (40.6)	1,859 (56.0)
White	2,152 (43.7)	849 (53.0)	1,303 (9.1)
Asian	39 (7.9)	23 (1.4)	16 (0.5)
Other	221 (4.5)	78 (4.9)	143 (4.3)
Hispanic	272 (5.5)	106 (6.6)	166 (5.0)	0.02
Comorbidities
Diabetes	321 (6.5)	175 (10.9)	146 (4.4)	<0.001
Hypertension	1,452 (29.5)	644 (40.0)	808 (24.3)	<0.001
Coronary artery disease	650 (13.2)	290 (18.1)	360 (10.8)	<0.001
Hepatitis C	1,856 (37.7)	606 (37.9)	1,250 (37.6)	0.43
Hepatitis B	429 (8.7)	135 (8.4)	294 (8.9)	0.81
BMI	23.1 (20.2, 26.9)	25.2 (21.2, 28.2)	22.7 (19.6, 26.0)	<0.001
Obesity	508 (10.3)	250 (15.6)	258 (7.8)	<0.001
Social risk factors
History of smoking	2,083 (42.3)	704 (44.0)	1,379 (41.5)	0.01
Alcohol, high risk	91 (1.8)	53 (3.3)	38 (1.1)	<0.001
Intravenous drug use, Ever	134 (2.7)	78 (4.9)	56 (1.7)	<0.001
Cocaine, Ever	330 (6.7)	195 (12.2)	135 (7.1)	<0.001
Kidney characteristics
Last Creatinine	1.0 (0.8, 1.4)	1.0 (0.8, 1.4)	0.99 (0.8, 1.4)	<0.001
Days since last creatinine	104 (20, 601)	95 (23, 407)	111 (19, 693)	0.08
Chronic kidney disease (eGFR < 60 mL/min/1.73 m2)	396 (8.0)	132 (8.3)	264 (7.9)	0.64
End-stage renal disease	261 (5.3)	95 (5.9)	166 (5.0)	0.17
KDPI	65 (44, 85)	70 (47, 90)	62 (42, 82)	<0.001
0–20	186	49 (3.1)	137 (4.1)
21–84	2,586	875 (54.7)	1,711 (51.5)
85+	955	442 (27.6)	513 (15.4)
Liver Characteristics
ALT	31 (19, 54)	30 (19, 52)	31 (19, 54)	0.34
AST	37 (25, 66)	33 (23, 59)	39 (26, 71)	<0.001
Bilirubin	0.5 (0.3, 0.8)	0.5 (0.4, 0.9)	0.5 (0.3, 0.8)	0.15
Platelet count	197 (134, 258)	208 (146, 267)	191 (129, 251)	<0.001
Fibrosis-4	1.69 (1.00, 3.29)	1.51 (0.93, 3.01)	1.76 (1.03, 3.45)	<0.001
0–1.45	2,076 (43.4)	760 (47.5)	1,316 (39.6)
1.45–3.25	1,493 (31.2)	451 (28.2)	1,042 (31.4)
3.25+	1,214 (25.4)	366 (22.9)	848 (25.5)
HIV history
VL < 200 copies/mL	1,600 (32.5)	1,600 (100.0)	0 (0.0)	<0.001
Time since VL measurement	155 (56, 729)	124 (53, 451)	180 (58, 855)	<0.001
OI
Past	2,311 (47.0)	660 (41.2)	1,651 (49.7)	<0.001
Current	66 (1.3)	9 (0.6)	57 (1.7)	<0.001

Figure 2.

Exclusion diagram for the CFAR Network of Integrated Clinical Systems (CNICS) cohort

Potential Kidney Donors

To estimate the number of potential deceased kidney donors, 751 PWH were excluded from the cohort of 4,921 due to a CKD diagnosis or ESKD diagnosis. An additional 63 were excluded because of an opportunistic infection at the time of death. An additional 866 were excluded due to a missing KDPI value attributable to missing BMI values. Those remaining were considered potential deceased kidney donors. In total, we identified 3,241 potential deceased kidney donors, which represents 65.9% of our cohort of 4,921 (Figure 1). Viral load and KDPI were used to further designate a subset of this group as lower-risk. Of the 3,241 potential deceased kidney donors, 2,067 (63.8%) were viremic and 1,174 (36.2%) were non-viremic. The group of potential deceased kidney donors had a median CD4 count of 201 (IQR: 44, 420). The viremic group had a median CD4 count of 102 (IQR: 20, 308) and the non-viremic group having a CD4 count of 371 (IQR: 192, 593)(p-value<.0001). Fewer viremic donors had a KDPI ≥85 compared to non-viremic donors (18.8% versus 30.0%, p<.0001). Considering those who were non-viremic with a low to moderate KDPI as lower-risk donors, we identified 821 potential deceased kidney donors, which represents 16.7% of our cohort of 4,921 (Table 2).

Figure 1:

Exclusion diagram for the CNICS cohort

Table 2:

Characteristics of ideal potential kidney and liver non-viremic cohorts

	Ideal Potential Deceased Kidney Donors (N=821)	Ideal Potential Deceased Liver Donors (N=1,206)
Demographics
Age (Years)	48 (42, 53)	51 (44, 58)
Male, Present Sex	637 (78.2)	835 (69.2)
Transgender	6 (0.7)	5 (0.4)
Race
African American/Black	216 (25.3)	482 (40.0)
White	538 (67.4)	653 (54.1)
Asian	20 (2.2)	16 (1.3)
Other	47 (5.0)	55 (4.6)
Hispanic	73 (8.8)	74 (6.1)
Comorbidities
Diabetes	46 (4.5)	122 (10.1)
Hypertension	230 (27.4)	488 (40.5)
Coronary artery disease	97 (9.8)	212 (17.6)
Hepatitis C	237 (30.0)	384 (31.8)
Hepatitis B	75 (8.8)	88 (7.3)
BMI	24.6 (21.6, 28.3)	24.7 (21.5, 28.5)
Obesity	152 (18.0)	198 (16.4)
Social risk factors
History of smoking	374 (45.5)	525 (43.5)
Alcohol, high risk	39 (4.8)	43 (3.6)
Intravenous drug use, Ever	40 (5.5)	60 (5.0)
Cocaine, Ever	117 (14.7)	159 (13.2)
Kidney characteristics
Last Creatinine	0.93 (0.8, 1.1)	1.0 (0.8, 1.3)
Days since last creatinine	99 (27, 402)	112 (29, 428)
Liver Characteristics
ALT	30 (20, 50)	27 (18, 45)
AST	31 (22, 54)	27 (21, 40)
Bilirubin	0.5 (0.3, 0.8)	0.5 (0.3, 0.7)
Platelet count	215 (153, 273)	231 (184, 289)

Nationally, 233,051 deaths were identified among PWH between 1999 and 2018 and were captured in the CDC Wonder database. Exclusions included death occurring in a location outside of care (n=92,857) and an underlying cause of death due to HIV or neoplasms (n=128,146). This left a national cohort of 12,048. Applying the percentage of potential donors estimated using the CNICS cohort (65.9%) to the national cohort of 12,048, there were 7,939 potential deceased donors with HIV available during this twenty-year timeframe, which equates to about 396 donors or 792 kidneys annually. Applying the percentage of lower-risk transplantable kidneys estimated using the CNICS cohort (16.7%) to the national cohort of 12,048, we estimate there were 2,012 lower-risk potential deceased donors with HIV available during this twenty-year time period, which equates to approximately 100 donors or 200 kidneys annually.

Sensitivity Analyses

After removal of the HCV inflation factor from the KDPI formula, lower-risk potential deceased donors with HCV had a KDPI score that was an average of 7.25 points lower, resulting in 141 potential donors falling below a KDPI of 85. Subsequently, the number of lower-risk potential kidney donors in the CNICS cohort increased from 821 (16.7%) to 962 (19.5%). Applied to the national cohort of 12,048, this would result in 2,349 lower-risk potential deceased kidney donors during the twenty-year period or approximately 117 donors or 234 kidneys annually, an increase of 34 kidneys per year beyond estimates that penalized co-infection with HCV.

When using the race-based CKD-EPI equation to calculate eGFR and subsequently diagnose CKD, the estimated number of lower-risk potential kidney donors increased from 821 to 850. This represented 17.3% of our eligible cohort. Applied to the national cohort of 12,048, this would result in 2,084 potential deceased kidney donors over the twenty-year period, or 104 donors or 208 kidneys annually, an increase of 8 kidneys per year beyond estimates based on race-less eGFR. This increase in potential kidney donors largely comes from those of African American/Black race, accounting for 51.7% of those who would be classified as having CKD under the 2021 CKD-EPI formula but not the 2009 CKD-EPI formula.

Using the restricted timeframe of 2005–2018, 72.3% of our cohort would have been designated a potential kidney donor compared with 65.9% estimated in our main results. Using the CDC Wonder database, a national cohort of 8,216 was identified. Applying the percentage from CNICS to this national cohort, there were an estimated 424 donors or 848 kidneys available for donation annually, an increase of 56 kidneys. Our primary findings are slightly underestimated, although comparable with those found in our sensitivity analyses.

Potential Liver Donors

To estimate the number of potential deceased liver donors, 1,214 PWH were excluded from the cohort of 4,921 due to a FIB-4 score greater than 3.25, indicating severe fibrosis. An additional 34 were excluded because of an opportunistic infection at the time of death. An additional 137 were excluded due to a missing FIB-4 score. Those remaining were considered potential deceased liver donors. In total, we identified 3,536 potential deceased liver donors, which represents 71.9% of our final cohort of 4,921. Similar to the kidney analysis, viral load was used to further designate a subset of this group as lower-risk based on the risk of superinfection. Of the 3,536 potential deceased liver donors, 2,330 (65.9%) were viremic and 1,206 (34.1%) were non-viremic. The total group of potential deceased liver donors had a median CD4 count of 222 (IQR: 50, 453). The viremic sub-group had a median CD4 count of 129 (IQR: 22,347) and the non-viremic sub-group a median CD4 count of 401 (IQR: 224, 625) (p-value <.0001). Considering those who were non-viremic as lower-risk donors, we identified 1,206 potential deceased liver donors, which represents 24.5% of our cohort of 4,921 (Table 2).

Applying the percentage of transplantable organs estimated using the CNICS cohort (71.9%) to the national cohort of 12,048 described above, we estimate there were 8,662 potential deceased donors with HIV during this twenty-year period, which equates to about 433 livers annually. Applying the percentage of lower-risk transplantable livers estimated using the CNICS cohort (24.5%) to the national cohort of 12,048, we estimate there were 2,951 lower-risk potential deceased donors with HIV during this twenty-year period, which equates to about 147 livers annually.

Sensitivity Analysis

Using the restricted timeframe of 2005–2018, 76.4% of our cohort would have been designated a potential kidney donor compared with 71.9% estimated in our main results. Again, using the CDC Wonder database, the restricted timeframe, and the same exclusions discussed in our main Results, we estimated a national cohort of 8,216. Extrapolated nationally, we estimate there were 448 liver donors annually, an increase of 15 livers over our main estimate of 433. Although comparable, this estimate shows that our main results were slightly underestimated.

Discussion

The HOPE Act was passed with the intention to mitigate the organ shortage by allowing donors with HIV to donate to PWH with end-stage organ disease. This provides an organ source exclusively designated for PWH, a vulnerable population who are less likely to receive a transplant than their HIV-negative counterparts^5,6. However, transplant clinicians have not observed the large number of organs from donors with HIV that were expected. In 2018, 1,636 PWH were on the kidney transplant waitlist and 437 PWH were on the liver transplant waitlist nationally, revealing a demand for these donor organs²⁷. We estimated there were 792 eligible kidneys and 433 available livers annually, yet only 170 kidney and 53 liver transplants have ever been performed as of December 2020 in the United States²⁸. Our results suggest there may be more organs available for candidates living with HIV than current transplant numbers indicate. While recent research suggests that concerns about superinfection attributable to viral load may not be valid^14,17,18, donors were characterized by presence of active viremia to determine what proportion would meet such a threshold. There were an estimated 200 lower-risk kidneys (based on KDPI and viral load) and 147 lower-risk livers (based solely on viral load) annually from 1999–2018. Compared to the total estimates of 792 kidneys and 433 livers, this distinction highlights the large difference in the total number of potential donors and those who would be considered lower-risk based on viremia and KDPI.

Our estimates for lower-risk donors are conservative, as high viral loads and high KDPI are not necessarily disqualifications for organ donation. For example, we chose to classify those with a high viral load separately to minimize the risk of HIV superinfection; however, the true risk of HIV superinfection is difficult to predict as it is unknown how much of the virus remains in a donor organ after it has been prepared for transplantation²⁹. South Africa, which is known to have low ART resistance rates, has been using organs from donors with high viral loads, with results indicating that the chances of superinfection as a result of a new strain of HIV being passed from the donor are low¹³. Similarly, early results in the United States suggest superinfection may not be a clinical concern as long as recipients maintain a suppressed viral load and adhere to an ART treatment regimen¹⁸. While the largest single exclusion criteria for both lower-risk groups was viral load (2,067 potential deceased kidney donors and 2,330 potential deceased liver donors), our data suggest the proportion of PWH who would confer such a risk are decreasing. More intensive management of HIV including strict adherence to ART regimens could lead to a larger pool of donors with HIV in the future. The other criterion for determination of a lower-risk kidney donor was KDPI<85, yet kidneys from donors with KDPI>85are routinely transplanted, albeit at a lower rate than higher quality organs³⁰. As the practice of transplantation with donor organs from PWH expands, refinement of donor criteria will also occur as evidenced by the consideration of superinfection risk^,15,17,18.

Further education is needed to increase the efficacy of the HOPE Act. In a study of 114 PWH from 2018, only 21.1% of PWH were registered organ donors, however 79.8% of PWH said they were willing to be deceased donors³¹. Although most PWH were aware of the ongoing organ shortage (80.7%), only 24.6% were aware of the HOPE Act in 2016. Of the participants who were aware of the HOPE Act, most learned of it from the news instead of from a healthcare provider³¹. Nguyen et. al. suggests that education should emphasize the legality of donations between PWH and that the process of organ donor registration is identical for PWH and people without HIV. Additionally, education campaigns should be targeted toward both PWH and the medical community about the possibility of organ donation for PWH³¹. Direct discussions about organ donation with a primary care provider may be associated with an increased willingness to donate among patients in the general population³¹. Despite this, discussions about organ donation are uncommon between primary care physicians and their patients, with less than four percent reporting they had conversed with the majority of their patients about the possibility³². By broaching this topic with patients and providing educational resources on the topic, physicians may be able to encourage more patients to consent to organ donation. This is especially important among PWH, as many remain unaware that they are legally able to donate their organs. Current criteria for organ donation can be found on the United Network for Organ Sharing website³³.

A significant portion of the responsibility to implement the HOPE Act falls on Organ Procurement Networks (OPOs) which face several barriers. A total of fifty-eight OPOs evaluate deceased donors in the United States within their respective designated service areas. Predmore et al. identified several barriers to implementation including financial and ethical reasons as well as fears of HIV infection among OPO and hospital staff³⁴. Legal barriers also exist, such as whether a deceased donor’s HIV status should be disclosed to their next of kin. Some participants in their study reported that referrals of donors with an HIV diagnosis were treated differently due largely to stigmatization and assumptions about an HIV diagnosis. The participants suggested that educational efforts and building relationships with local HIV advocacy groups are the best ways to address and overcome these barriers³⁴.

Our study expands on previous work by Boyarsky et al. and Richterman et al. to address important limitations in each. We projected fewer available organs than Boyarsky. However, organ quality was not assessed in their study¹⁶. We estimated a greater number of available kidneys but fewer livers than Richterman et. al. However, their study was limited to data from one city and thus, may have underestimated the national donor pool¹⁷. These highlight the strengths of our study. The utilization of data from a multicenter national cohort of PWH accounts for a large degree of geographic diversity. Detailed clinical information on HIV treatment was used in order to make estimates based on viral load as well as organ quality. These factors allowed us to create a large, comprehensive cohort of potential deceased donors with HIV. Further, our study utilized national death data to extrapolate the estimates from the CNICS cohort to a national scale, resulting in a total of 7,886 potential kidney donors and 7,542 potential liver donors.

There were also limitations to this study. Beyond PWH, the HOPE Act has also allowed people who falsely tested positive for HIV to donate organs to PWH, individuals whose organs would historically have been discarded. Durand et. al estimates an additional 50–100 HIV false-positive donors per year based on accuracy of testing methods and the number of deceased donors annually screened³⁵ and, as CNICS only captures PWH, we have likely underestimated the total pool of donor organs available to PWH as a result of the HOPE Act. Additionally, we assumed that CNICS is representative of the national population of PWH. However, all PWH enrolled in the CNICS cohort were receiving HIV treatment, and therefore may not accurately represent the entire population of potential donors with HIV in the United States. The retrospective design of this study relies on historical data based on practices and treatments that are not reflective of current HIV treatment standards. Prospective studies could elucidate provider acceptance patterns, potential donor consent rates, and candidate willingness to accept such organs while also reflecting contemporary clinical care of PWH. As many PWH are diagnosed at time of death, we are likely missing a pool of potential donors with HIV. We cannot ascertain what percentage of our cohort would have consented to donation. We do not know how many individuals in the CNICS cohort would have been excluded from donation due to death occurring outside of a hospital setting as this data was unavailable. Our timeframe of 1999–2018 includes years when ART was available but had not yet become widespread, and therefore it is possible that our results underestimate the number of potential donors that would be available today. Our sensitivity analyses confirmed this, though the magnitude of such underestimation was minimal (56 kidneys and 15 livers). Assessing donor quality was contingent upon data elements collected within CNICS. While KDPI only has moderate power in predicting kidney graft survival³⁶, this is the measure that is used clinically to determine the allocation of kidney donations. KDPI, however, was missing for a number of PWH due to missing BMI data. As PWH missing KDPI were excluded, we may have subsequently under-estimated the number of potential donors. Last, we were unable to calculate the liver donor risk index because it requires variables that are not recorded in the CNICS cohort. We used FIB-4 as an alternative measure, which has been validated in PWH³⁸. While other organs such as hearts and lungs may be transplanted under the HOPE Act, our study focuses solely on kidneys and livers. The HOPE Act stresses that clinical expertise performing transplants between donors without HIV and recipients with HIV be gained prior to transplants between both donors and recipients with HIV. Greater research on heart and lung transplantation among PWH is warranted to spur expansion of transplantation among PWH, including utilization of organs from donors with HIV. Despite the inherent limitations of this study, these data provide a meaningful update on the estimated number of available donor organs from PWH nationally.

In conclusion, we estimated there were a substantial number PWH who would meet criteria for kidney and/or liver organ donation. Donations from PWH are now legal through the HOPE Act, however these data suggest there are more PWH who could potentially serve as donors than has been realized in current practice. Because of the ongoing organ shortage and the additional vulnerabilities faced by PWH, efforts should be made to educate the medical community and the HIV-positive community on the importance and legality of organ donation.

ABBREVIATION LIST:

A1C

Glycated Hemoglobin

AIDS

Acquired Immunodeficiency Syndrome

ALT

Alanine Aminotransferase

ART

Antiretroviral Therapy

AST

Aspartate Aminotransferase

CDC

Centers for Disease Control

CFAR

Center for AIDS Research

CKD

Chronic Kidney Disease

CKD-EPI

Chronic Kidney Disease Epidemiology Collaboration

CNICS

Center for AIDS Research Network of Integrated Clinical Systems

eGFR

Estimated Glomerular Filtration Rate

ESKD

End-Stage Kidney Disease

FIB-4

Fibrosis 4

HBV

Hepatitis B Virus

HCV

Hepatitis C Virus

HIV

Human Immunodeficiency Virus

HOPE

HIV Organ Policy Equity

ICD10

International Classification of Diseases

IRB

Institutional Review Board

KDPI

Kidney Donor Profile Index

OI

Opportunistic Infection

OPO

Organ Procurement Network

LDKT

Living Donor Kidney Transplantation

PWH

People with HIV