HOPE in Action: A Prospective Multicenter Pilot Study of Liver Transplantation from Donors with HIV to Recipients with HIV

Abstract

Liver transplantation (LT) from donors-with-HIV to recipients-with-HIV (HIV D+/R+) is permitted under the HOPE Act. There are only 3 international single case reports of HIV D+/R+ LT, each with limited follow-up. We performed a prospective multicenter pilot study comparing HIV D+/R+ to donors-without-HIV to recipients-with-HIV (HIV D−/R+) LT. We quantified patient survival, graft survival, rejection, serious adverse events (SAEs), HIV breakthrough, infections, and malignancies, using Cox and negative binomial regression with inverse probability of treatment weighting. Between March 2016-July 2019, there were 45 LTs (8 simultaneous liver-kidney) at 9 centers: 24 HIV D+/R+, 21 HIV D−/R+ (10 D− were false-positive). Median follow-up time was 23 months. Median recipient CD4 was 287 cells/μl with 100% on antiretroviral therapy; 56% were hepatitis C virus(HCV)-seropositive, 13% HCV-viremic. Weighted one-year survival was 83.3% vs. 100.0% in D+ vs D− groups (p=0.04). There were no differences in one-year graft survival (96.0% vs. 100.0%), rejection (10.8% vs. 18.2%), HIV breakthrough (8% vs 10%) or SAEs (all p>0.05). HIV D+/R+ had more opportunistic infections, infectious hospitalizations and cancer. In this multicenter pilot study of HIV D+/R+ LT, patient and graft survival were better than historical cohorts, however, a potential increase in infections and cancer merits further investigation.

INTRODUCTION

End-stage liver disease (ESLD) remains a leading cause of death among people living with HIV (PLWH)^1,2 due in part to co-infection with hepatitis C virus (HCV),³ hepatitis B virus (HBV),⁴ as well as non-alcoholic fatty liver disease ⁵. Liver transplantation has favorable outcomes for selected PLWH. ^6–11 However, there is a serious organ shortage. PLWH in particular have a greater risk of death on the transplant waiting list,^12,13 and lower access to this life-saving treatment.¹⁴

Transplantation of organs from donors with HIV to recipients with HIV (HIV D+/R+) is one strategy to expand the donor pool.¹⁵ There is emerging evidence to support HIV D+/R+ kidney transplantation,^16,17 yet there are only 3 international case reports of HIV D+/R+ liver transplantation with limited follow-up.^18–20 If feasible and safe, HIV D+/R+ liver transplantation could mitigate mortality related to ESLD and the organ shortage for PLWH directly, and all members of the liver waitlist indirectly. However, potential concerns with D+/R+ transplantation include the risk of HIV superinfection,²¹ increased donor-derived infections, increased rejection and/or inferior organ quality due to HIV-associated liver disease in donors.²²

Following passage of the HIV Organ Policy Equity (HOPE) Act in 2013, HIV D+/R+ transplantation can be performed under research protocols in the United States.^23,24 The HHS Secretary is mandated to review research outcomes to determine whether this practice may continue. Thus, in order to inform policy and practice, the objective of this multicenter study was to explore the feasibility and safety of HIV D+/R+ deceased donor liver transplantation, and to directly compare outcomes in HIV D+ vs D− liver recipients.

METHODS

Study Design and Oversight

This was a multicenter prospective observational pilot study at 9 centers (Supplementary Table 1). Five centers shared a pilot protocol (ClinicalTrials.gov number, NCT02602262). More than 400 candidates (kidney plus liver) were consented under this pilot protocol until July 2019 when our group opened an NIH-funded study of HIV D+/R+ LT; after that, candidates consented under the pilot who were still waiting for an organ transitioned into this study, which is ongoing ClinicalTrials.gov number, NCT03734393). Four centers had independent protocols with identical federally-mandated eligibility and outcome requirements²⁴ and shared data but not biospecimens (i.e. data-sharing centers). All transplant centers met HOPE Act Safeguards and Research Criteria (Supplementary Table 2).²⁴ Institutional review boards at each center approved the study. All participants provided written informed consent. Safety monitoring committee review occurred annually.

Study Population

Individuals with HIV ≥ 18 years of age with ESLD had to meet institutional criteria for liver transplantation and HOPE Act Research Criteria²⁴ including (i) CD4 ≥ 100 cells/μL within 16 weeks of transplant and (ii) receiving antiretroviral therapy (ART) with HIV RNA < 50 copies/mL. Liver-only and simultaneous liver-kidney (SLK) transplantation were permitted. Recipient exclusion criteria included active opportunistic infection, prior progressive multifocal leukoencephalopathy, or central nervous system lymphoma.

Treatment Groups

Transplant teams received both HIV D+ and D− liver offers according to national allocation policies. Donors could not have active opportunistic infections or cancer. There were no specific criteria for donor HIV RNA or CD4 count, however teams had to describe effective ART for the recipient post-transplant.²⁴ HIV D− were evaluated according to institutional criteria. Organs from HCV-viremic donors were only utilized for HCV-viremic recipients.

Donor HIV screening included HIV antibody (Ab) or antigen/Ab and HIV qualitative nucleic acid test (NAT) per Organ Procurement and Transplantation Network (OPTN) policy. Donors with no history of HIV and discordant HIV Ab and NAT were suspected to have false positive (FP) tests.²⁵ Confirmatory testing was done by the OPTN or HOPE in Action Core Laboratory (Johns Hopkins School of Medicine, Baltimore, MD); confirmatory testing was negative in all suspected FP cases, thus recipients of FP organs were included in the D− group.

Testing for human herpesvirus 8 (HHV8) infection ²⁶ was performed in donors and recipients by detection of antibodies to one latent and one lytic protein, ORF73 and K8.1, respectively; positivity was defined as reactivity to one or both.²⁷ In those who developed HHV8 disease, we also report results of a more sensitive multiplex assay including additional HHV8 antigens.²⁸ HIV− donors and recipients from data-sharing centers did not have biospecimens available for HHV8 testing.

Immunosuppression varied (Table 1). All participants received prophylaxis for Pneumocystis jirovecii. Prophylaxis for cytomegalovirus (CMV) varied by center (Supplementary Table 3).

Table 1.

Baseline characteristics of liver and simultaneous liver-kidney recipients of organs by donor HIV status

Characteristics	Total (N=45)	HIV D+/R+ (N=24)	HIV D−/R+ (N=21)a	P value
Age (years), median (IQR)	57 (50, 62)	61 (52, 63)	55 (46, 61)	0.17
Male sex, no. (%)	35 (78)	20 (83)	15 (71)	0.48
Race, no. (%)				0.74
White	31 (69)	18 (75)	13 (62)
Black	12 (27)	5 (21)	7 (33)
American Indian	2 (4)	1 (4)	1 (5)
Latino ethnicity, no. (%)	11 (24)	8 (33)	3 (14)	0.18
Blood type, no. (%)				0.79
O	21 (47)	10 (42)	11 (52)
A	17 (38)	10 (42)	7 (33)
B	7 (16)	4 (17)	3 (14)
BMI (kg/m2), median (IQR)	27 (22, 31)	27 (22, 31)	28 (22, 31)	0.86
Hypertension, no. (%)	15 (33)	8 (33)	7 (33)	>0.99
Diabetes, no. (%)	11 (24)	6 (25)	5 (24)	0.93
Cause of ESLDb, no. (%)
HBV	7 (16)	3 (13)	4 (19)	0.69
HCV	25 (56)	14 (58)	11 (52)	0.69
HCC	10 (22)	5 (21)	5 (24)	0.81
NASH	6 (13)	3 (13)	3 (14)	>0.99
Alcoholic liver disease	4 (9)	3 (13)	1 (5)	0.61
Other/cryptogenic/idiopathic	5 (11)	2 (8)	3 (14)	0.65
Biologic MELD-Na, median (IQR)c	20 (14, 24)	17 (15, 22)	22 (14, 26)	0.25
Allocation MELD-Na, median (IQR)d	24 (18, 30)	22 (17, 30)	24 (21, 31)	0.28
Received SLK, no. (%)	8 (18)	3 (13)	5 (24)	0.44
HIV and co-infection characteristics
Duration of HIV infection (years), median (IQR)	23 (16, 28)	23 (16, 27)	23 (16, 30)	0.49
Prior opportunistic infection, no. (%)	9 (20)	5 (21)	4 (19)	>0.99
HIV acquisition risk, no. (%)b
Men who have sex with men	19 (42)	11 (46)	8 (38)	0.60
Heterosexual sex	11 (24)	6 (25)	5 (24)	0.93
Injection drug use	11 (24)	8 (33)	3 (14)	0.18
Other/unknown	10 (22)	4 (17)	6 (29)	0.48
CD4+ T cell count, median (IQR)	287 (154, 461)	287 (168, 447)	320 (144, 476)	0.88
HIV RNA<200 (copies/mL), no. (%)	44 (98)	23 (96)e	21 (100)	>0.99
HBV core IgG positive, no. (%)	28 (62)	14 (58)	14 (67)	0.57
HBV sAg positive, no. (%)	7 (16)	3 (13)	4 (19)	0.69
HCV Ab positive, no. (%)	25 (56)	14 (58)	11 (52)	0.69
HCV NAT positive, no. (%)	6 (13)	2 (8)	4 (19)	0.40
CMV Ab positive, no. (%)	35 (78)	17 (71)	18 (86)	0.30
CMV D+/R−	8 (18)	6 (25)	2 (10)	>0.99
CMV D−/R−	2 (4)	1 (4)	1 (5)
HHV8 Ab positive, no. (%)f	7 (21)	4 (24)	3 (19)	>0.99
ART regimen
INSTI-containing, no. (%)	42 (93)	22 (92)	20 (95)	>0.99
NNRTI-containing, no. (%)	11 (24)	8 (33)	3 (14)	0.18
Protease or cobicistat-containing, no. (%)	2 (4)	2 (8)	0 (0)	>0.49
CCR5-antagonist, no. (%)	3 (7)	0 (0)	3 (14)	0.09
Immunosuppression
Induction regimeng, no. (%)				0.22
ATG + steroids	1 (2)	1 (4)	0 (0)
Basiliximab + steroids	7 (16)	2 (8)	5 (24)
Steroids only	37 (82)	21 (88)	16 (76)
Maintenance
Tacrolimus, no. (%)	42 (93)	23 (96)	19 (90)	0.59
Cyclosporine, no. (%)	3 (7)	1 (4)	2 (10)	0.59
Prednisone, no (%)	42 (93)	22 (92)	20 (95)	>0.99
Mycophenolate mofetil/MMF, no. (%)	42 (93)	23 (96)	19 (90)	0.59

^aOf 21 D−/R+ recipients, 10 donors had a false positive HIV test

^bPercentages may add up to more than 100 since an individual can fall under more than 1 category

^cModel for end-stage liver disease calculated using serum sodium, international normalized ratio, bilirubin, creatinine

^dModel for end-stage liver disease including exception points for medical conditions. Of 24 D+/R+ recipients, 7 had exception points (1 for portopulmonary hypertension, 5 for HCC, 1 for other reasons); of 21 D−/R+ recipients, 7 had exception points (1 for hepatopulmonary syndrome, 4 for HCC, 2 for other reasons)

^eOne participant had a viral load of 214 copies/mL

^fMissing in 12 recipients (7 D+ and 5 D−) as no biospecimens were available

^gSLK recipients: 4 received basiliximab and steroids, and 1 received ATG and steroids

Ab, antibody; ART, ATG, anti-thymocyte globulin; ESLD, end-stage liver disease; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HHV8, Human Herpesvirus-8; HIV, human immunodeficiency virus; IQR, interquartile range; INSTI, Integrase Strand Transfer Inhibitor; MELD, Model For End-Stage Liver Disease; NASH, non-alcoholic steatohepatitis; NNRTI, non-nucleoside reverse transcriptase inhibitors, RNA, ribonucleic acid; sAg, surface antigen; SLK, simultaneous liver-kidney transplant.

Recipient assessments

Recipient and donor characteristics, virologic and immunologic measures were recorded. Model for End Stage Liver Disease score (MELD) was reported both as biologic MELD (laboratory values)²⁹ and allocation MELD (including additional points for cancer or medical complications).³⁰ At follow-up, medications, hospitalizations, infections, and laboratory values were collected. Serious adverse events (SAEs) defined by Division of AIDS (DAIDS) Table for Grading SAEs, Version 2.0 were collected in recipients through 1/24/19; thereafter SAEs were only collected in D+ and FP recipients. Opportunistic infections included Centers for Disease Control AIDS-defining infections³¹ and CMV viremia, aspergillosis and mucormycosis; Kaposi’s sarcoma (KS) was defined as a cancer. Quantitative HIV PCR was performed at baseline, week 1, 2, 3, 4, 13, 26, then every 6 months. CD4+ T-cells were measured at baseline, weeks 4, 8, 13, 26, then every 6 months.

Statistical analysis

Characteristics of donors and recipients were compared using the Kruskal-Wallis test for non-normally distributed continuous variables, and Fischer’s exact or Pearson’s chi-squared tests for categorical variables.

Since this was not a randomized controlled trial, there was potential for imbalance in donor and/or recipient characteristics. Recipient age, HCV viremia, MELD score, and SLK have been associated with increased mortality. Therefore, we used inverse probability of treatment weighting (IPTWs) to ensure balance of these factors between groups. Balance of the weighted population is found in Supplementary Table 4.

Graft failure was defined as liver failure resulting in death, re-listing or re-transplantation. Time-to-event outcomes (all-cause mortality, cause-specific graft failure, and cancer) were estimated using weighted Kaplan-Meier method and compared with weighted Cox regression. Recurrent event outcomes (rejection, HIV breakthrough, SAEs, hospitalizations, and opportunistic infections) were quantified using weighted negative binomial regression. For SAEs and infectious hospitalizations, we followed all recipients until death, study removal, or administrative censorship on January 24, 2019 when SAE collection stopped in the D− arm. For all other recurrent events and time-to-event analyses, we followed recipients until death, study removal, or administrative censorship on March 6, 2020. All analyses were done by two independent analysts to confirm accuracy.

There was no missing data. All analyses were two-tailed (α=0.05) and were performed using Stata 16.0/MP for Linux (College Station, Texas).

RESULTS

Recipient Characteristics

Between March 2016 and July 2019, there were 45 liver transplants performed including 8 SLKs: 24 D+/R+ and 21 D−/R+ (Table 1) with follow-up until March 2020. Overall, median recipient age was 57 years (interquartile range [IQR]: 50, 62), 78% were male, and 69% were white. The most common indication for liver transplant was HCV (56%), 22% had hepatocellular carcinoma (HCC), 16% had hepatitis B virus and 13% had non-alcoholic steatohepatitis. Overall, median biologic and allocation MELD scores were 20 and 24, respectively.

Median duration of HIV was 23 years and 20% had a prior opportunistic infection. All recipients were on ART, median CD4 was 287 cells/μL and 98% had HIV RNA < 200 copies/mL. At baseline, 93% were on an integrase-strand transferase inhibitor (INSTI), 24% non-nucleoside reverse transcriptase inhibitor (NNRTI), and 4% on boosted protease inhibitor or cobicistat-containing regimen. Among recipients, 78% were CMV IgG seropositive and 21% were HHV8 seropositive with antibodies to ORF73 and/or K8.1.

For the 8 SLK recipients, the median age was 54 years (IQR: 50, 62), 6 were male, and 4 were white. The most common indication for SLK was HCV (n=5) and 1 had HCC. Median biologic and allocation MELD scores were 25 and 26, respectively. Additional characteristics stratified by donor HIV status for this group are included in Supplementary Table 9.

Deceased Donor Characteristics

Among 45 donors, 24 were D+ and 21 were D− (10 had FP HIV tests) (Table 2). Median age was 36 and 39 years for D+ and D−, respectively. Among D+, median CD4 was 329 cells/μL. No donors had active opportunistic infections or cancer. Six were discovered to have HIV during donor evaluation. Among 18 with prior HIV diagnoses, 16 (89%) were on ART, of whom 13 (81%) had HIV<400 copies/mL. Among D+ not on ART, median HIV RNA was 46,524 copies/mL. At the time of transplant, genotypes were not done in real time for clinical decision making; however post-transplant genotypes were successfully obtained from 17 of 24 D+ and results are provided in Supplementary Table 5.

Table 2.

Characteristics of liver/liver-kidney donors, according to HIV status.

Characteristics	HIV D+ (N=24)	HIV D−a (N=21)	p-value
Age (years), median (IQR)	36 (29, 44)	39 (28, 44)	0.63
Male sex, no. (%)	19 (79)	14 (67)	0.34
Race, no. (%)			0.42
White	10 (40)	10 (48)
Black	12 (50)	7 (33)
Other	2 (4)	1 (5)
Hispanic ethnicityb, no. (%)	2 (8)	4 (19)	0.40
BMI (kg/m2), median (IQR)	24 (22, 26)	26 (24, 33)	0.14
Hypertension, no. (%)	6 (25)	6 (29)	0.79
Diabetes, no. (%)	2 (8)	2 (10)	>0.99
Cancer, no. (%)	1 (4)	0 (0)	>0.99
Blood type, no. (%)			0.80
O	15 (63)	13 (62)
A	8 (33)	6 (29)
B	1 (4)	2 (10)
Brain death donors, no. (%)	24 (100)	20 (95)	0.47
Cause of death, no. (%)			>0.99
Anoxia	12 (50)	10 (48)
Cerebrovascular/Stroke	5 (21)	5 (24)
Head Trauma	6 (25)	5 (24)
Other	1 (4)	1 (5)
HIV and co-infection characteristics
HIV acquisition risk		NA	-
Men who have sex with men, no. (%)	9 (38)
Heterosexual sex	7 (29)
Injection drug use	4 (17)
Other/unknown	7 (29)
CD4 T cell count, median (IQR)	329 (121, 609)	NA	-
HIV Ab positive, no. (%)	24 (100)	7 (33)	<0.001
HIV qualitative NAT positive, no. (%)c	17 (71)	2 (10)d	<0.001
HBV qualitative NAT positive, no. (%)d	0 (0)	0 (0)	-
HCV qualitative NAT positive, no. (%)	1 (4)	4 (19)	0.17
CMV IgG positive, no (%)	21 (88)	13 (62)	0.08
HHV8 Ab positive, no. (%)e	6 (26)	0 (0)	0.15
New HIV diagnosis, no. (%)	6 (25)	NA	-
Of those with known HIV, on ART, no. (%)f	16 (89)
Of those on ART, HIV RNA<400 (copies/mL), no. (%)	13 (81)
Of those off ART, HIV RNA (copies/mL), median (IQR)	46,524 (15,488, 184,182)
Injection drug use, no. (%)	3 (13)	5 (24)	0.44

^aIncludes 10 donors with false positive HIV screening tests that were negative on confirmatory testing

^bTwo recipients identified as race other and ethnicity Hispanic/Latino

^cNegative in 6 donors on ART who had positive HIV Ab

^dConfirmatory quantitative HIV NAT was undetectable

^eMissing in 1 D+ and 12 D−

^f1 D+ with known HIV but unknown ART status

Ab, antibody; ART, antiretroviral therapy; BMI, body mass index; HCV, hepatitis C virus; HIV, human immunodeficiency virus; HHV8, Human Herpesvirus-8 IQR, interquartile range; IV, NA, not applicable; NAT, nucleic acid testing, ND, not done; RNA, ribonucleic acid.

For 10 HIV false-positive donors, 7 had a false-positive Ab, 2 had a false-positive NAT and 1 was falsely reported positive on next-of-kin interview; all confirmed negative by testing. Five donors were HCV-viremic: 1 D+, 4 D−. HHV8 seropositivity was 25% among 20 D+ and 0% among 9 FP with available testing; biospecimens were not available for the D− group.

Patient survival, graft survival and rejection

Median follow-up time was 23 months (IQR: 15.3, 32.4). There were 8 deaths, 6 in D+ (1 SLK) and 2 in D− recipients (0 SLK). Details are provide in Supplementary Table 6. Causes of death in D+ recipients included: cardiac arrest and aspiration (day 9 post-transplant); graft ischemia after embolization procedure for arterial bleeding and sepsis (day 30 post-transplant); pulmonary hypertension and sepsis (day 140 post-transplant); metastatic myoepithelial parotid gland carcinoma (day 258 post-transplant); HHV8-associated lymphoma and visceral Kaposi sarcoma (KS) (1.1 years post-transplant); allograft rejection and graft failure due to medication non-adherence (3.3 years post-transplant). Deaths among D− recipients included: recurrent HCC (1.4 years post-transplant); recurrent HCC and cholangiocarcinoma (1.5 years post-transplant). Weighted one-year survival was 83.3% for D+ vs. 100.0% for D− recipients (Figure 1A). This translated into a 6.55-fold higher mortality risk for D+ compared to D− recipients (weighted hazard ratio [wHR]=6.55, 95%CI:1.11-36.60, p=0.04) (Table 4).

Figure 1.

Post-transplant survival (A), liver rejection-free survival (B), opportunistic infection free-survival (C), and cancer-free survival (D) for liver and simultaneous liver-kidney recipients by donor HIV status, after weighting.

Table 4.

Post-transplant time-to-event and recurrent outcomes for liver and simultaneous liver-kidney recipients by donor HIV status, after weighting.

Outcomes	Estimate (95% CI)	p-value
Time-to-event
Mortality, wHR	6.55 (1.11-38.60)	0.04
Cancer, wHR	7.34 (1.48-36.39)	0.02
Recurrent events
Rejection episodes, wIRR	0.91 (0.19-4.23)	0.90
IR HIV D+/R+ (per 100 person-years)	15.3 (6.3-37.2)
IR HIV D−/R+ (per 100 person years)	16.8 (5.0-56.2)
SAEs, wIRR	1.60 (0.83-3.06)	0.16
IR HIV D+/R+ (per 100 person-years)	188.6 (118.1-301.1)
IR HIV D−/R+ (per 100 person years)	118.2 (75.0-186.3)
Infectious hospitalization, wIRR	4.25 (1.43-12.62)	0.009
IR HIV D+/R+ (per 100 person-years)	73.8 (39.0-139.9)
IR HIV D−/R+ (per 100 person years)	17.4 (7.3-41.1)
Opportunistic infection, wIRR	5.37 (1.32-21.83)	0.02
IR HIV D+/R+ (per 100 person-years)	20.8 (8.8-48.8)
IR HIV D−/R+ (per 100 person years)	3.9 (1.3-11.8)
HIV breakthrough, wIRR	0.90 (0.12-6.58)	0.92
IR HIV D+/R+ (per 100 person-years)	4.2 (1.1-16.3)
IR HIV D−/R+ (per 100 person years)	4.7 (1.1-19.5)

CI, confidence interval; IR, incidence rate; SAE, serious adverse event; wHR, weighted hazard ratio; wIRR, weighted incidence rate ratio.

There were 2 graft failures among D+ recipients who died due to ischemia and rejection, respectively; in all other deaths, recipients died without liver failure. (Table 3) There was 1 graft failure in a D− recipient immediately post transplant due to hepatic artery thrombosis with re-transplant with a D− liver at day 30. Weighted one-year graft survival was 96.0% for D+ vs. 100.0% for D− recipients. Rejection occurred in 17% and 19% of D+ and D− recipients, respectively (Table 3), with similar incidence rates between groups (weighted incidence rate ratio [wIRR]=0.91, 95%CI:0.19-4.23, p=0.90) (Table 4).

Table 3.

Post-transplant outcomes for liver and simultaneous liver-kidney recipients by donor HIV status

Outcomes	HIV D+/R+ (N=24)	HIV D−/R+ (N=21)	p-value
Median follow-up time (months), (IQR)	18 (12, 24)	28 (21, 40)	0.002
Deaths, no. (%)	6 (25)	2 (10)	0.25
Graft failure, no. (%)	2 (8)	1 (5)a	>0.99
Recipients with any liver rejectionb, no. (%)	4 (17)	4 (19)	>0.99
SLK recipients with any kidney rejection, no. (%)	1 (33)	0 (0)	0.38
Recipients with a SAEa, no. (%)	15 (68)	16 (80)	0.66
Recipients with an infectious hospitalizationa, no. (%)	8 (36)	5 (25)	0.43
Recipients with an opportunistic infection, no. (%)	6 (25)	3 (14)	0.47
Opportunistic infection episodesc, no.	8	3	0.049
Pulmonary aspergillosis, no.	1	0
Candida esophagitis, no.	0	1
CMV,d no.	7	2
Recipients with HIV breakthrough, (%)	2 (8)	2 (10)	>0.99
Recipients with cancer, no. (%)	6 (25)	2 (10)	0.25
Bowen’s disease (squamous cell carcinoma in situ), no.	1	0
Kaposi’s sarcoma and/or HHV8-related lymphoma,e no.	3	0
Myoepithelial carcinoma of right parotid gland, no.	1	0
Anal cancer, no.	1	0
Recurrent hepatocellular carcinoma, no.	0	2

^a1 D− recipient experienced hepatic artery thrombosis immediately post-transplant and was re-transplanted at day 30

^b11 episodes among 8 recipients, 5 episodes in D+/R+ (1 SLK recipient had both a kidney and liver rejection); 6 episodes in D+/R−

^c2 D+ and 1 D− recipients transplanted after SAE censorship date 1/24/19 were excluded from this analysis

^dp-value obtained using unweighted negative binomial regression

^e9 episodes among 7 recipients, 2 HIV D+ recipients had 2 episodes each; 6 (5 HIV D+, 1 HIV D−) of 7 recipients were CMV D+/R−; 7 episodes of CMV viremia without end-organ disease, 2 episodes of probable CMV colitis

^fAmong 2 D+: 1 episode of cutaneous-only KS treated with sirolimus, 2 with visceral disease treated with chemotherapy.

^g1 D+ recipient with Kaposi’s sarcoma of skin, 1 D+ recipient with Kaposi’s sarcoma of liver and lung, 1 D+ recipient with HHV8-related diffuse large B cell lymphoma and Kaposi’s sarcoma of lymph nodes

CMV, cytomegalovirus; SAE, serious adverse event; SLK, simultaneous liver-kidney transplant.

Serious adverse events, ART changes, CD4 over time, HIV breakthrough, infection, and malignancies

SAEs were common, 68% and 80% of D+ and D− recipients, respectively, had ≥1 SAE (Table 3). In the weighted analysis, there was no difference in the incidence rate of SAEs between groups (wIRR=1.60, 95%CI:0.83-3.06, p=0.16) (Table 4).

ART regimens were changed post-transplant in 14 of the 45 recipients, primarily to reduce pill burden or avoid drug-drug interactions. In one case, a change was made from abacavir, dolutegravir, lamivudine to bictegravir, emtricitabine, tenofovir alafenamideto for clinically suspected ART resistance in the donor. HIV breakthrough occurred in 2 D+ and 2 D− recipients due to ART non-adherence (Table 3). In the weighted analysis, there was no difference in breakthrough between groups (wIRR=0.90, 95%CI:0.12-6.58, p=0.92) (Table 4). In 3 cases of breakthrough no sample was available for genotype and HIV viral load was suppressed with re-initiation of ART. In 1 D+ recipient, a sample at the time of breakthrough was available which has been previously published.³⁷ In this recipient, the HIV viral load was 2,080,000 copies per mL and phylogenetic analysis demonstrated that only recipient viral sequence were detected. CD4 counts over time are shown in Supplementary Figure 1.

Infections requiring hospitalization occurred in 36% and 25% of D+ and D− recipients, respectively (p=0.43) and opportunistic infections occurred in 25% and 14% of D+ and D− recipients, respectively (p=0.47) (Table 3). Specifically, there were 11 opportunistic infections among 9 recipients: aspergillosis n=1, candida esophagitis n=1, CMV viremia n=7, CMV colitis n=2. The 9 episodes of CMV occurred in 7 recipients: 5 D+ (4/5 CMV D+/R−) and and 2 D− (both CMV D+/R−); details are provided in Supplementary Table 8.

In the weighted analysis, incidence of infectious hospitalizations was higher in D+ vs D− recipients (wIRR=4.25, 95%CI:1.43-12.62, p=0.009) as was incidence of opportunistic infections (wIRR=5.37, 95%CI:1.32-21.83, p=0.02) (Table 4). There were 6 recipients with untreated HCV pre-transplant. Five were successfully treated with direct acting antivirals (DAAs) post-transplant. One was treated with 12 weeks of DAAs (ledipasvir/sofosbuvir) but had viral relapse; shortly thereafter recurrent HCC was diagnosed and due to poor life expectancy, the decision was made not to retreat. None experienced severe HCV recurrence or graft failure from HCV.

Post-transplant cancers occurred in 6 D+ and 2 D− recipients (p=0.25) (Table 3). KS occurred in 3 D+ recipients with HHV8-seropositive donors and included KS of the skin, KS of liver and lungs, and an HHV8-related large B cell lymphoma with KS in lymph nodes. These recipients had no prior KS; baseline HHV8 serology was unavailable in the recipient with cutaneous KS; the other 2 recipients were seronegative to ORF73 and K8.1, but reactive on the multiplex assay, suggesting prior infection.²⁸ Four cancers were fatal, 2 in D+ (HHV8-associated lymphoma, myoepithelial parotid gland carcinoma) and 2 in D− (HCC, HCC plus cholangiocarcinoma). Among all cancer cases, median time to diagnosis was 7.3 months (IQR: 4.4, 17.5). In the weighted analysis, risk of cancer was higher in D+ vs D− recipients (wHR=7.34, 95%CI:1.48-36.4, p=0.02) (Table 4; Figure 1D).

DISCUSSION

In this multicenter prospective study, we report the first results of HIV D+/R+ liver transplantation in the United States, the first series of HIV D+/R+ LT beyond 3 single case reports, and the first comparing outcomes according to donor HIV status in order to assess the attributable risk. Overall, one-year patient survival was good and graft survival was excellent in the D+/R+ group. However, there were a higher rate of infections and more cancer in the HIV D+/R+ group, which translated to higher mortality. These findings raise important questions.

Historically, liver transplantation in PLWH has been controversial. Early trials demonstrated lower survival, particularly for recipients with HIV/HCV co-infection compared to HCV monoinfection.^6,7 The NIH HIV Multisite Transplant Recipient study (HIV-TR) reported one-year patient and graft survival of 76% and 72%, respectively, in a cohort of HIV/HCV-co-infected recipients from 2003-2010.⁶ National registry studies reported one-year survival from 81-86% in HIV-mono-infected liver recipients.^8,11 A combined US and European series of liver transplantation, with 60% HCV co-infected recipients, reported a one-year survival of 86% for those transplanted after 2012, with significant improvement compared to prior eras.⁹ In particular, our cohort included 8 SLK recipients, of whom 7 survived with close to 2 year follow-up time. This is notable as SLK recipients with HIV were reported to have worse outcomes in NIH HIV-TR⁶ and as a result, some transplant centers may not offer SLK transplantation to individuals with HIV based on these older series. The robust one-year survival we observed (83% D+, 100% D−) likely reflects improving outcomes with advances in HIV and transplant care and the profound impact of curative HCV therapies. Although 56% of our recipients were HCV-seropositive in our study, most had been treated pre-transplant. Thus, only 13% were HCV-viremic and post-transplant we saw no episodes of severe HCV recurrence or HCV-related graft failure with successful DAA treatment.

Rejection occurred in 18% of recipients, with only 1 rejection-related graft failure and no difference in D+ vs D− recipients. The relatively low rejection rate is reassuring, particularly since a trend towards higher rejection has been observed in HIV D+/R+ kidney transplantation.¹⁷

Notably, we observed a higher hazard of death among D+ recipients. One death, a fatal HHV8-associated lymphoma, might be related to donor HIV status since HHV8 infection is likely higher in this group. HHV8 infection is increased in those with HIV, men who have sex with men, and those who use injection drugs,^32–34 though it can occur in transplant recipients without these risks due to immunosuppression.³² In the Multisite HIV-TR study, there were 4 non-fatal cutaneous KS in kidney recipients.¹⁰ In milder cases, KS can be treated by changing immunosuppression from tacrolimus to sirolimus.³⁵ Our experience suggests HIV D+/R+ liver transplantation might pose an increased risk of severe KS post-transplant, not observed in our study of HIV D+/R+ kidney transplantation.¹⁷ More data from a larger cohort are needed to better understand this risk. If confirmed, investigation into whether KS risk could be mitigated with donor selection, HHV8 monitoring, or expanded use of sirolimus. Currently, there are no FDA-approved HHV8 serology tests in the United States,³⁶ which limits screening.

We observed that the incidence of hospitalizations due to infection and opportunistic infections was higher in D+ recipients which may have contributed to higher mortality. Sawinski et al. observed infectious deaths in 20% and 30% of HIV and HIV/HCV liver transplant recipients, respectively in the US national registry.¹¹ Of note, we did not observe conventional AIDS-related infections such as Pneumocystis jirovecii or cryptococcosis; rather we observed 9 cases of CMV viremia in 7 recipients, 1 case of candida esophagitis, and 1 case of pulmonary aspergillosis. The CMV viremia primarily occurred in CMV D+/R− recipients which is not surprising and is one of the most common post-transplant infectious observed generally in transplant recipients. Within NIH Multisite HIV-TR liver study, CMV infection was associated with graft loss,⁶ which we did not observe.

HIV breakthrough was infrequent and not higher in D+ recipients. In most cases, transient viremia was attributable to interruption of ART and suppression was reestablished with medications. In one D+ recipient with high-level viremia after ART interruption, sequence analysis did not show donor HIV.³⁷ This suggests, as seen in HIV D+/R+ kidney transplantation, that clinically significant HIV superinfection is rare.³⁸

Overall, our findings should be taken into consideration with regards to the decision anticipated from the Secretary of Health and Human Services regarding whether or not HIV D+/R+ liver transplantation can be expanded beyond research. The increases we observed among D+ recipients in mortality, CMV viremia, infectious hospitalizations, and cancer in this first pilot study should be weighed carefully against mortality on the liver transplant waitlist for individuals with HIV and advanced chronic liver disease in the context of the current organ shortage.

Our study has important limitations that should be taken into account. It was a relatively small pilot study, yet this multicenter experience expands significantly upon knowledge from the published literature of 3 single case reports of HIV D+/R+ liver transplantation.^18–20 The number of HOPE donors per year has been lower than prior estimates of the full potential.^15,39 Several barriers to HOPE implementation have been identified ⁴⁰ and like any innovation, gradual adoption over time is expected as more organ procurement organizations ⁴¹ and transplant centers ⁴² expand their practice. Another limitation of our study is that it was not randomized; there may have been distinct factors among recipients who accepted D+ organs. Randomization was not feasible given constraints of national organ allocation, matching, and ethics of assigning a donor type when the counterfactual donor is not immediately available. Finally, HHV8 seroprevalence was incomplete as we did not have biospecimens from donors without HIV, though we did test HIV false-positive donors which are reasonable surrogates.

There are many strengths of this study. It is a national study that includes all HIV D+/R+ liver transplants performed in the United States in the first three years of the HOPE Act. There is a control group of HIV D−/R+ liver recipients, which is critical since recipients with HIV have unique risks of infection, rejection, and cancer. Although not randomized, our design approximated “natural randomization” whereby D+ and D− groups were determined by organ availability, outside of investigator control which reduces confounding by indication. Furthermore, we used IPTW to minimize confounding. Donors with suspected false-positive HIV tests²⁵ were considered HIV-positive until confirmatory results were available post-transplant; this is an ideal counterfactual to true HIV+ donors, as provider and participant accepted the liver without knowing the final HIV status. Importantly, our study is multicenter, including 9 centers with diverse immunosuppressive and prophylaxis strategies. Thus, it provides a more generalizable assessment of rejection and infection risk. Finally, we collected all SAEs and hospitalizations to provide a comprehensive safety assessment.

In this multicenter national study, we found that HIV D+/R+ liver transplantation was associated with good short-term patient survival and graft survival. However in D+ recipients, we observed higher mortality, infections, and cancer. These findings warrant further investigation and perhaps consideration of additional donor and recipient infection and malignancy monitoring.

ABBREVIATIONS

ART

antiretroviral therapy

ATG

Antithymocyte globulin

BMI

body mass index

D+/R+

donor-positive recipient-positive

ESLD

end stage liver disease

FP

false positive

HCV

hepatitis C virus

HHV8

human herpes virus 8

HIV

human immunodeficiency virus

HR

hazard ratio

INSTI

integrase strand transferase inhibitor

IQR

interquartile range

KS

Kaposi sarcoma

MELD

Model for end stage liver disease

OPTN

Organ Procurement and Transplantation Network

SRTR

Scientific Registry of Transplant Recipients