Outcomes of Simultaneous Liver–Kidney Transplant Recipients According to Pre-Transplant Transjugular Intrahepatic Portosystemic Shunt (TIPS) in the United States

Abstract

Background: Previous data suggested that the outcomes for liver-alone transplant recipients following transjugular intrahepatic portosystemic shunt (TIPS) insertion were comparable to those without TIPS. This study investigates the association between TIPS and outcomes among simultaneous liver–kidney (SLK) recipients in the United States. Methods: Utilizing the Scientific Registry for Transplant Recipients (SRTR) standard analysis file from 2003 to 2022, we examined 9717 adult SLK recipients, among whom 858 had undergone TIPS before transplantation. Kaplan–Meier curves were generated to assess recipient and death-censored liver and kidney graft survival. Mixed-effects Cox proportional hazard models were employed to analyze the association between TIPS and the outcomes of interest, where the transplant center was treated as a random effect. The models were adjusted for recipient age, sex, MELD score, diabetes, duration of listing, induction, steroid maintenance, hepatitis C status, donor age, donor sex, cold ischemia time, local vs. shipped organs, and allocation era. Results: Overall, the two groups were comparable, with minor differences. Notably, the median liver waiting time was significantly longer in the TIPS group compared to the non-TIPS group (4.1 vs. 2 months, p < 0.001). One-year rejection rates for liver and kidney allografts did not differ significantly between groups. Univariable Cox regression analyses demonstrated no association between TIPS and worse outcomes for recipient, liver, and kidney survival (p = 0.65, p = 0.22, and p = 0.54, respectively). TIPS did not emerge as a predictor of recipient or death-censored liver or kidney graft survival in multivariable models. Conclusion: In this extensive national cohort of SLK transplant recipients, pre-transplant TIPS was not linked to adverse outcomes for recipients or their allografts.

1. Introduction

Portal hypertension is a common sequela of cirrhosis that contributes markedly to morbidity and mortality [1]. A variety of medical therapies are available for managing the complications of portal hypertension, such as diuretics, nonselective beta-blockers, or octreotide. However, advanced stages of portal hypertension, marked by the development of recurrent ascites requiring large volume paracenteses or recurrent variceal bleeding, are often refractory to medical therapy and negatively impact patient quality of life. A common option in these cases is the insertion of a transjugular intrahepatic portosystemic shunt (TIPS) within the liver parenchyma, which decreases portal hypertension by shunting blood from the portal vein to the systemic circulation [2,3]. TIPS is often used in this setting as a bridge to liver transplant, improving portal hypertension-related symptoms [4,5]. Apart from transplant surgery, preoperative TIPS in individuals with chronic liver failure was associated with better postoperative outcomes, especially visceral surgeries [6]. In a meta-analysis [7], TIPS prior to liver transplantation (LT) was not associated with increased bleeding complications perioperatively, such as increased red blood cell transfusion or fresh frozen plasma requirements. Nonetheless, TIPS can be associated with surgical challenges during transplantation, especially if mispositioned [8,9].

In patients with concordant prolonged kidney dysfunction, such as patients with hepatorenal syndrome, simultaneous liver and kidney transplantation (SLK) has been shown to have substantial survival benefits [10]. These patients have superior graft survival compared to those who require post-liver transplantation dialysis or those who receive kidney after liver transplantation (KALT). Single-center and national registry reports on the short-term outcomes of LT recipients with a history of TIPS placement are not different or are slightly better than those without [7,11,12,13]. SLK candidates who require TIPS may be sicker than those who did not require TIPS and have required more blood transfusions than their counterparts, which can lead to a higher degree of sensitization [14]. TIPS has been associated with worsening pulmonary hypertension [15,16], a known predictor of poor outcomes after liver [17] and kidney transplant [18].

Long-term data on outcomes for simultaneous liver–kidney recipients with a history of transjugular intrahepatic portosystemic shunt are currently unavailable. This study examines the long-term outcomes of SLK recipients categorized by their TIPS status. Our analysis relies on the Scientific Registry of Transplant Recipients, chosen for its consistent tracking of pretransplant TIPS status and comprehensive records of graft and recipient survival.

2. Materials and Methods

2.1. Data Source

This study used data from the Scientific Registry of Transplant Recipients (SRTR). The SRTR data system includes data on all donors, wait-listed candidates, and transplant recipients in the US, submitted by the members of the Organ Procurement and Transplantation Network (OPTN). The Health Resources and Services Administration (HRSA), U.S. Department of Health and Human Services, provides oversight to the activities of the OPTN and SRTR contractors.

The study authors were included in the project plan and were approved to participate by the SRTR. Organs from executed prisoners are not used in the United States. This study was exempted by the Mayo Clinic Institutional Review Board and compliant with the SRTR and HRSA data use agreement.

2.2. Study Population

Using the Scientific Registry for Transplant Recipients (SRTR) standard analysis file, we identified all adult primary SLK recipients between 2003 and 2022. We excluded those with unknown TIPS status. The cohort consisted of 9717 adult simultaneous liver–kidney recipients. The cohort was dichotomized according to TIPS into the TIPS group (n = 858) and no TIPS group (8859).

2.3. Outcomes of Interest

The primary outcomes of interest are recipient survival and death-censored liver and kidney graft survival. Our secondary outcomes of interest include liver and kidney allograft rejection rates at six- and twelve-month intervals.

2.4. Statistical Analyses

Data was summarized using means ± standard deviations or medians and interquartile ranges (IQRs) for continuous data, and counts and percentages for categorical data. Baseline characteristics were compared between those with/without TIPS using a t-test or Wilcoxon rank sum test for continuous data, and a Chi-square test for categorical data. Survival methods including the Kaplan–Meier method and mixed-effects Cox proportional hazards regression models were employed to assess the association between TIPS and the outcomes of interest, where the transplant center was treated as a random effect. The multivariable models were adjusted for recipient age, sex, MELD score, diabetes, duration of listing, induction, steroid maintenance, hepatitis C status, donor age, donor sex, cold ischemia time, local vs. shipped organs, and allocation era. Associations were summarized using hazard ratios (HRs) and 95% confidence intervals (CIs). All analyses were performed using SAS version 9.4 software (SAS Institute, Inc.; Cary, NC, USA) and R version 4.2.2 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria), and p-values < 0.05 were considered statistically significant.

3. Results

3.1. Baseline Characteristics

The cohort consisted of 9717 SLK transplant recipients, of whom 858 had pretransplant TIPS. Recipients in the TIPS group were slightly older, 56.9 (±9.2) years vs. 55.6 (±10.5) years in the group without TIPS (p = 0.01). More recipients were males and were distributed equally between groups. Black recipients were less likely to have received TIPS. The proportion of recipients with MELD scores above 30 was lower in the TIPS group (41.5% vs. 45.2%, p = 0.013). On average, the waitlisting time was nearly doubled in the TIPS group (4.1 vs. 2 months, p < 0.001).

Compared to recipients without TIPS, polycystic liver disease was a less frequent indication for liver transplantation in recipients with TIPS, while non-alcoholic steatohepatitis was a more frequent indication. Otherwise, the liver disease etiologies were similar between groups. Hepatitis C infection and diabetes were observed more frequently in the TIPS group. At the time of transplantation, the average creatinine was lower in recipients with TIPS (3.1 mg/dL vs. 3.6 mg/dL, p < 0.001). However, the proportions of recipients on dialysis were similar between groups (66.2% vs. 68.3%, p = 0.31). Antibody induction immunosuppression use was more frequent in the TIPS group. Crossmatch-positive transplants were more frequently observed in recipients with TIPS. Noteworthy, the proportion of recipients who experienced encephalopathy prior to transplantation was higher in the TIPS group (77.4% vs. 67.3%, p < 0.001) compared to that in the group without TIPS. The remainder of the characteristics were similar between groups and are detailed in

Table 1. Baseline clinical characteristics according to TIPS status prior to transplantation.

	TIPS n = 858	Without TIPS n = 8859	p-Value
Recipient Age, years	56.9 ± 9.2	55.6 ± 10.5	0.010
Male recipient gender	554 (64.6%)	5557 (62.7%)	0.29
Recipient Race			<0.001
White	729 (85.0%)	7090 (80.0%)
Black	80 (9.3%)	1282 (14.5%)
Other	49 (5.7%)	487 (5.5%)
BMI	28.3 ± 9.6	27.6 ± 8.4	0.001
MELD			0.013
0–19	76 (8.9%)	608 (6.9%)
20–29	435 (50.7%)	4242 (47.9%)
30–39	264 (30.8%)	2943 (33.2%)
40+	83 (9.7%)	1066 (12.0%)
Encephalopathy			<0.001
None	194 (22.6%)	2896 (32.7%)
1–2	516 (60.1%)	4707 (53.1%)
3–4	148 (17.3%)	1255 (14.2%)
Kidney List, months, median (IQR)	1.8 (0.4, 6.4)	1.1 (0.3, 4.8)	<0.001
Liver List, months, median (IQR)	4.1 (0.9, 13.2)	2.0 (0.5, 7.6)	<0.001
Diagnosis of Liver Disease			<0.001
Alcohol	237 (27.6%)	2328 (26.3%)
Autoimmune	33 (3.8%)	321 (3.6%)
Cryptogenic	52 (6.1%)	544 (6.1%)
HCC	76 (8.9%)	763 (8.6%)
Metabolic	10 (1.2%)	231 (2.6%)
NASH	185 (21.6%)	1423 (16.1%)
PBC	13 (1.5%)	183 (2.1%)
PSC	12 (1.4%)	212 (2.4%)
Polycystic	3 (0.3%)	398 (4.5%)
Other	237 (27.6%)	2456 (27.7%)
Hepatitis C			0.024
Positive	262 (30.5%)	2480 (28.0%)
Negative	583 (68.0%)	6120 (69.1%)
Unknown	13 (1.5%)	259 (2.9%)
On Dialysis	568 (66.2%)	6050 (68.3%)	0.31
Last Creatinine	3.1 (2.1, 4.5)	3.6 (2.4, 5.2)	<0.001
Diabetes			<0.001
No	427 (49.8%)	5205 (58.8%)
Yes	423 (49.3%)	3601 (40.6%)
Unknown	8 (0.9%)	53 (0.6%)
Induction			0.006
Depletional	144 (16.8%)	1376 (15.5%)
Non-Depletional	500 (58.3%)	4809 (54.3%)
None Noted	214 (24.9%)	2674 (30.2%)
Steroid Maintenance	663 (77.3%)	6995 (79.0%)	0.25
Donor Age, years	36.3 ± 14.1	35.6 ± 13.8	0.25
Male donor gender	537 (62.6%)	5409 (61.1%)	0.38
Donor Race			0.25
White	688 (80.2%)	7155 (80.8%)
Black	130 (15.2%)	1389 (15.7%)
Other	40 (4.7%)	315 (3.6%)
Donor BMI	26.8 ± 5.9	26.9 ± 5.9	0.92
Local Organ	612 (71.3%)	6079 (68.6%)	0.10
Kidney Cold Ischemia Time a	10.3 (8.2, 15.0)	10.5 (8.0, 14.8)	0.53
Liver Cold Ischemia Time b	6.0 (4.8, 7.5)	6.1 (5.0, 7.6)	0.27
Number of HLA Mismatches			0.23
0–2	25 (2.9%)	243 (2.7%)
3–4	282 (32.9%)	2847 (32.1%)
5–6	426 (49.7%)	4665 (52.7%)
Unknown	125 (14.6%)	1104 (12.5%)
Crossmatch Status			0.041
Negative	165 (74.3%)	1987 (79.1%)
Positive	57 (25.7%)	525 (20.9%)
Missing	636	6347

Values are presented as median [IQR] or N (%). Abbreviations: HCC, hepatocellular carcinoma; MELD, model for end-stage liver disease; NASH, non-alcoholic steatohepatitis. ^a Available in 819 with TIPS and 8252 without TIPS. ^b Available in 841 with TIPS and 8652 without TIPS.

.

3.2. Univariable Outcomes

The six- and twelve-month kidney rejection rates did not differ statistically by TIPS status (2.7% vs. 3.5% and 4.2% vs. 5.5%, in TIPS vs. no TIPS groups, respectively; p > 0.05). Similarly, six- and twelve-month liver rejection rates did not vary by TIPS status (5.3% vs. 5.9%; p = 0.51 and 7.5% vs. 8.7%, respectively; p = 0.27). The rates of re-hospitalization and post-transplant lymphoproliferative disorder (PTLD) and the average creatinine at one year were not different between groups. (

Table 2. Recipient and allograft short-term outcomes.

	TIPS	Without TIPS	p-Value
Kidney Rejection within 6 months	19/704 (2.7%)	252/7205 (3.5%)	0.27
Kidney Rejection within 12 months	28/660 (4.2%)	373/6733 (5.5%)	0.16
Liver Rejection within 6 months	38/719 (5.3%)	432/7336 (5.9%)	0.51
Liver Rejection within 12 months	50/669 (7.5%)	605/6923 (8.7%)	0.27
Hospital Readmission within 12 months	226/704 (32.1%)	2459/7146 (34.4%)	0.22
PTLD within 12 months	1/661 (0.2%)	18/6718 (0.3%)	0.57
12-month eGFR a mL/min/1.73 m2	63.9 (21.2)	64.3 (21.0)	0.97

Values are presented as median and interquartile range (IQR) or N (%). Abbreviations: PTLD, post-transplant lymphoproliferative disorder; Cr., creatinine. ^a Available in 669 with TIPS and 6849 without TIPS.

).

In the univariate analyses for recipient survival and death-censored liver and kidney graft loss, TIPS status did not emerge as a predictor of outcomes (p = 0.65, p = 0.22, p = 0.54, respectively) (Figure 1, Figure 2 and Figure 3).

3.3. Determinants of Recipient Death

TIPS was not associated with worse recipient survival in the multivariable analysis. Older recipient age, male sex, higher MELD scores, hepatitis C infection, diabetes, and older donor age were predictors of lower recipient survival. Steroid maintenance was associated with lower recipient mortality [HR 0.76, 95% C.I. (0.69–0.82), p < 0.001]. Recipients of local organs had 14% better survival [HR 0.86, 95% C.I. (0.79–0.94), p = 0.001]. Recipient survival was 20% better for those who received their transplants between 2013 and 2022 compared to the era between 2003 and 2012 [HR 0.80, 95% C.I. (0.74–0.88), p < 0.001] (

Table 3. Multivariable Cox proportional hazard model for patient death.

	Hazard Ratio	95% Confidence Interval	p-Value
TIPS	0.98	0.86–1.12	0.79
Recipient age (per 10 years)	1.19	1.14–1.24	<0.001
Male Recipient Gender	1.10	1.02–1.19	0.019
MELD
<20	Reference
20–29	1.12	0.96–1.30	0.16
30–39	1.24	1.06–1.45	0.007
40+	1.30	1.09–1.55	0.004
Liver Waitlist (per 6 months)	1.00	0.98–1.01	0.77
Immunosuppression Induction
Depletional	1.10	0.99–1.22	0.077
Non-Depletional	Reference
None	1.04	0.96–1.13	0.35
Steroid Maintenance	0.76	0.69–0.82	<0.001
Hepatitis C
Positive	1.18	1.09–1.28	<0.001
Negative	Reference
Unknown	1.20	0.99–1.45	0.065
Donor age (per 10 years)	1.12	1.09–1.15	<0.001
Male Donor Gender	0.98	0.91–1.06	0.59
Local Organ	0.86	0.79–0.94	0.001
Liver Cold Ischemia Time	1.00	0.99–1.01	0.80
New Allocation Era
2003–2012	Reference
2013–2022	0.80	0.74–0.88	<0.001
Diabetes	1.38	1.28–1.49	<0.001

Abbreviations: MELD, model for end-stage liver disease.

).

3.4. Determinants of Death-Censored Graft Failure

In the multivariable models for death-censored liver or kidney failure, TIPS status was not associated with increased liver or kidney graft loss. Older recipient age, steroid maintenance, and a newer transplant era were associated with a lower risk of liver graft failure. Older donor age and recipient hepatitis C infection and diabetes were associated with an increased risk of liver or kidney graft failure. A higher MELD score was associated with an increased risk of long-term kidney graft failure but was not a predictor of liver graft failure (

Table 4. Multivariable Cox proportional hazard model for death -censored liver allograft failure.

	Hazard Ratio	95% Confidence Interval	p-Value
TIPS	0.75	0.51–1.09	0.13
Recipient age (per 10 years)	0.76	0.70–0.84	<0.001
Male Recipient Gender	0.98	0.80–1.19	0.80
MELD
<20	Reference
20–29	0.81	0.57–1.16	0.25
30–39	0.85	0.59–1.23	0.40
40+	1.03	0.68–1.56	0.88
Liver Waitlist (per 6 months)	0.97	0.93–1.01	0.18
Immunosuppression Induction
Depletional	0.82	0.61–1.09	0.18
Non-Depletional	Reference
None	1.05	0.85–1.30	0.63
Steroid Maintenance	0.78	0.62–0.99	0.039
Hepatitis C
Positive	1.82	1.49–2.23	<0.001
Negative	Reference
Unknown	1.45	0.90–2.33	0.13
Donor age (per 10 years)	1.18	1.10–1.26	<0.001
Male Donor Gender	0.91	0.75–1.10	0.33
Local Organ	0.96	0.76–1.22	0.74
Liver Cold Ischemia Time	1.01	0.98–1.03	0.61
New Allocation Era
2003–2012	Reference
2013–2022	0.57	0.46–0.71	<0.001
Diabetes	1.28	1.05–1.56	0.016

Abbreviations: MELD, model for end-stage liver disease.

and

Table 5. Multivariable Cox proportional hazard model for death-censored kidney allograft failure.

	Hazard Ratio	95% Confidence Interval	p-Value
TIPS	0.91	0.71–1.16	0.44
Recipient age (per 10 years)	0.88	0.82–0.94	<0.001
Male Recipient Gender	0.92	0.80–1.07	0.28
MELD
<20	Reference
20–29	1.41	1.04–1.93	0.030
30–39	1.48	1.07–2.03	0.017
40+	1.54	1.08–2.18	0.016
Liver Waitlist (per 6 months)	0.99	0.96–1.02	0.45
Immunosuppression Induction
Depletional	0.86	0.70–1.06	0.17
Non-Depletional	Reference
None	1.05	0.90–1.22	0.53
Steroid Maintenance	0.80	0.68–0.95	0.010
Hepatitis C
Positive	1.30	1.13–1.51	<0.001
Negative	Reference
Unknown	1.46	1.05–2.03	0.025
Donor age (per 10 years)	1.33	1.26–1.39	<0.001
Male Donor Gender	0.96	0.83–1.10	0.54
Local Organ	0.76	0.64–0.90	0.001
Liver Cold Ischemia Time	1.01	0.99–1.03	0.46
New Allocation Era
2003–2012	Reference
2013–2022	0.71	0.60–0.83	<0.001
Diabetes	1.50	1.31–1.73	<0.001

Abbreviations: MELD, model for end-stage liver disease.

).

4. Discussion

While numerous analyses have explored the association between TIPS and waitlist survival in liver transplant candidates, the long-term outcomes post-transplant, particularly for SLK recipients, have not been well described. In this large cohort of transplant recipients with cirrhosis and renal failure requiring simultaneous liver–kidney transplants, we examined the long-term outcomes of recipient and grafts survival according to TIPS status. Our results can be summarized as follows: (1) the long-term outcomes of SLK recipients who had TIPS prior to transplant were not worse than those without TIPS; (2) SLK recipients who had TIPS prior to transplantation did not have an increased risk of rejection of either organ, PTLD or rehospitalization by one-year post-transplant compared to those without TIPS.

The use of a transjugular intrahepatic portosystemic shunt (TIPS) can act as a transitional solution to address severe complications associated with portal hypertension, such as refractory ascites, portal hypertensive bleeding, hepatic hydrothorax, and portal vein thrombosis. TIPS placement also holds the potential to extend life, enhance outcomes, and improve patient comfort. Berry et al. [5] demonstrated that patients awaiting liver transplantation who underwent TIPS had a lower mortality rate compared to those without intervention. Nonetheless, SLK candidates who require TIPS are more complex due to the risk of worsening pulmonary hypertension [15,16] and the higher immunologic risk evident by more positive crossmatch transplants, likely related to blood transfusion and potential sensitization. Despite the acknowledged benefits of TIPS, the procedure comes with inherent risks, including the development of hepatic encephalopathy, stent dysfunction or occlusion, and mispositioning, necessitating further intervention. In our analysis, we observed a 10% higher frequency of encephalopathy prior to transplantation in patients with TIPS. Additionally, more severe grades of encephalopathy were observed more frequently in the TIPS group. These risks underscore the critical importance of careful patient selection, meticulous procedural technique, and ongoing monitoring to mitigate complications and optimize patient outcomes following TIPS placement.

The specific reasons for undergoing TIPS placement are unavailable in the SRTR database. However, the indications for TIPS placement are similar between patients who receive liver-only transplants and those who undergo SLK. Interestingly, the MELD score was initially developed [19,20] to predict survival in cirrhosis patients undergoing the TIPS procedure, with patients with higher MELD scores (≥18) showing worse survival after TIPS placement. Serum creatinine is a component of the MELD score calculation. Patients with significant renal dysfunction are likelier to have higher MELD scores, leading hepatologists to be reluctant to offer TIPS in this context. However, most recipients in the TIPS group had a MELD score ≥ 20 immediately before transplantation. Nonetheless, a MELD score of 40 or higher was less frequently observed in the TIPS group.

Our results complement those of Berry and colleagues, who found that TIPS is associated with lower waitlist mortality. We also show that long-term outcomes post-transplantation are similar in those who underwent simultaneous liver–kidney transplantation with or without prior TIPS.

One of the known complications of the TIPS procedure is mispositioning, which may result in additional surgical challenges. Our study analyzed the cold ischemia time before liver and kidney transplantation. The liver cold ischemia time was similar between the two groups. The kidney cold ischemia in the SLK recipients was also similar, irrespective of TIPS status, indicating a similar operative liver course between the two groups.

Strengths and Limitations

Our study is the largest and has the longest follow-up, describing the association between TIPS and the long-term outcomes of simultaneous liver–kidney recipients. We utilized the SRTR and chose well-tracked outcomes of recipient and allograft survival. However, there are a few limitations to point out. The standard analysis file did not include the indication for TIPS or granular data on the severity of portal hypertension to ascertain those who might have needed TIPS and did not receive it. Also, we acknowledge the lack of data pertaining to the potential worsening of pulmonary hypertension after TIPS.

Additionally, the crossmatch data were missing in a significant portion of the study population, which did not allow us to include it in the models. However, our group previously reported on the protective effect of the liver in SLK recipients [21,22] and similar outcomes of crossmatch-positive and crossmatch-negative SLK recipients have been reported [23]. Data on misposition, type of stent, or other TIPS-related complications were not available to analyze.

5. Conclusions

In conclusion, TIPS placement in patients who subsequently received a simultaneous liver and kidney transplant was not associated with increased mortality or allograft rejection. Additional research aimed at analyzing the appropriateness of TIPS placement, indications and incorporating the severity of TIPS-related complications into the database, would enhance our understanding and help to elucidate further the role of TIPS placement in both clinical and surgical management strategies for these patients.