Liver Transplantation in Elderly Recipients: Is Age Just a Number?

Key Takeaways

Despite a growing demand for liver transplantation in older patients, many are excluded from referral and/or waitlisting due to concerns about comorbidity burden and general frailty.

Introduction

Liver transplant (LT) is the only curative treatment for end-stage liver disease. The shortage of organs and the complexity of LT has necessitated selection criteria, of which advanced age has historically been an absolute contraindication. Progress in perioperative and surgical care of LT recipients has permitted transplantation in increasingly older patients with acceptable outcomes. However, there is no well-defined acceptable recipient upper age limit. Amidst this debate, the current transplant waitlist is older than it has ever been despite a relative stability in length. At the end of 2020, 21.7% of the adult waiting list was 65 years or older, a marked increase from 9.4% in 2010. ¹ As the population ages and life expectancy increases, the need for LT continues to grow among elderly patients.^2,3

Recipient age has not been established as an independent risk factor for poor outcomes,^4-6 and morbidity in elderly patients undergoing LT is often attributed to medical comorbidities and poor physical function. Post-transplant mortality is greater in all-comers with low body mass index (BMI), ⁷ cardiovascular disease, ⁸ coronary artery disease (CAD), diabetes, chronic obstructive pulmonary disease (COPD), and chronic kidney disease (CKD). ⁹ Frailty, the physiologic decline commonly associated with advanced age, can prognosticate need for delisting and 90-day mortality in transplant candidates. ⁵ This is closely tied to malnutrition and may present as underweight BMI (<18.5), muscle wasting, and weight loss; however, nutrition status may be masked by ascites and obesity. Discerning selection of older patients without these risk factors is at the crux of more favorable outcomes.

The lack of established guidelines for judicious patient evaluation and selection of elderly patients has led to high variability in center-specific selection criteria. We aimed to study specific selection practices and outcomes of patients ≥70 years old undergoing LT. We hypothesized that 1-year graft and patient outcomes would not differ between well-selected elderly patients and their younger peers.

Methods

Patients referred for LT between January 1, 2018 and December 31, 2020 were retrospectively identified using a single center institutional transplant database. Patients were divided into 2 cohorts based on age at time of referral, an “elderly” group (age ≥ 70) and a “young” group (age < 70). Baseline demographics included sex, race, laboratory MELD score, presence of specific complications of cirrhosis, pre-operative functional status, pre-transplant dialysis, ICU admission, or mechanical ventilator, anatomic and surgical factors including multi-organ transplant, presence of portal vein thrombosis, prior abdominal surgery or TIPS, and time to transplant. Donor characteristics including cold ischemia time and donor risk index (DRI) were also collected. DRI is a metric of donor quality which helps predict graft failure after LT and utilizes 7 risk factors, which include age, height, race, donor cause of death, DCD donor, partial/split liver graft, organ geographic location, and cold ischemia time. ¹⁰ Recipients were followed for a minimum period of 12 months. Institutional Review Board (IRB) exemption was obtained from our institution for this study including need for consent.

The primary outcomes were 1-year patient and graft survival. Secondary outcomes included intensive care unit (ICU) and hospital length of stay (LOS), respiratory failure rate (defined as the need for high flow nasal cannula or non-invasive positive pressure ventilation), rate of re-intubation, rate of major post-operative complications (ie, Clavien-Dindo grade ≥ III), and 30-day re-admissions. Pre-operative variables included recipient characteristics, laboratory model for end-stage liver disease (MELD) score, end-stage liver disease complications, pre-transplant dialysis or ventilator need, time to transplant, and poor functional status (based on the Karnofsky Performance Status Scale). The Karnofsky Performance Status Scale measures functional status on an 11-point scale ranging from 100% (no symptoms) to 0% (death), based on a patient’s ability to carry out their normal work and activities, the presence of symptoms, need for assistance, or how much time is spent bed-bound. ¹¹

For univariate analyses, a comparison of continuous variables was done using the Mann-Whitney U test and Student’s t test as appropriate. The Pearson chi-squared coefficient was applied to categorical and binary data elements with Fisher’s exact test used where appropriate. Variables were identified as parametric or non-parametric using the Shapiro-Wilk test where a P > .05 identified a normally distributed sample. Median values with interquartile ranges (IQR) were reported for non-parametric continuous variables and mean values with standard deviations were reported for parametric continuous variables. A P < .05 was considered statistically significant. Kaplan-Meier survival analysis was used to compare 1-year patient and graft survival rates. All analyses were performed using Stata 15 (StataCorp, College Station, TX).

Results

Patient and Donor Characteristics

After rigorous evaluation and successful listing, 322 patients underwent LT during the study period; the most common underlying diagnoses were alcoholic liver disease (42.9%), hepatocellular carcinoma (25.5%), and non-alcoholic fatty liver disease (23.6%). Both the young and elderly cohorts were predominantly white male (61% vs 70%), without serious sequelae of hepatic disease or other medical comorbidities, such as CKD, hepatopulmonary syndrome, or previous abdominal surgery (Table 1). A minority of recipients in both cohorts had poor functional status (49% vs 30%, P = .11). The only significant age-associated pre-operative findings in the elderly group were lower laboratory MELD scores (19 vs 24, P = .02) and a higher prevalence of cancer (60% vs 23%, P < .001). Donor characteristics were similar between groups, namely, DRI (1.47 vs 1.55, P = .57) and cold ischemia time of the liver grafts (3.0 vs 3.6 hours, P = .07).

OPEN IN VIEWER

Table 1.

Comparison of Liver Transplant Recipient and Donor Characteristics in Young (Age <70 Years Old) and Elderly (Age ≥70 Years Old) Cohorts.

Recipient Characteristics	Recipient Age <70 y (n = 302)	Recipient Age ≥70 y (n = 20)	P-Value
Male Gender, n (%)	184 (61)	14 (70)	.42
White race, n (%)	211 (70)	12 (60)	.35
Lab MELD, median (IQR)	24 (17-31)	19 (9-25)	.02
BMI, (kg/m2), median (IQR)	28.2 (24.4-32.4)	25.9 (24.3-29.6)	.22
Living donor, n (%)	5 (2)	0 (0)	1
Re-transplant, n (%)	5 (2)	0 (0)	1
HCC/CCA, n (%)	70 (23)	12 (60)	<.001
History of hepatic hydrothorax, n (%)	28 (9)	3 (15)	.42
History of hepatopulmonary syndrome, n (%)	20 (7)	0 (0)	.62
CRRT/Dialysis pre-transplant, n (%)	46 (15)	1 (5)	.32
ICU pre-transplant, n (%)	47 (16)	2 (10)	.75
Ventilator pre-transplant, n (%)	7 (2)	0 (0)	1
Status 1A, n (%)	3 (1)	0 (0)	1
Multi-organ, n (%)	35 (12)	4 (20)	.28
PVT, n (%)	38 (13)	3 (15)	.73
Previous abdominal surgery, n (%)	135 (45)	12 (60)	.18
TIPS, n (%)	31 (10)	0 (0)	.24
Time to transplant, days, median (IQR)	73 (8-246)	225 (30-281)	.16
Poor functional status, n (%)	148 (49)	6 (30)	.11
Private insurance, n (%)	177 (59)	3 (15)	<.001
Donor characteristics	Recipient age <70 y (n = 302)	Recipient age ≥70 y (n = 20)	P-value
DRI, median (IQR)	1.55 (1.31-1.87)	1.47 (1.24-1.88)	.57
Cold ischemia time, hours, median (IQR)	3.6 (2.6-4.8)	3.0 (2.6-3.8)	.07

MELD, Model for End Stage Liver Disease; BMI, body mass index; HCC, hepatocellular carcinoma; CCA, cholangiocarcinoma; CRRT, continuous renal replacement therapy; ICU, intensive care unit; PVT, portal vein thrombosis; TIPS, Transjugular intrahepatic portosystemic shunt; DRI, donor risk index. Bold used to denote significant findings.

Post-Operative Outcomes

Compared to the young cohort, there was no increase in the rate of severe complications (Clavien-Dindo grade ≥ III) in the elderly group (35% vs 23%, P = .2). Further, on assessment of overall complications during the initial hospital stay, no difference was seen between young and elderly groups (45% vs 60%, P = .2). There were only two outcomes on which recipient age had a negative impact, respiratory failure within the first 24 hours after transplant (1% young vs 10% elderly, P = .02) and ICU re-admissions (6% young vs 20% elderly, P = .04). On Kaplan-Meier analysis, graft and patient survival at 1-year was equivalent across cohorts. Specifically, the 1-year patient survival rate was 94.7% in the young cohort and 90.9% in the elderly cohort (P = .88), and the 1-year graft survival rate was 93.3% in the young cohort as compared to 90.9% in the elderly cohort (P = .72) (Table 2). The median follow-up for those over age 70 was 12 months (IQR 7-16) and for those under 70 was 17 months (IQR 9-25).

OPEN IN VIEWER

Table 2.

Comparision of Post-Operative Outcomes of Young (Age <70 Years Old) and Elderly (Age ≥70 Years Old) Patients Undergoing Liver Transplant.

Post-Operative Outcomes	Recipient Age <70 (n = 302)	Recipient Age >70 (n = 20)	P-Value
Post-transplant LOS, days, median (IQR)	5.4 (4.3-7.9)	5.0 (4.3-7.9)	.78
Total hospital LOS a , days, median (IQR)	7.2 (5.0-15.9)	5.4 (4.6-9.1)	.08
Index post-transplant ICU LOS, days, median (IQR)	1.5 (1.1-2.5)	1.2 (1.1-2.3)	.35
Total post-transplant ICU LOS, days, median (IQR)	1.6 (1.1-2.7)	1.8 (1.1-2.8)	0.5
Total ventilator hours, median (IQR)	0 (0-2.8)	0 (0-0)	.32
Extubation rate, n (%)	231 (76)	19 (95)	.056
Re-intubation within 24hrs, n (%)	0 (0)	0 (0)	-
Respiratory failure b in 24hrs, n (%)	2 (1)	2 (10)	.02
ICU readmission, n (%)	19 (6)	4 (20)	.04
Pneumonia, n (%)	8 (3)	0 (0)	.46
Any complication during hospitalization, n (%)	137 (45)	12 (60)	0.2
Severe complications (Clavien-Dindo ≥ grade III), n (%)	68 (23)	7 (35)	0.2
30 Day readmission, n (%)	97 (32)	8 (40)	.47
Direct cost, $, median (IQR)	126,183 (112,757-161,806)	113,729 (104,957-151,922)	.07
1-year patient survival, % (95%CI)	94.7 (91.1-96.8)	90.9 (50.8-98.7)	.88
1-year graft survival, % (95%CI)	93.3 (89.5-95.8)	90.9 (50.8-98.7)	.72

LOS, length of stay; ICU, intensive care unit.

^aTotal hospital LOS includes pre-transplant and post-transplant days admitted.

^bDefined as the need for high flow nasal cannula or non-invasive positive pressure ventilation. Bold used to denote significant findings.

Discussion

In this study, advanced age alone was not a significant risk factor for poor patient or graft outcomes in LT recipients. Specifically, age ≥70 years did not lead to differences in 1-yearpatient and graft survival, or worsening of secondary post-operative outcomes including hospital or ICU LOS, in-hospital complications, or Clavien-Dindo complications ≥ grade III. Despite higher rates of post-operative high flow nasal cannula or non-invasive positive pressure ventilation, elderly patients had higher primary extubation rates (95% vs 76%, P = .056) and furthermore, none required re-intubation. An increased rate of ICU re-admissions was observed and was likely in part due to a lowered clinical threshold to transfer elderly patients to a higher level of care.

Within a conflicted body of literature, our findings support the cautious transplantation of well-selected elderly patients. Wilson et al came to similar conclusions in one of the largest US multicenter studies of elderly LT patients. In their case-controlled analysis, recipient age was not identified as an independent risk factor for graft loss or patient mortality at 5-year follow-up. ¹² However, other studies have failed to come to the same conclusion. For example, Sharpton et al found that age was associated with 1-year graft loss, though elderly graft loss was attenuated at lower MELD scores. ¹³ Significantly lower 5-year survival rates in patients over age 70 were reported by Sharma et al based on 2002-2012 UNOS data, ¹⁴ and studies from Chen et al and Gil et al have reported similar findings in Taiwanese and Korean populations, respectively.^15,16 A more recent systematic review and meta-analysis of 10 studies including 4,752 elderly (age ≥70) LT recipients demonstrated higher perioperative mortality and graft failure in this population, though with considerable heterogeneity noted upon analysis. ¹⁷ Whether this variance across studies stems from center-specific factors, progressive improvements in surgical and intensive care over the last decade, or a wide variation in selection practices and patient populations, it underlines that advanced recipient age should not be an absolute contraindication for referral or transplant without characterization of physiologic age.

Over the last several years, many physical frailty screening tools have been developed to better objectively assess patient physiologic age and medical vulnerability. Common factors of these measures include recent weight loss, diminished capacity for physical activity, and presence of certain chronic illnesses. In addition, sarcopenia, or the loss of muscle mass and skeletal strength, is central to the functional decline seen in older adults and can also be objectively accounted for. In patients awaiting LT, frailty has been firmly established as an independent risk factor for poor outcomes. Post-transplant patients with higher Liver Frailty Index scores had higher morality rates and health care utilization costs, including hospital length of stay and non-home discharge. ¹⁸ Measurement of frailty provides rich data regarding functional status and suitability for transplantation more than chronologic age, and should be part of the work-up for all transplant candidates, especially those who are older.

We found elderly patients were more likely to undergo LT for malignancy, representing 60% of our elderly study population in comparison to 23% of younger patients. While locoregional therapies exist for hepatocellular carcinoma (HCC), some patients may not be eligible due to tumor distribution or burden, or underlying cirrhosis. Recent data from Shah et al reported a survival advantage for elderly patients with HCC undergoing LT vs liver-directed therapy alone. ¹⁹ Patients undergoing LT had more significant comorbidities and higher MELD scores, and yet 5-year survival rates in the LT group were 70% compared to 31% in the directed therapy group. This study also found a yearly 3% increase in risk of mortality in patients over age 70 regardless of treatment option. Taken together these findings emphasize that while LT is beneficial in these patients, careful counseling of elderly patients prior to listing remains paramount.

While many patients may be adequate candidates at the time of initial evaluation, an extended waiting period prior to LT may result in changes to clinical status, especially in older patients. In our study, elderly patients waited almost three times as long for LT (225 vs 73 days). This longer wait for elderly patients has been seen across centers and regions, and likely due to lower MELD scores, with higher rates of HCC and preserved liver function. Almost one-third of the current waitlist is over age 65, and increased wait times may challenge the feasibility of LT and necessitate removal in declining patients. ²⁰ Thus, optimizing outcomes in elderly patients requires appropriate selection and ongoing re-evaluation. Specifically, older patients should be held to more rigorous and stringent selection criteria than their younger peers who may be listed with higher MELD scores as they typically have less physiologic reserve. Gilbo et al found elderly patients to have a low-risk profile when they demonstrated a more stable disease, that is, HCC within Milan criteria, lower MELD (<15), and not admitted to the ICU or on continuous renal replacement therapy prior to organ offer. However, the only significant predictor of poor patient survival was hospitalization in the ICU prior to LT. ²¹

Limitations of this study include sample size and the single-institution population. Additionally, it is important to note that although it is now a standard practice at our institution, during the study period we did not conduct a formal frailty or sarcopenia assessment in our patient cohorts. Lastly, due to a median overall follow-up time of 16.4 months, we were unable to examine more remote metrics such as 5-year outcomes, which are critical to understand the benefits and durability of LT in this population. Despite this, because of the paucity of granular data on patients over age 70 in the literature, this study addresses gaps and helps inform selection guidelines in these patients.

Many questions remain unanswered with regards to liver transplantation of elderly recipients. Indeed, the burden of overall medical comorbidities, frailty, and nutritional status should be objectively assessed, and multidisciplinary judgment applied in determining if LT is likely to confer survival benefit to a patient. This includes the potential utility of pre-habilitation protocols in expanding indications and even further improving outcomes of this higher risk group of patients. Considering our finding of a correlation between age of recipients, ICU readmission, and respiratory failure, dedicated post-operative protocols to help mitigate these risks are being established. Importantly, as most research has been focused on graft and patient survival rather than the quality of life after transplant, further investigation into this as well as functional status remains imperative.