Outcome after liver transplantation: Predicting the unpredictable

Abstract

A good outcome after liver transplantation (LT) is a major target of the procedure. Several scoring systems have been evaluated in this setting. The Charlson Comorbidity Index was evaluated in living donor LT. The findings of this study are commented upon together with the most recent data on this field.

Key Words: Liver transplant outcomes; Prognostic scoring systems; Frailty; Charlson Comorbidity Index; Liver transplantation

Core Tip: Over the past two decades, liver transplantation has undergone remarkable evolution. Advances in surgical technique and perioperative care have expanded eligibility to increasingly complex recipients—older patients and those with extrahepatic comorbidity. In this context, traditional models of risk assessment, centered almost exclusively on the severity of liver disease, are no longer sufficient. There is a growing need for prognostic tools that also capture the burden of systemic disease and its impact on postoperative outcomes.

INTRODUCTION

In recent years, increasing life expectancy and the rising prevalence of metabolic-related cirrhosis[1] have significantly reshaped the profile of liver transplant candidates[2,3]. We more frequently encounter candidates of advanced age, nowadays[4,5]. They usually present a substantial burden of comorbidities together with varying degrees of frailty[6-8]. While these conditions frequently coexist, they represent distinct entities: (1) Chronological age reflects time-dependent biological consumption[9,10]; (2) Comorbidity defines the simultaneous/cumulative burden of coexisting diseases[11,12]; and (3) Frailty captures a state of reduced physiological reserve in reacting to stressors[13,14]. This evolving scenario has prompted a reassessment of transplant eligibility criteria[15,16], with growing emphasis on prognostic scoring systems aimed at improving prediction of perioperative risk as well as early and long-term outcomes, related to nonhepatic conditions[17,18]. One of the first investigations in this field was a retrospective study of 624 deceased-donor liver transplantation (LT) recipients by Volk et al[19]; who evaluated the Charlson Comorbidity Index (CCI)[20-22], as a predictor of post-transplant survival. They found that 40% of patients had at least one comorbidity, and that CCI was an independent predictor of survival. Coronary artery disease, diabetes, chronic obstructive pulmonary disease, connective tissue disorders and renal failure, were particularly influential. A recalibrated version, the CCI orthotopic LT (CCI-OLT), which incorporated additional data on the etiology of liver disease, further improved predictive performance, highlighting its potential utility to assess post-transplant survival. While the authors acknowledged that this score could benefit from the inclusion of qualitative variables (such as graft characteristics or type of clinical care), on the other hand, all the parameters that a clinician can evaluate at the time of transplant eligibility assessment, are included. Raszeja-Wyszomirska et al[23] developed a modified Child-Pugh-Turcotte (CPT) score[24,25] incorporating creatinine (Crea) levels to predict early mortality after LT. This index demonstrated a good predictive capacity for early mortality after orthotopic LT (OLT), with an area under the curve (AUC) 0.748 (P value = 0.011) in the receiver operating characteristic analysis. This study published in 2009, showed the best predictive value of CPT + Crea, compared to 0.659 for model for end-stage liver disease (MELD)[26-28] and 0.689 for MELD-to-serum sodium ratio[29,30]. However, given the small sample size (48 patients), results had some limitations. In the same year, Wasilewicz et al[31] proposed to use the modified CCI-OLT to assess 1-month mortality after transplant. The study population was primarily affected by viral (39%) and alcohol-related (23%) cirrhosis, which were the most common indications for transplantation. The results indicated that CCI-OLT did not appear to be a reliable predictor of early mortality (1-month) after LT. An Italian study further confirmed the relevance of the CCI as a simple but effective tool to assess the burden of comorbidity and its impact on post-transplant survival[32]. Higher preoperative comorbidity was associated with a higher risk of graft loss and mortality at one year, with both patient and graft survival showing significant correlations with the score. It should be noted that the study population was not homogeneous with respect to the type of organ allocated, as it included recipients of both living donor LT (LDLT)[33,34], and deceased donor LT (DDLT)[35,36]. A multicentric study conducted by Alim et al[37] retrospectively applied the CCI in a cohort of LDLT recipients, demonstrating its usefulness as a selection criteria for elderly candidates, in whom comorbidity burden more strongly affects survival. This study examined an older population compared to previous ones, with a mean age of 63 years, making the use of the score more specifically targeted to a multimorbid population. The value of CCI-OLT in assessing the impact of pretransplant comorbidities and stratifying the risk for short and long-term outcomes after OLT was also demonstrated by Niewiński et al[38]. They reported that recipient age and diabetes strongly influenced short-term post-transplant survival, while metabolic and vascular complications were the leading causes of death five years after OLT. In particular, the study population included only patients with viral cirrhosis, excluding those with metabolic-associated liver disease, a group typically characterized by a greater comorbidity burden. More recently, Choi et al[39] examined the independent predictive value of CCI-OLT, age, sex, CTP score, MELD score, and a modified-Frailty Index[40,41], on post-LT survival. They reported that both CCI-OLT and MELD scores effectively predicted survival in older LT recipients (median age 69 years). Main heterogeneities in including both LDLT and DDLT[42,43], regard post-surgical complication such as: (1) Small-for-size syndrome[44,45] in LDLT; and (2) Prolonged ischemia-time[46] in DDLT. The list of studies examined in this manuscript is reported in Table 1.

Table 1 summary of studies evaluating the performance of Charlson Comorbidity index in liver transplantation.

Ref.	Number of patients included	Type of study	Primary outcome	Median age at transplant (year)	LDLT/DDLT recipients	Score	Main findings
Volk et al[19]	624	Retrospective	Long-term mortality (1 year and 5 years)	51	DDLT	CCI; CCI-OLT	CCI-OLT is useful in predicting post-LT long-term survival
Raszeja-Wyszomirska et al[23]	48	Retrospective	Short-term mortality (1 month)	51	DDLT	CPT + Crea; CPT + Na; MELD; MESO	CPT-Crea is a good predictor of 1-month mortality after LT
Wasilewicz et al[31]	197	Retrospective	Short-term mortality (1 month)	50	DDLT	CCI-OLT	CCI-OLT is a valuable predictor of long-term survival after OLT, but its prognostic power is reduced in predicting 1-month survival
Grosso et al[32]	221	Retrospective	Long-term mortality (1 year)	52	DDLT and LDLT	CCI	CCI is significantly associated with post-LT outcomes
Alim et al[37]	951	Retrospective	Long-term mortality (1 year)	63	LDLT	CCI	CCI is useful in predicting survival after 12 months post-LDLT in patients older than 60 years
Niewiński et al[38]	248	Retrospective	Short and long-term mortality (90 days and 5 years)	54	DDLT	CCI-OLT	CCI-OLT is a good prognostic score to assess post-LT survival. Age and diabetes are the main components of CCI able to affect 90-day mortality post-OLT while metabolic and vascular disease are risk factors for 5-year mortality
Choi et al[39]	155	Retrospective	Short and long-term mortality (1 month and 3 months, 1 year, 3 years and 5 years)	69	DDLT and LDLT	CCI-OLT; CPT; MELD; mFI-5	CCI-OLT and MELD are more accurate than CPT and mFI-5 in predicting post-LT survival

LDLT: Living donor liver transplantation; DDLT: Deceased donor liver transplantation; CCI: Charlson Comorbidity Index; OLT: Orthotopic liver transplantation; CPT: Child-Pugh-Turcotte; Crea: Creatinine; Na: Sodium; MELD: Model for end stage liver disease; MESO: Model for end stage liver disease-to-serum sodium ratio; mFI-5: Modified-frailty index-5; LT: Liver transplantation.

FINDINGS FROM RAJAN G et al

The retrospective study “Application of modified Charlson Comorbidity Index for predicting outcomes following adult living donor liver transplantation”, recently published in the World Journal of Hepatology and conducted by Rajan et al[47], evaluates the CCI - and, more importantly, a modified Charlson Comorbidity Index (mCCI) - as predictors of outcomes in adult LDLT recipients. A key methodological contribution is the removal of the standard malignancy weighting of the CCI when applied to hepatocellular carcinoma (HCC). In the transplant setting, HCC does not have the same short-term prognostic disadvantage as in general oncology populations[48] and including it could distort risk stratification. This recalibration enhances the relevance of mCCI as a transplant-specific assessment tool. In a robust cohort of 497 patients, the mCCI emerged as a significant and independent predictor of postoperative morbidity, outperforming MELD in capturing the burden of extrahepatic disease. Patients with high mCCI scores experienced markedly higher rates of prolonged mechanical ventilation, wound complications, cardiac and neurologic events, and postoperative renal replacement therapy, diabetes, chronic kidney disease, and other systemic comorbidities that mCCI effectively accounts for. The authors also explored a composite score that integrates MELD and mCCI, which improved discrimination for 90-day mortality (AUC 0.70) compared with either score alone. Although the composite score did not perform better than mCCI alone as an independent predictor in multivariate analysis, the idea of combining liver and extrahepatic risk into a single score is promising and should be validated in prospective multicenter studies, stratifying the population by age, cirrhosis etiology and type of organ allocated (by living-donor or deceased-donor).

DISCUSSION

The study could be further enhanced by incorporating data on the impact of portal hypertension, as clinically significant portal hypertension can influence both surgical outcomes and post-OLT survival[49,50]. One potential limitation of the study is that it included only recipients of living donor grafts: Smaller graft volumes in this setting increase the risk of small-for-size complications, which can significantly affect post-transplant outcomes. As with any retrospective analysis, limitations exist - including restricted preoperative variables and potential confounders[51], but the data set is large, contemporary, and clinically relevant. Perhaps the most striking result is that the severity of liver disease contributed less to postoperative morbidity than the burden of systemic comorbidity. This observation is particularly pertinent, as metabolic dysfunction-associated steatotic liver disease[52-54] is becoming relevant among LT candidates. It remains unclear whether the performance of this (and other) scoring systems is valid for assessing short- or long-term survival. The perioperative period is largely influenced by surgical and infectious complications[55-57], which are not fully captured by comorbidity indices such as the CCI. These indices primarily reflect baseline chronic conditions and overall patient vulnerability, providing valuable information for long-term risk prediction. However, they do not account for procedure- or anesthesia-specific factors that may critically affect early outcomes. Repeated assessment of the CCI over time after transplantation could further support the monitoring of patient prognosis, including complications related to immunosuppression, such as metabolic or renal problems[58-60]. Although the mCCI shows promise in predicting outcomes, its generalizability to patients with different cirrhosis etiologies, age groups, or comorbidity profiles remains unclear. Differences in reported cut-off values across studies further complicate clinical application, highlighting the need for prospective validation and standardized criteria in diverse patient populations. Future liver transplant prognostic models should aim to integrate multi-dimensional indicators that capture not only comorbidity and liver function, but also nutritional condition and procedure-specific risk factors. Incorporating dynamic assessment over time, both pre- and post-transplant, could improve the ability to monitor evolving risk and guide individualized clinical decision-making.

CONCLUSION

This study underscores the potential of the mCCI to enrich traditional liver-focused scores by reflecting recipient frailty and systemic disease burden. Prospective, multicenter studies are needed to validate these findings, establish standardized cut-off points and enabling effective integration of comorbidity assessment into routine practice. In the future, machine learning-based models could further enhance predictive accuracy by capturing complex interactions among variables[61-63], always maintaining a final clinical interpretation/evaluation.