Pretransplant immunotherapy increases acute rejection yet improves survival outcome of HCC patients with MVI post-liver transplantation

Abstract

Immune checkpoint inhibitor (ICI)-based immunotherapy has emerged as the most promising strategy for hepatocellular carcinoma (HCC) downstaging prior to liver transplantation (LT). However, further evidence is required to assess the feasibility and safety of pretransplant ICI exposure. We retrospective analyzed 159 HCC patients who underwent LT at our institution from June 2019 to December 2023, and 39 recipients (39/159, 24.5%) received pretransplant ICI therapy. The perioperative acute rejection rate and rejection-related mortality rate in the ICI group were 23.1% (9/39) and 12.8% (5/39), respectively, which were significantly higher than those in the non-ICI group, at 5% (6/120, P = 0.002) and 0% (0/120, P = 0.001). There was no significant difference in the 90-day post-transplant overall survival (OS) (P = 0.447) and recurrence-free survival (RFS) (P = 0.723) between these two groups. We found 37.1% (59/159) recipients were found to have microvascular invasion (MVI), no matter whether the HCC tumor is within Milan criteria or not. Notably, though MVI was identified as a risk factor for the LT recipients, pretransplant ICI exposure appeared to be a protective factor for HCC patients with MVI which benefits its overall survival. Besides, the RFS and OS in the ICI exposure recipients with MVI were comparable to the non-ICI exposure recipients without MVI. However, no synergistic anti-tumor effects were observed with pretransplant ICI immunotherapy when combined with locoregional of TACE, HAIC, RFA and systematic of lenvatinib or sorafenib downstaging treatments, nor with post-transplant adjuvant of systematic or FOLFOX chemotherapy. Further comprehensive studies are needed to balance the dual natural effects of immunotherapy by optimizing downstaging protocols and patient selection to reduce acute rejection and improve long-term survival.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00262-024-03853-9.

Introduction

Hepatocellular carcinoma (HCC) is one of the most common cancers worldwide and its mortality rates are increasing globally [1]. HCC often arises from a background of chronic cirrhosis, which may result from hepatitis virus infection, autoimmune liver disease, alcohol-associated liver disease or metabolic dysfunction-related steatotic liver disease. Liver transplantation (LT) is the definitive curative option for liver cirrhosis patients with HCC who are not eligible for resection due to compromised liver function, as LT simultaneously removes the entire tumor burden and replaces the diseased liver with a new healthy one [2, 3]. However, the shortage of liver donors necessitates downstaging interventions for most HCC patients to control tumor progression and reduce tumor burden, thereby minimizing the risk of delisting while awaiting LT.

Currently, numerous HCC downstaging treatment options are available [4, 5]. Studies have demonstrated that the overall survival (OS) and recurrence-free survival (RFS) in advanced HCC liver transplant (LT) recipients who have undergone successful downstaging treatment are comparable to those of patients with HCC tumors within the Milan criteria [6–8]. However, HCC recurrence remains a significant concern post-transplantation, even when successful downstaging treatment has been implemented prior to LT, necessitating effective downstaging and adjuvant therapies to improve long-term outcomes. Among HCC downstaging treatments, immune checkpoint inhibitor (ICI)-based immunotherapy has emerged as one of the most powerful and promising strategies, by activating the anti-tumor potential of the immune system and reactivating exhausted effector immune cells to target and eliminate tumor cells [5]. Nonetheless, a sufficiently suppressed immune system is required to prevent allograft rejection post-transplantation. Pretransplant ICI exposure may increase the risk of acute rejection (AR) or even mortality post-transplantation, thereby limiting its widespread application in patients with advanced HCC [9]. Although some studies have reported that ICI administration indeed increases the AR rate, it has not been associated with increased graft loss [4, 10].

Although strict LT criteria have been applied, many HCC recipients still experience tumor recurrence or metastasis post-transplantation [11, 12]. This is mainly attributed to micrometastatic tumor cells that were present before LT. Considering the immune remodeling effect and the potent anti-tumor activity induced by ICI prior to LT, whether these remodeled immune cells retain a memory capable of targeting residual tumor cells post-transplantation remains largely unexplored. Moreover, effector T cells activated by successful immunotherapy will undergo a range of differentiation states, with a subset differentiating into memory T cells. These memory T cells exhibit long-lasting and highly proliferative characteristics, which are considered crucial determinants in cancer immunotherapy [13, 14]. Furthermore, the potential benefits of combining pretransplant immunotherapy with post-transplant adjuvant regimens, such as targeted multi-kinase inhibitors or FOLFOX (including fluorouracil, leucovorin and oxaliplatin) chemotherapy, on the survival prognosis of LT recipients are an area that warrants further investigation and discussion.

Study design and methods

Patients and data collection

All HCC patients aged 18–75 years who underwent LT at Sun Yat-sen Memorial Hospital from June 2019 to December 2023 were retrospectively assessed. The primary indication for LT in HCC patients is that the tumor falls within the UCSF criteria, regardless of whether recipients have undergone downstaging treatments. For HCC patients who exceed the UCSF criteria, downstaging can be achieved through locoregional therapies (such as TACE, HAIC and RFA), systemic therapies (including lenvatinib and sorafenib), ICI immunotherapy or a combination of these approaches. A multi-disciplinary clinical committee will assess the eligibility of HCC patients who respond well to downstaging treatment during the waiting period, even if they have not yet met the UCSF criteria. LT recipients who received one or more cycles of ICI therapy for tumor downstaging prior to LT are regarded as pretransplant ICI exposure. The exclusion criteria were as follows: (1) exceeded one organ transplantation, (2) histologically proven cholangiocarcinoma or combined hepatocellular cholangiocarcinoma, and (3) primary tumor burden exceeding UCSF criteria or portal vein tumor thrombi that are resistant to pretransplant downstaging treatments. Patient general information, test results, tumor characteristics, perioperative management and their survival outcomes were analyzed. Data were extracted from the hospital's electronic medical record system.

Perioperative management

All LT recipients, including the ICI exposure and non-ICI exposure recipients, both received routine immune suppression regimens including basiliximab and steroids immune induction, as well as the daily maintenance dose alone or in combination with calcineurin inhibitors, mammalian target of rapamycin inhibitor or mycophenolate mofetil. During the post-LT recovery period, AR is defined if alanine aminotransferase (ALT) or aminotransferase (AST) abnormally elevated to 1.5 times the normal upper limit range and continues to rise for 48 h, while excluded from the vascular or bile duct complication, drug-related liver injury and hepatitis virus relapse. Liver biopsy is not mandatory for AR diagnosis, while if a biopsy is performed, the rejection activity index is defined according to the Banff score [15]. Enhanced immunosuppression, including intensified oral regimens and intravenous administration of corticosteroids, basiliximab and ATG, alone or in combination, was provided to recipients upon diagnosis of AR. The specific regimens were personalized based on the recipient’s clinical condition. The reverse of AR is regarded as the ALT and AST return to their baseline within four weeks.

Post-transplant adjuvant administration protocol

The administration of adjuvant regimens, including the targeted multi-kinase inhibitor or FOLFOX chemotherapy, was reviewed and approved by the multi-disciplinary clinical experts’ assessment one month post-LT to reduce tumor recurrence risk based on the recipient's tumor characteristics, as well as the tolerability and willingness to receive the adjuvant regimens due to its perceived side effects. Generally, the multi-kinase inhibitor was administrated as sorafenib (400 mg twice a day) or lenvatinib (8 mg every day) for 6 months. The FOLFOX chemotherapy was monthly administrated as 6 cycles of fluorouracil (400 mg/m² as a 46-h continuous infusion), leucovorin (400 mg/m² on day 1) and oxaliplatin (100 mg/m² on day 1).

Follow-up and outcome

Regular follow-up was performed at the outpatient clinic post-LT. Recipients with a minimum of 3-month follow-up post-transplant were included for 90-day post-transplant tumor recurrence and mortality analysis. The last follow-up was on June 1st, 2024. The short-term primary endpoints were the AR rate and AR-related death, and the long-term primary endpoints were the 90-day post-transplant overall survival (OS) and recurrence-free survival (RFS).

Organ procurement and ethics statement

No organs from executed prisoners were used in the present study, and all organs were procured with informed consent. The study was performed following the Declaration of Helsinki and current ethical guidelines. This study was approved by the Ethics Committee of the Sun Yat-sen Memorial Hospital, Sun Yat-sen University (SYSKY-2022-145-01). The work has been registered in the Chinese Clinical Trial Registry Center (MR-44-23-017535).

Statistical analysis

For statistical analysis, continuous variables of normal distribution were expressed as mean ± standard deviation (Mean ± SD) and the comparison between groups was made by the t-test, while data of non-normally distributed were expressed as medians and ranges or quartiles, and the Mann–Whitney test was used. Classified variables were expressed as frequency and percentages, and comparisons between groups were made using Pearson’s Chi-squared test or Fisher's exact test. Kaplan–Meier curves were used for survival analysis, and P values were assessed using the log-rank test. Factors associated with acute rejection and overall survival outcomes were assessed using univariate and multivariate logistic regression or Cox proportional hazards analysis, with a 95% confidence interval applied to the estimates. Data analysis was performed using SPSS software (version 25.0), and P < 0.05 was considered statistically significant.

Result

Clinical characteristics of the LT recipients

A total of 159 patients with hepatocellular carcinoma who underwent LT in our institute from June 2019 to December 2023 were enrolled in this study. The general clinical characteristics in terms of liver cirrhosis and tumoral characteristics, as well as the perioperative characteristics, are summarized in Table 1. The cohort's overall median age was 55 (31–74) years, with 93.1% (148/159) being male and 6.9% (11/159) female, and the BMI was 22.9 (15.2–37.5). Among the 159 recipients, liver cirrhosis was mainly caused by chronic hepatitis B virus infection, with a median MELD score of 11 which ranges from 1 to 31, while 74.2% (118/159), the majority of recipients, present a score of less than 14. 45.9% (73/159) and 56% (89/159) of the LT recipients were within the Milan and UCSF criteria, respectively. 64.8% (103/159) of the recipients presented did not exceed 3 tumor nodules, and the AFP level of 84.3% (134/159) recipients was less than 400 ng/L. Moderately differentiated phenotype is the dominant (102/159, 64.2%) tumor-grade characteristic via post-transplant pathology examination. Although performing LT for HCC patients with portal vein tumor thrombi or tumor thrombi resistant to downstaging treatments pretransplant has been avoided at our center, it is impossible to identify recipients with MVI pretransplant. Fifty-nine recipients (59/159, 37.1%) were found to have MVI invading the surrounding microvascular adjacent to the primary tumor lesions, regardless of whether the HCC tumor was within Milan criteria or not (Table 1).

Table 1.

Clinical characteristics of HCC LT recipients

	Value
General characteristics
Age, year, median (range)	55 (31–74)
Male, n (%)/Female, n (%)	148(93.1%)/11(6.9%)
BMI, kg/m2, median (range)	22.9 (15.2–37.5)
ABO blood type, n (%)
A	39(24.5%)
B	40(25.2%)
AB	13(8.2%)
O	67(42.1%)
Liver cirrhosis characteristics
Etiology, n (%)
Hepatitis B	133 (83.6%)
Hepatitis C	5 (3.1%)
Alcoholic	12 (7.5%)
Other	9 (5.7%)
Child–Pugh grade, n (%)
A	70 (44.0%)
B	54 (34.0%)
C	35 (22.0%)
MELD score, median (range)	11 (1–31)
≤ 14, n (%)	118 (74.2%)
> 14, n (%)	41 (25.8%)
Tumor characteristics
Tumor grade
Within Milan, n (%)	73 (45.9%)
Within UCSF, n (%)	89 (56.0%)
Tumor nodules, n (%)
1	50 (31.5%)
2–3	53 (33.3%)
> 3	56 (35.2%)
AFP level (ng/L), median (range)	8.9 (0.6–25,893)
≤ 400, n (%)	134 (84.3%)
> 400, n (%)	25 (15.7%)
Tumor pathology characteristics
Poorly differentiated, n (%)	26 (16.4%)
Moderately differentiated, n (%)	102 (64.2%)
Well differentiated, n (%)	31 (19.5%)
MVI, n (%)	59 (37.1%)
Pretransplant downstaging treatment, n (%)	104 (62.6%)
Locoregional therapy	94 (59.1%)
TACE	69 (43.4%)
HAIC	32 (20.1%)
RFA	16 (10.1%)
Systematic therapy	43 (26.5%)
Lenvatinib	40 (25.2%)
Sorafenib	3 (1.8%)
ICI Immunotherapy	39 (24.5%)
In combination	68 (42.8%)
Adjuvant regimens, n (%)	39 (24.5%)
Lenvatinib	31 (19.5%)
FOLFOX chemotherapy	26 (16.4%)
Lenvatinib + FOLFOX chemotherapy	18 (11.3%)
Perioperative characteristics
Surgery data
LT technique, n (%)
Modified piggyback	76 (47.8%)
Conventional	83 (52.2%)
CIT, minutes, range	312 (120–612)
AHP, minutes, range	54 (32–130)
TST, minutes, range	432 (258–918)
Post-transplant complications, n (%)	89 (53.6%)
Early allograft dysfunction	55 (34.6%)
Bile duct	23 (14.5%)
Acute rejection	19 (11.9%)
Vascular	8 (5.0%)
Death, n (%)	8 (4.8%)
Acute rejection	5 (3.0%)
Infection	1 (0.6%)
Early allograft dysfunction	1 (0.6%)
Intracerebral hemorrhage	1 (0.6%)
Survival characteristics
OS, Day, IQR	721 (410–1013)
RFS, Day, IQR	677 (275–958)

AHP, anhepatia period; BMI, body mass index; CIT, cold ischemia time; HCC, hepatocellular carcinoma; HAIC, hepatic artery infusion chemotherapy; LT, liver transplantation; MELD, model for end-stage liver disease; MVI, microvascular invasion; OS, overall survival; RFS, recurrence-free survival; RFA, radiofrequency ablation; TACE, transarterial chemoembolization; TST, total surgery time

Within the LT recipients, 104 (104/159, 62.6%) received pretransplant downstaging treatment, 94 (94/159, 59.1%) received locoregional therapy of TACE, HAIC and RFA, 43 (43/159, 26.5%) received systematic therapy of lenvatinib and sorafenib, and 39 (39/159, 24.5%) received pretransplant ICI Immunotherapy, respectively. 68 recipients (68/159, 42.8%) of which received more than one therapy mentioned above. Thirty-nine recipients (39/159, 24.5%) received post-transplant adjuvant therapies including 19.5% (31/159) of lenvatinib, 16.4% (26/159) of FOLFOX chemotherapy and 11.3% (18/159) of its combination, respectively. It is necessary to clarify that each HCC LT recipient may have undergone multiple pretransplant downstaging treatments as well as various post-transplant adjuvant therapies.

LT surgery of modified piggyback and conventional techniques performed within the recipients were similar (47.8% vs 52.2%, P > 0.05). The median time for cold ischemia time (CIT) is 312 (120–612) minutes, anhepatia period (AHP) time is 54 (32–130) minutes, and total surgery time (TST) is 432 (258–918) minutes. Early allograft dysfunction (EAD) is the most common complication (55/159, 34.6%) related to the transplanted liver; however, most EAD will be restored by adequate treatment interventions. Within the 55 EAD recipients, only 1 recipient progressed to liver failure and eventually died. The remaining post-transplant liver complications include 23 bile duct (23/159, 14.5%), 19 AR (19/159, 11.9%) and 8 vascular (8/159, 5%) complications. Five recipients died subsequently to AR, 1 recipient died of severe microbial infection, 1 died because of postoperative intracerebral hemorrhage and 1 died due to EAD-related liver failure, respectively. Surprisingly, out of the 8 mortalities, 7 were observed in the ICI group, while only 1 was noted in the non-ICI group.

Pretransplant immunotherapy increased perioperative acute rejection and mortality rate

To evaluate the impact that pretransplant ICI exposure may have on recipients post-LT, we investigated the related factors of LT recipients with or without pretransplant ICI exposure. Overall, 39 recipients received ICI regimens (ICI group) while 120 recipients had no ICI exposure (non-ICI group) pretransplant. Their general clinical characteristics, laboratory parameters and perioperative outcomes are summarized in Table 2. The ICI group had a younger recipient population with a similar BMI index (P = 0.059). Notably, the tumor characteristics associated with the Milan and UCSF criteria in the ICI group were both less favorable than those in the non-ICI group (P < 0.01), indicating a higher tumor burden in patients receiving LT within the ICI group, while the AFP level has no difference (P = 0.505) within these two groups. The post-transplant pathology examination shows that the ICI group had a lower rate of well differentiated tumor type, further indicating an intrinsic aggressive tumor feature in the ICI group. Besides, the MVI detection rate (P = 0.035) was increased in the ICI group, which also further confirmed its aggressive phenotype in the ICI group.

Table 2.

Pretransplant ICI exposure increase perioperative acute rejection and mortality rate

Variable	ICI group (n = 39)	Non-ICI group (n = 120)	P value
General characteristics
Age (y), median (range)	51 (36–71)	55 (31–74)	0.004
Male/Female, n (%)	36/3 (92.3%/7.7%)	112/8 (93.3%/6.7%)	0.826
BMI (kg/m2), median (range)	22.5 (16.9–26.3)	23.0 (15.2–37.5)	0.059
Tumor characteristics
Tumor Within Milan, n (%)	9 (23.1%)	64 (53.3%)	0.001
Tumor Within UCSF, n (%)	11 (28.2%)	78 (65.0%)	< 0.001
AFP level, median (range)	9.9 (1.2–10,429.0)	8.9 (0.6–24,475.0)	0.505
Tumor pathology characteristics
Pathology grade, n (%)
Poorly differentiated	10 (25.6%)	16 (13.3%)	0.071
Moderately differentiated	26 (66.7%)	76 (63.3%)	0.706
Well differentiated	3 (7.7%)	28 (23.3%)	0.032
MVI, n (%)	20 (51.3%)	39 (32.5%)	0.035
Acute rejection characteristics
AR rate, n (%)	23.1% (9/39)	5% (6/120)	0.002
AR reverse rate, n (%)	4/9 (44.4%)	6/6 (100%)	0.044
AR occurrence post-LT, median (range)	10 (7–16)	11 (7–26)	0.463
AR duration, median (range)	23 (11–55)	19 (4–40)	0.272
Extra AR reverse regimens, median (range)
Basiliximab doses	1 (0–2)	0 (0–2)	0.012
ATG doses	1 (0–10)	0 (0–9)	0.388
Corticosteroid amount (mg)	2340 (40–5292)	1682 (88–3040)	0.315
Peak laboratory parameters, median (range)
ALT (U/L)	816 (165–9360)	569 (70–8549)	0.128
AST (U/L)	1148 (223–10,975)	982 (142–13,550)	0.367
GGT (U/L)	287 (64–953)	275 (44–1291)	0.590
TBIL (mmol/L)	86.3 (17.5–520.4)	80.8 (27.2–491)	0.339
Outcome
Perioperative rejection-related death, n (%)	5 (12.8%)	0	0.001
ICU stay days, median (range)	7 (3–63)	6 (2–47)	0.167
Postoperative stay days, median (range)	21 (6–68)	19 (5–86)	0.079
Total length of hospital stay, median (range)	24 (7–81)	21 (12–87)	0.022

ALT, alanine aminotransferase; AR, acute rejection; AST, aspartate aminotransferase; ATG, anti-thymocyte globulin; BMI, body mass index; GGT, γ-glutamyl transferase; ICI, immune checkpoint inhibitor; LT, liver transplantation; MVI, microvascular invasion; TBIL, total bilirubin

The median interval period of ICI exposure prior to LT is 50 (3–840) days, and a total of 174 ICI doses were received with a median of 4 (1–24) cycles of ICI treatments in the 39 ICI exposure recipients (Supplementary Table 1). The most commonly used ICI is tislelizumab (11/39), followed by sintilimab (10/39) and camrelizumab (7/39), while the least commonly used ICI is atezolizumab (1/39). There are 9 and 6 recipients in the ICI group (9/39) and non-ICI group (6/120) who experienced perioperative AR (Table 2), respectively. Not surprisingly, the AR rate was significantly increased in the ICI group (P = 0.002), while the timing of AR diagnosis, the AR duration and the peak laboratory parameters of ALT, AST, GGT and TBIL were similar (P > 0.05) between the two groups. The timing of AR occurrence and its duration were 10 (7–16) days and 23 (11–55) days post-LT in the ICI group, while they were 11 (7–26) days and 19 (4–40) days in the non-ICI group, respectively. The AR is diagnosed by liver biopsy or clinical signs and hematological tests, and increased immunosuppression regimens of calcineurin inhibitors, basiliximab, ATG and steroids were administrated alone or in combination when an AR is diagnosed, and the managements of AR in the ICI and non-ICI group are summarized in Supplementary Table 2.

The general characteristics and survival outcomes of the nine AR recipients in the ICI group are summarized in Supplementary Table 3, with five of these patients having biopsy-proven AR. As expected, the ICI group required more intensive immunosuppressive regimens to reverse AR (Table 2). Multivariate logistic regression analysis, as shown in Supplementary Table 7 (The supplementary tables are not displayed in serial numbers when it first appeared in the text, this supplementary table 7 should be renumbered as supplementary table 4. Same issues are presented in the following supplementary table numbers.), indicates that pretransplant ICI immunotherapy is an independent risk factor for post-transplant AR (odds ratio = 8.742, 95% confidence interval 1.815–42.103; P = 0.007). Additionally, an ICI interval of fewer than 30 days before LT is a significant risk factor for post-transplant AR (odds ratio = 7.751, 95% confidence interval 1.322–45.426; P = 0.023). Furthermore, all AR instances in the non-ICI group were reversed (6/6, 100%), whereas less than half of the AR cases (4/9, 44.4%) were reversed in the ICI group (P = 0.044, Table 2). Additionally, AR-related mortality was significantly higher in the ICI group compared to the non-ICI group (P = 0.001). Although no statistical difference was found in ICU admissions (P = 0.167) or postoperative duration (P = 0.079), the ICI group experienced a prolonged hospital stay compared to the non-ICI group (P = 0.022, Table 2), suggesting that patients with pretransplant ICI exposure require more intensive medical interventions.

Pretransplant immunotherapy improves survival outcomes of HCC patients with microvascular invasion

Resection of the primary lesions may restore the recognition ability of the immune system to eradicate the residual metastasized extrahepatic tumor cells since the unique hepatic immunotolerant microenvironment is responsible for the poor liver cancer immunotherapy response [16]. Besides, the enhanced and persistent immune activity is evidenced by the increased AR rate post-LT (Table 2). Therefore, we assumed that the pretransplant ICI immunotherapy may benefit the long-term survival outcome of the HCC recipient post-LT. In this hypothesis, the perioperative mortality (< 90d post-transplant) which included 7 deaths occurring in the ICI group and 1 death occurring in the non-ICI group was excluded (Tables 1 and 2). Thirty-two ICI exposure and 119 non-ICI exposure recipients were included for further long-term survival analysis. We found the 90-day post-transplant RFS (P = 0.723) and OS (P = 0.447) were similar in the ICI and non-ICI groups, and no survival difference (P > 0.05) was found between these two groups related to the tumor stage within or beyond Milan and UCSF criteria (Table 3).

Table 3.

Survival comparison of LT recipients with or without ICI exposure

Primary endpoint	Variable	ICI (n = 32)	Non-ICI (n = 119)	P value
RFS (days)	Overall RFS	734 (396–938)	677 (361–1020)	0.723
	Tumor characteristics
	Within Milan	842 (758–967)	845 (435–1205)	0.472
	Beyond Milan	673 (244–888)	571 (263–881)	0.743
	Within UCSF	795 (519–939)	724 (400–1135)	0.817
	Beyond UCSF	724 (239–913)	623 (330–949)	0.399
	MVI	773 (397–943)	443 (208–719)	0.084
	Pretransplant neoadjuvant
	+ Systematic therapy	624 (259–784)	566 (251–871)	0.783
	+ Locoregional therapy	972 (840–1053)	671 (408–1031)	0.492
	+ Systematic and locoregional therapy	619 (244–773)	543 (251–774)	0.826
	Post-transplant adjuvant
	+ Systematic therapy	773 (396–938)	467 (168–716)	0.119
	+ FOLFOX chemotherapy	794 (152–1001)	457 (145–816)	0.511
	+ Systematic and FOLFOX chemotherapy	815 (396–949)	396 (198–611)	0.219
OS (days)	Overall OS	773 (624–938)	719 (408–1040)	0.447
	Tumor characteristics
	Within Milan	842 (758–967)	845 (480–1205)	0.539
	Beyond Milan	749 (614–895)	657 (329–955)	0.199
	Within UCSF	795 (668–939)	724 (408–1135)	0.459
	Beyond UCSF	769 (619–916)	698 (398–971)	0.153
	MVI	815 (649–943)	640 (292–944)	0.019
	Pretransplant neoadjuvant
	+ Systematic therapy	721 (581–805)	622 (325–871)	0.452
	+ Locoregional therapy	972 (840–1053)	771 (525–1142)	0.621
	+ Systematic and locoregional therapy	718 (549–795)	590 (325–774)	0.414
	Post-transplant adjuvant
	+ Systematic therapy	776 (577–938)	716 (343–1035)	0.149
	+ FOLFOX chemotherapy	795 (614–1001)	703 (283–1179)	0.199
	+ Systematic and FOLFOX chemotherapy	815 (773–949)	730 (339–1179)	0.216

LT, liver transplantation; ICI, immune checkpoint inhibitor; RFS, recurrence-free survival; OS, overall survival; MVI, microvascular invasion; FOLFOX chemotherapy: folinic acid, fluorouracil and oxaliplatin

After excluding perioperative mortality, 14 recipients in ICI group and 39 recipients in non-ICI group, a total of 53 recipients with MVI were included for further analysis. The clinical characteristics of LT recipients with and without MVI are summarized in Supplementary Table 4. Statistical analysis revealed that the MVI group exhibited a higher tumor burden and poorer survival outcomes. Furthermore, MVI status was identified as a significant risk factor for both OS and RFS among the LT recipients studied (Fig. 1A, B). Notably, although there was no significant difference in RFS (P = 0.084), the OS for pretransplant ICI exposure recipients with MVI was better than that of the non-ICI group (P = 0.018, Fig. 1C, D). The pretransplant tumor factors, liver function and post-transplant tumor recurrence sites, along with the treatments administered for recurrence in MVI-positive recipients, are summarized in Supplementary Table 5. The results indicate that the general conditions were comparable, while recipients in the non-ICI group exhibited poorer survival outcomes than those in the ICI group (Fig. 1D and Supplementary Table 5). Though MVI status is a risk factor for both OS and RFS, the survival curves show that ICI exposure recipients with MVI had comparable RFS and OS to non-ICI group recipients without MVI (P > 0.05, Fig. 1E, F). However, due to the limited number of MVI recipients in this cohort (Fig. 1D), specifically 14 recipients in the ICI group and 39 recipients in the non-ICI group, the multivariate Cox regression analysis did not demonstrate that ICI pretransplant exposure benefited the OS of MVI recipients (Supplementary Table 6). Since lots of systematic, locoregional and immunotherapy neoadjuvant therapies as well as post-transplant adjuvant therapies with systematic therapy and FOLFOX chemotherapy are applied in HCC recipients, we also investigated the potential synergistic anti-tumor effect of the pretransplant ICI immunotherapy with the neoadjuvant therapies and post-transplant adjuvant therapies. However, no statistical clinical benefit was found in this scenario (Table 3).

Fig. 1.

Survival outcomes of HCC LT recipients with MVI in the current study. (A, B). RFS and OS curves of HCC LT recipients with or without MVI. (C, D). RFS and OS curves of HCC LT recipients containing MVI with or without pretransplant ICI exposure. (E, F). Comparison of the RFS and OS curves between pretransplant ICI exposure HCC LT recipients with MVI and those non-ICI exposure recipients without MVI. Abbreviation: MVI, microvascular invasion; RFS, recurrence-free survival; OS, overall survival; ICI, immune checkpoint inhibitor

In summary, pretransplant downstaging with ICI therapy indeed improves the survival outcome of HCC patients with MVI, while no synergistic anti-tumor effect was found when combined with the commonly used neoadjuvant therapies or post-transplant adjuvant therapies.

Discussion

LT represents the ideal curative option for achieving complete tumor eradication and restoring liver function. However, due to the critical shortage of liver donors, the majority of HCC patients will lose transplant opportunities if strict criteria are enforced, as most of them are diagnosed at an advanced stage [11, 17]. Numerous efforts have been made to modify the criteria to expand the eligibility LT pool to meet the increasing demands of HCC patients [8, 18]. Investigations have focused on optimizing HCC therapeutic strategies, especially in cases where transplantation is considered. Recent advancements have significantly progressed in the downstaging interventions for advanced HCC, including improved locoregional treatments and increased systemic treatment options [19, 20], especially the emerging application of ICI-based immunotherapy [5, 8, 18, 21]. Notably, studies have documented that downstaging with anti-PD1 immunotherapy plus lenvatinib exhibited promising efficacy and tolerable mortality in LT recipients [22]. The downstaging strategies aim to reduce tumor burden, increase tumor necrosis and minimize the risk of post-transplant recurrence, which transform initially ineligible HCC patients into suitable candidates for LT and improve their prognosis [8, 21, 23].

Currently, atezolizumab plus bevacizumab and tremelimumab plus durvalumab immunotherapy regimens have been approved as the first-line treatment option for advanced HCC [24, 25]. Lots of HCC patients benefited from the immunotherapy-based downstaging treatment prior to LT [8]. Despite the potential benefits of pretransplant immunotherapy, concerns have emerged regarding its influence on post-transplant outcomes, especially the increased AR and mortality rate. However, investigations have shown conflicting results, with some studies reporting an increased risk of rejection after pretransplant immunotherapy [26–28], while others suggest it is not related to graft loss between the immunotherapy-treated and untreated patients [4, 29]. Nevertheless, the conclusions drawn from these studies were derived from retrospective analyses, whereas the heterogeneity of the HCC patient populations, variations in the interval period between the last ICI therapy and LT surgery, as well as differences in ICI dosage and treatment cycles, introduce considerable variability. We found that pretransplant ICI exposure significantly increased both the AR rate and mortality rate. Multivariate analysis was conducted to assess the effects of pretransplant ICI exposure while controlling for other confounding variables, and it was determined that pretransplant ICI exposure is an independent risk factor for AR. Furthermore, extending the ICI exposure interval to 30 days prior to LT will significantly reduce the post-transplant AR rate. Although pretransplant ICI exposure significantly heightened the perioperative AR rate and mortality rate in this small-scale retrospective investigation, the long-term safety and efficacy of pretransplant ICI therapy remain subject to ongoing debate. Thus, prospective randomized controlled trials are necessary to elucidate these issues. The ongoing pending results of trials (NCT05411926, NCT05475613, NCT05879328 and NCT04425226) may provide additional insights into the potential benefits and risks of pretransplant ICI therapy.

HCC with MVI presents a challenging scenario post-LT due to the increased risk of tumor recurrence and metastasis [30]. Vascular invasion, particularly involving major vessels such as the portal vein or hepatic vein, is associated with a higher likelihood of tumor spread and reduced survival post-transplantation [31]. Indeed, we found that 59 recipients exhibited MVI upon post-transplant pathological examination, no matter whether the HCC tumor is within Milan criteria or not (Table 1). Though MVI status significantly increased the recurrence and mortality rate (P = 0.0015 and 0.0025, Fig. 1A, B), the pretransplant ICI exposure appears to be a protective factor for HCC patients with MVI (Fig. 1 and Table 3). This protective effect may be due to the potent anti-tumor activity induced by pretransplant ICI treatment. Thus, the entire diseased liver, which contains the primary tumor responsible for the tumor-immunosuppressed microenvironment, has been removed, and the immune-remodeled tumor-targeting memory immune cells may exhibit sustained killing activity against the micrometastasized tumor cells post-transplantation.

We conclude that pretransplant immunotherapy presents a double-edged sword for LT recipients. While it increases the AR rate, it enhances the prognosis for HCC patients with MVI. However, no previous studies have reported these similar results, this may be due to the relatively recent adoption of pretransplant immunotherapy in LT patients. Furthermore, the conclusion that MVI HCC recipients with pretransplant ICI exposure had better survival outcomes compared to non-ICI exposure recipients is based on a small retrospective study with a short observation period. Specifically, the study included only 14 recipients in the ICI group and 39 recipients in the non-ICI group with MVI. While univariate statistical analysis indicated potential survival benefits, multivariate Cox regression analysis did not demonstrate a significant benefit of pretransplant ICI exposure on the overall survival of MVI recipients. Further verification through long-term survival follow-up and large-scale prospective studies is needed to address the potential selection bias and confounding variables. Moreover, specific mechanisms underlying these findings, as well as novel animal models, are needed to fully elucidate this encouraging result in deep sequencing cosmic studies.

HCC recurrence is a major concern post-LT, as it significantly reduces long-term survival for HCC recipients. Tumor number and size, tumor grade according to Milan and UCSF criteria, and pretransplant AFP levels are crucial factors contributing to HCC recurrence. In this study, we observed a comparable RFS of 734 (396–938) days in the ICI group and 677 (361–1020) days in the non-ICI group post-LT (Table 3). The RFS was consistent with reported studies indicating a high peak in recurrence approximately 2 years post-LT, which supports our findings [7, 36]. While adhering strictly to LT indications with Milan criteria for patient selection can indeed reduce recurrence, only 8% of LT recipients experienced HCC recurrence four years post-LT [2]. However, since most HCC patients are diagnosed at advanced stages, transplantation opportunities may be lost if strict LT eligibility criteria are applied. Notably, recent studies have demonstrated that successful downstaging of HCC beyond Milan criteria to within Milan criteria is associated with a low rate of HCC recurrence and excellent post-transplant survival [6, 8].

Adjuvant therapy following LT for HCC remains an area of active investigation aimed at reducing tumor recurrence and improving patient survival outcomes. Currently, there are no approved targeted therapies or chemotherapy regimens for LT adjuvant therapy in the clinical guidelines. Yet, systemic treatment strategies with sorafenib and lenvatinib have demonstrated efficacy in advanced HCC and are currently being explored in the adjuvant setting post-transplantation in several studies [32–34]. Additionally, FOLFOX chemotherapy has shown promise in certain studies as a post-transplant regimen for HCC patients [35, 36]. For instance, a small sample size RCT study demonstrated that adjuvant FOLFOX chemotherapy was associated with an improved 1-year survival rate for HCC patients beyond Milan criteria [37]. Additionally, for certain HCC LT recipients who exceed the Milan or UCSF criteria and are ineligible for multi-kinase inhibitors, the administration of adjuvant chemotherapy may help reduce the risk of HCC recurrence [36]. Thus, adjuvant therapy strategies, including FOLFOX chemotherapy and systemic treatment approaches, hold promise in reducing tumor recurrence and improving long-term outcomes for HCC LT recipients. However, no additional survival benefits were observed from the adjuvant therapies combined with pretransplant ICI exposure in the current study. This may be due to the small sample size and the intrinsic limitations of a retrospective study. Besides, the enrolled recipients received mixed pretransplant downstaging protocol and post-transplant adjuvant regimens. Therefore, continued research efforts are needed to optimize adjuvant regimens and identify potential biomarkers for patient selection in this context.

In conclusion, LT offers a promising curative option for HCC patients, with ongoing research focusing on optimizing patient selection and improving post-transplant outcomes. Downstaging strategies, including immunotherapy, hold the potential for expanding the pool of eligible candidates for transplantation. However, further studies are needed to elucidate the optimal use of immunotherapy in downstaging protocols and its impact on post-transplant rejection and patient prognosis. Additionally, newly generated combination therapies associated with improved survival outcomes on ICI-based immunotherapy combined with additional regimens for pretransplant downstaging or post-transplant adjuvant therapies warrant further validation.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

AHP

Anhepatia period

ALT

Aminotransferase

AR

Acute rejection

AST

Aminotransferase

ATG

Anti-thymocyte globulin

BMI

Body mass index

CIT

Cold ischemia time

EAD

Early allograft dysfunction

FOLFOX

Fluorouracil, leucovorin and oxaliplatin

GGT

γ-Glutamyl transferase

HAIC

Hepatic artery infusion chemotherapy

HCC

Hepatocellular carcinoma

ICI

Immune checkpoint inhibitor

LT

Liver transplantation

MELD

Model for end-stage liver disease

MVI

Microvascular invasion

OS

Overall survival

RFA

Radiofrequency ablation

RFS

Recurrence-free survival

TACE

Transarterial chemoembolization

TBIL

Total bilirubin

TST

Total surgery time

Author contribution

X.L. and C.L. were involved in conceptualization; X.L., Q.Z., J.C. and Z.Y. assisted with methodology; X.L., Q.Z., G.G., M.S., F.Z., H.L. and L.P. collected the data; X.L., Q.Z., W.W. and L.X. wrote the manuscript; and C.L., L.X. and H.L. were responsible for supervision. All authors have read and agreed to the publication of the manuscript.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request. No datasets were generated or analyzed during the current study.

Declarations

Conflict of interest

The authors of this manuscript have no conflicts of interest to disclosure.