Articles

Liver transplantation plus chemotherapy versus

chemotherapy alone in patients with permanently
unresectable colorectal liver metastases (TransMet): results
from a multicentre, open-label, prospective, randomised
controlled trial
René Adam, Céline Piedvache, Laurence Chiche, Jean Philippe Adam, Ephrem Salamé, Petru Bucur, Daniel Cherqui, Olivier Scatton, Victoire Granger,
Michel Ducreux, Umberto Cillo, François Cauchy, Jean-Yves Mabrut, Chris Verslype, Laurent Coubeau, Jean Hardwigsen, Emmanuel Boleslawski,
Fabrice Muscari, Heithem Jeddou, Denis Pezet, Bruno Heyd, Valerio Lucidi, Karen Geboes, Jan Lerut, Pietro Majno, Lamiae Grimaldi, Francis Levi,
Maïté Lewin, Maximiliano Gelli on behalf of the Collaborative TransMet group*

Summary
Background Despite the increasing efficacy of chemotherapy, permanently unresectable colorectal liver metastases are Lancet 2024; 404; 1107–18
associated with poor long-term survival. We aimed to assess whether liver transplantation plus chemotherapy could See Comment page 1078
improve overall survival. *Collaborative TransMet Group
listed at the end of the Article
Methods TransMet was a multicentre, open-label, prospective, randomised controlled trial done in 20 tertiary centres Department of Hepatobiliary
in Europe. Patients aged 18−65 years, with Eastern Cooperative Oncology Group performance score 0−1, permanently Surgery and Transplantation,
AP-HP Hôpital Paul-Brousse,
unresectable colorectal liver metastases from resected BRAF-non-mutated colorectal cancer responsive to systemic University of Paris-Saclay,
chemotherapy (≥3 months, ≤3 lines), and no extrahepatic disease, were eligible for enrolment. Patients were Villejuif, France
randomised (1:1) to liver transplantation plus chemotherapy or chemotherapy alone, using block randomisation. The (Prof R Adam PhD,
liver transplantation plus chemotherapy group underwent liver transplantation for 2 months or less after the last Prof D Cherqui PhD);
Clinical Research Unit,
chemotherapy cycle. At randomisation, the liver transplantation plus chemotherapy group received a median of AP-HP Hôpital Kremlin Bicêtre,
21∙0 chemotherapy cycles (IQR 18∙0−29∙0) versus 17∙0 cycles (12∙0−24∙0) in the chemotherapy alone group, in up to University of Paris-Saclay,
three lines of chemotherapy. During first-line chemotherapy, 64 (68%) of 94 patients had received doublet Le Kremlin Bicêtre, France
chemotherapy and 30 (32%) of 94 patients had received triplet regimens; 76 (80%) of 94 patients had targeted therapy. (C Piedvache PhD,
Prof L Grimaldi PhD);
Transplanted patients received tailored immunosuppression (methylprednisolone 10 mg/kg intravenously on day 0; Department of Hepatobiliary
tacrolimus 0·1 mg/kg via gastric tube on day 0, 6−10 ng/mL days 1–14; mycophenolate mofetil 10 mg/kg intravenously Surgery and Transplantation,
day 0 to <2 months and switch to everolimus 5−8 ng/mL), and postoperative chemotherapy, and the chemotherapy Hôpital Haut-Lévêque,
Bordeaux, France
group had continued chemotherapy. The primary endpoint was 5-year overall survival analysed in the intention to
(Prof L Chiche PhD,
treat and per-protocol population. Safety events were assessed in the as-treated population. The study is registered JP Adam MD); Department of
with ClinicalTrials.gov (NCT02597348), and accrual is complete. Digestive, Hepatobiliary and
Pancreatic Surgery, Regional
University Hospital,
Findings Between Feb 18, 2016, and July 5, 2021, 94 patients were randomly assigned and included in the intention-to-
Tours, France
treat population, with 47 in the liver transplantation plus chemotherapy group and 47 in the chemotherapy alone (Prof E Salamé PhD,
group. 11 patients in the liver transplantation plus chemotherapy group and nine patients in the chemotherapy alone P Bucur PhD); Department of
group did not receive the assigned treatment; 36 patients and 38 patients in each group, respectively, were included in Hepatobiliary Surgery, AP-HP
Hôpital Pitié-Salpêtrière,
the per-protocol analysis. Patients had a median age of 54∙0 years (IQR 47∙0−59∙0), and 55 (59%) of 94 patients were
Paris, France
male and 39 (41%) were female. Median follow-up was 59∙3 months (IQR 42∙4−60∙2). In the intention-to-treat (Prof O Scatton PhD);
population, 5-year overall survival was 56∙6% (95% CI 43∙2−74∙1) for liver transplantation plus chemotherapy and Department of
12∙6% (5∙2–30∙1) for chemotherapy alone (HR 0·37 [95% CI 0·21−0·65]; p=0·0003) and 73∙3% (95% CI 59∙6–90∙0) Gastroenterology and
Digestive Oncology, University
and 9∙3% (3∙2–26∙8), respectively, for the per-protocol population. Serious adverse events occurred in 32 (80%) of
Hospital Grenoble, Université
40 patients who underwent liver transplantation (from either group), and 69 serious adverse events were observed in Grenoble Alpes Grenoble,
45 (83%) of 54 patients treated with chemotherapy alone. Three patients in the liver transplantation plus chemotherapy France (V Granger MD);
group were retransplanted, one of whom died postoperatively of multi-organ failure. Department of Medical
Oncology, Gustave Roussy,
University of Paris-Saclay,
Interpretation In selected patients with permanently unresectable colorectal liver metastases, liver transplantation Villejuif, France
plus chemotherapy with organ allocation priority significantly improved survival versus chemotherapy alone. These (Prof M Ducreux PhD);
results support the validation of liver transplantation as a new standard option for patients with permanently Hepatobiliary Surgery and
Liver Transplant Unit,
unresectable liver-only metastases. University Hospital Padua,
Padua, Italy (Prof U Cillo PhD);
Funding French National Cancer Institute and Assistance Publique–Hôpitaux de Paris. Hepatobiliary Surgery Unit,
AP-HP Hôpital Beaujon,

www.thelancet.com Vol 404 September 21, 2024 1107

 Articles

Clichy, France (F Cauchy PhD); Copyright © 2024 Elsevier Ltd. All rights reserved, including those for text and data mining, AI training, and similar
Department of Hepatobiliary technologies.
Surgery and Liver
Transplantation, Hôpital de la
Croix-Rousse, Lyon, France Introduction 1980–90s had a 5-year survival of only 18%.6 Because of
(Prof J-Y Mabrut PhD); Complete surgical resection plus systemic chemotherapy organ scarcity and high rates of survival in common
Department of Oncology, represents the best treatment for liver metastases from indications for liver transplantation, colorectal liver
University Hospitals Leuven,
Leuven, Belgium
colorectal cancer to achieve long-term survival metastases were considered a contraindication to liver
(Prof C Verslype PhD); (5-year survival ~40%)1,2 and possible cure. However, less transplantation. At that time, chemotherapy was based
Department of Abdominal than 30% of patients are considered to have initially on fluorouracil (≤20% response rate), and a retrospective
Surgery and Transplantation, resectable metastases and be suitable for up-front analysis by one of the authors (RA) of this study showed
University Hospital Saint Luc,
Brussels, Belgium
surgery,1 and the potential cure from surgery remains that 44% of deaths were unrelated to tumour recurrence.7
(L Coubeau MD); Department of poor.2 In light of this, plus the increasing efficacy of
Digestive, Hepatobiliary and Tumour downstaging is a primary treatment objective chemotherapy, increased expertise of transplantation
Transplantation Surgery, to facilitate secondary resection in patients with initially teams, improved knowledge of metastatic disease, and
Marseille University Hospital
Timone, Marseilles, France
unresectable colorectal liver metastases. Complete improvements in imaging and immunosuppression, we
(Prof J Hardwigsen PhD); resection after response to chemotherapy was found to proposed revisiting liver transplantation for colorectal
Department of Digestive be associated with a 5-year survival of 33% in a large liver metastases in carefully selected patients who
Surgery and Transplantation,
retrospective series.3 Although response rates were high responded to chemotherapy. This idea was trialled in
University Hospital Lille,
Lille, France (54−80%) in phase 3 randomised controlled trials using Norway, which showed promising preliminary results.7
(Prof E Boleslawski PhD); doublet or triplet regimens with targeted therapy,4,5 only These results were supported 3 years later in the first
Department of Digestive 13–50% of patients had secondary curative-intent surgery. proof-of-concept pilot study of 21 consecutive patients
Surgery, Hôpital Rangueil,
Therefore, systemic chemotherapy represents the transplanted for colorectal liver metastases with a
University Hospitals Toulouse,
Toulouse, France standard of care for most patients with permanently 60% 5-year survival rate.8 These findings motivated
(Prof F Muscari PhD); Institute unresectable colorectal liver metastases. further research into outcomes after liver transplantation
of Experimental and Clinical In this setting, liver transplantation initially emerged for colorectal liver metastases via alternative routes to
Research, Catholic University of
as a promising treatment. However, preliminary results cadaveric donation, such as living-donor liver
Louvain, Louvain, Belgium
(Prof J Lerut PhD); Department from 50 patients with colorectal liver metastases who transplantation, and increased the use of liver
of Oncology, UPR underwent liver transplantation in Europe in the transplantation in experimental programmes worldwide.
Chronotherapy, Cancers and
Transplantation, Faculty of
Medicine, University of Paris- Research in context
Saclay, Villejuif, France
(Prof F Levi PhD, Prof R Adam, Evidence before this study current standard of care in patients with permanently
Prof M Lewin PhD); Department Complete resection of liver metastases is the best option for unresectable colorectal cancer and liver metastases. Our
of Hepatobiliary and Digestive long-term survival in patients with liver metastases from findings show that patients with permanently unresectable
Surgery, University Hospital,
colorectal cancer. However, this surgical treatment is only liver metastases from colorectal cancer have better overall
Rennes, France (H Jeddou PhD);
Department of General suitable for a small proportion of patients, and systemic survival after liver transplantation following chemotherapy
Surgery, University Hospital chemotherapy remains the standard of care for patients with than patients receiving chemotherapy alone. This is the
Clermont-Ferrand, Clermont- unresectable liver metastases. Recent advances in liver first comparative study demonstrating a notable benefit of
Ferrand, France
transplantation, including living-donor transplantation and use transplantation in liver metastases from an aggressive digestive
(Prof D Pezet PhD); Department
of Digestive and Oncological of partial grafts, as well as positive findings from pilot, cancer, expanding the concept of transplant oncology. In the
Surgery, Regional University non-controlled studies (NCT01311453 and NCT01479608), absence of evidence from randomised controlled trials, the role
Hospital Besançon, Besançon, have reignited interest in liver transplantation for patients with of liver transplantation in addition to systemic chemotherapy
France (Prof B Heyd PhD);
permanently unresectable colorectal liver metastases. in patients with permanently unresectable liver metastases
Department of Digestive
Surgery and Transplantation, We searched PubMed from database inception to April 17, 2024, from colorectal cancer has not been scientifically shown. Strong
Hôpital Universitaire de with the search terms “colorectal cancer”, “colorectal evidence of clinical benefit is especially important in this
Bruxelles, Université Libre de carcinoma”, “rectal cancer”, “rectal carcinoma”, “colon cancer”, setting, given the demand for, and scarcity of organs as well as
Bruxelles, Belgium
“colon carcinoma”, “liver metastasis”, and “transplant” for the competition with standard indications.
(V Lucidi MD); Department of
Gastroenterology/Digestive randomised trials comparing liver transplantation plus
Implications of all the available evidence
Oncology, University Hospital chemotherapy with chemotherapy alone in patients with
Ghent, Ghent, Belgium The TransMet trial shows that liver transplantation plus
metastatic colorectal cancer. No randomised controlled trials
(Prof K Geboes PhD); chemotherapy considerably improves outcomes, achieving a
comparing systemic chemotherapy plus liver transplantation
Department of Biomedical potential of cure in patients with permanently unresectable
Science, University of Italian versus chemotherapy alone were identified.
colorectal cancer liver metastases compared with chemotherapy
Switzerland, Lugano,
Switzerland (Prof P Majno PhD); Added value of this study alone. These findings support liver transplantation plus
Department of Radiology, To our knowledge, the TransMet trial is the first randomised chemotherapy as a new standard option for carefully selected
Hôpital Paul-Brousse, study to prospectively compare liver transplantation plus patients with permanently unresectable liver metastases from
University of Paris-Saclay,
chemotherapy versus chemotherapy alone as the colorectal cancer.
Villejuif, France (Prof M Lewin);

1108 www.thelancet.com Vol 404 September 21, 2024

 Articles

Despite international recommendations,9 liver trans­ absence of local recurrence on colonoscopy performed Department of Anaesthesia,
plantation plus chemotherapy versus chemotherapy 12 months or less before enrolment (unless primary Surgery and Interventional
Radiology, Gustave Roussy
alone needs to be validated in light of consistent tumour resection was performed within the past 12 Hospital, University of
improvements in systemic treatments. Outcomes of liver months), renal function within normal limits, white blood Paris-Saclay, Villejuif, France
transplantation plus chemotherapy for colorectal liver cell count higher than 2500 cells per mL and platelet count (M Gelli PhD)
metastases should be assessed within validated higher than 80 000 cells per mL, receipt of informed Correspondence to:
indications for liver transplantation, in which a 5-year consent, and expected patient co-operation for treatment Prof René Adam, Department of
Hepatobiliary Surgery and
survival rate of 70−80% can be expected. A recent and follow-up (appendix p 4). Patients were excluded if
Transplantation, AP-HP Hôpital
systematic review concluded that further evidence from they had general contraindication to liver transplantation, Paul Brousse, University of Paris-
ongoing prospective trials is needed to determine had active alcohol or substance misuse, had active Saclay, 94800 Villejuif, France
whether, and to what extent, liver transplantation has a infection or uncontrolled sepsis, had no psychosocial [email protected]

role in liver-only, surgically unresectable, metastatic support from social services or were unable to comply
colorectal cancer.10 The TransMet trial (NCT02597348) with medical treatment, had other malignancies, either
was therefore initiated to assess the potential clinical concomitant or within 5 years before inclusion in
benefit of liver transplantation plus chemotherapy versus TransMet, had not implemented the recommended
chemotherapy alone in patients with permanently guidelines for primary colorectal cancer surgery, had
unresectable colorectal liver metastases. previous or concomitant extrahepatic metastases or local
recurrence, were pregnant, did not provide signed
Methods consent, and had no health insurance. Eligible participants
Study design were selected by the local multidisciplinary tumour board
TransMet was a multicentre, open-label, prospective, at each centre. Eligibility was assessed by an independent
randomised controlled trial that compared curative- multidisciplinary committee of international expert
intent liver transplantation plus chemotherapy versus oncologists, radiologists, and liver surgeons via monthly
chemo­ therapy alone in selected patients with per­ videoconferences in the presence of local investigators.
manently unresectable colorectal liver metastases. This The TransMet trial completed recruitment on July 5, 2021,
trial was conducted in 20 tertiary centres in Europe (14 in and the database was frozen on Jan 18, 2024, once the
France, four in Belgium, and two in Italy). A list of study predefined number of events was reached. Patients
centres, including the principal investigators and provided written informed consent.
number of patients in each, is provided in the appendix See Online for appendix
(pp 2–3). The study was conducted in accordance with Randomisation and masking
the Declaration of Helsinki, Good Clinical Practice Patients were randomly assigned (1:1) to liver
guidelines, and the relevant French and European laws. transplantation plus chemotherapy or chemotherapy
The study was approved by the Ethical Committee of Île alone, using block randomisation with randomly selected
de France VII. The study protocol amendments are block sizes, and stratified by centre or centre cluster for
provided in the appendix (p 5). The results of a feasibility those with an expected low recruitment rate.12 An
study within the TransMet trial have been previously independent statistician prepared the randomisation list
published.11 with NQuery Advisor (version 7.0) using a pseudo-
random numbers generator and randomly assigned
Participants patients using an interactive web-response system.
Eligible patients were adults (aged 18–65 years) with an Patients were enrolled by their treating physician after
Eastern Cooperative Oncology Group (ECOG) perfor­ validation by the expert panel committee. Investigators,
mance status score of 0–1, histologically proven colorectal clinicians, participants, caregivers, and the expert panel
cancer adenocarcinoma, BRAF wild–type colorectal were not masked to treatment allocation.
cancer, permanently unresectable colorectal liver meta­
stases centrally confirmed by an independent valida­tion Procedures
panel, and an objective response (stable disease or partial Radiological evaluation was centrally reviewed by an
response) according to Response Evaluation Criteria in expert radiologist (ML). Definitive technical
Solid Tumours (RECIST) criteria for at least 3 months unresectability of colorectal liver metastases was assessed
during the last chemotherapy line. Additional criteria by at least two expert liver surgeons, two expert
were up to three chemotherapy lines for metastatic oncologists, and one radiologist (ML) on imaging at
disease, carcinoembryonic antigen concentration of less diagnosis, and was confirmed on imaging after
than 80 ng/mL at inclusion or a decrease of 50% or more chemotherapy, integrating the fact that missing
of the highest serum concentration of carcinoembryonic metastases were considered as potentially persistent
antigen observed during disease, no extrahepatic disease tumours for defining unresectability. Primary tumour
on CT scan and PET–CT imaging, high standard resection (ie, modality and timing) and chemotherapy
oncological surgical resection of the primary tumour (ie, before inclusion were done according to local practice at
safe margins of resection, with adequate TNM staging), participating centres. In selected patients with primary

www.thelancet.com Vol 404 September 21, 2024 1109

 Articles

treatment by chemotherapy, the primary tumour was Oncological follow-up was based on thoraco-abdominal
resected if the patient was eligible for enrolment. Disease CT scan and tumour markers performed every 3 months
control by postoperative chemo­ therapy for at least during the first 2 years, and every 6 months thereafter.
2 months was mandatory to definiti­vely validate patient Additionally, in the liver transplantation plus chemo­
eligibility. Previous hepatic resection was not a therapy group, a PET–CT was performed at 6 months,
contraindication to eligibility provided there was 12 months, and 24 months, and every year thereafter up
technical unresectability. to 5 years. This was not performed in the chemotherapy
In the liver transplantation plus chemotherapy group, group because it would not have changed the treatment
local contributing centres assessed patient transplant strategy.
eligibility before randomisation. If extrahepatic pro­ Standard transplantation follow-up included physical
gression was detected during pre-transplantation evalua­ examination, complete blood cell count, and blood
tion, the patient was ineligible and not included. As chemical and liver function tests according to local
colorectal liver metastases were not a validated indication practice in association with the oncological follow-up.
for liver transplantation, the TransMet trial was the only
potential access to liver transplantation in the participating Outcomes
countries. Once eligibility for liver transplantation was The primary endpoint was 5-year overall survival, defined
confirmed, investigators registered patients on organ as the time from random assignment to death from any
donor waiting lists via a specific prioritisation process cause. Patients alive at the time of database freezing were
with national organ-sharing organisations. This process censored at their last assessment. Secondary outcomes
ensured liver transplantation within 2 months from the were 3-year overall survival, 3-year and 5-year progression-
last chemotherapy cycle to reduce the risk of progression free survival, 3-year and 5-year recurrence rate, and
while minimising the risk of postoperative complications. health-related quality of life. Progression-free survival
In cases of progression while on the waiting list, was defined as the time from random assignment to the
chemotherapy was restarted, and the patient was first event (ie, disease progression, defined as recurrence
temporarily contraindicated to liver transplantation until in the liver transplantation plus chemotherapy group and
disease control was achieved. All liver transplantation progression, according to RECIST version 1.1, in the
procedures were preceded by complete abdominal chemotherapy alone group, or death from any cause). A
exploration for occult extrahepatic disease, with frozen post-hoc analysis was performed for secondary
section of any suspicious deposit or pedicle lymph node. progression-free survival to assess the effect of curative
Orthotopic liver transplantations were done using whole surgery or local ablation of recurrence following liver
cadaveric liver grafts and conventional reconstruction transplantation. Secondary progression-free survival was
techniques. After liver transplantation, patients were calculated as the time from random assignment to failure
managed in the intensive care unit for 3−4 days and then of curative-intent treatment of disease progression, by
on hepatology wards for 1−4 weeks. Tailored immuno­ surgery or ablation. In the chemotherapy alone group, a
suppression was recommended for transplant recipients. post-hoc exploratory analysis was done to describe overall
The initial immunosuppressive regimen (day 0) consisted survival in patients who had undergone resection after
of methylprednisolone (10 mg/kg intravenously), random assignment. However, no crossover was allowed
tacrolimus (0·1 mg/kg via gastric tube), and myco­ in the study.
phenolate mofetil (1 g twice a day intravenously). Postoperative complications in the liver transplantation
Thereafter, the recommendation was to maintain the plus chemotherapy group were assessed within 90 days
tacrolimus dose at trough levels of 6−10 ng/mL during of surgery according to Clavien–Dindo grading;13 severe
the first 14 days, to switch mycophenolate mofetil to morbidity was defined as grade 3b or higher compli­
everolimus at trough levels of 5−8 ng/mL within 2 months cations in relation to liver transplantation within 90 days
of liver transplantation, to reduce the tacrolimus dose to of surgery. Adverse events and serious adverse events
3−5 ng/mL after introduction of mammalian target of were investigator-assessed throughout the study by a
rapamycin inhibitors, and to taper the steroid dose during member of the central research unit to each participating
the first 3−6 months. Administration of post-transplan­ centre. The seriousness and causal relationship between
tation systemic chemo­therapy (usually doublet) regimens, serious adverse events and the procedures were centrally
shown to be effective in pre-transplant period, was reviewed during the final analysis. Toxicity related to
recommended in the absence of post-liver transplantation systemic chemotherapy after inclusion was evaluated
complications but not mandatory and at the discretion of according to the Common Terminology Criteria for
the medical oncologist in charge of the patient. Adverse Events version 4.0.
In the chemotherapy group, systemic chemotherapy Health-related quality of life was assessed using the
(mainly doublet) regimens were continued. Chemo­ European Organisation for Research and Treatment of
therapy type, duration, and modality of administration Cancer Core Quality of Life questionnaire (QLQ-C30).
were at the discretion of the team in charge of the patient Global health scores were calculated from the Global
according to tumour response and toxicity. Health Status and Quality of Life scale in the QLQ-C30;

1110 www.thelancet.com Vol 404 September 21, 2024

 Articles

a standardised score (range 0–100) was calculated from per-protocol population according to the type of treatment.
the item responses. Questionnaires were completed at Safety was described in the as-treated population, which
baseline, months 6, 12, 18, and 24, and years 3, 4, and 5. comprised patients who had liver transplantation and
Translational research, as specified in the protocol, and those treated by chemotherapy (without liver trans­
central analysis of pathological response are ongoing and plantation). Descriptive statistics were reported as median
will be published separately. (IQR) for quantitative variables and frequency (%) for
categorical variables. For statistical significance, a two-
Statistical analysis tailed p value of less than 0·05 was used. All statistical
The trial was designed to detect a 40% difference in 5-year analyses were done independently using SAS, version
overall survival rates (from 10% for the chemotherapy 9.4, except for restricted mean survival time, which was
alone group to 50% for liver transplantation plus calculated using the survRM2 package in R, version 4.3.2.
chemotherapy group) with 90% power and a two-sided α Data monitoring was done by the Clinical Research
level of 0·05. In the survival analysis, the number of Unit of Assistance Publique–Hôpitaux de Paris, Paris-
participants needed was derived from the number of Saclay University, which centralised and controlled, via
deaths. Initially, 29 deaths were needed in the combined routine monitoring visits, all data collected by the
groups. However, as some patients in the liver research teams at each participating centre (eg, patient
transplantation plus chemotherapy group did not undergo consent, reliability and completeness of collected data,
liver transplantation, the required number of deaths was and checking for serious adverse events). An independent
re-estimated to 50 deaths at the final analysis to preserve data and safety monitoring board controlled the safety
the intended power of the trial. data. The trial is registered with ClinicalTrials.gov
Overall survival was compared between the two groups (NCT02597348) and accrual is complete.
using the log-rank test and Cox regression analysis to
obtain a hazard ratio (HR) and 95% CI. If the proportional
hazards assumption was not met, a restricted mean
survival time analysis was performed, although this was 157 patients assessed for eligibility
not prespecified in the protocol. The difference in
restricted mean survival time represents the gain or loss
63 ineligible
in event-free survival time with 95% CI in the liver 36 tumour progression
transplantation plus chemotherapy group versus the 18 extrahepatic
16 intrahepatic
chemotherapy alone group, up to a prespecified clinical 2 both
point (60 months in this instance). 13 equivocal unresectability
Analysis of overall survival was done in the intention- 5 ≥3 lines of chemotherapy
9 mixed contraindications
to-treat population and per-protocol population, which
included patients who received allocated treatment
without major protocol deviation (appendix p 6). Patients 94 enrolled
without major protocol deviation were defined, in the
liver transplantation plus chemotherapy group, as those
94 randomised
without liver transplant, those with disease progression
at pre-liver transplantation CT scan or with 3 months or
longer between the last chemotherapy cycle and liver
transplantation (appendix p 7), and in the chemotherapy 47 assigned to liver transplantation 47 assigned to chemotherapy alone
plus chemotherapy
alone group as patients who did not receive chemotherapy
or who had undergone liver transplantation or resection
after random assignment (appendix p 8). For har­ 47 included in intention-to-treat 47 included in intention-to-treat
monisation, we considered recurrence (liver trans­ analysis analysis
plantation plus chemotherapy group) and progression
(chemotherapy alone group) to be equivalent and
11 did not receive assigned 9 did not receive assigned
therefore comparable. treatment treatment
Progression-free survival and quality-of-life analyses 9 tumour progression 9 received liver
1 transplantation on transplantation or resection
were done in the per-protocol population to evaluate the progression
effect of treatments delivered according to the study 1 time between last
chemotherapy and liver
design. The recurrence rate and secondary progression- transplantation ≥3 months
free survival (calculated using Kaplan–Meier estimation)
were presented in the liver transplantation plus chemo­
therapy group. Quality-of-life analyses were descrip­tive. 36 included in per-protocol analysis 38 included in per-protocol analysis
Postoperative complications and post-randomis­ ation
chemotherapy-related toxicity were assessed in the Figure 1: Trial profile

www.thelancet.com Vol 404 September 21, 2024 1111

 Articles

Liver transplantation plus Chemotherapy alone (n=47) Role of the funding source
chemotherapy (n=47)
The funders of the study had no role in study design,
Primary tumour data collection, data analysis, data interpretation, or
Primary tumour site* writing of the report.
Right 7 (15%) 7 (15%)
Left 25 (53%) 29 (62%) Results
Rectum 15 (32%) 11 (23%) Between Feb 18, 2016, and July 5, 2021, 157 patients were
(y)pT3-T4 assessed for eligibility and 63 were excluded (figure 1). Of
Yes 37 (79%) 38 (81%) four patients whose eligibility was requested before
No 9 (19%) 9 (19%) resection of the primary tumour, three were excluded
Missing 1 (2%) 0 due to extrahepatic tumour (n=2) or potentially resectable
(y)pN status colorectal liver metastases (n=1), and one was randomly
N0 21 (45%) 16 (34%) assigned 4 months after resection of the primary tumour.
N+ 26 (55%) 31 (66%) Consequently, 94 patients were included in the intention-
RAS mutation status to-treat population, with 47 in the liver transplantation
Yes 17 (36%) 13 (28%) plus chemotherapy group and 47 in the chemotherapy
No 29 (62%) 32 (69%) alone group. The per-protocol population included
Missing 1 (2%) 2 (4%) 36 patients from the liver transplantation plus
Mismatch repair status chemotherapy group and 38 from the chemotherapy
Proficient mismatch repair 47 (100%) 46 (98%) alone group, after excluding those with major deviations.
Deficient mismatch repair 0 1 (2%) Details of patients excluded from the intention-to-treat
Liver metastases at diagnosis population are in the appendix (pp 7–8). There were no
Timing of metastases patients with missing data for the primary outcome.
Synchronous† 47 (100%) 45 (96%) Baseline disease characteristics at diagnosis (table 1)
Metachronous 0 2 (4%) and randomisation (table 2) were similar in both groups.
Number of colorectal liver metastases 20·0 (14·0–25·0) 20·0 (12·0–25·0) Patients had a median age of 54∙0 years (IQR 47∙0−59∙0),
<10 5 (11%) 7 (15%) and 55 (59%) of 94 patients were male and 39 (41%) were
10−20 19 (40%) 18 (38%)
female. All patients who underwent liver transplantation
>20 23 (49%) 22 (47%)
plus chemotherapy had synchronous metastases versus
Diameter of largest colorectal liver 55·0 (43·0–76·0) 50·0 (27·0–83·0)
all but two who received chemotherapy only. The median
metastases, mm number of colorectal liver metastases at diagnosis was
CEA level, ng/mL 305·0 (32·9– 762·0) 81·0 (20·0, 530·0) 20 in both groups. At randomisation, the liver
CA19–9 level, UI/mL 96·0 (19·7–800·0) 193·0 (20·9–1949·0) transplantation plus chemotherapy group received a
Systemic chemotherapy after diagnosis median of 21∙0 chemotherapy cycles (IQR 18∙0−29∙0)
Type of chemotherapy (first line) versus 17∙0 cycles (12∙0−24∙0) in the chemotherapy
Fluorouracil alone 0 0 alone group, in up to three lines of chemotherapy.
Oxaliplatin based 22 (47%) 22 (47%)
During first-line chemotherapy, 64 (68%) of 94 patients
Irinotecan based 9 (19%) 11 (23%)
had received doublet chemotherapy and 30 (32%) of
Triplet chemotherapy 16 (34%) 14 (30%)
94 patients had received triplet regimens; 76 (81%) of
Targeted therapy (first line)
94 patients had targeted therapy. 14 (15%) patients had
previous liver resection. The median delay between
None 8 (17%) 10 (21%)
diagnosis and randomisation was 15·9 months
Anti-VEGF only 21 (45%) 16 (34%)
(11·8−25·7) for liver transplantation plus chemotherapy
Anti-EGFR only 18 (38%) 21 (45%)
versus 13·5 months (9·0−19·4) for chemotherapy alone.
Tumour response (first line)‡
Grade 3 or worse toxicity during the last chemotherapy
Complete response 1 (2%) 0
line before randomisation occurred in six (13%) patients
Partial response 27 (57%) 27 (57%)
in the liver transplantation plus chemotherapy group and
Stable disease 14 (30%) 14 (30%)
eight (17%) patients in the chemotherapy alone group
Progression 5 (11%) 5 (11%)
(table 2). Baseline characteristics of the per-protocol
Missing 0 1 (2%)
population are presented in the appendix (p 10).
CA19–9=carbohydrate antigen 19–9. CEA=carcinoembryonic antigen. EGFR=epidermal growth factor receptor. 38 (81%) of 47 patients in the liver transplantation plus
VEGF=vascular endothelial growth factor. Data are n (%) or median (IQR). *Right=primary tumour located proximally to chemotherapy group underwent liver transplantation at a
the colic flexure. Left=primary tumour located distally to the colic flexure. Rectum=primary tumour located within 15 cm
median of 50∙5 days (IQR 30∙0–65∙0) post-randomisation
of the anal verge. †Synchronous is defined as metastases diagnosed within 1 month of diagnosis of the primary tumour.
‡Tumour response was evaluated according to the Response Evaluation Criteria in Solid Tumours criteria version 1.1. with 30 (79%) patients having surgery within 2 months
from the last chemotherapy cycle. All but one (5%) patient
Table 1: Baseline characteristics at diagnosis of colorectal liver metastases in the intention-to-treat
had a low Oslo score in the liver transplantation plus
population
chemotherapy group. Nine patients (19%) in the liver

1112 www.thelancet.com Vol 404 September 21, 2024

 Articles

Liver transplantation Chemotherapy Liver transplantation Chemotherapy

plus chemotherapy alone (n=47) plus chemotherapy alone (n=47)
(n=47) (n=47)
Age, years 52·0 (47·0–59·0) 55·0 (47·0–59·0) (Continued from previous column)
Sex Grade ≥3 toxicity (CTCAE)
Male 27 (57%) 28 (60%) Yes 6 (13%) 8 (17%)
Female 20 (43%) 19 (40%) No 39 (83%) 34 (72%)
ECOG performance status Missing 2 (4%) 5 (11%)
0 38 (81%) 37 (79%) Cumulative chemotherapy
1 9 (19%) 10 (21%) Total number of
Number of colorectal liver 14·0 (8·0–25·0) 15·0 (5·0–25·0) chemotherapy lines
metastases 1 18 (38%) 23 (49%)
<10 12 (26%) 16 (34%) 2 21 (45%) 17 (36%)
10−20 20 (43%) 17 (36%) 3 8 (17%) 7 (15%)
>20 15 (32%) 14 (30%) Cumulative number of 21·0 (18·0–29·0) 17·0 (12·0–24·0)
Diameter of largest colorectal 27·0 (18·0–42·0) 27·0 (16·0–45·0) chemotherapy cycles (total
liver metastases, mm lines)
CEA, ng/mL 3·6 (2·2–12·4) 3·6 (2·0–22·1) ≤12 3 (6%) 14 (30%)
CA19–9, IU/mL 11·4 (5·9–30·0) 15·0 (6·5–28·7) 13−23 25 (53%) 21 (45%)
Fong’s clinical risk score* ≥24 19 (40%) 12 (26%)
Low (0–2) 20 (43%) 13 (28%) Previous curative intent surgery
High (3–5) 27 (57%) 34 (72%) None 43 (91%) 37 (79%)
Time between diagnosis and 15·9 (11·8–25·7) 13·5 (9·0–19·4) Minor hepatectomy 2 (4%) 5 (11%)
randomisation, months Major hepatectomy 2 (4%) 5 (11%)
Ongoing chemotherapy Delay between primary resection and randomisation >24 months
Type of chemotherapy Yes 5 (11%) 7 (15%)
Fluorouracil alone 7 (15%) 1 (2%) No 42 (89%) 40 (85%)
Oxaliplatin based 12 (26%) 11 (23%)
CA19–9=carbohydrate antigen 19–9. CEA=carcinoembryonic antigen.
Irinotecan based 20 (43%) 27 (57%) CTCAE=Common Terminology Criteria for Adverse Events. ECOG=Eastern
Triplet 8 (17%) 8 (17%) Cooperative Oncology Group. EGFR=epidermal growth factor receptor.
Targeted therapy agent IU=international units. VEGF=vascular endothelial growth factor. Data are n (%)
or median (IQR). *Despite the decrease in size and tumour markers levels, patients
None 2 (4%) 4 (9%)
remained with multiple metastases impossible to resect completely. †Tumour
Anti-VEGF 17 (36%) 16 (34%) response was evaluated according to the Response Evaluation Criteria in Solid
Anti-EGFR 28 (60%) 27 (57%) Tumors criteria, version 1.1.
Number of chemotherapy 14·0 (8·0–20·0) 11·0 (7·0–14·0) Table 2: Baseline characteristics at randomisation in the intention-to-
cycles (last line)
treat population
Tumour response†
Partial response 26 (55%) 21 (45%)
Stable disease 21 (45%) 26 (55%) (14·0–24·0) in the hospital (appendix p 13). Early post-liver
(Table 2 continues in next column) transplantation immunosuppressive regimens for the
liver transplantation plus chemotherapy group are
presented in the appendix (p 14). Post-transplantation
transplantation plus chemotherapy group had no liver chemotherapy was administered overall in 26 (68%) of
transplantation because of intercurrent hepatic or 38 patients, with 15 (58%) receiving for more than six
extrahepatic progression while waiting for liver transplan­ cycles (appendix p 15).
tation (n=5), intraoperative discovery of extrahepatic Of 46 patients in the chemotherapy alone group who
disease (n=3), or identification of prostate cancer during continued chemotherapy for a median of 16 cycles
pre-transplantation checks (n=1; appendix p 7). The (IQR 5−45), six patients (13%) had one line, three (7%)
median time to progression for patients who progressed had two lines, and 37 (80%) had three lines of systemic
while waiting for liver transplantation was 30∙0 days treatment from randomisation. Three patients
(IQR 14∙0–48∙0). Waiting times for three patients excluded underwent local ablation and 11 had radioembolisation to
at laparotomy were 15 days, 26 days, and 168 days. increase local control. Characteristics and outcome of
Recipient data are presented in the appendix (p 13). nine patients who unexpectedly underwent partial
Severe morbidity (34%), retransplantation (8%), and hepatectomy (n=7) or liver transplantation (n=2) during
mortality rates (3%) within 3 months have been previously the study period after a median interval of 20∙7 months
reported.11 Median length of stay was 6·0 days (IQR 12∙2−25∙8) from randomisation are shown in the
(IQR 5·0–8·0) in the intensive care unit and 16·0 days appendix (pp 8–9).

www.thelancet.com Vol 404 September 21, 2024 1113

 Articles

In the intention-to-treat population, after a median group and was 29·7 months (95% CI 20·8–39·4) in the
follow-up of 59∙3 months (IQR 42∙4−60∙2) at database chemotherapy group.
cutoff, 56 deaths had been reported. Median survival was 5-year overall survival was 56∙6% (95% CI 43∙2−74∙1)
not reached for the liver transplantation chemotherapy for liver transplantation plus chemotherapy and 12∙6%
(5∙2–30∙1) for chemotherapy alone (HR 0·37 [95% CI
A 0·21−0·65]; p=0·0003; figure 2A).
100 HR 0·37 (95% CI 0·21−0·65); p=0·0003 The estimated restricted mean survival time up to
90 60 months was 43·9 months (95% CI 38·0–49·9) for the
80 liver transplantation plus chemotherapy group and 31·3
70 months (26·3−36·3) for the chemotherapy alone group,
Overall survival (%)

60 corresponding to a gain of 12·6 months (95% CI

50 4∙9−20∙4 months; p=0·0014). In the intention-to-treat
40 analysis, 3-year overall survival was 65∙5% (95% CI
30 53∙2−80∙8) for the liver transplantation plus chemo­
20
Chemotherapy only
therapy group and 38∙9% (26∙9−56∙2) for the
10 LT plus chemotherapy chemotherapy alone group. In the per-protocol analysis,
0 5-year overall survival was 73∙2% (95% CI 59∙6−90∙0;
0 6 12 18 24 30 36 42 48 54 60
nine events) for the liver transplantation plus
Number at risk
(number censored)
chemotherapy group and 9∙3% (3∙2–26∙8; 33 events) the
Chemotherapy only 47 (0) 47 (0) 41 (0) 33 (0) 28 (0) 20 (3) 16 (3) 8 (6) 6 (6) 5 (6) 2 (8) chemotherapy alone group. Median survival was not
LT plus chemotherapy 47 (0) 45 (0) 41 (0) 38 (0) 35 (0) 30 (2) 28 (3) 23 (6) 18 (11) 14 (14) 10 (18) reached for the liver transplantation plus chemotherapy
group and was 26·6 months (95% CI 16·5−35·7) for the
B
100 HR 0·16 (95% CI 0·07−0·33); p<0·0001
chemotherapy alone group (HR 0·16 [95% CI 0·07−0·33];
90
p<0·0001; figure 2B).
80 The median progression-free survival was 17·4 months
70 in the liver transplantation plus chemotherapy group
Overall survival (%)

60 versus 6·4 months in the chemotherapy alone group,

50 with a 3-year progression-free survival rate of 32∙9%
40 (95% CI 20∙6–52∙7) versus 3∙9% (0∙7–23∙0) and a 5-year
30 progression-free survival rate of 19∙9% (9∙0–44∙1) versus
20 0%, respectively (HR 0·34 [95% CI 0·20–0·57]; p<0·0001;
10 figure 2C).
0 Among the 36 patients in the per-protocol analysis who
0 6 12 18 24 30 36 42 48 54 60
underwent transplantation, 26 (72%) presented an isolated
Number at risk recurrence in the liver (one [4%]), the lungs (14 [54%]), the
(number censored)
Chemotherapy only 38 (0) 38 (0) 32 (0) 24 (0) 20 (0) 15 (1) 11 (1) 6 (2) 5 (2) 4 (2) 2 (3) lymph nodes (three [12%]), in other sites (five [19%]), or in
LT plus chemotherapy 36 (0) 35 (0) 35 (0) 34 (0) 32 (0) 28 (2) 26 (3) 21 (6) 17 (10) 14 (13) 10 (17) multiple sites (three [12%]). 19 (73%) of these 26 patients
received chemotherapy. Surgery or local ablation with
C curative intent was performed in 12 (46%) patients (liver
100 HR 0·34 (95% CI 0·20–0·57); p<0·0001
in one patient, lung in eight patients, colorectal in one
90
patient, and other in two patients; appendix p 16). No
80
Progression-free survival (%)

70
patient in the liver transplantation plus chemotherapy
60
group received best supportive care for first recurrence.
50 The median secondary progression-free survival in the
40 liver transplantation plus chemo­ therapy group was
30 35∙4 months, with a 5-year secondary progression-free
20 survival rate of 36∙1% (95% CI 21∙9–59∙4; figure 3). At the
10 last follow-up in the per-protocol population, 15 (42%) of
0 36 patients in the liver transplantation plus chemotherapy
0 6 12 18 24 30 36 42 48 54 60
group were alive without disease compared with one (3%)
Time since randomisation (months)
Number at risk in the chemo­therapy alone group.
(number censored)
Chemotherapy only 38 (0) 20 (0) 10 (0) 6 (0) 4 (0) 2 (1) 1 (1) 0 (1) ·· ·· ··
During the study period in the as-treated population,
LT plus chemotherapy 36 (0) 32 (0) 24 (0) 17 (0) 14 (0) 13 (1) 11 (1) 8 (2) 6 (4) 4 (6) 2 (7) 110 serious adverse events were observed in 32 (80%) of
40 patients who underwent liver transplantation (from
Figure 2: Survival outcomes in chemotherapy alone and chemotherapy plus liver transplantation
either group), and 69 serious adverse events were
Overall survival in the intention-to-treat population (A) and per-protocol population (B), and progression-free
survival in the per-protocol population (C). Shaded areas represent 95% CIs. Tick marks represent censored observed in 45 (83%) of 54 patients treated with
patients. LT=liver transplantation. HR=hazard ratio. chemotherapy alone. In the liver transplantation plus

1114 www.thelancet.com Vol 404 September 21, 2024

 Articles

chemotherapy group, three (8%) of 36 patients who 100 PFS 19·9 (9·0-44·1%)
received a transplant were retransplanted because of 90
Secondary PFS 36·1 (21·9-59·4%)
primary non-function, post-transplantation discovery of 80
a gallbladder cancer in the graft, or caval obstruction 70
related to a large-for-size liver graft. One patient died 60
postoperatively of multi-organ failure.

PFS (%)
50
The most frequent grade 3 or worse complications with 40
liver transplantation plus chemotherapy were biliary 30
complications (n=4), pulmonary complications (n=3), 20
early allograft dysfunction (n=3), primary non-function 10
(n=2), postoperative haemorrhage (n=2), and superficial 0
site infection (n=2; table 3). In the liver transplantation 0 6 12 18 24 30 36 42 48 54 60
Time since randomisation (months)
plus chemotherapy group (per protocol), among patients Number at risk
who received post-liver transplantation chemotherapy, (number censored)
PFS 36 (0) 32 (0) 24 (0) 17 (0) 14 (0) 13 (1) 11 (1) 8 (2) 6 (4) 4 (6) 2 (7)
grade 3 or 4 toxicity was observed in eight (36%) of Secondary PFS 36 (0) 32 (0) 28 (0) 24 (0) 22 (0) 21 (1) 17 (1) 14 (3) 12 (5) 9 (8) 5 (10)
22 evaluable patients (table 4). In the chemotherapy
group (per protocol), chemotherapy-related grade 3 or 4 Figure 3: Secondary PFS in liver transplantation plus chemotherapy group who had a liver transplantation in
the per-protocol population
toxicity after randomisation occurred in 17 (47%) of
A significant proportion of the 28 patients who presented with recurrence after liver transplantation were
36 patients (table 4).14,15 considered eligible for potentially curative treatment of the recurrent disease. These patients were therefore
Health-related quality-of-life data were available for censored from the PFS curve at the time of recurrence. As they become free from disease after resection or ablation
55 (74%) of 74 patients at inclusion. The median global of their recurrence, they were no longer considered to be censored in the corrected secondary PFS. Secondary PFS
shows the effect of the treatment of recurrence when compared with primary PFS. Shaded areas represent 95% CIs.
health score from QLQ-C30 was 75 (IQR 58–83) in the
Tick marks represent censored patients. PFS is defined as the time to first recurrence after liver transplantation.
liver transplantation plus chemotherapy group and Secondary PFS is the time to first recurrence without secondary remission occurring after liver transplantation.
71 (58–83) in the chemotherapy alone group at baseline, PFS=progression-free survival.
with no significant difference between groups in
QLQ-C30 scores (appendix pp 17–18). The global health therapy. This study is the first randomised controlled
scores did not vary substantially between the groups over trial to show a clear benefit in overall survival for liver
time. However, we observed a trend towards physical transplantation added to chemotherapy versus chemo­
functioning decline over time in the chemotherapy alone therapy alone. This clear survival advantage could be
group. Similarly, there were declines in fatigue, pain, related to three key factors. The first factor was strict
dyspnoea, and loss of appetite in the chemotherapy group patient selection by using more rigorous eligibility
(appendix p 22). criteria than the initial Norwegian study,8 as shown by
the low Oslo score shared by all but one transplant
Discussion recipients. Partial response or stable disease after
In the TransMet trial, liver transplantation plus chemotherapy was a prerequisite, as patients progressing
chemotherapy in patients with permanently unresectable on chemotherapy have poor outcomes after surgery or
colorectal liver metastases was associated with liver transplantation.17,18 BRAF mutation was an exclusion
significantly better 5-year survival than chemotherapy criterion in light of reduced progression-free survival and
alone. These results were observed in the intention-to- overall survival in these patients after liver resection.19,20
treat population although 19% of patients allocated to Also, no more than three lines of chemotherapy were
liver transplantation plus chemotherapy dropped out and permitted to avoid transplants in patients with no further
a similar proportion of patients allocated to chemotherapy active treatment options.
alone subsequently underwent liver resection or The second factor was the implementation of an
transplantation. The per-protocol analysis yielded a independent expert committee, which excluded 40% of
greater 5-year survival benefit for liver transplantation patients considered as potentially eligible by local tumour
plus chemotherapy versus chemotherapy alone, under­ boards. Empathy for patients without curative options
lining the success of this approach. might lead local medical teams to propose liver
Since liver transplantation was reconsidered as a transplantation as a compassionate indication, even with
potential treatment for patients whose only other option potentially worse outcomes.21 In addition, the committee’s
is palliative chemotherapy and who have a poor prognosis expertise in radiological assessments and surgical
for long-term survival, several non-comparative studies assessment of unresectability of colorectal liver metas­
have suggested that, with improved patient selection, tases was a key factor for accurate and homogeneous
liver transplantation combined with chemotherapy patient selection.
might have better outcomes than chemotherapy alone.7,16 A third factor was prioritising patients for transplant.
Historically, however, confounding criteria might have Rapid access to an organ ready for transplantation was
resulted in patients with better prognoses undergoing essential to avoid long waiting times and risk of tumour
liver transplantation than those selected for chemo­ progression in the context of multinodular bilobar

www.thelancet.com Vol 404 September 21, 2024 1115

 Articles

disease. The objective agreed with national organ-

Any grade (n=36) Grade ≥3b (n=36)
sharing organisations to do transplantations in these
Hepatic patients within 2 months of interruption of
Biliary 5/36 (14%) 4/36 (11%) chemotherapy was achieved in 79% of patients, with
Arterial 6/36 (17%) 1/36 (3%) only one patient waiting for more than 3 months after
Early graft dysfunction* 4/36 (11%) 3/36 (8%) his last cycle of chemotherapy. Meeting this timeframe
Collection 3/36 (8%) 1/36 (3%) was especially important in countries with organ
Primary non-function† 2/36 (6%) 2/36 (6%) shortages. National organ-allocation policies, such as the
Haemorrhage 2/36 (6%) 2/36 (6%) sickest-first approach, are based on the model for end-
Hepatic or caval 1/36 (3%) 1/36 (3%) stage liver disease (MELD) score.22 However, patients
Ascites 2/35 (6%) 0 with colorectal liver metastases have preserved liver
Portal 1/36 (3%) 0 function and a low MELD score, and are unlikely to be
Rejection 3/35 (8%) 0 prioritised over patients with end-stage liver disease.
Digestive Therefore, MELD exception points should be allocated
Ileus 3/36 (8%) 1/36 (3%) to increase the priority for these patients, as in this study.
Malnutrition 1/36 (3%) 0 An alternative approach to supplementing the limited
Other 7/33 (19%) 1/35 (3%) donor pool is to use living donors23–25 or technical
General condition 2/34 (6%) 0 refinements that enable the use of partial grafts
Haematological (segments 2 and 3), such as in the RAPID concept (ie,
Anaemia 1/36 (3%) 0 resection and partial liver segment 2–3 transplantation
Other or not defined 3/34 (8%) 0 with delayed total hepatectomy).26,27
Pulmonary Survival and toxicity data in the chemotherapy alone
Pleural effusion 6/36 (17%) 1/36 (3%) group were consistent with known regimens for
Other 4/36 (11%) 3/36 (8%) metastatic colorectal cancer28 and previously reported
Cardiovascular mortality and severe complication rates with liver
DVT 2/35 (6%) 0 transplantation plus chemotherapy were consistent
Other 7/34 (19%) 1/34 (3%)
with liver transplantation in common indications (eg,
Renal 8/36 (22%) 1/36 (3%)
cirrhosis and other validated liver primary malig­
Superficial site infection 3/36 (8%) 2/36 (6%)
nancies).29 Although few data supported the combination,
adjuvant chemotherapy was delivered with immuno­
Infection
suppression showing acceptable toxicity rates. However,
CMV 3/36 (8%) 0
no clinical benefit could be observed due to the limited
Other 7/33 (19%) 1/35 (3%)
number of patients.
Diabetes 6/36 (17%) 0
From an oncological perspective, these results support
Data are n/N (%). CMV=cytomegalovirus. DVT=deep vein thrombosis. *Early graft liver transplantation as the best option for chemotherapy-
dysfunction was defined according to Olthoff and colleagues.13 †PNF was defined controlled, liver-only, unresectable colorectal liver
according to Makowka and colleagues.14
metastases. This finding might represent a major
Table 3: Postoperative complications in the liver transplantation plus change in clinical practice and in perceiving the role
chemotherapy group according to the per-protocol population that transplantation could play in prolonging survival
and offering a cure for patients with metastatic
dissemination. From a transplantation perspective, our
Liver transplantation plus Chemotherapy alone (n=38)
chemotherapy (n=24)
findings support unresectable colorectal liver metastases
as an indication for transplantation, with the 73% 5-year
Any grade Grade ≥3 Any grade Grade ≥3
overall survival rate being in line with the survival rate
Any toxicity 23/24 (96%) 8/22 (36%) 35/36 (97%) 17/36 (47%) seen in patients with common indications for liver
Haematological disorders 10/24 (42%) 4/23 (17%) 16/29 (55%) 6/25 (24%) transplantation.30 A 19% dropout rate was observed
Gastrointestinal disorders 15/23 (65%) 3/21 (14) 26/29 (90%) 3/27 (11%) during the wait for transplantation or at laparotomy,
Nervous system disorders 4/23 (17%) 1/22 (5%) 21/31 (68%) 3/27 (11%) which is similar to that seen in liver transplantation for
General disorders 15 (63%) 0 29/34 (85%) 2/24 (8%) perihilar cholangio­carcinoma and, to a lesser extent, for
Renal disorders 0 0 1/23 (4%) 0 hepatocellular carcinoma. Whenever contraindication
Infectious disorders 0 0 2/28 (7%) 0 was detected during surgical exploration, however, no
Immune system disorders 0 0 3/29 (10%) 0 graft loss was observed, and once transplanted, graft use
Other disorders 15/24 (63%) 3/22 (14%) 31/33 (94%) 10/27 (37%) was optimal.
From both perspectives, post-transplantation outcomes
Data are n/N (%).
were notable for overall survival but less optimal for
Table 4: Toxicity related to systemic chemotherapy after randomisation in the per-protocol population progression-free survival because of significant recurrence
(74%). This was lower than the recurrence rate in the

1116 www.thelancet.com Vol 404 September 21, 2024

 Articles

SECA-I study,8 but similar to that in the SECA-II study,31 Toulouse, France); Prof Julien Taieb PhD (Department of
which was more rigorous in its patient selection. These Gastroenterology and Digestive Oncology, AP-HP Hôpital Européen
Georges Pompidou, Paris, France); Prof Jean Baptise Bachet PhD
results suggest that patient selection can be improved, by (Department of Oncology, AP-HP Hôpital Pitié-Salpêtrière, Paris,
molecular biology and detection of microscopic residual France); Mohamed Bouattour MD (Department of Digestive Oncology,
disease by circulating tumour DNA. Recurrence was AP-HP University Hospital Beaujon, Clichy, France); Marine Jary MD
confined exclusively to the lungs in more than 50% of (Department of Oncology, University Hospital Clermont-Ferrand,
Clermont-Ferrand, France); Patrizia Burra PhD (Department of Surgery,
cases; 46% of recurrences were treated by surgery or local Oncology and Gastroenterology, University Hospital Padua, Padua,
ablation. Both approaches led to secondary remission in Italy); Philippe Rougier PhD (Department of Gastroenterology and
25% of patients, and a secondary 5-year progression-free Digestive Oncology, European Hospital Georges Pompidou, Paris,
survival rate after resection or ablation of recurrence of France); Joan Figueras PhD (Hepatobiliary and Pancreatic Surgery Unit,
Department of Surgery, Dr Josep Trueta Hospital, Biomedical Institute
36%. Altogether, 42% of transplanted patients were of Research, IdIBGi [Institut d’Investigació Biomèdica de Girona],
disease-free at the last follow-up, arguing for a real Girona, Spain). *Prof Hebbar died in June 2024.
potential for cure with liver transplantation. Contributors
Some limitations of the study should be considered. RA designed the study. RA and MG coordinated the study and wrote the
Heterogeneity could have occurred in the chemotherapy study protocol. RA, ML, MD, MG, FL, JL, and PM participated in the
regimens (as well as immunosuppressive therapies) expert panel that conducted patient evaluations for eligibility. CP and LG
oversaw the methodology and statistical analysis. RA, MG, and CP
depending on the different policies of local oncologists supervised the data analysis and wrote the manuscript. RA, EB, PB, DC,
and hepatologists. Post-transplant chemotherapy was not FC, LCh, LCo, UC, MD, KG, MG, VG, BH, JH, HJ, VL, FM, J-YM, DP,
possible for all patients, mainly because of complications ES, OS, JPA, and CV participated in the recruitment of patients and
or suboptimal recovery after liver transplantation. verified the local raw data. RA, MG, and CP verified the raw data
according to local investigators. All authors were involved in drafting the
However, these variations reflect real-life practice, and work or reviewing it critically for important intellectual content.
the strength of the study relies on the homogeneous and All authors had full access to all the data in the study and had final
rigorous patient selection, the reproducibility of which responsibility for the decision to submit for publication.
ensured strict adherence to the eligibility criteria and Declaration of interests
guaranteed prioritisation of patients for organ allocation. We declare no competing interests.
This approach could conceivably be generalised to Data sharing
transplant centres, thus expanding the indications for Data are available on reasonable request. The Commission Nationale de
l’Informatique et des Libertés (French data privacy authority) and the
transplant oncology.
European Union’s General Data Protection Regulation mean that the
In summary, this study has shown that liver database and information and consent documents signed by the patients
transplantation combined with chemotherapy signifi­ cannot be transmitted. De-identified individual participant data
cantly improves overall survival and offers potential for underlying the results might be considered by the editorial board for
availability to interested researchers subject to terms and conditions of
cure in selected patients with unresectable colorectal liver such consultation and compliance with the applicable regulations. For
metastases versus chemotherapy alone. Patient survival all enquiries, please contact: [email protected].
after liver transplantation, when effectively performed, is Acknowledgments
similar to that observed in common indications for liver We would like to thank all the patients, their families, and the hospitals
transplantation. These results support validating liver and their research teams for participating in the TransMet study.
transplantation as a new standard option that might The TransMet trial was supported by the French National Cancer
Institute, the French Ministry of Health, and the Assistance Publique—
change current practice for liver-only, permanently Hôpitaux de Paris (AP–HP) and was endorsed by the French Chapter of
unresectable, colorectal liver metastases. the International Hepato-Pancreato-Biliary Association (Association de
Collaborative TransMet group Chirurgie Hépato-Bilio-Pancréatique et Transplantation). We are grateful
Prof Olivier Soubrane PhD (Department of Hepatobiliary Surgery to Nadjia Benarab, Ikrame Ramdhani, Felix Sandjo, and Sonia Makhlouf
and Transplant, AP-HP University Hospital Beaujon, Clichy, France); from the AP–HP Central Research Unit for the trial operational activities
Prof Mickael Lesurtel PhD (Department of Hepatobiliary Surgery (centre set-up, follow-up, and closure; data monitoring; data
and Transplant, AP-HP University Hospital Beaujon, Clichy, France); management; retrieval; and quality control), and to Patrizia Burra,
Prof Mohamed Hebbar MD* (Department of Oncology, Hôpital Claude Philippe Rougier, and Joan Figueras from the surveillance committee for
Huriez/University Hospital Lille, Lille, France); Prof Karim Boudjema their contribution. We would also like to acknowledge the national
PhD (Department of Hepatobiliary Surgery and Liver Transplantation, organ-sharing organisations for allowing the study to be realised.
Pontchaillou Hospital/University Hospital Rennes, Rennes, France); Editing support was provided by Deirdre Carman, Fiona Weston, and
Denis Smith MD (Department of Oncology, Haut-Lévêque Hospital, Lee Miller of Miller Medical Communications. This study was funded
Southern Group, University Hospital Bordeaux, Bordeaux, France); by the Programme Hospitalier de Recherche Clinique en
Prof Mircea Chirica PhD (Department of Digestive and Emergency Cancérologie-PHRC-K 2013 French National Cancer Institute and the
Surgery, University Hospital Grenoble, Grenoble, France); French Minister of Health, Directorate General of Care Provision.
Prof Thierry Lecomte PhD (Department of Liver Gastroenterology References
and Digestive Oncology, Trousseau Hospital, Regional University 1 Osterlund P, Salminen T, Soveri L-M, et al. Repeated centralized
Hospital Tours, Tours, France); Sara Lonardi PhD (Department of multidisciplinary team assessment of resectability, clinical behavior,
Oncology, Veneto Institute of Oncology IOV-IRCCS, Padua, Italy); and outcomes in 1086 Finnish metastatic colorectal cancer patients
Samuel Lesourd MD (Department of Medical Oncology, Centre Eugène (RAXO): a nationwide prospective intervention study.
Marquis, Rennes, France); Prof Christophe Borg PhD (Department of Lancet Reg Health Eur 2021; 3: 100049.
Medical Oncology, Jean Minjoz Hospital, University Hospital Besançon, 2 Tomlinson JS, Jarnagin WR, DeMatteo RP, et al. Actual 10-year
Besançon, France); Prof Rosine Guimbaud PhD (Digestive Medical survival after resection of colorectal liver metastases defines cure.
Oncology Unit, Hôpital Rangueil, University Hospitals Toulouse, J Clin Oncol 2007; 25: 4575–80.

www.thelancet.com Vol 404 September 21, 2024 1117

 Articles

3 Adam R, Delvart V, Pascal G, et al. Rescue surgery for unresectable 18 Dueland S, Smedman TM, Syversveen T, Grut H, Hagness M,
colorectal liver metastases downstaged by chemotherapy: a model to Line P-D. Long-term survival, prognostic factors, and selection of
predict long-term survival. Ann Surg 2004; 240: 644–57, discussion patients with colorectal cancer for liver transplant: a non-
657–58. randomized controlled trial. JAMA Surg 2023; 158: e232932.
4 Heinemann V, von Weikersthal LF, Decker T, et al. FOLFIRI plus 19 Margonis GA, Buettner S, Andreatos N, et al. Association of BRAF
cetuximab versus FOLFIRI plus bevacizumab as first-line treatment mutations with survival and recurrence in surgically treated
for patients with metastatic colorectal cancer (FIRE-3): patients with metastatic colorectal liver cancer. JAMA Surg 2018;
a randomised, open-label, phase 3 trial. Lancet Oncol 2014; 153: e180996.
15: 1065–75. 20 Cohen R, Liu H, Fiskum J, et al. BRAF V600E mutation in first-line
5 Bond MJG, Bolhuis K, Loosveld OJL, et al. First-line systemic metastatic colorectal cancer: an analysis of individual patient data
treatment strategies in patients with initially unresectable colorectal from the ARCAD database. J Natl Cancer Inst 2021; 113: 1386–95.
cancer liver metastases (CAIRO5): an open-label, multicentre, 21 Toso C, Pinto Marques H, Andres A, et al. Liver transplantation for
randomised, controlled, phase 3 study from the Dutch Colorectal colorectal liver metastasis: Survival without recurrence can be
Cancer Group. Lancet Oncol 2023; 24: 757–71. achieved. Liver Transpl 2017; 23: 1073–76.
6 Hoti E, Adam R. Liver transplantation for primary and metastatic 22 Freeman RB Jr, Wiesner RH, Harper A, et al. The new liver
liver cancers. Transpl Int 2008; 21: 1107–17. allocation system: moving toward evidence-based transplantation
7 Foss A, Adam R, Dueland S. Liver transplantation for colorectal policy. Liver Transpl 2002; 8: 851–58.
liver metastases: revisiting the concept. Transpl Int 2010; 23: 679–85. 23 Hernandez-Alejandro R, Ruffolo LI, Sasaki K, et al. Recipient and
8 Hagness M, Foss A, Line PD, et al. Liver transplantation for donor outcomes after Living-donor liver transplant for unresectable
nonresectable liver metastases from colorectal cancer. Ann Surg colorectal liver metastases. JAMA Surg 2022; 157: 524–30.
2013; 257: 800–06. 24 Sasaki K, Ruffolo LI, Kim MH, et al. The current state of liver
9 Bonney GK, Chew CA, Lodge P, et al. Liver transplantation for transplantation for colorectal liver metastases in the United States:
nonresectable colorectal liver metastases: the International Hepato- a call for standardized reporting. Ann Surg Oncol 2023; 30: 2769–77.
Pancreato-Biliary Association consensus guidelines. 25 Rajendran L, Claasen MP, McGilvray ID, et al. Toronto
Lancet Gastroenterol Hepatol 2021; 6: 933–46. management of initially unresectable liver metastasis from
10 Chávez-Villa M, Ruffolo LI, Line P-D, Dueland S, Tomiyama K, colorectal cancer in a living donor liver transplant program.
Hernandez-Alejandro R. Emerging role of liver transplantation for J Am Coll Surg 2023; 237: 231–42.
unresectable colorectal liver metastases. J Clin Oncol 2024; 26 Settmacher U, Ali-Deeb A, Coubeau L, et al. Auxilliary liver
42: 1098–101. transplantation according to the RAPID procedure in noncirrhotic
11 Adam R, Badrudin D, Chiche L, et al. Safety and feasibility of patients: technical aspects and early outcomes. Ann Surg 2023;
chemotherapy followed by liver transplantation for patients with 277: 305–12.
definitely unresectable colorectal liver metastases: insights from the 27 Line P-D, Hagness M, Berstad AE, Foss A, Dueland S. A novel
TransMet randomised clinical trial. EClinicalMedicine 2024; concept for partial liver transplantation in nonresectable colorectal
72: 102608. liver metastases: the RAPID concept. Ann Surg 2015; 262: e5–9.
12 European Medicines Agency. Guideline on adjustment for baseline 28 Cremolini C, Antoniotti C, Rossini D, et al. Upfront FOLFOXIRI
covariates in clinical trials. London: European Medicines Agency, plus bevacizumab and reintroduction after progression versus
2015. mFOLFOX6 plus bevacizumab followed by FOLFIRI plus
13 Clavien PA, Barkun J, de Oliveira ML, et al. The Clavien–Dindo bevacizumab in the treatment of patients with metastatic colorectal
classification of surgical complications: 5-year experience. Ann Surg cancer (TRIBE2): a multicentre, open-label, phase 3, randomised,
2009; 250: 187–96. controlled trial. Lancet Oncol 2020; 21: 497–507.
14 Olthoff KM, Kulik L, Samstein B, et al. Validation of a current 29 Battistella S, Grasso M, Catanzaro E, et al. Evolution of liver
definition of early allograft dysfunction in liver transplant recipients transplantation indications: expanding horizons. Medicina (Kaunas)
and analysis of risk factors. Liver Transpl 2010; 16: 943–49. 2024; 60: 412.
15 Makowka L, Gordon RD, Todo S, et al. Analysis of donor criteria for 30 Adam R, Karam V, Cailliez V, et al. 2018 Annual Report of the
the prediction of outcome in clinical liver transplantation. European Liver Transplant Registry (ELTR)—50-year evolution of
Transplant Proc 1987; 19: 2378–82. liver transplantation. Transpl Int 2018; 31: 1293–317.
16 Dueland S, Guren TK, Hagness M, et al. Chemotherapy or liver 31 Dueland S, Syversveen T, Solheim JM, et al. Survival following liver
transplantation for nonresectable liver metastases from colorectal transplantation for patients with nonresectable liver-only colorectal
cancer? Ann Surg 2015; 261: 956–60. metastases. Ann Surg 2020; 271: 212–18.
17 Adam R, Pascal G, Castaing D, et al. Tumor progression while on
chemotherapy: a contraindication to liver resection for multiple
colorectal metastases? Ann Surg 2004; 240: 1052–61, discussion
1061–64.

1118 www.thelancet.com Vol 404 September 21, 2024