Meta-analysis finds 75% five-year survival after selected living-donor liver transplants

A systematic review and meta-analysis found post-transplant outcomes among selected patients with hepatocellular carcinoma who were beyond the Milan criteria at presentation but successfully downstaged before receiving a living donor liver transplant. Across 17 studies, the investigators analyzed 511 recipients who achieved downstaging and underwent transplantation, reporting a pooled recurrence rate of 25.0% and five-year overall survival of 75.0%. In a pre-specified primary analysis restricted to nine studies using locoregional therapy and targeting Milan or University of California San Francisco criteria, five-year overall survival was 75.8% and recurrence was 22.1%.

These estimates describe outcomes among patients who responded to treatment and reached transplant, rather than the full population that attempted downstaging. Hepatocellular carcinoma, or HCC, is a cancer that develops within the liver. Liver transplantation can remove both a tumor and the diseased liver, but conventional selection criteria have restricted access.

The benchmark Milan criteria permit a single tumor no larger than 5 centimeters or no more than three tumors, each no larger than 3 centimeters, provided there is no macrovascular invasion or extrahepatic spread. Expanded systems, including UCSF and several Asian criteria, have widened eligibility, while downstaging uses locoregional treatment to reduce tumor burden so that an initially ineligible patient may qualify for transplantation. Most evidence for this approach comes from deceased-donor transplantation, where patients generally undergo a period of observation after treatment.

The new analysis asks whether outcomes reported for living donor transplantation, a paradigm with different selection and timing practices, are similarly promising. Researchers searched PubMed and Embase through April 8, 2026, and supplemented the search with Google Scholar, reference-list review and citation tracking. They included studies that enrolled patients beyond Milan criteria at presentation, administered downstaging therapy with the intent of enabling transplantation, provided separable data for living donor recipients, and reported at least one survival, recurrence or downstaging-success outcome.

Fifteen included studies were retrospective and two were prospective. They came from South Korea, Taiwan, Japan, India, China, Spain and Pakistan, and their downstaged living donor cohorts ranged from 4 to 144 patients. The therapies were diverse.

Most cohorts included transarterial chemoembolization, including conventional, drug-eluting bead and transarterial chemo-lipiodolization variants, either alone or combined with radiofrequency ablation, radiation therapy or hepatic resection. Three studies specifically addressed tumors with portal vein tumor thrombosis, and two recent studies examined immune checkpoint inhibitor regimens. Eight studies targeted Milan criteria, four targeted UCSF criteria, and five used institution-specific systems.

Nine studies met the primary-analysis definition because they used locoregional therapy and a standard radiologic endpoint of either Milan or UCSF criteria, without a predominant focus on portal vein tumor thrombosis or macrovascular invasion. Among the 511 downstaged transplant recipients, the overall pooled recurrence rate was 25.0% with a 95% confidence interval of 17.4% to 34.5%. Individual recurrence estimates ranged from zero in two cohorts with only four patients each to 51.4% in one Korean study, illustrating the marked variation associated with patient selection, treatment protocols and follow-up.

Pooled overall survival was 90.5% at one year, 77.7% at three years and 75.0% at five years. The five-year recurrence-free survival estimate was 70.8%. Statistical heterogeneity was low for one-year survival but substantial for three- and five-year survival, as well as for five-year recurrence-free survival.

Recurrence therefore appeared to be the most comparable outcome across studies, although even that estimate was not entirely consistent. The more narrowly defined primary analysis produced similar results. Across five studies contributing recurrence data, the pooled recurrence rate was 22.1% with a 95% confidence interval of 9.7% to 42.6%, based on 173 patients.

Six studies with 337 patients contributed to the primary five-year overall survival estimate of 75.8%, while four studies with 221 patients contributed a pooled five-year recurrence-free survival of 72.8%. Heterogeneity remained high in each analysis, showing that the restricted protocols did not create a uniform evidence base. A leave-one-out analysis of five-year overall survival found that no single study drove the overall estimate, while analyses excluding studies with fewer than 10 or 20 patients produced results close to the main findings.

The investigators also reconstructed patient-level data from published Kaplan-Meier curves to account for censoring and uneven follow-up. Ten studies contributing 328 patients were used for overall survival, and four studies contributing 226 patients for recurrence-free survival. The reconstructed analysis estimated overall survival at 91.6% after one year, 79.7% after three years and 75.7% after five years.

Reconstructed recurrence-free survival was 85.6%, 78.1% and 75.6% at the same intervals, with the median not reached. These time-to-event estimates were close to those obtained by pooling proportions, although the reconstruction depended on usable survival curves and sometimes on reported time-point rates where risk tables were unavailable. A central limitation lies in the two different denominators behind the results.

Survival and recurrence were measured among people who completed downstaging and then received a transplant. By contrast, downstaging success was calculated among everyone who began treatment. In 12 studies with data on 1,186 attempted cases, only 47.5% achieved downstaging, with a 95% confidence interval of 36.1% to 59.1%.

Success ranged from 10.4% in one immune checkpoint inhibitor cohort to 73.9% in a drug-eluting bead chemoembolization cohort, and heterogeneity was considerable. None of the included studies comprehensively followed the outcomes of patients who failed downstaging, progressed or became ineligible, nor did they compare these patients with beyond-Milan recipients who received transplants without successful downstaging. The favorable transplant outcomes are therefore conditional on response and carry selection and immortal-time biases; they do not establish the benefit of downstaging itself.

Timing is another important distinction between living and deceased donor transplantation. The U.S. deceased-donor framework includes an “ablate-and-wait” period of at least six months after successful downstaging, allowing aggressive tumors to reveal themselves and be removed from the waiting list. Living donor transplantation has little or no such interval because donor availability rather than an allocation queue determines when surgery occurs.

A recent response may consequently provide less evidence about durable tumor control than a response maintained during an extended observation period. The review noted that neither target criteria nor treatment modality significantly separated recurrence risk, but the small number of studies and their clinical diversity prevented firm comparisons. Subgroup estimates for expanded or institution-specific criteria were 28% recurrence, compared with 21% for Milan or UCSF criteria.

Portal-vein-thrombosis cohorts had 30.4% pooled recurrence, while separate analyses did not detect a statistically significant difference between studies with and without portal vein tumor thrombosis. Overall, the findings support further evaluation of downstaging before living donor liver transplantation but not a claim that it has been shown to improve survival for all eligible patients. The evidence remains retrospective, geographically concentrated and clinically heterogeneous, with small cohorts and inconsistent reporting of alpha-fetoprotein, des-gamma-carboxy prothrombin, explant pathology, treatment success and patient dropouts.

Most patients who attempted downstaging did not proceed to transplant, making those unreported outcomes a decisive gap. The authors call for prospective multicenter intention-to-treat studies that enroll patients when downstaging begins, track every dropout and its cause, standardize response assessment and success definitions, and report fixed-interval oncologic endpoints. Such studies would determine whether the encouraging results among successful responders persist when all attempted patients and the harms of unsuccessful treatment are included.

Subject of Research: Post-transplant outcomes after downstaging for hepatocellular carcinoma before living donor liver transplantation

Article Title: Post-transplant oncologic outcomes after locoregional downstaging before living donor liver transplantation for hepatocellular carcinoma beyond Milan criteria: a systematic review and meta-analysis of selected cohorts

Article References: Jo, Y., Park, Y., Lee, H.-K., Lee, S., Jung, H.-S., Choi, S. J. N., & Kim, H.-S. (2026). Post-transplant oncologic outcomes after locoregional downstaging before living donor liver transplantation for hepatocellular carcinoma beyond Milan criteria: a systematic review and meta-analysis of selected cohorts. Transplant International, 39, Article 17284. https://doi.org/10.3389/ti.2026.17284

Image Credits: AI Generated

DOI: 10.3389/ti.2026.17284

Keywords: hepatocellular carcinoma, living donor liver transplantation, downstaging, locoregional therapy, systematic review, meta-analysis, overall survival, recurrence-free survival, cancer recurrence, Post-transplant, oncologic, outcomes