Prognostic significance of pretreatment ¹⁸F-FDG PET/CT in hepatocellular carcinoma: A systematic review and meta-analysis

Abstract

Background & Aims

¹⁸F-fluorodeoxyglucose positron emission tomography/computed tomography (¹⁸F-FDG PET/CT) is not routinely recommended for hepatocellular carcinoma (HCC) diagnosis or surveillance, and its prognostic role in patients with HCC remains unclear. This meta-analysis aims to evaluate the prognostic significance of ¹⁸F-FDG PET/CT in HCC.

Methods

A systematic literature search was conducted until August 2024 to identify studies evaluating the prognostic value of ¹⁸F-FDG PET/CT in patients with HCC. Studies were included if they reported an association between pre-treatment PET/CT and clinical outcomes, including overall survival (OS), recurrence-free survival (RFS), and progression-free survival (PFS). A random-effects model was employed with heterogeneity assessed using the I² statistic.

Results

We identified 59 relevant studies (6 prospective and 53 retrospective) including 8,585 patients. Pre-treatment ¹⁸F-FDG PET/CT was positive in 30.8% of patients with HCC. Increased FDG activity was associated with poorer OS (hazard ratio [HR] 2.05, 95% CI 1.77–2.38, I² = 71.6%), RFS (HR 3.18, 95% CI 2.08–4.86, I² = 96.8%), and PFS (HR 1.64, 95% CI 1.36–1.98, I² = 71.8%), respectively. FDG activity was significantly associated with OS for both curative (HR 2.10, 95% CI 1.58–2.80, I² = 62%) and non-curative (HR 2.28, 95% CI 1.59–3.25, I² = 75.3%) treatments. The prognostic significance of ¹⁸F-FDG PET/CT remained consistent across various PET/CT parameters, including SUVmax, tumor-to-liver ratio, and visual assessment of PET/CT positivity, and curative/non-curative therapeutic modalities, all demonstrating that higher FDG uptake is associated with poorer clinical outcomes.

Conclusions

Pre-treatment ¹⁸F-FDG PET/CT is positive in 30.8% of patients with HCC and is significantly associated with adverse clinical outcomes. These findings support its prognostic relevance in HCC, while further studies may help refine its clinical role and standardize ¹⁸F-FDG PET/CT-based risk stratification.

PROSPERO ID

CRD42025633333.

Impact and implications

This meta-analysis demonstrates that pre-treatment ¹⁸F-fluorodeoxyglucose positron emission tomography/computed tomography (¹⁸F-FDG PET/CT) has prognostic value in hepatocellular carcinoma (HCC), with higher FDG uptake associated with worse overall, recurrence-free, and progression-free survival. Although not routinely recommended for HCC staging, FDG PET/CT appears to identify a biologically more aggressive subgroup. These findings support its potential role as a complementary tool for risk stratification alongside conventional imaging and clinical markers. Incorporating PET/CT-derived metabolic information may help refine prognostication and guide treatment selection in both curative and non-curative settings. Prospective studies are needed to standardize its use in clinical decision-making.

Graphical abstract

Highlights

• •Increased 18F-FDG tumor uptake is associated with poorer outcomes in patients with HCC, including shorter OS, RFS, and PFS.
• •The prognostic significance of PET-CT is evident across a wide range of treatment strategies.
• •PET-CT demonstrated consistent prognostic significance across various PET-CT parameters (SUVmax, TLR, visual assessment).

Introduction

Hepatocellular carcinoma (HCC) is one of the most common and aggressive malignancies worldwide.¹ The disease is highly heterogeneous, and its management involves a wide range of therapeutic strategies, including surgical resection, liver transplantation (LT), radiofrequency ablation, transarterial chemoembolization (TACE), transarterial radioembolization (TARE), external beam radiation therapy, and systemic therapies like tyrosine kinase inhibitors (TKIs) and immune checkpoint inhibitors (ICIs).^2,3 The choice of treatment depends on factors such as tumor stage, liver function, and patient performance status. Despite these diverse therapeutic options, the prognosis of HCC remains poor, with a 5-year survival rate of less than 20% for advanced stages.¹

To improve patient outcomes, accurate prognostic tools are essential for guiding treatment decisions. Over the years, various prognostic markers have been identified, such as tumor size, number, presence of vascular invasion, and serum tumor markers.³ More recently, imaging-based markers, particularly metabolic parameters derived from ¹⁸F-fluorodeoxyglucose positron emission tomography/computed tomography (¹⁸F-FDG PET/CT), have gained attention for their ability to reflect tumor aggressiveness and predict survival outcomes.⁴ Although ¹⁸F-FDG PET/CT is not included as a routine modality in current major international guidelines for HCC diagnosis or surveillance^2,5 because of its limited sensitivity, especially in well-differentiated tumors with low FDG uptake,⁶ it has already been incorporated into clinical practice in multiple centers and jurisdictions – particularly for treatment stratification and candidate selection for liver transplantation. To overcome these diagnostic limitations, alternative tracers such as ¹¹C-acetate and ¹¹C-methionine, as well as newer agents targeting FAP (fibroblast activation protein), PSMA (prostate-specific membrane antigen), and GPC3 (glypican-3), are being actively explored for more precise tumor characterization.^7,8 Beyond lesion detection alone, PET/CT provides both anatomical and functional information, making it a promising tool for assessing tumor metabolism and treatment response. In this context, increasing attention has been directed toward the prognostic role of ¹⁸F-FDG PET/CT in HCC. Although many studies have reported an association between increased FDG uptake and worse survival, the findings have not been entirely consistent across study designs, patient populations, and imaging protocols.^4,9

Here, we performed a comprehensive systematic review and meta-analysis to assess the impact of ¹⁸F-FDG PET/CT in this setting. We aimed to evaluate the prognostic value of ¹⁸F-FDG PET/CT for overall survival (OS), recurrence-free survival (RFS), and progression-free survival (PFS) in patients with HCC.

Materials and methods

Study design

This study adhered to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines.¹⁰ The protocol was registered with PROSPERO (Prospective Register of Systematic Reviews) under registration number CRD42025633333.

Data sources and search approach

A systematic literature search was conducted in PubMed, Scopus, Web of Science (including SCI-EXPANDED, CPCI-S, and ESCI), ProQuest™ Dissertations & Theses Citation Index, Preprint Citation Index, and the Cochrane Central Register of Controlled Trials (CENTRAL) from inception to August 26, 2024 without language limitation (Table S1). The full search strategy is detailed in Appendix 1.

Eligibility criteria

Inclusion criteria were: 1) retrospective or prospective cohort studies including more than 10 patients with HCC; 2) studies evaluating any treatment approach for HCC, including surgical resection, LT, locoregional therapies (LRTs), chemotherapy, ICIs, TKIs, and radiation therapy; 3) availability of pre-treatment ¹⁸F-FDG PET/CT data; 4) reporting of outcomes of interest, including OS, RFS, and/or PFS; and 5) provision of hazard ratios (HRs) with 95% CIs. Exclusion criteria were: 1) focusing on liver cancers other than HCC; 2) using ¹⁸F-FDG PET/CT for diagnosis or staging; 3) insufficient survival information; 4) meta-analyses, systematic reviews, case reports, or animal studies; and 5) overlapping study populations. If overlapping patient cohorts were used, we retained the latest or the larger study population.

Study screening and data extraction

All identified studies were imported into Covidence. The titles and abstracts were screened by two reviewers independently, and then the full texts were reviewed for eligibility. Any disagreements were resolved by a third reviewer (MSR-Z, JHM, or JDY). From the finally included studies, the following data were extracted: the first author, year of publication, study design, sample size, follow-up period, tumor stage, treatment type, treatment outcome, PET/CT parameters, cut-off values, and HRs with 95% CIs.

Quality assessment

The methodological quality and risk of bias of included studies were assessed using the Newcastle-Ottawa Scale. Detailed scoring criteria for selection, comparability, and outcome domains are provided in Appendix 2.

Evaluated outcomes and statistical analysis

The primary endpoint was OS, and secondary endpoints were RFS and PFS, defined as time from treatment initiation to death, recurrence, or first progression/death, with censoring at last follow-up.

The PET/CT parameters included maximum standardized uptake value (SUVmax), tumor-to-liver ratio (TLR, defined as the ratio of the tumor SUVmax to the liver SUVmean), Tmax/Lmax (defined as the ratio of the tumor SUVmax to the liver SUVmax), and metabolic tumor volume, among others. We used binary classifications (high- vs. low-uptake) based on study-specific cut-offs. When ¹⁸F-FDG PET/CT positivity was analyzed as an imaging parameter, it was typically defined by visual assessment.

HRs and 95% CIs were extracted, prioritizing multivariable over univariable HRs. An HR >1 indicated worse survival outcomes for PET/CT-positive or high-uptake groups. Heterogeneity was assessed using the I² statistic (0%: none, <25%: low, 25.1–50%: moderate, and >50%: substantial heterogeneity).¹¹ Subgroup analyses were performed on PET/CT parameters, including SUVmax, TLR, and visual PET/CT positivity when available, as well as according to treatment types. Treatments were grouped as curative-intent (LT, resection, and ablation) vs. non-curative-intent (TKIs, ICIs, radiation, and transarterial) and also analyzed individually. For the treatment subgroup analysis, only studies reporting HRs for a single treatment were included. Sensitivity analyses consisted of leave-one-out analyses to assess the stability of the pooled estimates. Publication bias was assessed via a funnel plot, Egger test, and adjusted with the trim-and-fill method.¹² Meta-regression assessed the association between prognostic value and a 10% increase in the proportion of intermediate-advanced stage (beyond Milan, BCLC C/D), or cirrhosis. p <0.05 was considered statistically significant. All analyses were conducted using the meta package in R software (version 4.4.1).

Results

Search results and study selection process

Initially, 7,047 studies were identified through database searches, with two additional studies^13,14 added from previous meta-analyses.⁴ After removing 1,466 duplicates, 5,583 studies were screened by title and abstract. Among these, 5,378 studies were excluded as irrelevant. Of the remaining 205 studies, one was excluded due to inaccessible full text, leaving 204 for full-text review. After full-text review, 59 studies involving 8,585 patients were included (Fig. 1).^8,9,[13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40], [41], [42], [43], [44], [45], [46], [47], [48], [49], [50], [51], [52], [53], [54], [55], [56], [57], [58], [59], [60], [61], [62], [63], [64], [65], [66], [67], [68], [69]

Fig. 1.

PRISMA diagram of all included studies.

CT, computed tomography; FDG, fluorodeoxyglucose; HR, hazard ratio; PET, positron emission tomography.

Study characteristics

As summarized in Table 1, sample sizes ranged from 20 to 896 patients (median: 92). Most studies (n = 36, 61%) were published between 2015 and 2020. The majority were conducted in South Korea (n = 29, 49.2%), followed by China, Japan, Europe, and the United States. Median follow-up was 30.8 months. Most studies were retrospective (n = 53, 89.8%). The primary outcomes were OS in 45 studies (76.3%), RFS in 19 studies (32.2%), and PFS in 14 studies (23.7%). The treatments included LT (n = 18, 30.5%), surgical resection (n = 13, 22.0%), multiple treatments (n = 15, 25.4%), TACE (n = 5, 8.5%), sorafenib (n = 3, 5.1%), and other therapies. PET/CT parameters included SUVmax (n = 25, 42.4%), TLR (n = 22, 37.3%), visual positivity (n = 20, 33.9%), and Tmax/Lmax (n = 4, 6.8%), with cut-offs ranging from 2.0–11.7 (median 4.9) for SUVmax and 1.35–4.8 (median 2.0) for TLR. Of the 59 included studies, 44 provided extractable patient-level counts of pre-treatment PET/CT findings classified as positive or negative according to each study definition. Across these studies, 2,026 of 6,569 patients (30.8%) were reported to have positive PET/CT findings.

Table 1.

Characteristics of included studies.

	First author	Year	Sample size	Follow-up (months)∗	Stage	Treatment	Outcome	PET/CT parameters
1	Abuodeh15	2016	34	12.0	BCLC A-D	TARE	PFS	SUVmax
2	Ahn16	2013	189	28.6	TNM I-IIIB	Resection	OS	SUVmax, TLR
3	Bauschke17	2023	103	79.0	Milan	LT	OS	TLR
4	Besson18	2023	62	12.0	NA	Sorafenib	OS	MTV
5	Bhangui19	2021	300	47.0	Milan	LT	OS, RFS	PET positivity
6	Blanc-Durand20	2018	47	11.0	A-C	TARE	OS, PFS	PET-RadScore
7	Cho21	2020	259	42.2	NA	Resection	OS, RFS	PET positivity
8	Cho22	2015	104	8.0	A-D	Multiple	OS	TSUVmax/mean mediastinum SUV
9	Ding23	2020	93	18.3	Milan	LT	RFS	Uptake-volume product
10	Han13	2014	298	32.0	NA	Resection	OS	SUVmax
11	Hiraoka24	2010	53	—	NA	Resection	RFS	Tmax/Lmax
12	Hwang25	2017	132	38.1	0-A	Resection	OS	TLR
13	Hwang26	2018	526	37.6	NA	Multiple	OS, RFS	PET positivity
14	Hyun27	2016	317	46	0-A	Multiple	OS, RFS	TLR
15	Ida28	2020	121	42.2	NA	RFA	RFS	PET positivity
16	Jia29	2022	34	—	TNM I-IV	Multiple	OS	SUVmax
17	Jiang30	2022	72	30.2	NA	Resection	RFS	SUVmax, TLR
18	John31	2020	148	—	0-D	Multiple	OS	PET positivity
19	Jreige32	2017	48	16.2	A-C	TARE	OS, PFS	SUVmax, TLR
20	Kan33	2024	131	72.0	NA	Multiple	RFS	PET positivity
21	Kang34	2022	49	47.3	Milan	LT	RFS	TLR
22	Kawamura14	2014	64	—	0-D	Multiple	OS	SUVmax
23	Kawamura35	2021	48	—	A-C	Lenvatinib	OS	TLR
24	Kawamura36	2022	20	—	A-C	Atezo/Bev	PFS	TLR
25	Kim37	2011	107	—	B-C	Multiple	OS, PFS	SUVmax
26	Kim9	2018	226	39.45	NA	Resection	OS	PET positivity
27	Kim38	2015	77	22.2	B	TACE	OS, PFS	TLR
28	Kim39	2021	627	79.0	Milan	LT	RFS	PET positivity
29	Kim40	2014	256	33.6	Milan	LT	OS, RFS	PET positivity
30	Kornberg41	2015	103	—	Milan	LT	RFS	PET positivity
31	Kornberg42	2017	116	43.0	A-C	LT	RFS	PET positivity
32	Lee43	2011	29	—	A, C	Sorafenib	OS, PFS	SUVmax
33	Lee44	2017	166	9.5	C	Multiple	OS, PFS	TLR
34	Lee45	2017	242	54.7	0-A	Resection	OS, RFS	TLR, MTV
35	Lee46	2016	214	10.7	B-C	Multiple	OS, PFS	TLR
36	Lee47	2015	59	31.5	A-C	Multiple	OS, PFS	TLR
37	Lee48	2014	93	30.8	Milan	LT	RFS	PET positivity
38	Lee49	2016	280	—	NCCK criteria	LT	OS	PET positivity
39	Lee50	2020	896	89.0	T1-T3	LT	OS, RFS	PET positivity
40	Lim51	2019	78	20.0	NA	Resection	RFS	PET positivity
41	Na52	2017	291	6.3	C	Multiple	OS	TLR
42	Pan53	2009	56	10.9	NA	NA	OS	TLR
43	Park54	2018	92	—	NA	Resection	OS	PET positivity
44	Park55	2018	42	30.1	0-D	TACE	RFS	TLR
45	Paudyal56	2008	31	—	T1-T4	Resection	OS	SUVmax
46	Rhee57	2017	228	12.7	T2-T4	Radiotherapy	OS, PFS	SUVmax
47	Seo58	2007	70	19.9	NA	Resection	OS	SUVmax, TLR
48	Shin59	2011	25	9.6	NA	Multiple	OS, PFS	SUVmax
49	Song60	2015	73	24.7	NA	TACE	OS	TLR
50	Song61	2013	58	31.3	NA	TACE	OS	TLR
51	Song62	2012	83	—	A-C	TACE	OS	TLR
52	Sun63	2018	61	13.2	0-C	NA	OS	SUVmax, TLR
53	Sung64	2018	35	6.1	C	Sorafenib	OS, PFS	TLR
54	Takeuchi65	2017	56	5.3	0-D	NA	OS	SUVmax, TLR
55	Wang66	2023	77	16.8	B-C	Multiple	OS	SUVmax
56	Wu8	2023	22	—	B-C	Multiple	OS, PFS	SUVmax
57	Yang67	2020	47	19.1	C-D	LT	OS	SUVmax
58	Ye68	2017	103	25.7	Milan	LT	OS	TLR
59	Yoh69	2021	314	57.9	Milan	Resection	OS	Tmax/Lmax

BCLC, Barcelona Clinic Liver Cancer; FDG, fluorodeoxyglucose; LT, liver transplantation; MTV, metabolic tumor volume; OS, overall survival; PET, positron emission tomography; PFS, progression-free survival; RFA, radiofrequency ablation; RFS, recurrence-free survival; SUVmax, maximum standardized uptake value; TACE, transarterial chemoembolization; TARE, transarterial radioembolization; TLR, tumor-to-liver ratio.

^∗Follow-up period is summarized as mean or median.

Quality assessment

The methodological quality of the 59 included studies was moderate to high, with 59.3% rated as "Good" and 40.7% as "Fair" based on NOS; none were rated "Poor" (Tables S2 and S3). In the selection domain, 32.2% included only a subset of HCC stages, and 81.4% did not evaluate multiple interventions, limiting generalizability. In the comparability domain, most studies (84.7%) adjusted for both tumor stage and other factors (e.g. alpha-fetoprotein levels). In the outcome domain, all studies performed multivariable analysis, but only 6.8% reported treatment-specific outcomes. Additionally, 6.8% had insufficient follow-up (Table S4).

Prognostic value of ¹⁸F-FDG PET/CT on OS

Analysis of 45 studies comprising 6,953 patients demonstrated that elevated metabolic activity on ¹⁸F-FDG PET/CT was associated with poorer OS, with a pooled HR of 2.05 (95% CI 1.77–2.38, I² = 71.6%) (Fig. 2). Although the funnel plot exhibited significant asymmetry (Egger’s test, p <0.001), the association remained after trim-and-fill adjustment (adjusted HR 1.43, 95% CI 1.22–1.66), with no substantial change in the pooled result on leave-one-out analysis (Figs S1A and S2). All ¹⁸F-FDG PET/CT parameters were significantly associated with poor OS: SUVmax (HR 1.81), TLR (HR 1.90), and PET/CT positivity (HR 2.14) (Fig. 3A). The heterogeneity was primarily driven by differences in SUVmax-based categorization. ¹⁸F-FDG PET/CT showed prognostic significance across both curative (HR 2.10) and non-curative (HR 2.28) settings (Fig. 3B). Treatment-specific analysis revealed significant HRs for TACE (HR 2.08), resection (HR 1.59), and LT (HR 2.85). Meta-regression revealed no significant effect of cirrhosis or advanced stage (BCLC C/D) on the prognostic performance of PET/CT (Table S5).

Fig. 2.

Pooled HR for overall survival.

HR, hazard ratio. Pooled estimates were calculated using a random-effects meta-analysis.

Fig. 3.

Subgroup analysis for pooled HR for overall survival.

(A) PET/CT parameters. (B) Treatment type (curative-intent vs. non-curative-intent treatment). HR, hazard ratio; SUVmax, maximum standardized uptake value; TLR, tumor-to-liver ratio. Pooled estimates were calculated using a random-effects meta-analysis.

Prognostic value of ¹⁸F-FDG PET/CT on RFS

In 19 studies involving 4,374 patients, higher FDG uptake was significantly associated with poorer RFS, with a pooled HR of 3.18 (95% CI 2.08–4.86, I² = 96.8%) (Fig. 4). Although funnel plot asymmetry was detected (Egger’s test, p = 0.020), the pooled estimate remained statistically significant after trim-and-fill adjustment (adjusted HR 4.67, 95% CI 3.12–6.99), and leave-one-out sensitivity analysis supported the stability of the findings (Figs S1B and S4).

Fig. 4.

Pooled HR for recurrence-free survival.

HR, hazard ratio. Pooled estimates were calculated using a random-effects meta-analysis.

Subgroup analyses demonstrated significant associations for PET/CT positivity (HR 2.91) and TLR (HR 2.41). Treatment-specific analysis showed significant prognostic value of PET/CT in patients undergoing LT (HR 5.04) and resection (HR 1.84) (Figs S5 and S6). Meta-regression revealed no significant association between PET/CT performance and the proportion of cirrhosis or advanced stage disease (beyond Milan criteria) (Table S5).

Prognostic value of ¹⁸F-FDG PET/CT on PFS

Among 14 studies with 1,111 patients, higher FDG metabolic activity was associated with a shorter PFS, with a pooled HR of 1.64 (95% CI 1.36–1.98, I² = 71.8%) (Fig. 5). Despite evidence of publication bias (Egger’s test, p <0.001), the direction and significance of the association were maintained after trim-and-fill adjustment (HR 1.22, 95% CI 1.02–1.46); leave-one-out analysis demonstrated that the association was generally preserved (Figs S1C and S7).

Fig. 5.

Pooled HR for progression-free survival.

HR, hazard ratio. Pooled estimates were calculated using a random-effects meta-analysis.

Subgroup analyses showed that both SUVmax (HR 1.47) and TLR (HR 1.73) were significant predictors of PFS. Treatment-specific analysis demonstrated prognostic significance across modalities, with TARE showing the highest pooled HR of 4.60 (95% CI 1.46–14.48) (Figs S8 and S9). Meta-regression revealed no significant association between PET/CT performance and the proportion of advanced stage disease (BCLC C/D) (Table S5).

Discussion

Our meta-analysis evaluated the prognostic value of pre-treatment ¹⁸F-FDG PET/CT in HCC, using data from 59 studies involving 8,585 patients. Elevated FDG uptake was consistently associated with poorer OS, RFS, and PFS, supporting the value of PET/CT as a prognostic biomarker.

Although ¹⁸F-FDG PET/CT is not currently a recommended modality for HCC^7,70 because of its limited sensitivity, with approximately 50% of HCCs showing low or no FDG uptake,⁶ it may still offer prognostic value. A recent multicenter study found that PET/CT altered BCLC staging in 5% of cases and influenced treatment decisions in only 2%, highlighting its limited impact on management.⁷¹ Despite these limitations, high FDG uptake generally reflects more aggressive tumor biology,⁷² and our meta-analysis reveals a consistent association between higher PET/CT parameters and poorer survival outcomes across studies. Beyond prognostication, PET/CT may also assist in treatment response assessment following locoregional therapies, such as TACE or TARE, where residual metabolic activity can identify incomplete responses not captured by conventional anatomical imaging.^32,61 To enhance diagnostic sensitivity and overcome the limitations of glucose metabolism-based imaging, alternative tracers have been explored: ¹¹C-acetate demonstrates higher sensitivity for well-differentiated HCC,⁷³ while more recent tracers targeting FAP, PSMA, and GPC3 show promise for more precise tumor characterization.^7,8

Importantly, these associations were consistent regardless of treatment strategy. Elevated FDG uptake predicted poorer OS in both curative-intent (HR 2.10) and non-curative-intent (HR 2.28) settings, and was most pronounced in LT recipients (HR 2.85 for OS; HR 5.04 for RFS), suggesting it may help identify LT candidates at higher recurrence risk.⁷⁴

A prior meta-analysis of 22 studies (n = 1,721) by Sun et al. suggested that high FDG uptake was associated with poorer OS and disease-free survival.⁴ However, the limited sample size restricted subgroup analyses, with no significant association observed for surgical resection. Our study extends this evidence by including more studies and patients, incorporating additional outcomes (RFS and PFS), and evaluating a broader spectrum of treatments including ICIs and lenvatinib. Beyond the larger sample size, an important contribution of the present study is that the prognostic association of pretreatment FDG PET/CT was observed across multiple treatment settings, including both curative-intent and non-curative-intent therapies, although the magnitude of association varied by treatment type. This comprehensive update provides stronger and clinically relevant support for the prognostic role of PET/CT in HCC.

Beyond OS, elevated FDG uptake was also associated with higher risks of recurrence and progression. A high SUVmax before TACE may require more intense follow-up and an earlier switch to systemic or combined therapies.^46,61,62 Similarly, post-resection patients with high TLR may benefit from more frequent surveillance.¹⁶ These findings highlight the potential of PET/CT in identifying high-risk patients and guiding personalized treatment strategies to improve outcomes.

Despite consistent trends, substantial heterogeneity was observed, partly attributed to variability in PET/CT parameters (notably SUVmax-based categorizations) and treatment types. Variability in imaging techniques, scan timing, cut-off values, and patient and tumor characteristics likely contributed but could not be fully analyzed due to limited reporting. Although these differences may partially account for the inconsistent effect sizes, they also highlight the complexity of applying PET/CT uniformly across diverse clinical settings. Nevertheless, the direction of the prognostic association was generally similar across various PET/CT parameters, although the magnitude of effect varied and should be interpreted in the context of substantial heterogeneity. Still, the lack of standardized cut-off values and parameters limits clinical use and cross-study comparability. Future research should establish unified ¹⁸F-FDG PET/CT criteria and integrate them with biomarkers like alpha-fetoprotein for better risk stratification and treatment planning in HCC.⁷

Some limitations should be acknowledged. First, significant heterogeneity was observed across studies. Second, the presence of publication bias suggests that studies with negative or non-significant results may be underrepresented in the literature; however, the trim-and-fill method confirmed the robustness of our findings. Third, most studies were conducted in East Asia, limiting generalizability to Western populations with differing HCC etiologies and management practices. Fourth, the predominance of retrospective studies introduces potential biases, and prospective trials are needed for validation. Fifth, most studies focused on surgical interventions, with fewer studies addressing contemporary ICI-based systemic therapies. Sixth, variability in PET/CT parameter cut-offs and the lack of raw data prevented continuous variable analysis and standardization. As a result, no unified FDG uptake threshold could be established, limiting direct clinical application. Seventh, individual patient data meta-analysis was not feasible. Nevertheless, most included studies reported multivariable-adjusted HRs. Therefore, our pooled estimates likely reflect effect sizes that are at least partially adjusted for prognostic variables. Finally, due to limited sensitivity in PET-negative tumors, the standalone prognostic value of PET/CT may be restricted. Future studies should identify clinical scenarios or tumor subtypes where ¹⁸F-FDG PET/CT offers the greatest prognostic value.

In summary, this meta-analysis demonstrated that ¹⁸F-FDG PET/CT is a significant prognostic indicator for OS, RFS and PFS across various HCC therapies. Further studies may help refine its optimal clinical role, standardize PET/CT parameters and cut-offs, and identify settings in which it provides the greatest incremental value for treatment planning and risk stratification (e.g. to identify high-risk patients who might benefit from neoadjuvant therapy or combination approaches).

Abbreviations

AFP, alpha-fetoprotein; BCLC, Barcelona Clinic Liver Cancer; CT, computed tomography; FDG, fluorodeoxyglucose; HCC, hepatocellular carcinoma; HR, hazard ratio; ICI, immune checkpoint inhibitor; LT, liver transplantation; OS, overall survival; PET, positron emission tomography; PFS, progression-free survival; RFS, recurrence-free survival; SUVmax, maximum standardized uptake value; TACE, transarterial chemoembolization; TARE, transarterial radioembolization; TKI, tyrosine kinase inhibitor; TLR, tumor-to-liver ratio; TNM, tumor-node-metastasis.

Authors’ contributions

Concept and Design: JHM, MSR-Z, JDY. Data Acquisition and/or Interpretation: JHM, MSR-Z, AMA, SYH, PJC, TY, HA, RH, HK, JDY. Drafting the Manuscript: JHM. Critical Revision: MSR-Z, AMA, SYH, PJC, TY, HA, RH, HK, BT, AGS, JDY. Final Approval and Agreement: All Authors.

Data availability

All data supporting the findings of this study are available within the article and its supplementary materials. This study did not generate new individual-level patient data.

Financial support

This research was supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (RS-2023-00305161) and the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (>RS-2025-24523339).

Conflicts of interest

Ju Dong Yang provides a consulting service for Eisai, Exact Sciences, Fujifilm Medical Sciences. Amit Singal has served as a consultant or on advisory boards for Genentech, AstraZeneca, Eisai, Bayer, Exelixis, Merck, Elevar, Boston Scientific, Sirtex, Histosonics, FujiFilm Medical Sciences, Exact Sciences, Roche, Glycotest, Abbott, and GRAIL. Bachir Taouli receives research support from Bayer, Takeda, Regeneron, Helio Health, Echosens, Guerbet, and Siemens. Additionally, BT provides consulting services for Bayer, Guerbet, and Ascelia. Other authors declare no conflict of interest.

Please refer to the accompanying ICMJE disclosure forms for further details.

Supplementary data

The following are the Supplementary data to this article: