Residual Hepatocellular Carcinoma Risk After Bariatric Surgery: Fibrosis Regression, Metabolic Improvement, and Implications for Surveillance—A Narrative Review

1. Introduction

Metabolic dysfunction-associated steatotic liver disease (MASLD) describes hepatic steatosis accompanied by cardiometabolic risk factors; metabolic dysfunction-associated steatohepatitis (MASH) denotes its inflammatory form. These terms replace nonalcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH), although historical studies used earlier definitions. The original terminology is retained when describing study populations because retrospective relabeling does not prove that every participant met current criteria [1].

Hepatocellular carcinoma (HCC) is an important consequence of metabolic liver disease. A large Veterans Health Administration cohort demonstrated that HCC risk varied markedly with cirrhosis status, while some cancers occurred without documented cirrhosis. This distinction matters: the existence of noncirrhotic HCC does not imply that surveillance is justified for every person with MASLD. Individual risk and population-level absolute incidence must be considered separately [2].

Metabolic and bariatric surgery offers a means of changing the underlying disease trajectory. Hepatological guidance recognizes its potential benefits in appropriately selected patients with obesity, while emphasizing assessment of advanced liver disease [3]. Histological studies show improvement in steatohepatitis and fibrosis, but these outcomes answer a different question from whether HCC risk is eliminated [4,5].

The central question of this review is therefore not simply whether surgery is associated with fewer cancers than nonsurgical management. It is whether postoperative improvement identifies patients whose remaining risk is sufficiently low to change surveillance. Residual risk is defined here as the risk of incident HCC after bariatric surgery in patients without pre-existing HCC. Recurrence after cancer treatment, cholangiocarcinoma, and liver metastases are outside the principal scope. Persistent advanced fibrosis provides a direct clinical reason to examine this question, although persistence of fibrosis alone does not quantify postoperative cancer incidence [6].

2. Scope and Literature Approach

This review was developed as a focused narrative synthesis rather than a systematic review or meta-analysis. A targeted literature search was conducted up to 1 October 2026 using PubMed/MEDLINE, supplemented by searches of relevant journal websites, reference lists, and official professional society guidelines. Search terms included combinations of “bariatric surgery,” “metabolic surgery,” “sleeve gastrectomy,” “Roux-en-Y gastric bypass,” “MASLD,” “NAFLD,” “MASH,” “NASH,” “hepatocellular carcinoma,” “liver fibrosis,” “fibrosis regression,” and “HCC surveillance.” Priority was given to randomized controlled trials, prospective and retrospective clinical studies, systematic reviews, meta-analyses, mechanistic investigations, and relevant clinical practice guidelines. Publications were selected according to their relevance to the objectives of this review, with particular emphasis on postoperative HCC risk, histological changes, long-term liver outcomes, and surveillance recommendations.

Evidence was selected for its relevance to postoperative HCC occurrence, liver histology, long-term clinical liver outcomes, mechanistic plausibility, or surveillance decisions. No exhaustive database export, duplicate independent screening, formal risk-of-bias scoring, or pooled analysis was performed. Accordingly, completeness cannot be assumed. Evidence is discussed in three distinct groups: studies directly reporting HCC; studies reporting histological or composite clinical endpoints; and guidance extrapolated from broader liver disease populations. Study populations are not summed because administrative datasets and institutional cohorts may overlap. Table 1 summarizes representative clinical evidence rather than an exhaustive study inventory.

3. Biological Pathways and Their Relevance After Surgery

Obesity-related hepatocarcinogenesis involves more than the amount of hepatic fat. In experimental models, dietary and genetic obesity enhanced interleukin-6 (IL-6) and tumor necrosis factor (TNF) signaling, hepatic inflammation, and tumor development. These observations support a connection between systemic metabolic dysfunction and a tumor-promoting hepatic environment. They do not establish that changing circulating cytokine concentrations after bariatric surgery prevents cancer in humans [13].

Endoplasmic reticulum stress provides another link between nutrient excess and hepatocyte injury. Nakagawa and colleagues demonstrated that hypernutrition and hepatocyte stress could cooperate to generate steatohepatitis and TNF-dependent HCC in a mouse model. Reduced nutrient stress after surgery is therefore biologically relevant, but the reversibility of an established carcinogenic process cannot be inferred from improved liver enzymes or reduced steatosis [14].

The gut–liver axis offers a further potential mechanism. Yoshimoto and colleagues linked obesity-associated microbial production of deoxycholic acid to DNA damage and a senescence-associated secretory phenotype that promoted liver cancer experimentally [15]. Separately, Ryan and colleagues identified a role for the bile acid receptor farnesoid X receptor (FXR) in the metabolic effects of sleeve gastrectomy in mice [16]. Taken together, these studies justify investigating postoperative bile acid and microbial changes; they do not demonstrate a direct causal chain from surgery to lower human HCC incidence.

Immune changes may also be important. Experimental work on NASH-associated HCC identified dysfunctional hepatic T-cell responses and context-dependent effects of immune checkpoint blockade [17]. For this review, the relevant implication is that metabolic liver injury can alter immune surveillance. Whether bariatric surgery reverses these changes sufficiently to affect future HCC remains an open research question; these data should not be used to select systemic cancer therapy according to prior bariatric surgery.

These mechanisms provide a coherent explanation for possible risk reduction while leaving room for persistent risk. Our interpretation is that removal of ongoing metabolic injury and reversal of established tissue damage should be treated as separate biological processes. A patient may improve metabolically before the liver has undergone substantial architectural remodeling. The proposed mechanistic distinctions are summarized in Table 2.

The postoperative questions are hypotheses proposed in this review. These studies do not prove that surgery-mediated changes prevent human HCC. FXR, farnesoid X receptor; IL-6, interleukin-6; TNF, tumor necrosis factor; HCC, hepatocellular carcinoma; NASH, nonalcoholic steatohepatitis.

4. Histological Improvement and the Limits of Fibrosis Regression

The strongest evidence for surgery concerns improvement in liver disease activity. In BRAVES, 288 participants with biopsy-confirmed NASH were randomized to lifestyle modification plus best medical care, Roux-en-Y gastric bypass, or sleeve gastrectomy. At one year, the intention-to-treat proportions achieving NASH resolution without worsening fibrosis were 16%, 56%, and 57%, respectively. This establishes efficacy for the specified histological endpoint, but the short follow-up and trial design do not establish an HCC prevention effect or oncological equivalence between procedures [5].

Longer follow-up supports the durability of histological benefit. Lassailly and colleagues reported NASH resolution without worsening fibrosis in 84% of available five-year biopsies and fibrosis reduction in approximately 70%. These findings derive from patients who underwent follow-up biopsy rather than every individual originally enrolled, making the biopsy denominator and potential attrition relevant to interpretation [4].

A different picture emerges when attention is restricted to severe baseline disease. Pais and colleagues studied 196 patients with advanced NASH; 66 underwent a repeat biopsy approximately six years after surgery. Among 36 participants with baseline advanced fibrosis, advanced fibrosis persisted in 17, or 47%, despite substantial improvement in other disease features. This is a selected paired-biopsy finding, not a persistence rate applicable to all bariatric patients. It nonetheless demonstrates that improvement in steatohepatitis can coexist with clinically important residual fibrosis [6].

The relevance of fibrosis is supported by prospective natural-history evidence outside bariatric populations. Sanyal and colleagues found higher mortality and liver-related complications in patients with advanced fibrosis than in those with less severe disease [18]. Applying this observation after surgery is an extrapolation: postoperative risk may change, but a lower weight or improved glycemic profile cannot independently establish that the risk associated with previous advanced disease has disappeared.

Postoperative assessment also requires care. Liver enzymes can normalize without excluding advanced disease, and noninvasive fibrosis tests are affected by the biological variables they measure [3]. In our interpretation, a postoperative fall in a fibrosis score should prompt longitudinal reassessment rather than an automatic change in cancer surveillance. Histological improvement, a decline in stiffness, and a fall in estimated HCC risk should be reported separately until their relationship has been validated specifically after surgery.

5. Clinical Evidence on HCC and Long-Term Liver Outcomes

Njei and colleagues used the US Nationwide Inpatient Sample to examine the cross-sectional association between previous bariatric surgery and HCC. After propensity matching, prior surgery was associated with lower HCC prevalence, with a prevalence ratio of 0.11 (95% confidence interval [CI], 0.03–0.48). The distinction between prevalence and incidence is essential: discharge-based data do not establish the prospective cancer risk of a consistently followed postoperative cohort. They also do not provide sufficiently detailed postoperative histology to guide surveillance cessation [7].

Rustgi and colleagues examined cancer outcomes among 98,090 adults with NAFLD and severe obesity, including 33,435 who underwent bariatric surgery. Cancer-specific analyses associated surgery with lower HCC risk. This supports a potential clinical benefit, but administrative definitions and the lack of serial histological characterization limit inference about patients with persistent or regressed advanced fibrosis. The study is more informative about comparative cancer occurrence than about surveillance after a documented hepatic response [8].

Ramai and colleagues synthesized nine observational studies. The pooled adjusted odds ratio from matched cohorts was 0.63 (95% CI, 0.53–0.75). This association is compatible with lower HCC occurrence after surgery, but neither an odds ratio nor a pooled rate establishes a stage-specific residual risk for an individual patient. Heterogeneity, possible dataset overlap, and differences in outcome ascertainment should temper interpretation. The pooled HCC incidence rates were 0.05 and 0.34 per 1,000 person-years in the bariatric surgery and control groups, respectively. However, these estimates were derived from heterogeneous observational populations and cannot establish the residual HCC incidence among patients with documented postoperative regression of advanced fibrosis [9].

Studies of composite liver outcomes provide complementary evidence. SPLENDOR included 1,158 patients with biopsy-proven fibrotic NASH without cirrhosis, of whom 650 underwent bariatric surgery. Surgery was associated with fewer major adverse liver outcomes. However, the endpoint combined progression to cirrhosis, HCC, transplantation, and liver-related death. A large reduction in this composite must not be reported as an equally large reduction in HCC alone [10].

SPECCIAL extended investigation to compensated MASH-related cirrhosis. Among 62 surgical patients and 106 nonsurgical controls, mean follow-up was approximately ten years. The estimated 15-year cumulative incidence of major adverse liver outcomes was 20.9% versus 46.4%, with an adjusted hazard ratio of 0.28 (95% CI, 0.12–0.64). These findings suggest that surgery can be associated with a better long-term clinical trajectory in selected patients, while the small cohort and composite endpoint restrict cancer-specific conclusions. They do not establish that HCC surveillance can be withdrawn after surgery [11].

A more recent retrospective network study by Stupalkowska and colleagues associated metabolic bariatric surgery with fewer adverse hepatic and extrahepatic outcomes over four years in patients with steatotic liver disease. However, the reduction in HCC hazard did not reach statistical significance, illustrating why favorable overall liver outcomes cannot substitute for a cancer-specific result. Coding-based cohorts also differ from studies designed to measure risk after verified fibrosis regression [12]. Overall, comparative evidence is encouraging; the specific evidence needed to redefine postoperative surveillance remains limited.

6. What Residual Risk Means in Clinical Practice

Residual risk should be considered as a longitudinal clinical problem rather than a binary label of surgical success or failure. A patient with longstanding cirrhosis, persistent diabetes, or recurrent metabolic dysfunction may have a different trajectory from a patient treated before advanced liver disease developed. Yet the literature reviewed here does not provide a validated postoperative model integrating these features with serial fibrosis measurements.

Three uncertainties are especially relevant. First, preoperative staging may be incomplete, so an apparently low postoperative score can be difficult to interpret without a reliable baseline. Second, the timescale of metabolic change differs from that of fibrosis remodeling and cancer development. Third, surveillance exposure may vary between surgical and nonsurgical patients, influencing the number and stage of detected cancers. These are methodological considerations, not established explanations for every observed association.

Surgery also does not remove unrelated causes of liver cancer risk. Viral hepatitis, alcohol exposure, and other chronic liver diseases need separate evaluation. The amount of postoperative weight loss cannot substitute for etiological assessment. Conversely, describing every patient as having substantial residual cancer risk would overstate the evidence. Most people without advanced liver disease will not develop HCC, and the absolute risk after surgery must be estimated within the appropriate disease stratum.

7. Surgical Selection and Continuity of Liver Care

Bariatric surgery in cirrhosis requires attention to liver function, clinically significant portal hypertension, sarcopenia, and transplant candidacy. The American Gastroenterological Association (AGA) expert review recommends restricting elective bariatric surgery in cirrhosis to carefully selected patients with compensated disease at experienced centers. Sleeve gastrectomy is considered the most likely preferred procedure because it preserves endoscopic access and avoids intestinal malabsorption; this preference concerns clinical management rather than proven HCC prevention superiority [19].

The practical implication is that an HCC-focused review should not create a new indication to operate solely for cancer prevention. Procedure selection must consider established obesity indications and hepatic risk. There is insufficient direct evidence to choose sleeve gastrectomy over bypass, or vice versa, on the basis of future HCC alone.

A proposed continuity strategy is to document the best-supported preoperative liver stage, communicate it to the postoperative team, and distinguish metabolic follow-up from liver-specific follow-up. When cirrhosis is present, improvement after surgery should trigger reassessment of clinical status without allowing the original diagnosis to disappear from the record. This is our care-process proposal, intended to reduce loss of surveillance during transfer between bariatric and hepatology services.

8. Surveillance After Bariatric Surgery

For patients with established cirrhosis who remain eligible for HCC-directed treatment, surgery alone does not justify stopping surveillance. The American Association for the Study of Liver Diseases (AASLD) recommends ultrasound plus alpha-fetoprotein (AFP) at approximately six-month intervals and advises against routine surveillance for noncirrhotic NAFLD with advanced fibrosis. It provides no bariatric-specific biochemical or histological criterion for surveillance withdrawal [20]. Our interpretation is therefore to retain the established cirrhosis-based indication while evaluating postoperative improvement.

Noncirrhotic bridging fibrosis (F3) is less straightforward. The 2024 MASLD guideline from the European Association for the Study of the Liver (EASL), European Association for the Study of Diabetes (EASD), and European Association for the Study of Obesity (EASO) allows surveillance to be considered after individual risk assessment in severe fibrosis [21]. The 2025 EASL HCC guideline states that surveillance cannot currently be recommended in advanced fibrosis without cirrhosis because evidence is insufficient [22]. The AGA expert review is more permissive about considering screening when noninvasive markers suggest advanced fibrosis or cirrhosis [23]. These statements differ in scope and timing. They should not be simplified into a universal recommendation to screen every postoperative patient with F3 fibrosis.

In patients without advanced fibrosis, prior bariatric surgery is not itself an indication for HCC surveillance. Continued metabolic and liver disease assessment is distinct from scheduled cancer surveillance [23]. Table 3 presents a proposed framework that combines existing guidance with explicit postoperative uncertainties; it is not a validated bariatric surveillance algorithm.

This framework is a narrative synthesis, not a tested decision rule. Surveillance eligibility also depends on competing illness and the ability to receive HCC treatment. F3, bridging fibrosis; CT, computed tomography; MRI, magnetic resonance imaging; HCC, hepatocellular carcinoma; MASLD, metabolic dysfunction-associated steatotic liver disease; NAFLD, nonalcoholic fatty liver disease.

Imaging quality also matters. When ultrasound visualization is inadequate, AASLD allows selected use of contrast-enhanced magnetic resonance imaging (MRI) or multiphase computed tomography (CT) [20]. The postoperative choice should depend on actual visualization, liver morphology, renal function, contraindications, local expertise, and access rather than assuming that weight loss guarantees an adequate examination. Abbreviated MRI and emerging biomarker strategies merit study, but cannot be described as validated specifically for patients after bariatric surgery.

A crucial research question is whether a previous diagnosis of cirrhosis continues to predict HCC after convincingly documented regression. Until postoperative cohorts establish an acceptably low absolute incidence and test a stopping strategy, surveillance cessation based solely on regression would remain unsupported. This uncertainty should be communicated without denying the substantial metabolic and hepatic benefits of surgery.

9. Evidence Gaps and Priorities for Research

Future investigations should be designed around the surveillance question. Prospective multicenter cohorts should record baseline fibrosis, portal hypertension, procedure type, diabetes, weight trajectories, alcohol exposure, and viral hepatitis. Follow-up should capture repeated hepatic assessments, surveillance modality and adherence, incident HCC, cancer stage, treatment, and competing mortality. Reporting only a composite endpoint will not resolve whether surveillance can be reduced.

Analyses should distinguish patients with persistent advanced fibrosis, documented regression, and no baseline advanced fibrosis. Time-dependent exposure definitions, landmark analyses, and competing-risk methods can reduce avoidable bias. Investigators should also report the completeness of baseline staging and the proportion of participants receiving repeat biopsy or elastography, because selective reassessment may distort apparent regression rates.

Mechanistic substudies could relate inflammatory mediators, bile acid profiles, metabolic markers, and immune phenotypes to clinical outcomes. Their value would lie in improving prediction or clarifying mediation, rather than simply documenting postoperative biomarker changes. A useful biochemical marker would need to add information beyond liver stage and established clinical variables, with external validation.

The immediate priority is a reliable estimate of HCC incidence in well-characterized postoperative cirrhosis and F3 cohorts. Only after this step would a comparative surveillance strategy become interpretable. Any proposed reduction in surveillance should assess delayed diagnosis, stage distribution, treatment eligibility, and cost-effectiveness, rather than relying on a lower fibrosis score as the sole outcome.

10. Conclusions

Bariatric surgery improves steatohepatitis and is associated with better long-term liver outcomes and lower HCC occurrence in observational studies. The remaining cancer risk after documented fibrosis regression is not adequately defined. Persistent advanced fibrosis demonstrates why metabolic improvement cannot be treated as proof that liver cancer risk has disappeared. Patients with established cirrhosis should retain guideline-based surveillance when appropriate for their clinical status; noncirrhotic advanced fibrosis requires individualized assessment within the limits of current guidance. The next advance should be a postoperative, stage-specific estimate of absolute HCC risk that can support surveillance decisions, rather than further conflation of weight loss, histological response, and cancer prevention.