# Lenvatinib vs. Atezolizumab Plus Bevacizumab for Liver Cancer That Cannot Be Surgically Removed: A Patient-Friendly Guide to a 6,620-Person Meta-Analysis This systematic review and meta-analysis pooled 12 studies and 6,620 patients. It compared two first-line treatments for unresectable hepatocellular carcinoma (HCC). HCC is the most common form of primary liver cancer, when the tumor cannot be removed by surgery. Lenvatinib (Lenvima) and the combination of atezolizumab plus bevacizumab (Tecentriq plus Avastin, often shortened to ATE/BEV) produced similar overall survival. The two treatments also produced similar progression-free survival, objective response rates and disease control rates. There were no statistically significant differences in any of these measures. The two treatments also caused similar rates of serious (grade 3 or higher) side effects, high blood pressure and protein in the urine. However, lenvatinib was linked to more loss of appetite, diarrhea, fatigue and hand-foot syndrome. Hand-foot syndrome is redness, peeling and pain on the palms and soles. ATE/BEV was linked to more cases of raised aspartate aminotransferase (AST). AST is a liver enzyme that leaks into the blood when liver cells are injured. # Efficacy and Safety of Lenvatinib versus Atezolizumab Plus Bevacizumab in the Treatment of Unresectable Hepatocellular Carcinoma: A Systematic Review and Meta-Analysis. ## Table of Contents - Key Points - Background: Why This Question Matters - The Two Treatments Being Compared - How the Researchers Conducted the Review - Which Studies Were Included - Who Was in These Studies - Overall Survival: No Clear Winner - Progression-Free Survival: No Clear Winner - Tumor Response and Disease Control - Serious Side Effects (Grade 3 or Higher) - Specific Side Effects: Where the Two Treatments Differ - Risk of Bias and Publication Bias - What This Means for Patients - Limitations: What This Review Could Not Prove - Practical Recommendations - Frequently Asked Questions - Source Information ## Key Points - In a review of 12 studies and 6,620 patients, lenvatinib and atezolizumab plus bevacizumab produced similar overall survival, progression-free survival, and tumor response, with no statistically significant differences. - Serious side effects, high blood pressure, and protein in the urine occurred at similar rates with both treatments. - Lenvatinib was linked to more loss of appetite, diarrhea, fatigue, and hand-foot syndrome, while atezolizumab plus bevacizumab was linked to more raised aspartate aminotransferase. - Most evidence was retrospective and came from Japan, Korea, and China, so results may not apply equally to all patients. - No head-to-head randomized trial exists; treatment choice should weigh liver function, overall health, and tolerance for specific side effects. ## Background: Why This Question Matters Hepatocellular carcinoma (HCC) is the most common type of primary liver cancer — cancer that starts in the liver itself rather than spreading there from elsewhere. It begins when hepatocytes (the main liver cells) turn malignant. HCC accounts for roughly 85% of liver cancers in people diagnosed with cirrhosis (permanent scarring of the liver). The disease is a major global health problem. In 2022 alone, doctors diagnosed more than **866,136 new cases** of liver cancer worldwide, making it a leading cause of cancer deaths. Despite real progress in understanding what causes HCC, the overall 5-year survival rate remains very low — only about **5% to 14%**. Survival depends heavily on how far the cancer has spread. The 5-year survival rate is: - **33%** when the cancer is localized (confined to one area) - **10%** when it is regional (spread to nearby lymph nodes or tissues) - **2%** when it is metastatic (spread to distant organs) Other factors also shape how long a patient lives. These include portal vein thrombosis, a blood clot in the liver's main vein. They also include the size of the tumor and alpha-fetoprotein (AFP) levels, a protein in the blood used as a tumor marker. The tumor stage and how much liver damage the patient already has also matter. Patients with more extensive cirrhosis tend to have shorter survival times and fewer treatment options. Doctors usually diagnose HCC using non-invasive criteria — blood tests and imaging — rather than surgery. Tissue biopsy is being used more often in clinical practice. Prevention is possible for some patients: vaccines and antiviral medicines can protect against hepatitis B (HBV) and hepatitis C (HCV), the two viruses most strongly linked to liver cancer. Other major risk factors include heavy alcohol use, obesity, and non-alcoholic fatty liver disease (NAFLD, now often called MASLD — fat buildup in the liver not caused by alcohol). Treatment depends on the stage. For early-stage HCC, doctors may perform liver resection (surgery to remove part of the liver) or liver transplantation, both of which are considered curative. Local ablation with radiofrequency (using heat to destroy the tumor) can be an option when surgery is not needed. For intermediate-stage disease, transarterial chemoembolization (TACE — injecting chemotherapy directly into the tumor's blood supply) has been the standard of care for the past two decades. ## The Two Treatments Being Compared For unresectable HCC — cancer that cannot be removed by surgery — systemic therapies (treatments that travel through the bloodstream) are now the main option. These include immune checkpoint inhibitors (ICIs), tyrosine kinase inhibitors (TKIs), and monoclonal antibodies (lab-made immune proteins that target specific molecules). **Lenvatinib (LEN)** is a multikinase inhibitor, a drug that blocks several enzymes at once. It targets pathways that HCC needs to grow and spread. One is the vascular endothelial growth factor (VEGF) pathway, which drives new blood vessel formation (angiogenesis). Another is the fibroblast growth factor (FGF) pathway, which drives tumor growth. In the phase III REFLECT trial, lenvatinib produced an average overall survival of **13.6 months** with an acceptable safety profile. **Atezolizumab plus bevacizumab (ATE/BEV)** is a combination therapy that pairs two different mechanisms. Atezolizumab is a PD-L1 inhibitor — it blocks a signal that cancer cells use to hide from the immune system, so the body's immune response against the tumor gets stronger. Bevacizumab is a monoclonal antibody that blocks the VEGF pathway, cutting off the new blood vessels the tumor needs. In the phase III IMbrave150 trial, this combination extended overall survival and progression-free survival compared with sorafenib (an older liver cancer drug), reaching an average overall survival of **19.2 months**. Importantly, the 13.6-month and 19.2-month figures come from two separate trials with different designs, not from a head-to-head comparison. Studies that directly compared lenvatinib with ATE/BEV have produced inconsistent results. That inconsistency is exactly why this meta-analysis was performed. ## How the Researchers Conducted the Review A meta-analysis is a study of studies. It combines the numbers from many separate studies into one overall result, which gives a more reliable answer than any single study alone. This review followed the PRISMA guidelines (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) and was registered in advance in PROSPERO, an international database of planned reviews (ID CRD42024624039). Registering the plan up front helps prevent researchers from cherry-picking results later. The team searched six databases — PubMed, ScienceDirect, Google Scholar, the Cochrane Library, SpringerLink, and Ebsco — for studies published up to July 2024. They used the keywords "unresectable hepatocellular carcinoma" or "unresectable HCC," combined with "Lenvatinib" and "Atezolizumab plus Bevacizumab" and "Efficacy" and "Safety." Two researchers performed the searches under the supervision of a senior investigator. **Studies were included if they met these criteria:** 1. They were randomized controlled trials (RCTs) with or without blinding, published in English, national or international — or observational studies (prospective or retrospective cohorts, case-control, or cross-sectional). 1. They compared lenvatinib with atezolizumab plus bevacizumab. 1. They enrolled adults aged 18 or older with a diagnosis of unresectable HCC. 1. They reported at least one relevant outcome: overall survival (OS), progression-free survival (PFS), objective response rate (ORR), disease control rate (DCR), or side effects. **Studies were excluded if they were:** - Duplicate publications - Missing a control group - Conference abstracts or case reports - Unpublished The researchers also attempted to obtain additional data by contacting the corresponding authors of the original studies. Study quality was graded with the Newcastle-Ottawa Scale (NOS). The NOS scores observational studies from 0 to 9 across three areas. These areas are how the study groups were selected, how comparable the groups were, and how outcomes were assessed. Randomized trials would have been assessed with the Cochrane Risk-of-Bias tool (Version 2). This tool examines the randomization process, adherence to the planned treatment, and handling of missing data. It also examines accuracy of outcome measurement and selection of reported outcomes. Data extracted from each study included: author name, publication year, country, study design, sample size, gender distribution, and age. It also included Child-Pugh class, a scoring system, A to C, for how well the liver is working. It also included ECOG score, a 0-to-5 scale for how much daily activity a patient can manage. It also included BCLC stage, the Barcelona Clinic Liver Cancer staging system. Statistical analysis used Review Manager (RevMan) 5.4 and RStudio version 2024.04.1. Results are reported as hazard ratios (HR) or odds ratios (OR) with 95% confidence intervals (CI). A confidence interval is the range in which the true value probably falls — if the range includes 1.0, the result is considered not statistically significant. Heterogeneity (how much the individual studies disagree with each other) was measured with the chi-squared (X²) and I² tests. When there was no meaningful heterogeneity (defined as P>0.1 or I²<50%), a fixed-effects model was used; otherwise a random-effects model was applied. Publication bias was checked with funnel plots, and the threshold for statistical significance was set at 0.05. ## Which Studies Were Included The initial search turned up **930 records**. The review team then narrowed them down step by step: 1. **39 articles** were removed because they were duplicates. 1. **686 studies** were discarded after the researchers read their titles and abstracts. 1. **165 articles** were excluded after full-text review because they did not match the inclusion and exclusion criteria. 1. **27 studies** were dropped because they lacked the relevant data. 1. **12 articles** were ultimately included in the analysis. ## Who Was in These Studies The 12 included studies enrolled a total of **6,620 participants**. Of these, **3,745** received lenvatinib and **2,875** received atezolizumab plus bevacizumab. The studies were published between 2022 and 2024. Every included study used a cohort design, following a group of patients over time. Most were retrospective. This means the researchers looked back at medical records rather than enrolling patients in advance. Only one study had a prospective design (following patients forward in time). Three studies were multicenter, meaning they collected data from more than one hospital or institution. Six studies came from Japan, two from Korea, and one from China. Most patients in the pooled studies had Child-Pugh class A, meaning their liver function was relatively preserved. Most also had an ECOG score of 0, meaning they were fully active and able to carry out normal daily activities without restriction. By BCLC category, most patients fell into category B (intermediate-stage disease). Quality scores on the Newcastle-Ottawa Scale ranged from **7 to 9**, which the authors describe as indicating high-quality data across all included studies. Studies were reported by Kimura (2024, Japan), Park (2024, Korea), Persano (2023, multicenter), Maesaka (2022, Japan, prospective), Gardini (2023, multicenter), Hatanaka (2023, Japan), Muto (2023, Japan), Niizeki (2022, Japan), Kim (2022, Korea), Hiraoka (2022, Japan), Su (2022, China), and Rimmini (2022, multicenter). ## Overall Survival: No Clear Winner The key finding on overall survival (how long patients lived) is that there was no statistically significant difference between the two treatments. Three studies reported OS. The combined hazard ratio was **0.72** with a 95% confidence interval of **0.44 to 1.18** and a p-value of **0.20**. In plain terms, a hazard ratio below 1.0 would favor lenvatinib, but because the confidence interval crosses 1.0 and the p-value is above 0.05, this difference could easily be due to chance. A random-effects model was used because the studies differed substantially from one another (p = 0.04, I² = 69%, indicating moderate-to-high heterogeneity). The researchers also looked separately at patients whose HCC was linked to a viral infection (hepatitis B or C) and those whose cancer was not virus-related. **In patients with viral infection** (3 studies), the hazard ratio was **0.75** (95% CI: 0.46–1.22, p = 0.24) — no significant difference. A fixed-effects model was used because the studies agreed closely (p = 0.63, I² = 0%). **In patients without viral infection** (3 studies), the hazard ratio was **0.81** (95% CI: 0.25–2.56, p = 0.72) — again, no significant difference. Here the confidence interval is very wide, spanning from strongly favoring lenvatinib to strongly favoring ATE/BEV. A random-effects model was used because the studies disagreed a great deal (p = 0.0005, I² = 87%). ## Progression-Free Survival: No Clear Winner Progression-free survival (PFS) measures how long patients lived without their cancer getting worse. Four studies reported this outcome. The pooled hazard ratio was **0.90** (95% CI: 0.75–1.07, p = **0.23**), meaning no statistically significant difference between lenvatinib and ATE/BEV. Once again, the confidence interval crosses 1.0. The authors report that a fixed-effects model was used, with heterogeneity present (p = 0.04, I² = 68%). **PFS in patients with viral infection** (3 studies): hazard ratio **0.84** (95% CI: 0.60–1.16, p = 0.29) — no significant difference and very consistent results across studies (p = 0.94, I² = 0%). **PFS in patients without viral infection** (3 studies): hazard ratio **0.89** (95% CI: 0.46–1.74, p = 0.74) — no significant difference, with substantial disagreement among the studies (p = 0.04, I² = 68%). Together, these subgroup results suggest that the cause of a patient's liver cancer — viral or non-viral — does not change how these two treatments compare. ## Tumor Response and Disease Control Two additional measures describe how the tumors themselves responded. **Objective response rate (ORR)** is the percentage of patients whose tumor shrank measurably or disappeared. Eight studies reported ORR. The pooled odds ratio was **1.16** (95% CI: 0.86–1.56, p = **0.34**) — no significant difference between the two drugs. A random-effects model was used because the studies varied (p = 0.008, I² = 63%). **Disease control rate (DCR)** is the percentage of patients whose cancer either shrank or stayed stable — that is, did not progress. Seven studies reported DCR. The pooled odds ratio was **1.14** (95% CI: 0.97–1.34, p = **0.12**) — no significant difference, though the result leaned slightly toward lenvatinib. A fixed-effects model was used (p = 0.07, I² = 49%). Because the confidence interval here runs from 0.97 to 1.34, a very small advantage for one drug cannot be ruled out — but the result did not reach statistical significance. ## Serious Side Effects (Grade 3 or Higher) The researchers graded side effects using standard cancer trial categories, where grade 1 is mild and grade 5 is fatal. Serious adverse events of grade 3 or higher are those that interfere significantly with daily life or require medical intervention. Three studies reported grade 3 or higher adverse events. The pooled odds ratio was **1.15** (95% CI: **0.29–4.55**, p = **0.84**) — essentially identical between the two groups. The confidence interval is extremely wide, which reflects how much the individual studies disagreed (p = 0.0007, I² = 86%). This wide range means the analysis cannot rule out meaningful differences in either direction for severe side effects overall. ## Specific Side Effects: Where the Two Treatments Differ This is where the two treatments look most different. Lenvatinib was associated with more of several common side effects, while ATE/BEV was associated with more of one liver-related side effect. **Side effects that were MORE common with lenvatinib:** - **Decreased appetite** (5 studies): odds ratio **2.95** (95% CI: 1.12–7.79, p = **0.03**). This was statistically significant, meaning the odds of losing appetite were roughly three times higher with lenvatinib. Heterogeneity was high (p = 0.0003, I² = 81%). - **Diarrhea** (6 studies): odds ratio **2.61** (95% CI: 2.06–3.32, p **< 0.00001**). Highly significant, with studies agreeing closely (p = 0.62, I² = 49%). - **Fatigue** (5 studies): odds ratio **1.48** (95% CI: 1.27–1.73, p **< 0.00001**). Highly significant and very consistent across studies (p = 0.92, I² = 0%). - **Hand-foot syndrome** (4 studies): odds ratio **7.73** (95% CI: 4.84–12.33, p **< 0.00001**). This was the largest difference found in the entire review. Hand-foot syndrome — also called palmar-plantar erythrodysesthesia — causes redness, swelling, peeling, blisters, and pain on the palms of the hands and soles of the feet. Results were very consistent (p = 0.78, I² = 0%). **Side effects that were MORE common with ATE/BEV:** - **Increased aspartate aminotransferase (AST)** (3 studies): odds ratio **0.44** (95% CI: 0.28–0.69, p = **0.0004**). An odds ratio below 1.0 favors lenvatinib here. This means patients on ATE/BEV were roughly twice as likely to show elevated AST. AST is a liver enzyme that rises when liver cells are damaged. Studies agreed closely (p = 0.41, I² = 0%). **Side effects that were SIMILAR between the two treatments:** - **Hypertension (high blood pressure)** (7 studies): odds ratio **1.39** (95% CI: 0.84–2.28, p = **0.20**). No significant difference. The studies disagreed substantially (p < 0.00001, I² = 85%). - **Proteinuria (protein leaking into the urine)** (5 studies): odds ratio **1.10** (95% CI: 0.75–1.60, p = **0.63**). No significant difference. Moderate inconsistency between studies (p = 0.07, I² = 54%). An important caution applies to all of these numbers. Odds ratios describe relative differences in risk — they do not tell you how many patients out of 100 will actually experience a side effect. The original studies did not consistently report absolute percentages, so this review cannot say "X in Y patients on lenvatinib developed diarrhea." ## Risk of Bias and Publication Bias Publication bias happens when studies with positive or dramatic results are more likely to be published than studies showing no difference. Funnel plot analysis was used to check for this. The analysis flagged three variables as having a potential for publication bias: - Progression-free survival (PFS) - Disease control rate (DCR) - Diarrhea The authors attribute part of this to the relative newness of the research — there simply are not yet enough studies to produce a reliable funnel plot for every outcome. Funnel plots need many studies to be trustworthy, and most comparisons here included only three to eight studies. For most variables, however, the funnel plots showed a symmetric distribution, which suggests a low risk of publication bias. The review's own quality assessment found Newcastle-Ottawa Scale scores of 7 to 9 out of 9 across all included studies. ## What This Means for Patients The headline conclusion is that lenvatinib is **non-inferior** to atezolizumab plus bevacizumab. "Non-inferior" is a research term meaning the treatment is not meaningfully worse — it may be roughly as good. Across four separate measures of how well the drugs fight cancer, the two performed similarly. That said, the individual hazard ratios for OS (0.72) and PFS (0.90) both fell below 1.0, numerically favoring lenvatinib, even though neither reached statistical significance. The authors do not draw a firm conclusion from this. The wide confidence intervals — especially the OS non-viral subgroup range of 0.25 to 2.56 — mean the data simply cannot distinguish a real difference from noise. The safety picture is more clearly separated. Lenvatinib showed higher rates of several bothersome side effects, particularly hand-foot syndrome, diarrhea, fatigue, and decreased appetite. These are often manageable but can affect quality of life and may lead to dose reductions. Atezolizumab plus bevacizumab showed a higher rate of elevated AST, a marker of liver irritation — a concern in patients whose livers are already compromised by cirrhosis. Notably, rates were similar for the most feared outcomes: severe grade 3 or higher adverse events, high blood pressure, and protein in the urine. These are the toxicities doctors watch most closely in this patient population. The practical takeaway is that there is no single "best" drug for everyone. Treatment choice should weigh tumor factors, liver function (Child-Pugh class), and overall fitness (ECOG score). It should also weigh the cause of the liver disease (viral versus non-viral) and each patient's tolerance for specific side effects. ## Limitations: What This Review Could Not Prove Several important limitations temper these findings. **Almost all the evidence was retrospective.** Only one of the 12 studies was prospective. Retrospective studies look backward at existing records, and doctors may have chosen one drug over the other based on how sick a patient was. This kind of selection bias can distort comparisons, and the review's statistical methods cannot fully correct for it. **The evidence was concentrated in specific countries.** Six studies came from Japan, two from Korea, and one from China. Results may not apply equally to patients in other regions with different patterns of liver disease, different genetic backgrounds, and different standards of care. **Several analyses showed high heterogeneity.** The I² statistic measures what percentage of the variation between studies is due to real differences rather than chance. Values above 50% are considered substantial. Some analyses here reached 81%, 85%, 86%, and even 87%. When studies disagree this much, the combined estimate is less trustworthy. **Some pooled estimates were imprecise.** The wide confidence interval for grade 3 or higher adverse events (0.29 to 4.55) means the review genuinely cannot tell whether severe side effects differ between the drugs. **Publication bias could not be fully excluded** for progression-free survival, disease control rate, and diarrhea. **No head-to-head randomized trial exists.** The 13.6-month survival figure for lenvatinib came from the REFLECT trial and the 19.2-month figure for ATE/BEV came from IMbrave150. These are different trials with different patient populations, so directly comparing them is misleading. Only a randomized trial that assigns patients to one drug or the other could answer that question definitively. ## Practical Recommendations If you or someone you care for is facing a decision about first-line treatment for unresectable HCC, here is what this review supports: 1. **Ask your oncology team why a specific drug is being recommended for you.** Because the two options performed similarly on survival and tumor response, the decision often comes down to your liver function. It also comes down to your overall health and which side effects you are best able to tolerate. 1. **Discuss side effect profiles in advance.** If you have a history of gut problems, ask about diarrhea risk with lenvatinib. If your liver function is already fragile, ask how your team will monitor for AST elevation on ATE/BEV. 1. **Report hand-foot syndrome early.** With an odds ratio of 7.73, this is by far the most common lenvatinib-specific problem. Early reports of tingling, redness, or tenderness on the palms and soles allow your team to adjust doses or recommend moisturizers and cooling measures before symptoms become severe. 1. **Expect monitoring for blood pressure and urine protein.** Both drugs can affect these, and both are checked routinely during treatment. 1. **Ask about clinical trials.** The authors note that head-to-head randomized data are still lacking. A trial that directly compares these two regimens would give patients far more certainty than the current retrospective evidence. 1. **Do not stop or change treatment on your own.** These findings describe population-level averages. Your individual situation — tumor stage, liver reserve, prior treatments — matters more than any pooled statistic. ## Frequently Asked Questions ### What is unresectable hepatocellular carcinoma? Hepatocellular carcinoma is the most common type of primary liver cancer, starting in the liver's main cells. It is called unresectable when the tumor cannot be removed by surgery. For these patients, systemic treatments that travel through the bloodstream are the main option, including lenvatinib and the combination of atezolizumab plus bevacizumab. ### How do lenvatinib and atezolizumab plus bevacizumab compare for survival? In a review of 12 studies and 6,620 patients, overall survival and progression-free survival were similar between lenvatinib and atezolizumab plus bevacizumab, with no statistically significant differences. The combined hazard ratios were 0.72 for overall survival and 0.90 for progression-free survival, but both confidence intervals crossed 1.0, meaning the differences could be due to chance. ### Which treatment has more side effects? Both treatments caused similar rates of serious side effects, high blood pressure, and protein in the urine. However, lenvatinib was linked to more loss of appetite, diarrhea, fatigue, and hand-foot syndrome. Atezolizumab plus bevacizumab was linked to more cases of raised aspartate aminotransferase. Aspartate aminotransferase is a liver enzyme that rises when liver cells are injured. ### Does it matter if my liver cancer is caused by hepatitis B or C? The review looked separately at patients whose cancer was linked to viral infection and those whose cancer was not. In both groups, overall survival and progression-free survival were similar between lenvatinib and atezolizumab plus bevacizumab, with no statistically significant differences. This suggests the cause of the liver cancer does not change how these two treatments compare. ### What does a hazard ratio of 0.72 mean? A hazard ratio compares the risk of an event, such as death, between two treatments. A value below 1.0 would favor lenvatinib, but the 95% confidence interval was 0.44 to 1.18, which crosses 1.0. Because the interval includes 1.0 and the p-value was 0.20, the difference in overall survival could easily be due to chance. ### Should I ask my doctor about clinical trials? Yes. The review authors note that no head-to-head randomized trial directly compares lenvatinib with atezolizumab plus bevacizumab. The 13.6-month and 19.2-month survival figures come from two separate trials with different designs, so comparing them directly is misleading. A trial that assigns patients to one regimen or the other would give far more certainty. ### When should a patient with inoperable liver cancer (unresectable HCC) seek a second opinion before starting lenvatinib or atezolizumab plus bevacizumab? Lenvatinib and atezolizumab plus bevacizumab produced similar overall survival, progression-free survival, tumor response and disease control. There were no statistically significant differences. The choice often rests on liver function, overall fitness and which side effects a patient can tolerate. A second opinion can help weigh these factors, since lenvatinib carries more loss of appetite, diarrhea, fatigue and hand-foot syndrome, while the combination carries more raised AST. Seeking one is reasonable before committing to a first-line regimen, particularly when liver function is fragile or the tumor and cirrhosis details are complex. Diagnostic Detectives Network provides independent expert second opinions. ## Source Information **Original article title:** Efficacy and Safety of Lenvatinib versus Atezolizumab Plus Bevacizumab in the Treatment of Unresectable Hepatocellular Carcinoma: A Systematic Review and Meta-Analysis. **Authors:** Ni Putu Sri Indrani Remitha, Ni Putu Rista Pradnya Dewi, Komang Wira Ananta Kusuma, I Gede Aswin Parisya Sasmana, I Gede Putu Supadmanaba, Dwijo Anargha Sindhughosa, and I Ketut Mariadi **Affiliations:** Faculty of Medicine, Udayana University, Denpasar, Bali, Indonesia; Department of Biochemistry, Faculty of Medicine, Udayana University; Division of Gastroenterology and Hepatology, Department of Internal Medicine, Faculty of Medicine, Udayana University; and Division of Gastroenterology and Hepatology, Department of Internal Medicine, Faculty of Medicine, Udayana University / Prof. Ngoerah Hospital, Bali, Indonesia **Journal:** Asian Pacific Journal of Cancer Prevention (Asian Pac J Cancer Prev), Volume 26, Issue 5, pages 1529–1542 **DOI:** 10.31557/APJCP.2025.26.5.1529 **Publication timeline:** Submitted January 2, 2025; accepted May 15, 2025 **Registration:** PROSPERO ID CRD42024624039 **Keywords:** Atezolizumab, bevacizumab, hepatocellular carcinoma, lenvatinib *This patient-friendly article is based on peer-reviewed research. It is intended for educational purposes and does not replace personalized medical advice. Treatment decisions for hepatocellular carcinoma should always be made in consultation with a qualified oncology and hepatology team.* --- Publisher: Diagnostic Detectives Network (https://diagnosticdetectives.com) — independent multi-expert medical second opinions, worldwide, private-pay. Author byline: Anton Titov, MD, PhD. Contact: https://diagnosticdetectives.com/pages/contact Canonical page: https://diagnosticdetectives.com/products/lenvatinib-vs-atezolizumab-plus-bevacizumab-for-liver-cancer-that-cannot-be-surgically-removed-a-patient-friendly-guide-to-a-6-620-person-meta-analysis