Impact of Comorbidity and Drug Use Patterns on Recurrence and Mortality Risks in Primary Hepatocellular Carcinoma: A Chinese Cohort Network Analysis Study

Background & AimsComorbidities in hepatocellular carcinoma (HCC) impair liver function, limit treatment choices, disrupt drug metabolism, and adversely impact survival. This study identifies comorbidity and multidrug combination patterns to evaluate their relationships with HCC outcomes. MethodsThis cohort study comprised 15,998 patients (62.7% male, age at HCC diagnosis: 69.2 [IQR: 58.7-78.7] years old) from the Hong Kong Hospital Authority, enrolled between January 2008 and December 2019 with follow-up until May 2024. Primary endpoints were HCC recurrence and liver cancer-related mortality, while secondary endpoints included cancer-related mortality, non-cancer-related mortality, and all-cause mortality. We identified the most frequent double, triple, quadruple, quintuple patterns of both comorbidities and multidrug combinations using UpSet plots, then examined their associations with clinical outcomes using network analyses. Risk factors were evaluated through Cox proportional hazards regression models, with robustness confirmed via sensitivity analyses and subgroup assessments. ResultsIn this cohort, 93.0% of patients had at least one pre-existing comorbidity (mean 3.0 per patient), of whom 87.8% (N=13,084) had 1-5 comorbidities. Diabetes mellitus, hypertension, and existing chronic liver diseases were the predominant comorbidities across all combinations, followed by colorectal cancer and renal diseases. Males demonstrated greater comorbidity burden, particularly involving metabolic disorders and advanced liver diseases. Notably, males with moderate-to-severe liver diseases had elevated HCC recurrence risk compared to females. Multidrug use rate among those with HCC recurrence was 79.5% (N=9999), of whom 25.7% with double-drug exposure, 13.8% with triple-drug exposure. Anti-diabetic drugs, ACEI/ARB, and diuretics for heart failure, anti-cancer drugs, and NSAIDs were the most common combination use. Both adjusted Cox regression models and subgroup analyses confirmed multi-comorbidity as a robust predictor of clinical outcomes. ConclusionOur gender and age-specific characterization of comorbidity and multidrug patterns in HCC provides clinicians and patients with actionable insights to guide proactive management and patient-centered care, enabling earlier comorbidity detection and prevention strategies. HighlightsO_LIA Chinese population-level characterization of comorbidity profiles, multidrug combinations, and their interactions in HCC. C_LIO_LIDiabetes mellitus, hypertension, and existing chronic liver diseases were significant risk factors of HCC recurrence and mortality. C_LIO_LINetwork analyses revealed sex/age/CCI-stratified associations between comorbidity-drug clusters and clinical outcomes. C_LIO_LIIdentified comorbidity and multi-drug patterns showed prognostic utility for HCC recurrence and mortality risk prevention. C_LI Research in contextO_ST_ABSEvidence before this studyC_ST_ABSAccording to our PubMed search (without language restriction) on March 30, 2025, using ("hepatocellular carcinoma" and "comorbidity" and "network") or ("hepatocellular carcinoma" and "drug" and "network") as search criteria, we found only two preliminary studies. However, they do not consider the heterogeneity of the network, nor the relationship between comorbidities or drugs and outcomes such as HCC recurrence or all-cause mortality. Therefore, to the best of our knowledge, there is no research study investigating age, sex, and CCI stratified network of multimorbidity and multi-drugs exposure, nor exploring the associations between co-comorbidities or co-drugs exposure and follow-up HCC recurrence and mortality risk. Added value of this studyWe proposed a dual perspective (comorbidity and drug) on primary and secondary outcomes by exploring the most frequent comorbidity and multidrug patterns in terms of patients initially diagnosed with HCC. In addition, we explored heterogeneity of comorbidities and drugs with outcome based on sex, age, and CCI. Followed by identifying risk factors associated with outcomes. Impact and implicationsUsing Chinese population-based cohort data, we implement visualization and network analysis to identify comorbidity and multidrug combinations in patients with initial HCC. It can encourage health-care servers and people to provide proactive search for the presence of comorbidities and supply patient-centered treatment. Clinicians can consider a patients co-morbidities and medication history when evaluating patients with HCC, leading to safer and more effective treatment regimens. Graphical illustration abstracts O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=153 SRC="FIGDIR/small/25330803v1_ufig1.gif" ALT="Figure 1"> View larger version (62K): org.highwire.dtl.DTLVardef@1679a2org.highwire.dtl.DTLVardef@18ba7dcorg.highwire.dtl.DTLVardef@1b972d9org.highwire.dtl.DTLVardef@2334a1_HPS_FORMAT_FIGEXP M_FIG C_FIG