Impact of Diabetes Mellitus and Obesity Comorbidities on Survival Outcomes after Hepatocellular Carcinoma Resection: A Multicenter Retrospective Study

Hepatocellular carcinoma (HCC) is the sixth most prevalent malignancy and the third leading cause of cancer-related deaths globally [1]. Hepatic resection is the standard curative treatment for early stage HCC [2]. However, postoperative long-term survival remains unsatisfactory because of the high recurrence rate [3]. As previously described, HCC recurrence is considered to be derived from primary tumor spread and newly developed tumors [4]. Therefore, establishing a treatment strategy for both recurrences is necessary to improve long-term outcomes after HCC resection.

Obesity and diabetes mellitus (DM), which are closely associated with metabolic syndrome development, are well known to induce steatotic liver diseases, potentially causing liver cirrhosis and HCC development [5, 6]. Recently, a new concept named metabolic dysfunction-associated steatotic liver disease (MASLD) has replaced nonalcoholic fatty liver disease. MASLD is diagnosed by the presence of either of the cardiometabolic risk factors regardless of other liver disease comorbidities [5]. Obesity and DM are considered to increase liver carcinogenesis even in the presence of other liver disease comorbidities [6‒8]. Therefore, obesity and DM are assumed to be risk factors for postoperative HCC recurrence. However, no studies revealed consistent results on the risk of postoperative recurrence of obesity and DM. Furthermore, obesity and DM are mutually related in their pathogenesis and progression of steatotic liver disease [9]. Several studies revealed that obesity and DM have a synergistic effect on HCC development [7, 10, 11]. A previous study reported that the coexistence of obesity and DM (compared to no obesity and no DM) increased the risk of HCC by >6 times, whereas the hazard ratios (HRs) for obesity alone and DM alone were 2.1 and 1.4, respectively [11]. The synergistic effects of obesity and DM should be considered in HCC recurrence analysis. However, a few reports evaluated the association of the synergistic effect of obesity and DM with recurrence and survival after HCC resection. The current study evaluated the association of obesity and DM comorbidity with recurrence and survival after HCC resection, considering their synergistic effects.

Data were collected from seven medical centers specializing in liver resection surgeries for HCC. This study included patients undergoing initial and curative hepatic resection for solitary HCC without vascular invasion based on computed tomography from January 2007 to December 2019. We identified 1,653 patients at the seven centers. Among them, we excluded 9 patients who had a surgery-related death within 30 days postoperatively. This study enrolled the remaining 1,644 patients. The Ethics Review Committees of the participating institutions approved this study (approval number: 2022-121; Osaka Metropolitan University). All patients or their family members at each institution provided informed consent using an opt-out form.

The metabolic syndrome-related factors assessed were obesity, DM, dyslipidemia, and hypertension history. Obesity is a body mass index (BMI) of ≥25 kg/m² following the criterion for adult Asian populations [12]. DM is a fasting serum glucose of ≥126 mg/dL or the use of antidiabetic [13]. Dyslipidemia is a serum triglyceride of ≥150 mg/dL or the use of indicated medications [14]. A history of hypertension was determined by interview.

HCC-specific tumor marker levels were measured, and ultrasonography or dynamic computed tomography was performed every 3 months. Recurrence was a new tumor lesion appearance with radiological features of HCC and/or tumor marker elevation. The patient received further treatment by repeat hepatic resection, radiofrequency ablation or percutaneous ethanol injection therapy, transcatheter arterial chemoembolization, drug therapy, or other modalities as indicated when recurrence was detected.

The histological classification of the tumor and background liver damage was evaluated according to the Liver Cancer Study Group of Japan system [15].

The χ² test was used to compare categorical variables. The Kruskal-Wallis test was used to compare continuous data. The Kaplan-Meier method was used to evaluate recurrence-free survival (RFS), overall survival (OS), and cumulative recurrence rates in the early and late phases. The inflection point is 2 years postoperatively [16]; thus, recurrence within and beyond 2 years postoperatively was classified as early or late recurrence. The late recurrence rate peaks at 5 years postoperatively [17]. Additionally, cumulative recurrence rates after 5 years postoperatively were assessed. A Cox proportional hazard regression model was used to compare RFS, OS, early cumulative recurrence rate, and late cumulative recurrence rate among patients with or without obesity and with or without DM. Participants were limited to 910 and 393 who were confirmed to be recurrence-free at 2 years and 5 years postoperatively for late cumulative recurrence rate analysis, respectively. The multivariate Cox regression models were adjusted for the following covariates: age, sex, American Society of Anesthesiologists physical status classification (ASA; <3 or ≥3), comorbidity of obesity and DM, hypertension, dyslipidemia, alcohol use history (alcohol consumption of >60 g/day for >5 years), hepatitis B surface antigen positivity, preoperative hepatitis C virus (HCV) infection status (negative anti-HCV antibody [HCV-Ab], positive preoperative HCV-Ab with sustained virological response [SVR] by interferon-based treatment or direct acting antiviral agents [DAAs], and positive HCV-Ab without SVR), alanine aminotransferase (≤30 or >30 IU/L), platelet count (<10 × 10⁴ or ≥10 × 10⁴ mm³), α-fetoprotein (≤20 or >20 ng/mL), Child-Pugh grade (A or B), major hepatectomy (≥3 segments), intraoperative blood loss (≤1,500 or >1,500 g), surgical approach (open or laparoscopic approach), years of surgery (2007–2010, 2011–2014, and 2015–2019), tumor differentiation degree (poorly or well-differentiated/moderate), liver cirrhosis, tumor size, macroscopic satellite nodule, and microvascular invasion. All statistical analyses were performed using Statistical Package for the Social Sciences version 22.0 software (SPSS, Chicago, IL, USA) and EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan). A p value of <0.05 indicated significance.

We identified 860 patients without obesity and DM (OB[−]DM[−] group), 276 patients with obesity and without DM (OB[+]DM[−] group), 315 patients without obesity and with DM (OB[−]DM[+] group), and 202 patients with obesity and DM (OB[+]DM[+] group). Table 1 shows the background characteristics of the four groups.

Figure 1a and b show the postoperative RFS and OS in the four groups. Their corresponding 5- and 10-year RFS were 44% and 32%, 48% and 23%, 44% and 28%, and 33%, and 21%, and the corresponding 5- and 10-year OS were 71% and 52%, 74% and 56%, 68% and 48%, and 68% and 40%, respectively. The unadjusted Cox proportional hazards regression analysis revealed the unadjusted HRs of reduced RFS of 1.01 (95% confidence interval [CI]: 0.84–1.21; p = 0.90), 1.01 (95% CI: 0.85–1.21; p = 0.87), and 1.24 (95% CI: 1.02–1.51; p = 0.032) in the OB(+)DM(−), OB(−)DM(+), and OB(+)DM(+) groups, respectively, compared with the OB(−)DM(−) group. Further, the corresponding unadjusted HRs of reduced OS were 0.86 (95% CI: 0.67–1.07; p = 0.17), 1.09 (95% CI: 0.88–1.36; p = 0.41), and 1.27 (95% CI: 1.00–1.60; p = 0.047), respectively.

The multivariate Cox proportional hazards regression analysis revealed that the adjusted HRs of reduced RFS was 1.10 (95% CI: 0.91–1.33; p = 0.31), 0.94 (95% CI: 0.78–1.12; p = 0.48), and 1.24 (95% CI: 1.01–1.54; p = 0.045) in the OB(+)DM(−), OB(−)DM(+), and OB(+)DM(+) groups, respectively. Additionally, the corresponding adjusted HRs of reduced OS in the OB(+)DM(−), OB(−)DM(+), and OB(+)DM(+) groups were 0.93 (95% CI: 0.73–1.19; p = 0.57), 0.97 (95% CI: 0.78–1.21; p = 0.76), and 1.38 (95% CI: 1.07–1.77; p = 0.013), respectively (Table 2).

Recurrence within 2 years was reported in 913 patients. The 2-year cumulative recurrence rates in the early phase were 39%, 35%, 35%, and 41% in the OB(−)DM(−), OB(+)DM(−), OB(−)DM(+), and OB(+)DM(+) groups, respectively. The unadjusted HRs of increased early recurrence rates of the OB(+)DM(−), OB(−)DM(+), and OB(+)DM(+) groups were 0.87 (95% CI: 0.69–1.10; p = 0.26), 0.89 (95% CI: 0.71–1.11; p = 0.28), and 1.04 (95% CI: 0.81–1.34; p = 0.75), respectively (Fig. 2a). Postoperative late recurrence beyond 2 years was observed in 333 patients. The corresponding 5- and 10-year cumulative recurrence rates beyond 2 years postoperatively in the OB(+)DM(−), OB(−)DM(+), and OB(+)DM(+) groups, respectively, were 28% and 49%, 27% and 66%, 31% and 56%, and 43% and 64% in the four groups, respectively. The unadjusted HRs of increased recurrence rates beyond 2 years postoperatively were 1.30 (95% CI: 0.97–1.73; p = 0.080), 1.27 (95% CI: 0.95–1.69; p = 0.10), and 1.72 (95% CI: 1.24–2.37; p = 0.0010) in the OB(+)DM(−), OB(−)DM(+), and OB(+)DM(+) groups, respectively (Fig. 2b). Postoperative recurrence beyond 5 years was observed in 88 patients. The unadjusted HRs of increased recurrence rates beyond 5 years postoperatively were 2.88 (95% CI: 1.74–4.78; p < 0.001), 1.81 (95% CI: 1.02–3.21; p = 0.042), and 1.77 (95% CI: 0.85–3.69; p = 0.13) in the OB(+)DM(−), OB(−)DM(+), and OB(+)DM(+) groups, respectively (Fig. 2c).

The multivariate Cox proportional hazards regression analysis revealed the adjusted HRs of increased early recurrence rates of 1.00 (95% CI: 0.78–1.27; p = 0.97), 0.83 (95% CI: 0.66–1.05; p = 0.12), and 1.10 (95% CI: 0.84–1.44; p = 0.49) in the OB(+)DM(−), OB(−)DM(+), and OB(+)DM(+) groups, respectively. Additionally, the corresponding adjusted HRs of increased recurrence rates beyond 2 years postoperatively were 1.25 (95% CI: 0.92–1.69; p = 0.15), 1.11 (95% CI: 0.83–1.50; p = 0.49), and 1.51 (95% CI: 1.06–2.15; p = 0.024), respectively. Moreover, the corresponding adjusted HRs of increased postoperative recurrence rates beyond 5 years were 3.83 (95% CI: 2.17–6.73; p < 0.001), 1.95 (95% CI: 1.04–3.66; p = 0.037), and 2.53 (95% CI: 1.02–6.31; p = 0.046), respectively (Table 3).

Online Supplementary Table 1 shows the distribution of causes of death. The proportions of cancer-related death, liver-related death, and other causes were 54.7%, 11.7%, and 33.6% in the OB(−)DM(−) group, 58%, 14.8%, and 27.2% in OB(+)DM(−) group, 48.7%, 14.8%, and 36.5% in OB(−)DM(+) group, and 42.9%, 17.6%, and 39.5% in OB(+)DM(+)group, respectively.

Online Supplementary Table 2 shows the frequency of intrahepatic and extrahepatic recurrence and their treatment at the first recurrence. The proportions of intrahepatic, extrahepatic, and both intrahepatic and extrahepatic recurrences were 86%, 7.9%, and 6.1% in the OB(−)DM(−) group, 87.9%, 6.4%, and 5.7% in the OB(+)DM(−) group, 88.5%, 9.2%, and 2.3% in the OB(−)DM(+) group, and 89.7%, 7.1%, and 3.2% in the OB(+)DM(+) group, respectively. Curative treatment, transarterial chemoembolization, drug therapy, and other treatments were selected with similar frequency for HCC recurrence among the four groups.

Multivariate analysis revealed that patients with obesity and DM had an approximately 1.3-fold increased risk for RFS and OS compared with those without obesity and DM. No significant difference in the early recurrence rate was observed between the OB(−)DM(−) and OB(+)DM(+) groups, considering the timing of recurrence, but the OB(+)DM(+) group demonstrated an approximately 1.5-fold increased risk for late recurrence beyond 2 years postoperatively compared with the OB(−)DM(−) group. Further, the recurrence risks beyond 5 years postoperatively increased 3.8-fold, 2.0-fold, and 2.5-fold in the OB(+)DM(−), OB(−)DM(+), and OB(+)DM(+) groups, respectively.

Intrahepatic recurrences are classified into intrahepatic metastasis and multicentric recurrence. Intrahepatic metastasis may originate from primary cancer seeding via the portal vein, whereas chronic hepatitis or liver cirrhosis is considered to cause multicentric recurrence [4]. Previous reports described that early recurrence within 2 years may be from intrahepatic metastasis, whereas late recurrence occurring >2 years postoperatively was mainly derived from multicentric carcinogenesis, which is a result of liver inflammation and fibrosis [4, 16]. This study identified tumor-related factors, such as high AFP level, tumor size, macroscopic satellite nodule, and microvascular invasion, as risk factors for early recurrence, whereas liver-related factors, such as high alanine aminotransferase level, low platelet count, and liver cirrhosis, were determined as risk factors for late recurrence beyond 2 years postoperatively. This study revealed the coexistence of obesity and DM as a risk factor for the recurrence beyond 2 years postoperatively. Further, obesity and DM every single comorbidity were determined as risk factors for the recurrence beyond 5 years postoperatively. The coexistence of obesity and DM would increase the multicentric recurrence risk and thereby reduce survival outcomes in this study, considering the relationship between multicentric carcinogenesis and late recurrence.

Obesity was closely associated with DM development and progression [18]. Adipose tissue remodeling under obesity conditions would increase the release of pro-inflammatory adipokine by adipocytes, which induces the insulin resistance of DM. Long-term systemic inflammation in patients with obesity also induces the development of various cancers, such as breast, endometrial, colon, prostate, and esophageal cancers [19]. Insulin resistance in obesity may promote hepatocarcinogenesis [20]. Insulin increases insulin-like growth factor (IGF)-1 production in the liver. IGF-1 influences HCC proliferation and progression through anti-apoptosis, increased angiogenesis via vascular endothelial growth factor production, and improved cell proliferative activity [21]. Inflammatory cytokines, such as tumor necrosis factor-alpha and interleukin-6, have been promoted hepatocarcinogenesis. Interleukin-6 exerts its effects on cell proliferation and anti-apoptosis, whereas TNFα activates c-Jun-NH2-terminal kinase-1, nuclear factor-kappaB, mammalian target of rapamycin, and extracellular signal-regulated kinase, which have carcinogenic effects [21]. Ohki et al. revealed that a BMI of >25 kg/m² was an independent risk factor for HCC development in patients with chronic hepatitis C, with a relative risk of 1.86 [22]. A large population-based study of >1.2 million men revealed that a higher BMI in young adulthood was significantly associated with an increased risk of HCC [23]. JA Davila et al. revealed that DM increased the risk of HCC development by 2.5 times in a population-based case-control study for the USA [24]. Furthermore, N’Kontchou G, et al. revealed that the coexistence of obesity and DM increased the risk of developing HCC than obesity or DM alone in alcoholic or HCV-related cirrhosis [11]. Therefore, the carcinogenic mechanisms of obesity and DM may have synergistically increased the risk of late recurrence postoperatively. Another possible mechanism for the increased risk of late recurrence is the progression of steatotic liver disease. Hepatic steatosis caused lobular inflammation and hepatocellular injury, which promoted hepatocarcinogenesis [25]. The coexistence of obesity and DM may promote hepatocarcinogenesis through hepatic steatosis and increase late recurrence in this study. However, pathological evaluation along with pre- and postoperative assessment using imaging results is difficult for hepatic steatosis in this study because of the long-term, multicenter, retrospective design. Further investigation through prospective studies is warranted.

This study did not determine obesity and DM alone as risk factors for late recurrence beyond 2 years postoperatively but identified them as risk factors for the recurrence beyond 5 years postoperatively. This might be because the multicentric carcinogenic potential in obesity or DM alone was lower than the coexistence of both in synergistic hepatocarcinogenesis [11]. The peak of late recurrence is approximately 5 years postoperatively [17]. Limiting HCC recurrences to those beyond 5 years postoperatively when most recurrences are considered as multicentric recurrences [17] might indicate that obesity or DM alone is a risk factor for HCC recurrence. Moreover, this might be due to the change in body weight postoperatively and DM treatment. Lifestyle interventions, such as weight loss and exercise, were related to reducing the risk of several cancers, including breast, lung, bowel, and kidney cancers [26]. A previous study revealed that control of diabetes would reduce postoperative recurrence in patients with HCV-related HCC [27]. Collection of the clinical information on lifestyle intervention and diabetic control postoperatively and their effect assessment on HCC recurrence is difficult due to multicenter retrospective studies. Obesity and DM management might have affected the analysis. Meanwhile, the coexistence of obesity or DM alone was determined as a risk factor for the recurrence beyond 5 years postoperatively. Therefore, postoperative continuous surveillance for HCC recurrence would be required even after 5 years for patients with obesity or DM.

This study revealed that patients with positive HCV-Ab who did not achieve SVR preoperatively exhibited an increased risk for HCC recurrence, whereas those with positive HCV-Ab who achieved SVR preoperatively demonstrated a reduced risk for HCC recurrence and more favorable OS postoperatively compared with those with negative HCV-Ab. Overall, 55.9% patients with negative and 45% with positive HCV-Ab who achieved SVR preoperatively reported obesity or DM. HCV infection negatively affects glucose metabolism and increases insulin resistance while HCV eradication reduces fasting glucose levels and glycosylated hemoglobin values [28]. SVR achievement might reduce HCV-related hepatocarcinogenesis along with obesity and DM-related hepatocarcinogenesis. Meanwhile, SVR can be obtained in almost all patients with HCV due to DAA administration in recent years [29]. Clinical information regarding postoperative SVR achievement by HCV treatment, including DAAs, was not obtained in this study because of the retrospective multicenter study design although 31.1% of all patients demonstrated positive HCV-Ab but had not achieved SVR preoperatively. Therefore, evaluating the effect of postoperative SVR achievement for HCV on the risk of HCC recurrence and OS would be necessary in the future.

The 2011−2014 and 2015−2019 surgery periods exhibited reduced risks for RFS and late recurrence beyond 2 years postoperatively. The 2015−2019 surgery period demonstrated better OS. Recent advancements, such as DAAs therapy for HCV [29] and the emergence of various molecular targeted therapies and immune checkpoint inhibitors for HCC improved the treatment for liver diseases [30]. The advances in HCV and HCC treatment might be reasons for the reduced risk of recurrence and survival in recent years of surgery.

Patients with obesity and DM demonstrated a more unfavorable OS than those without obesity and DM. Obesity and DM are considered risk factors for developing hypertension, cardiovascular disease, and several malignancies, which are potentially life-threatening diseases [31]. However, cancer-related etiology is the most prevalent cause of death in patients with obesity and DM. Patients in the OB(+)DM(+) group received curative treatment for HCC recurrence, primarily intrahepatic, as predominantly as in the other three groups. Furthermore, the OS curve in the OB(+)DM(+) group is comparable with those of the other three groups up to 8 years postoperatively. This might indicate that treatment of HCC recurrence in the OB(+)DM(+) group was effective and contributed to long-term prognosis. Meanwhile, a previous study revealed that postoperative diabetic control might decrease HCC recurrence postoperatively [27]. Bariatric surgery for patients with severe obesity reduces HCC incidence by improving nonalcoholic steatohepatitis [21]. Higher physical activity levels were related to a reduced risk of development of liver cancer [21]. Therefore, a further therapeutic approach for obesity and DM might enhance postoperative survival outcomes in patients with obesity and DM.

This study had several limitations. First, potential bias in background characteristics, such as microscopic liver inflammation and steatosis, among the four groups could not be completely excluded despite adjusting for background clinicopathological factors by multivariate analyses because of its retrospective nature. Second, the long study duration may have influenced the current results, as medical techniques and treatment approaches may have changed over time. Third, detailed clinical information regarding postoperative management for HCV infection, obesity, DM, and alcohol abstinence in patients taking alcohol along with changes in hepatic steatosis pre- and postoperatively as evaluated with imaging results were not obtained because of the long-term, multicenter, retrospective study design. Finally, several pharmacological treatments for MASLD may prevent HCC development [32], but detailed clinical information regarding pharmacological treatment, including antidiabetic, antihypertensive, and dyslipidemic drugs, was not included in this study. However, the coexistence of obesity and DM would increase the late recurrence risk of HCC and thereby reduce survival outcomes. The present results might be crucial for establishing a treatment strategy and surveillance for HCC with obesity and DM.

In conclusion, the coexistence of obesity and DM increased late recurrence and worsened prognosis in patients with HCC undergoing hepatic resection. The results help surgeons develop possible different surveillance protocol and need to focus on diabetes/obesity control during life-long surveillance for patients with HCC.

The Human Research Committee of the Institution approved the study protocol (Approval No.: 2022-121; Osaka Metropolitan University). Opt-out informed consent protocol was used for use of participant data for research purposes. This consent procedure was reviewed and approved by [Osaka Metropolitan University Medical Research Ethics Review Committee], Approval No. [2022-121], date of decision [May/19/2023].

The authors have no conflicts of interest to declare.

We have no funding of any research relevant to our study for study design, execution and analysis, and manuscript conception, planning, writing and decision to publish.

Conception and design: Shinkawa and Ishizawa. Data acquisition and analysis: Kaibori, Ueno, Yasuda, Ikoma, Aihara, Nakai, Kinoshita, Kosaka, Hayami, Matsuo, Morimura, Nakajima, and Nobori. Data analysis and interpretation: Shinkawa, Kaibori, and Ueno. Drafting of the manuscript: Shinkawa. Critical revision of the manuscript: Ishizawa. Final approval: all authors have approved the final version of the manuscript.

All data generated or analyzed during this study are included in this article and its online supplementary material files. Further inquiries can be directed to the corresponding author.