A multicentre, multinational study of clinical characteristics and prognosis of hepatocellular carcinoma

462 Research article EMHJ – Vol. 29 No. 6 – 2023 Abstract Background: Hepatocellular carcinoma (HCC) is a significant health problem, and the associated mortality rate is increasing. Aim: We aimed to determine the clinical characteristics and prognosis for HCC in member countries of the OncoBridge Study Group. Methods: We recruited 630 patients diagnosed with HCC between 2013 and 2019 from 4 countries (Türkiye, Russia, Georgia, and Greece). Univariate and multivariate analyses were conducted to investigate clinical and laboratory prognostic factors. Receiver operating characteristic (ROC) analysis was used to determine the prognostic value of the neutrophil to lymphocyte ratio (NLR) and alpha-fetoprotein (AFP) value. Results: The 3 most common etiological factors were hepatitis B infection (39.7%), hepatitis C virus infection (17.0%) and non-alcoholic fatty liver disease (9.0%). Median overall survival for the whole group was 25 [95% confidence interval (CI): 15.7–34.2] months. Cut-off values for AFP and NLR were accepted as 200 ng/mL and 3.45, respectively. The area under the ROC curve values for AFP, NLR and NLR+AFP were 0.625 (95% CI: 0.547–0.704), 0.589 (95% CI: 0.512–0.667) and 0.657 (95% CI: 0.583–0.731). From the multivariate analysis, advanced tumour size, lymph node involvement and metastasis (TNM) stage, presence of cirrhosis, high AFP, and high NLR values were associated with poor survival. Conclusion: AFP, NLR, advanced TNM, and presence of cirrhosis may predict prognosis in patients with HCC. Studies involving more countries are needed to corroborate these findings. Keywords: hepatocellular carcinoma, prognosis, neutrophil to lymphocyte ratio, Türkiye, Russia, Georgia, Greece Citation: Dirican A, Uncu D, Sekacheva M, Artaç M, Aladashvil A, et al. A multicentre, multinational study of clinical characteristics and prognosis of hepatocellular carcinoma. East Mediterr Health J. 2023;29(6):462–473. https://doi.org/10.26719/emhj.23.087 Received: 09/10/22; accepted: 08/12/22 Copyright © Authors 2023; Licensee: World Health Organization. EMHJ is an open access journal. This paper is available under the Creative Commons Attribution Non-Commercial ShareAlike 3.0 IGO licence (CC BY-NC-SA 3.0 IGO; https://creativecommons.org/licenses/by-nc-sa/3.0/igo). A multicentre, multinational study of clinical characteristics and prognosis of hepatocellular carcinoma Ahmet Dirican,1 Dogan Uncu,2 Marina Sekacheva,3 Mehmet Artaç,4 Archil Aladashvil,5 Atike Erdogan,6 Muhammet Kaplan,7 Ahmet Alacacıoğlu,8 Ioannis Boukovinas9 and Nazim Turhal10 1Department of Medical Oncology, Medicalpark Hospital, Izmir University of Economics, Izmir, Türkiye (Correspondence to Ahmet Dirican: ahmetdirican@yahoo.com). 2Department of Medical Oncology, Ankara Numune Education and Research Hospital, Health Sciences University, Ankara, Türkiye. 3World-Class Research Center “Digital Biodesign and Personalized Healthcare”, Sechenov First Moscow State Medical University, Moscow, Russia. 4Department of Medical Oncology, Meram Medical Faculty, Necmettin Erbakan University, Konya, Türkiye 5Cancer Research Unit and Oncology Surgery, National Cancer Center, Tbilisi State Medical University, Tbilisi, Georgia. 6Department of Medical Oncology, Medical Faculty, Manisa Celal Bayar University, Manisa, Türkiye. 7Department of Medical Oncology, Medical Faculty, Dicle University, Diyarbakır, Türkiye. 8Department of Medical Oncology, Medical Faculty, Izmir Katip Celebi University, Izmir, Türkiye 9Medical Oncology Unit, Bioclinic Thessaloniki, Thessaloniki, Greece. 10Department of Medical Oncology, Anatolia Medical Center, Istanbul, Türkiye. Introduction According to the surveillance, epidemiology, and end results database, rates for new liver cancer cases increased by an average of 2.1% annually in the past decade (1). Mortality rates increased by an average of 2.4% between 2007 and 2016. The incidence of hepatocellular carcinoma (HCC) varies among ethnic groups and even among regions in the same country (1,2). Although it is more common in Asian and African countries, the incidence is thought to have increased in developed countries because of the increased access to more effective diagnostic screening and immigration to these countries from high risk regions (2). Risk factors for HCC are well defined compared with some other cancers. Viral hepatitis, chronic alcoholism, non-alcoholic fatty liver disease (NAFLD) are among the main factors. Although the frequency of risk factors varies from region to region, on the whole, chronic viral hepatitis is the most common cause (1). The Child–Turcotte–Pugh classification is the oldest assessment tool for evaluating HCC prognosis, but many other prognostic factors have been studied recently because of the limitations of this commonly used scoring system, which is based on liver function tests (3,4). The other prognostic factors include tumour histology, serum alpha-fetoprotein (AFP) level, hepatitis B and C, antiviral therapy, diabetes mellitus and inflammation markers (5–10). In the bone marrow, the production and release of neutrophils increase in response to inflammation (11,12). The chemokines playing a significant role in cancer- related inflammation induce neutrophil release from the bone marrow with the accumulation of neutrophils in peripheral tissue (13). The effects of inflammation 463 Research article EMHJ – Vol. 29 No. 6 – 2023 increases the neutrophil count in the peripheral blood while the lymphocyte count decreases. Consequently, these changes increase the neutrophil to lymphocyte ratio (NLR). Considering that increased NLR may be a marker for cancer-related inflammation, it may also be associated with prognosis. Research has shown that high NLR may be associated with poor prognosis in various types of cancer (14). As part of the OncoBridge Working Group, our aim in this study was to clarify the etiological and prognostic factors of HCC and to provide basic data for health practitioners and healthcare professionals. Methods OncoBridge Study Group The OncoBridge Project started in 2014 with the participation of 8 countries (Azerbaijan, Armenia, Egypt, Greece, Israel, Morocco, Saudi Arabia and Türkiye) with the aim of increasing scientific and educational collaboration on various platforms, including international congresses, cooperative study groups, fellow exchanges, etc. The participants agreed to organize scientific meetings annually in Türkiye. The initiative is open to all other neighbouring countries interested in the aims of the platform and willing to participate. Eventually, representatives of several other countries, including Bulgaria, Georgia, Kazakhstan, Lebanon and Romania, joined the various activities of OncoBridge. Study design and patient cohort This was a retrospective study conducted by the OncoBridge Study Group in which the etiological and prognostic factors of HCC were studied. Patients who were diagnosed with HCC between 2013 and 2019 in 9 centres in 4 countries (Georgia, Greece, Russia and Türkiye) were studied. Patients aged 18 years or older were eligible for enrolment if they had been histologically or radiologically diagnosed according to the European Association for the Study of the Liver consensus recommendations on HCC (15) or had any stage of tumour size lymph node involvement metastasis (TNM). After applying the eligibility criteria, 630 patients were eligible for the final analysis. Data collection An Excel spreadsheet template was used for data collection. Each centre entered their data into that template and the main database was developed. This was then converted and entered into SPSS statistical software. Each patient was represented by her/his initials and given a unique identification number representing their country, centre, researcher and case number systematically. Coding tables were kept separate from the data table from a third party within each centre in order to maintain patient anonymity. Coding rules On the coding table for patient identification, the code for the patient started from 1 and ended with the number of the last participant at each centre (e.g. 1, 2, 3, ... 205). The subsequent parts of the codes were the initial of the participating country, the initials of the participating centre and the initials of the researcher and initials of the participant. Each researcher recorded the national ID and hospital ID of the patients on their own computer. These data were not shared in the dataset. Demographic and clinical variables included age at diagnosis, sex, geographical region (rural or urban), liver cirrhosis, viral hepatitis, smoking, alcohol intake, NAFLD and other cirrhosis-related factors, Child–Pugh score, treatment for hepatitis B virus/hepatitis C virus (HBV/ HCV), diabetes and its treatment, statin and aspirin use, blood type, height and weight, sorafenib treatment, TNM staging, local ablative therapy and surgical treatment. Neutrophil and lymphocyte counts and serum AFP levels were also recorded. Ethical approval The study protocol conformed with the ethical guidelines of the 2008 Declaration of Helsinki and was approved by Manisa Celal Bayar University (No. 20478486). Statistical analysis All statistical analyses were done using SPSS, version 22. All statistical assessments were 2-sided. Statistical significance was set at P = 0.05. Overall survival was calculated from the diagnosis of the patient up to either the date of death from any cause or the date of the last follow-up. Differences in cumulative survival were determined using the Kaplan–Meier method and a log-rank test. Univariate and multivariate analyses for survival difference were carried out using the Cox proportional hazards model, and were expressed as hazard ratio and 95% confidence interval (CI). The area under the receiver operating characteristic (ROC) curve for AFP and NLR to determine the prognosis of HCC patients were calculated using SPSS along with the optimal cut-off value, sensitivity, specificity, positive predictive value, and negative predictive value. Binary logistic regression is a statistical analysis that determines how much (if any) variance is explained on a dichotomous dependent variable by a set of independent variables. We used binary logistic regression analysis to determine the combined use of AFP and NLR parameters and consequently to determine the prognosis for HCC. Results Patients The characteristics of the patient cohort are presented in Table 1. The largest group, 453 patients (71.9%), were from Türkiye, 111 (17.6%) were from Russia, 55 (8.7%) from Georgia and 11 (1.7%) from Greece. Hepatocellular carcinoma was more common among male patients. 464 Research article EMHJ – Vol. 29 No. 6 – 2023 Mean age at the time of diagnosis was 60.0 (SD 12.8) years. The 3 most common etiological factors were HBV infection (39.7%), HCV infection (17.8%) and NAFLD (9.0%). Prognostic efficacy of neutrophil to lymphocyte ratio and alpha-fetoprotein in hepatocellular carcinoma patients We used the ROC analysis method to evaluate the prognostic ability at the time of diagnosis, NLR, AFP and the combination (NLR+AFP) in HCC patients. The results are presented in Table 2 and Figure 1. According to the Table 1 Characteristics of patients diagnosed with hepatocellular carcinoma (n = 630) during 2013–2019 from 4 countries in the OncoBridge Project Characteristic Mean (SD) Age (years) 60.0 (12.8) No. (%) Sex Female 136 (21.6) Male 494 (78.4) Centre Türkiye 453 (71.9) Ankara Numune Training and Research Hospital 233 (37) Necmettin Erbakan University 101 (16) Manisa Celal Bayar University 40 (6.3) Dicle University 36 (5.7) Izmir Katip Celebi University 35 (5.6) Anatolia Medical Centre 8 (1.3) Russia First Moscow State Medical University 111 (17.6) Georgia National Cancer Institute, Tbilisi 55.0 (8.7) Greece Bioclinic Thessaloniki 11 (1.7) Residence Rural 136 (21.6) Urban 481 (76.3) Unknown 13 (2.1) Child–Pugh classification A 335 (53.1) B 85 (13.4) C 115 (18.2) Unknown 95 (15.3) TNM stage Stage I-II 100 (15.8) Stage III-IV 344 (54.6) Unknown 186 (29.6) Smoking status Never 158 (25.0)) Former 121 (19.2) Current 292 (46.3) Unknown 59 (9.5) Alcohol drinker None 404 (64.1) Regular 134 (21.2) Unknown 92 (14.7) Presence of cirrhosis Yes 326 (51.7) No 213 (33.8) Unknown 91 (14.5) Diabetes mellitus No 427 (67.7) Characteristic Mean (SD) Yes 132 (20.9) Unknown 71 (11.4) Presence of metabolic syndrome No 358 (56.8) Yes 29 (4.6) Unknown 243 (38.6) Body mass index Underweight or normal (< 25 kg/m2) 158 (25) Overweight (25 to < 30 kg/m2) 86 (13.6) Obese (≥ 30 kg/m2) 57 (9.0) Unknown 329 (52.4) Etiology of liver disease Hepatitis B virus 250 (39.7) Hepatitis C virus 112 (17.8) Hepatitis B/hepatitis C co-infection 9 (1.4) Non-alcoholic fatty liver disease 57 (9.0) Alcohol-related liver disease 20 (3.2) Aflatoxin 14 (2.2) Cryptogenic cirrhosis 13 (2.1) Other 58 (9.2) Unknown 97 (15.4) Treatment Transplant 11 (1.7) Resection 74 (11.7) Local ablative procedures 138 (21.9) Sorafenib 228 (35.6) Chemotherapy 364 (57.8) Mean (SE) Laboratory parameter Alpha fetoprotein (ng/mL) 9418.3 (2092.7) Neutrophils (cells/mm3) 4803.5 (328.4) Lymphocytes (cells/mm3) 1230.2 (65.9) Neutrophil to lymphocyte ratio 3.21 (0.12) SD = standard deviation. SE = standard error. TNM = tumour size, lymph node involvement, metastasis. Table 1 Characteristics of patients diagnosed with hepatocellular carcinoma (n = 630) during 2013–2019 from 4 countries in the OncoBridge Project (concluded) 465 Research article EMHJ – Vol. 29 No. 6 – 2023 ROC analysis, the cut-off value for NLR was 3.45. The cut- off value for AFP was accepted as 200 ng/mL (16). Both AFP and NLR had high specificity (but lower sensitivity) for prognosis; AFP gave a larger area under the ROC curve (0.625, 0.547–0.704; P = 0.003) than NLR (0.589; 0.512– 0.667; P = 0.036). Further evaluation of the combination of NLR+AFP had the highest area under the ROC curve (0.657; 0.583–0.731; P < 0.001) with a significantly higher specificity (88.3%) and a lower sensitivity (29.8%) compared with AFP or NLR alone. Patients who had a high NLR (≥ 3.45) had significantly shorter survival than low NLR (< 3.45) patients (Figure 2). In the low AFP group (AFP < 200 ng/mL) survival was not statistically significant, and in the high AFP group (AFP ≥ 200 ng/mL) median overall survival was 12 months (P < 0.01) (Figure 3). They were grouped as follows to show the relationship between AFP and NLR use and survival: AFP and NLR both higher (combined high group), AFP and NLR both low (combined low group), only patients with high NLR (high NLR group). None of the patients had high AFP only (i.e. without high NLR). The median survival time of HCC patients in the combined high, combined low and high NLR groups were 8.0 (95% CI: 5.232–10.768) months, 15.0 (95% CI: 9.332–20.668) months and 11.0 (95% CI: 8.136– 13.864) months, respectively (Figure 4). Sorafenib treatment outcomes Sorafenib was used in the treatment of 228 patients in 2007–2019. It was administered until disease progression, unacceptable toxicity or withdrawal of consent. The treatment resulted in 5 (2.2%) complete responses, 31 (13.7%) partial responses, 116 (50.4%) with stable disease, and 76 (33.6%) with progressive disease. Median progression-free survival was 5 months (95% CI: 1.5–27) and median overall survival was 9 months (95% CI: 7.2–10.7). Univariate analysis Median overall survival for all patients after diagnosis was 25 months (95% CI: 15.7–34.2) (Figure 5). The 3-year median overall survival rate among TNM stage I–IV patients was 77%, 62%, 34% and 29%, respectively (Log Rank; P < 0.001). The hazard ratios for clinical and laboratory parameters for overall survival among HCC patients were determined using Cox regression analysis (Table 3). The univariate analysis showed that male sex, urban residence, advanced TNM stage, smoking, chronic HBV infection, chronic HCV infection, cirrhosis, diabetes mellitus, metformin use, insulin use, high AFP, high NLR and combined (NLR+AFP) high group were all associated with a high mortality risk while HBV treatment, HCV treatment, statin use, local ablative treatment and surgical treatment were associated with low mortality risk. Age, Child–Pugh grade, alcohol intake, sulfonylurea use, body mass index, aspirin use, chemotherapy treatment, sorafenib treatment and ABO blood type were not associated with overall survival in this analysis. Multivariate analysis The variables found to be associated with prognosis in the univariate analysis were evaluated using the Cox proportional hazard multivariate model (Table 3). The model demonstrated that high NLR was an independent prognostic factor for overall survival (P = 0.048). The other independent prognostic factors for poor overall survival included high AFP value (P = 0.040), high combined score (P = 0.009), cirrhosis (P = 0.035) and advanced TNM stage (P = 0.015). Discussion The distribution of HCC for age, sex and etiology varies by region. Overall, worldwide, the proportion of men with HCC is greater than for women, with a male:female ratio ranging from 2:1 to 4:1. Liver cancer is more common among men in Middle Eastern countries (17,18). The median age at diagnosis tends to peak in the 55–64 years age group, although it may vary by region (1). These age-specific differences are due to differences in the age at time of infection with hepatitis virus in this region. In rural areas, where there is high pollution from insecticides, HCC is more common but other environmental factors may also be contributors (19). In our study, the median age at diagnosis for HCC was 55 years and the male to female ratio was 3.6:1 (males 78.4%, females 21.4%). Unlike in some other research, our HCC patient population had more urban than rural residents, possibly due to the fact that the centres participating in the study were located in big cities. Chronic HBV infection is responsible for 50–80% of HCC cases worldwide (20). However, the ratio is 75–90% in areas where HBV is endemic, such as Türkiye. Like HBV infection, the contribution of HCV infection to HCC varies worldwide. In Europe and the United States of America, HCV is the main risk factor for HCC. However, HCV infection is the second most common infection in Asia. According to research from 12 countries in the WHO Eastern Mediterranean Region and the Middle East, HCC cases in the Islamic Republic of Iran, Lebanon, Türkiye and Yemen were mainly attributable to HBV, while in Table 2 Prognostic efficacy of neutrophil to lymphocyte ratio (NLR) and alpha-fetoprotein (AFP) in hepatocellular carcinoma patients (n = 630) diagnosed during 2013–2019 from 4 countries in the OncoBridge Project Factor Sensitivity (%) Specificity (%) PPV NPV AUROC (95% CI) P NLR 34.4 83.6 88.6 25.5 0.589 (0.512–0.667) 0.036 AFP (ng/mL) 47.3 87.0 86.2 52.1 0.625 (0.547–0.704) 0.003 NLR+AFP 29.8 88.3 89.7 26.9 0.657 (0.583–0.731) < 0.001 PPV = positive predictive value; NPV = negative predictive value; AUROC = area under the receiver operating characteristic (ROC) curve; CI = confidence interval. 466 Research article EMHJ – Vol. 29 No. 6 – 2023 Figure 1 Receiver operating characteristic (ROC) curves for neutrophil to lymphocyte ratio (NLR), alpha-fetoprotein (AFP) and NLR+AFP for the prognosis of hepatocellular carcinoma evaluating prognostic ability at time of diagnosis (2013–2019) among 630 patients from 4 countries in the OncoBridge Project Figure 2 Relationship between neutrophil to lymphocyte ratio (NLR) and su rvival among patients with hepatocellular carcinoma (n = 630) from 4 countries in the OncoBridge Project (mOS = median overall survival) Figure 3 Relationship between alpha-fetoprotein (AFP) and survival among patients with hepatocellular carcinoma (n = 630) from 4 countries in the OncoBridge Project (mOS = median overall survival) 467 Research article EMHJ – Vol. 29 No. 6 – 2023 Pakistan, Saudi Arabia and the North African countries, cases were predominantly related to HCV (21). Infection with HBV or HCV and excessive alcohol intake have been identified in 56.0%, 23.2% and 15.9% of Turkish HCC patients (22–24). Recent research suggests that NAFLD causes the development of HCC, especially in Western European countries and the United States of America (25). In our study, infection with HBV and HCV were the most important reasons for the development of HCC. Since vaccination against HBV is becoming more common, this may lead to more instances where HCV is the cause of HCC over time. Unfortunately, there is currently no vaccine for HCV. Therefore, measures such as better blood screening techniques and other harm reduction programmes will further reduce the risk of HCC. We found that NAFLD was the third most common cause of HCC; the high NAFLD rate may be related to the high rates of diabetes and obesity in our study group. The TNM stage has been shown to be associated with poor prognosis among patients with HCC (26), however, it is the liver function that determines the prognosis of patients with underlying cirrhosis. The prognosis for HCC depends not only on tumour stage but also on liver failure due to cirrhosis. The Okuda and CLIP systems may be more useful than TNM stage to assess prognosis (27). However, irrespective of cirrhosis in our patients, TNM stage was statistically significantly associated with mortality (TNM stage III–IV vs I–II: hazard ratio:3.489, P = 0.015). The accumulating evidence indicates that outcomes for patients with HCC are significantly correlated with the level of tumour-associated inflammation (28,29). While these inflammatory changes lead to tumour growth, normal distribution in haematological parameters is disrupted (lymphopenia, granulocytosis, etc.). It is known that pro-inflammatory factors (IL–1, TNF-α, vascular endothelial growth factor, tissue factor and Figure 4 Survival of patients with hepatocellular carcinoma (n = 630) in accordance with the combined use of neutrophil to lymphocyte ratio (NLR) and alpha-fetoprotein (AFP): combined high group (AFP and NLR both high), combined low group (AFP and NLR both low) and high NLR group (only patients with high NLR) (mOS = median overall survival) Figure 5 Overall survival i n all hepatocellular carcinoma patients (n = 630) from 4 countries in the OncoBridge Project 468 Research article EMHJ – Vol. 29 No. 6 – 2023 Table 3.Univariate and multivariate analysis (Cox regression) of prognostic factors for overall survival among patients (n = 630) diagnosed with hepatocellular carcinoma during 2013–2019 from 4 countries in the OncoBridge Project Variable No. Univariate analysis Multivariate analysis HR (95% CI) P-value HR (95% CI) P-value Age (years) < 65 355 1.055 (0.846–1.316) 0.635 ≥ 65 225 Sex Female 127 1.387 (1.053–1.827) 0.020 Male 453 Residence Rural 124 0.674 (0.525–0.867) 0.002 Urban 444 Child–Pugh grade A 335 1.195 (0.879–1.626)a 0.999 (0.746–1.338)b 0.256 0.994B 85 C 115 TNM stage 1–2 84 3.260 (2.176–4.884) < 0.001 3.489 (1.271–9.580) 0.015 3–4 311 Smoking Non-smoker 141 0.938 (0.703–1.253)c 1.739 (1.316–2.299)d 0.665 < 0.001Ex-smoker 116 Smoker 266 Alcohol None 346 0.954 (0.730–1.245) 0.727 Regular 153 Chronic HBV No 332 1.773 (1.404–2.238) < 0.001 Yes 207 HBV treatment No 160 0.428 (0.299–0.612) 0.001 Yes 104 Chronic HCV No 407 1.430 (1.048–1.952) 0.024 Yes 85 HCV treatment No 376 0.787 (0.538–1.151) 0.217 Yes 62 HBV and HCV No 295 1.277 (0.565–2.884) 0.557 Yes 8 Cirrhosis No 326 1.529 (1.208–1.934) < 0.001 1.684 (1.038–2.731) 0.035 Yes 213 DM No 379 0.942 (0.710–1.248) 0.676 Yes 123 Metformin No 87 2.413 (1.357–4.290) 0.002 Yes 36 469 Research article EMHJ – Vol. 29 No. 6 – 2023 Variable No. Univariate analysis Multivariate analysis HR (95% CI) P-value HR (95% CI) P-value Sulfonylurea No 94 0.490 (0.223–1.074) 0.058 Yes 29 Insulin No 100 2.040 (1.161–3.583) 0.013 Yes 23 BMI Not obese 220 1.180 (0.806–1.726) 0.395 Obese 55 Statin use No 148 0.457 (0.218–0.960) 0.039 Yes 14 Aspirin No 142 0.939 (0.540–1.631) 0.823 Yes 21 Chemotherapy No 134 0.967 (0.745–1.256) 0.803 Yes 308 Surgical resection No 265 0.773 (0.480–1.243) 0.039 Yes 69 Local ablative treatment No 123 0.713 (0.511–0.995) 0.046 Yes 84 Sorafenib No 278 1.205 (0.959–1.513) 0.109 Yes 195 Blood type A 79 0.983 (0.644–1.499) 0.935 Other 134 AFPe Low 124 5.581 (3.802–8.192) < 0.001 1.773 (1.116–2.819) 0.028 High 221 NLRf Low 175 1.753 (1.300–2.364) < 0.001 2.023 (1.012–4.129) 0.014 High 73 Combined score Low 170 1.891 (1.362–2.626) < 0.001 2.128 (1.132–4.896) 0.009 High 56 HR = hazard ratio. CI = confidence interval. HBV = hepatitis B virus. HCV = hepatitis C virus. TNM = tumour size, lymph node involvement, metastasis. aB vs A. bC vs A. cNon-smoker vs ex-smoker. dNon-smoker vs smoker. eHigh AFP: group with AFP ≥ 200 ng/mL. Low AFP: group with AFP < 200 ng/mL. fHigh NLR: group with NLR ≥ 3.45. Low NLR: group with NLR values < 3.45. Table 3.Univariate and multivariate analysis (Cox regression) of prognostic factors for overall survival among patients (n = 630) diagnosed with hepatocellular carcinoma during 2013–2019 from 4 countries in the OncoBridge Project (concluded) 470 Research article EMHJ – Vol. 29 No. 6 – 2023 cancer procoagulant) are secreted by cancer cells (30,31) and cause changes in blood cell count. These changes in hematologic parameters among cancer patients have been studied as prognostic factors among HCC patients (32,33). The NLR can elucidate the complex prognostic information of these 2 conditions (granulocytosis and lymphopenia), and can be a very strong predictor of clinical outcome. Higher levels of NLR are an indication of poor survival for patients with HCC and NLR can be considered a biomarker in making clinical decisions about HCC treatment (14). There have been more studies on AFP than on NLR to determine the prognosis of HCC; high AFP values are associated with poorly differentiated tumour and have been reported to be an independent predictor of survival (34) but there is no generally accepted AFP cut-off value associated with prognosis. Our findings confirmed the prognostic importance of AFP and NLR values. Until now, NLR and AFP have not been evaluated together for diagnosis or prognosis (35). In our study, the prognostic values of AFP and NLR as well as their combinations were evaluated and compared and our data showed that AFP remained a good prognostic marker, and that NLR was also a prognostic marker and was comparable to AFP. Further evaluation of the biomarker combination showed that the combination of AFP and NLR had the highest prognostic accuracy. The area under the ROC curve for this combination was 0.657, with sensitivity 88.3% and specificity 29.8%, indicating prognostic accuracy comparable to AFP and NLR alone. The combination of AFP and NLR is a useful prognostic marker for HCC. We observed that the combined use of these markers was better at predicting survival than their use individually. However, it should be noted that sensitivity for these markers is low. In the Cox regression analysis, sorafenib did not reduce the risk of death. This may be because sorafenib can be used after local ablative and chemotherapy treatments in some of the countries in our study. However, the median overall survival and median progression-free survival value with sorafenib was compatible with the findings of previous research (36–38). In a study by Manghisi et al., median survival following diagnosis ranged from approximately 6 to 20 months in patients with HCC (39). It is obvious that, with targeting therapies and immunotherapies, survival from time of diagnosis will be extended, therefore, the results from long-term survival should be included . Similar to previous findings (40), the median survival among our HCC patients followed up between 2013 and 2019 was 25 months. According to the multivariate analysis, we identified advanced TNM stage and the presence of cirrhosis as independent prognostic factors, and again this agreed with the findings of previous research (41). In conclusion, the risk factors significantly associated with mortality among patients with HCC were advanced TNM stage, cirrhosis and high AFP and NLR levels. Combined use of NLR and AFP was a better prognostic indicator than either of these factors taken separately. Overall survival calculated from the time of diagnosis, and after sorafenib use, among our patients was consistent with previous reports (38,40). Although there were some limitations to this study, it was an important attempt at reflecting the HCC data of the 4 participating countries. The major limitation was the retrospective nature of the study. Another limitation was that we could not access some data for the patients. Lastly, the sample size from each country in our study was not proportionate with the population size; convenience sampling was used to determine the sample size used. There were no other criteria for the sampling method except that the centre be available and willing to participate. Our findings need to be validated by further studies with the participation of more OncoBridge member countries. Funding: None Competing interests: None declared. Étude multicentrique et multinationale des caractéristiques cliniques et du pronostic du carcinome hépatocellulaire Résumé Contexte : Le carcinome hépatocellulaire (CHC) représente un problème de santé majeur, et le taux de mortalité qui lui est associé est en augmentation. Objectifs : Nous avons cherché à déterminer les caractéristiques cliniques et le pronostic du CHC dans les pays membres du groupe d'étude OncoBridge. Méthodes : Nous avons recruté 630 patients diagnostiqués avec un CHC entre 2013 et 2019 dans quatre pays (Géorgie, Grèce, Russie et Türkiye). Des analyses univariées et multivariées ont été menées pour étudier les facteurs pronostiques cliniques et biologiques. L'analyse de la courbe ROC (caractéristique du fonctionnement du récepteur) a été utilisée pour déterminer la valeur pronostique du rapport neutrophiles/lymphocytes (NLR) et du taux d'alpha-fœtoprotéine (AFP). Résultats : Les trois facteurs étiologiques les plus courants étaient l'infection par le virus de l'hépatite B (39,7 %), l'infection par le virus de l'hépatite C (17,0 %) et la stéatose hépatique non alcoolique (9,0 %). La survie médiane 471 Research article EMHJ – Vol. 29 No. 6 – 2023 References 1. Brar G, Greten TF, Graubard BI, Mc Nell T, Petrick J, McGlynn K et al. Hepatocellular carcinoma survival by etiology: a SE- ER-Medicare Database analysis. Commun. 2020 Aug 9;4(10):1541–51. doi:10.1002/hep4.1564 2. Sharma R. Descriptive epidemiology of incidence and mortality of primary liver cancer in 185 countries: evidence from GLOBO- CAN 2018. Jpn J Clin Oncol. 2020 Dec 16;50(12):1370–9. doi:10.1093/jjco/hyaa130 3. Levy I, Sherman M, Liver Cancer Study Group of the University of Toronto. Staging of hepatocellular carcinoma: assessment of the CLIP, Okuda, and Child–Pugh staging systems in a cohort of 257 patients in Toronto. Gut. 2002 Jun;50(6):881–5. doi:10.1136/ gut.50.6.881 4. Johnson PJ, Berhane S, Kagebayashi C, Satomura S, Teng M, Reeves HL, et al. Assessment of liver function in patients with hepatocellular carcinoma: a new evidence-based approach-the ALBI grade. J Clin Oncol. 2015 Feb 20;33(6):550–8. doi:10.1200/ JCO.2014.57.9151 5. Okuda K, Musha H, Nakajima Y, Kubo Y, Shimokawa Y, Nagasaki Y, et al. Clinicopathologic features of encapsulated hepa- tocellular carcinoma: a study of 26 cases. Cancer. 1977 Sep;40(3):1240–5. doi:10.1002/1097-0142(197709)40:3<1240::aid-cn- cr2820400339>3.0.co;2-y 6. Tangkijvanich P, Anukulkarnkusol N, Suwangool P, Lertmaharit S, Hanvivatvong O, Kullavanijaya P, et al. Clinical character- istics and prognosis of hepatocellular carcinoma: analysis based on serum alpha-fetoprotein levels. J Clin Gastroenterol. 2000 Dec;31(4):302–8. doi:10.1097/00004836-200012000-00007 globale pour l'ensemble du groupe était de 25 mois [intervalle de confiance (IC) à 95 % : 15,7-34,2]. Les valeurs seuils pour l'AFP et le NLR ont été acceptées respectivement à 200 ng/mL et 3,45. L'aire sous les valeurs de la courbe ROC pour l'AFP, le NLR et le NLR+AFP était de 0,625 (IC à 95 % : 0,547-0,704), 0,589 (IC à 95 %: 0,512-0,667) et 0,657 (IC à 95 % : 0,583-0,731). L'analyse multivariée a montré que la taille avancée de la tumeur, l'atteinte ganglionnaire lymphatique et le stade métastatique (classification TNM), la présence d'une cirrhose, ainsi que des valeurs d'AFP et de NLR élevées étaient associés à une faible survie. Conclusion : L'AFP, le NLR, un stade TNM avancé et la présence d'une cirrhose peuvent permettre de prédire le pronostic chez les patients atteints de CHC. Des études incluant un plus grand nombre de pays sont nécessaires pour corroborer ces résultats. الهآمو ةَّيِِدبَكلا اَيلاَلخا َناطََسَرل ةيريسَرلا صئاصخلل تايسنلجاو زكارلما ةددعتم ةسارد سيناوي ،ولغوأ دحمأ ،نلاباك دممح ،ناغودرأ قيتع ،ليفشادلاأ ليشترأ ،كاشرأ دممح ،افيشتاكيس انيرام ،وكنوأ ناغود ،ناكيريد دحمأ لاحروت ميظن ،سانيفوكوب ةصلالخا .هب ةطبترلما تايفولا لدعم في ديازت كانهو ،ةيربكلا ةيحصلا لكاشلما نم ةَّيِِدبَكلا اَيلاَلخا َناط ََسَر دعُي :ةيفللخا .OncoBridge ةسارد ةعوممج في ءاضعلأا نادلبلا في الهآمو ةَّيِِدبَكلا اَيلاَلخا َناط ََسَرل ةيريسَرلا صئاصلخا ديدتح لىا ةساردلا هذه تفده :فادهلأا ،ايجروجو ،ايسورو ،ايكرت( نادلب 4 نم 2019و 2013 يَماع ينب ةترفلا في ةَّيِِدبَكلا اَيلاَلخا َناط ََسَرب اًصخشُم اًضيرم 630 انترخا :ثحبلا قرط صئاصخ ليلتح مدخُتساو .ا ًّيبرتمخو اًّيريسَر ضرلما لآمب ؤبنتلا لماوع ءاصقتسلا تايرغتلما ةددعتمو تايرغتلما ةيداحأ تلايلتح تيرجُأو .)نانويلاو .ينتورب وتيف افلأ شرؤم ةميقو ،ةيوافمللا ايلالخا لىإ تلادعلا ةبسنل ةيؤبنتلا ةميقلا ديدحتل تِلابْقَتْسُلما لعف دبكلا ضرمو ،)17.0%( C دبكلا باهتلا سويرفب ىودعلاو ،)39.7%( B دبكلا باهتلا ىودع يه اًعويش ةببسم لماوع 3 رثكأ تناك :جئاتنلا ميقلا تغلبو .اًرهش ]34.2-15.7 :%95 ةقثلا لصاف[ 25 اهلمكأب ةعومجملل ةايلحا ديق لىع ءاقبلا طسوتم ناكو .)%9.0( ليوحكلا يرغ ينهدلا تتح ةحاسلما تغلبو .لياوتلا لىع ،3.45و ،ترل لييم /مارجونان 200 ةيوافمللا ايلالخا لىإ تلادعلا ةبسنو ينتورب وتيف افلأ نم ٍّلكل ةلوبقلما ةي ِّدلحا ةيوافمللا ايلالخا لىإ تلادعلا ةبسنو ،ةيوافمللا ايلالخا لىإ تلادعلا ةبسنو ،ينتورب وتيف افلأ نم ٍّلكل ةبسنلاب تِلابْقَتْسُلما لعف صئاصخ ىنحنم ميق لصاف( 0.657و )0.667-0.512 :%95 ةقثلا لصاف( 0.589و ،)0.547-0.704 - 0.547 :%95 ةقثلا لصاف( 0.625 ينتورب وتيف افلأ + ةلحرلما نم ٍّلكب ةايلحا ديق لىع ءاقبلا لىع ةردقلا ضافخنا طابترا َ َّينبت ،تايرغتلما ددعتلما ليلحتلا للاخ نمو .)0.731-0.583 :%95 ةقثلا .ةيوافمللا ايلالخا لىإ تلادعلا ةبسن ميق عافتراو ،ينتورب وتيف افلأ عافتراو ،دبكلا فُّيلت دوجوو ،لئاقنلاو ةيفمللا دقعلاو مرولا مجلح ةمدقتلما ةيفمللا دقعلاو مرولا مجلح ةمدقتلما ةلحرلماو ،ةيوافمللا ايلالخا لىإ تلادعلا ةبسنو ،ينتورب وتيف افلأ نم ٍّلك مادختسا نكمي :تاجاتنتسلاا نم اًديزم لمشت تاسارد ءارجإ لىإ ةجاح كانهو .ابه ينباصلما ضىرلما ىدل ةَّيِِدبَكلا اَيلاَلخا َناط ََسَر لآمب ؤبنتلل ،دبكلا فُّيلت دوجوو ،لئاقنلاو .جئاتنلا هذه ديكأتل نادلبلا 472 Research article EMHJ – Vol. 29 No. 6 – 2023 7. Kim BK, Park JY, Kim DY, Kim JK, Kim KS, Choi JS, et al. Persistent hepatitis B viral replication affects recurrence of hepatocellu- lar carcinoma after curative resection. Liver Int. 2008 Mar;28(3):393–401. doi:10.1111/j.1478-3231.2007.01625.x 8. Wu JC, Huang YH, Chau GY, Su CW, Lai CR, Lee PC, et al. Risk factors for early and late recurrence in hepatitis B-related hepato- cellular carcinoma. J Hepatol. 2009 Nov;51(5):890–7. doi:10.1016/j.jhep.2009.07.009 9. Wang YG, Wang P, Wang B, Fu ZJ, Zhao WJ, Yan SL. Diabetes mellitus and poorer prognosis in hepatocellular carcinoma: a systematic review and meta-analysis. PLoS One. 2014 May 15;9(5):e95485. doi:10.1371/journal.pone.0095485 10. Wang D, Hu X, Xiao L, Long G, Yao L, Wang Z, et al. Prognostic Nutritional Index and Systemic Immune-Inflammation Index predict the prognosis of patients with HCC. J Gastrointest Surg. 2021 Feb;25(2):421–7. doi:10.1007/s11605-019-04492-7 11. Schmidt H, Bastholt L, Geertsen P, Christensen IJ, Larsen S, Gehl J, et a. Elevated neutrophil and monocyte counts in periph- eral blood are associated with poor survival in patients with metastatic melanoma: a prognostic model. Br J Cancer. 2005 Aug 8;93(3):273–8. doi:10.1038/sj.bjc.6602702 12. Cho H, Kim JH. Multiplication of neutrophil and monocyte counts (MNM) as an easily obtainable tumour marker for cervical cancer. Biomarkers. 2009 May;14(3):161–70. doi:10.1080/13547500902777616 13. Sadik CD, Kim ND, Luster AD. Neutrophils cascading their way to inflammation. Trends Immunol. 2011 Oct;32(10):452–60. doi:10.1016/j.it.2011.06.008 14. Cupp M, Cariolou M, Tzoulaki I, Aune D, Evangelou E, Berlanga-Taylor A. Neutrophil to lymphocyte ratio and cancer progno- sis: an umbrella review of systematic reviews and meta-analyses of observational studies. BMC Med. 2020 Nov 20;18(1):360. doi:10.1186/s12916-020-01817-1 15. European Association for the Study of the Liver. EASL Clinical Practice Guidelines: management of hepatocellular carcinoma. J Hepatol. 2018 Jul;69(1):182–236. doi:10.1016/j.jhep.2018.03.019 16. Hsu CY, Liu PH, Lee YH, Hsia CY, Huang YH, Lin HC, et al. Using serum α-fetoprotein for prognostic prediction in patients with hepatocellular carcinoma: what is the most optimal cutoff? PLoS One. 2015 Mar 4;10(3):e0118825. doi:10.1371/journal.pone.0118825 17. Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin. 2005 Mar–Apr;55(2):74–108. doi:10.3322/ canjclin.55.2.74 18. Poustchi H, Sepanlou S, Esmaili S, Mehrabi N, Ansarymoghadam A. Hepatocellular carcinoma in the world and the middle East. Middle East J Dig Dis. 2010 Jan;2(1):31–41. PMID:25197510 19. Abdel-Wahab M, El-Ghawalby N, Mostafa M, Sultan A, El-Sadany M, Fathy O, et al. Epidemiology of hepatocellular carcinoma in lower Egypt, Mansoura Gastroenterology Center. Hepatogastroenterology. 2007 Jan–Feb;54(73):157–62. PMID:17419252 20. McGlynn KA, Petrick JL, London WT. Global epidemiology of hepatocellular carcinoma: an emphasis on demographic and regional variability. Clin Liver Dis. 2015 May;19(2):223–38. doi:10.1016/j.cld.2015.01.001 21. Alavian SM, Haghbin H. Relative importance of hepatitis B and C viruses in hepatocellular carcinoma in EMRO countries and the Middle East: a systematic review. Hepat Mon. 2016 Mar 6;16(3):e35106. doi:10.5812/hepatmon.35106 22. Fasani P, Sangiovanni A, De Fazio C, Borzio M, Bruno S, Ronchi G, et al. High prevalence of multinodular hepatocellular carcino- ma in patients with cirrhosis attributable to multiple risk factors. Hepatology. 1999 Jun;29(6):1704–7. doi:10.1002/hep.510290604 23. Block TM, Mehta AS, Fimmel CJ, Jordan R. Molecular viral oncology of hepatocellular carcinoma. Oncogene. 2003 Aug 11;22(33):5093–107. doi:10.1038/sj.onc.1206557 24. Uzunalimoğlu O, Yurdaydin C, Cetinkaya H, Bozkaya H, Sahin T, Colakoğlu S, et al. Risk factors for hepatocellular carcinoma in Turkey. Dig Dis Sci. 2001 May;46(5):1022–8. doi:10.1023/a:1010705910858 25. Mittal S, El-Serag HB, Sada YH, Kanwal F, Duan Z, Temple S, et al. Hepatocellular carcinoma in the absence of cirrhosis in United States veterans is associated with nonalcoholic fatty liver disease. Clin Gastroenterol Hepatol. 2016 Jan;14(1):124–31.e1. doi:10.1016/j.cgh.2015.07.019 26. Kong J, Wang T, Shen S, Zhang Z, Wang W. A nomogram predicting the prognosis of young adult patients diagnosed with hepa- tocellular carcinoma: a population-based analysis. PLoS One. 2019 Jul 11;14(7):e0219654. doi:10.1371/journal.pone.0219654 27. Poon RT, Fan ST, Lo CM, Liu C L, Ng I O, Wong J. Long-term prognosis after resection of hepatocellular carcinoma associated with hepatitis B-related cirrhosis. J Clin Oncol. 2000 Mar;18(5):1094–101. doi:10.1200/JCO.2000.18.5.1094 28. Margetts J, Ogle LF, Chan SL, Chan AWH, Chan KCA, Jamieson D, et al. Neutrophils: driving progression and poor prognosis in hepatocellular carcinoma? Br J Cancer. 2018;118:248–57. doi:10.1038/bjc.2017.386 29. Hu Z, Chen H, Chen S, Huang Z, Qin S, Zhong J, et al.The value of neutrophil to lymphocyte ratio and gamma-glutamyl trans- peptidase to platelet ratio in patients with hepatocellular carcinoma. Medicine (Baltimore). 2019 Mar;98(9):e14749. doi:10.1097/ MD.0000000000014749 30. Hoffman R, Haim N, Brenner B. Cancer and thrombosis revisited. Blood Rev. 2001 Jun;15(2):61–7. doi:10.1054/blre.2001.0149 31. Lip GY, Chin BS, Blann AD. Cancer and the prothrombotic state. Lancet Oncol. 2002 Jan;3(1):27–34. doi:10.1016/s1470- 2045(01)00619-2 32. Zhang Y, Brodin P, Ohri N, Thibaud S, Kaubisch A, Kinkhabwala M, et al. Association between neutrophil–lymphocyte ratio, socioeconomic status, and ethnic minority with treatment outcome in hepatocellular carcinoma. Hepatol Int. 2019 Sep;13(5):609– 17. doi:10.1007/s12072-019-09965-0 473 Research article EMHJ – Vol. 29 No. 6 – 2023 33. Chon YE, Park H, Hyun HK, Ha Y, Kim MN, Kim BK, et al. Development of a new nomogram including neutrophil-to-lympho- cyte ratio to predict survival in patients with hepatocellular carcinoma undergoing transarterial chemoembolization. Cancers (Basel). 2019 Apr 10;11(4):509. doi:10.3390/cancers11040509 34. Matsumoto Y, Suzuki T, Asada I, Ozawa K, Tobe T, Honjo I. Clinical classification of hepatoma in Japan according to serial changes in serum alpha-fetoprotein levels. Cancer. 1982 Jan 15;49(2):354–60. doi:10.1002/1097-0142(19820115)49:2<354::aid-cn- cr2820490224>3.0.co;2-j 35. Hu J, Wang N, Yang Y, Ma L, Han R, Zhang W, et al. Diagnostic value of alpha-fetoprotein combined with neutrophil-to-lympho- cyte ratio for hepatocellular carcinoma. BMC Gastroenterol. 2018 Dec 13;18(1):186. doi:10.1186/s12876-018-0908-6 36. Abou-Alfa GK, Schwartz L, Ricci S, Amadori D, Santoro A, Figer A, et al. Phase II study of sorafenib in patients with advanced hepatocellular carcinoma. J Clin Oncol. 2006 Sep 10;24(26):4293–300. doi:10.1200/JCO.2005.01.3441 37. Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, Blanc JF, et al. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med. 2008 Jul 24;359(4):378–90. doi:10.1056/NEJMoa0708857 38. Cheng AL, Kang YK, Chen Z, Tsao CS, Qin S, Kim JS, et al. Efficacy and safety of sorafenib in patients in the Asia-Pacific region with advanced hepatocellular carcinoma: a phase III randomised, double-blind, placebo-controlled trial. Lancet Oncol. 2009 Jan;10(1):25–34. doi:10.1016/S1470-2045(08)70285-7 39. Manghisi G, Elba S, Mossa A Giorgio A, Aloisio V, Perrotta A et al. A new prognostic system for hepatocellular carcinoma: a ret- rospective study of 435 patients: the Cancer of the Liver Italian Program (CLIP) investigators. Hepatology. 1998 Sep;28(3):751–5. doi:10.1002/hep.510280322 40. Stuart KE, Anand AJ, Jenkins RL. Hepatocellular carcinoma in the United States. Prognostic features, treatment outcome, and survival. Cancer. 1996 Jun 1;77(11):2217–22. doi:10.1002/(SICI)1097-0142(19960601)77:11<2217::AID-CNCR6>3.0.CO;2-M 41. Vauthey JN, Lauwers GY, Esnaola NF, Do KA, Belghiti J, Mirza N, et al . Simplified staging for hepatocellular carcinoma. J Clin Oncol. 2002;20(6):1527–36. doi:10.1200/JCO.2002.20.6.1527