Liver cancer is the sixth most common incident cancer worldwide and the fourth most common cause of cancer death according to the Global Burden of Disease Liver Cancer Collaboration last report . In addition, hepatocellular carcinoma (HCC) is a leading cause of death among patients with cirrhosis . Survival of HCC is globally very poor, with overall 5-year survival around 10–15% . As other tumors, prognosis of HCC largely depends on cancer stage at diagnosis, as patients diagnosed at early stage and treated achieve 5-year survival rates of 70%, whereas those with advanced HCC have a median survival of less than 1 year . Since the cornerstone for improving survival of HCC in patients with cirrhosis is the application of a potential curative therapy , an early diagnosis of HCC is essential to achieve this goal and it has been the rationale for recommending HCC surveillance in high-risk patients [2,4,5].
Ultrasonography is the preferred test for HCC surveillance, based on the results of previous nonrandomized studies and a single randomized controlled study conducted in a Chinese population infected with hepatitis B virus . Ultrasound is well tolerated and widely available. Current international guidelines recommend HCC surveillance by ultrasound examination every 6 months in patients with cirrhosis, irrespective of its cause [4,5].
Individuals living with HIV are a high-risk population for developing HCC, which is mainly driven by hepatitis C virus (HCV) coinfection [7–11]. The burden of HCC in the HIV setting is substantial, as HCC is the second cause of death in HIV/HCV-coinfected patients with cirrhosis in Spain . As in other high-risk groups, HCC surveillance by biannual ultrasound examinations is recommended in HIV-infected patients with cirrhosis [4,5,17]. These recommendations are based on extrapolated data from non-HIV populations, as there are no specific studies addressing the performance of such strategy in HIV-infected patients. However, currently recommended ultrasound surveillance strategies could be less effective in these patients. With regard to this, it has been suggested that the time from HCV infection to the development of HCC is shorter in the presence of HIV and that HCC could have a more aggressive course in the HIV-infected patient [13–15]. Consequently, the effectiveness of this surveillance policy needs to be specifically evaluated in the scenario of HIV infection. Because of these, the objective of our study was to assess the performance of ultrasound surveillance for the diagnosis of HCC in HIV-infected patients.
Patients and methods
Study design and patients
The GEHEP-002 cohort (ClinicalTrials.gov ID: NCT02785835) recruits all the HCC cases diagnosed in HIV-infected patients from 32 centers in Spain. HCC cases occurring before 31 December 2010 were retrospectively included as previously reported . From this date, new HCC cases are recruited at HCC diagnosis. Patients are included in the cohort provided that they met the American Association for the Study of Liver Diseases criteria for the diagnosis of HCC . For this retrospective study, the information recorded in the last update of data made in 31 December 2017 was used.
To assess the performance of surveillance we selected cases diagnosed within a screening program among those included in the GEHEP-002 cohort. For this purpose, diagnosis by surveillance was considered when all scheduled ultrasound had been performed at least within 1 year prior to HCC diagnosis. Otherwise, HCC diagnosis was considered to be made out of the screening program.
A control population comprised of all HCC cases diagnosed in HCV-monoinfected patients during the study period (2000–2017) at the Liver Unit from the Hospital Universitario de Valme was used to evaluate the impact of HIV on surveillance. Information regarding these cases was recorded as previously described  and was revised and updated for the present analysis.
Clinical data, follow-up and surveillance protocol
Epidemiological, clinical and laboratory parameters are routinely collected in the GEHEP-002 cohort and were used for this analysis. Liver function was assessed by means of the Child–Turcotte–Pugh scoring system. HCC staging and treatment were established by the Barcelona Clinic Liver Cancer (BCLC) staging system . Management of HCC was done according to the European Association for the Study of the Liver recommendations  and the guidelines in force in Spain [17,18].
Surveillance of HCC was done by the performance of an abdominal ultrasound every 6 months following the recommendations of guidelines in force [4,5,17,18]. Surveillance was offered to all HIV-infected patients with cirrhosis, irrespective of its cause, attending the participant centers during the study period. In HIV/HCV-coinfected patients with cirrhosis who achieved sustained virological response, surveillance was maintained after viral eradication. Ultrasound examinations were performed by an experienced radiologist at each recruiting center in the GEHEP-002 cohort. Although it was not considered mandatory for surveillance, alpha-fetoprotein measurement every 6 months was also performed.
Surveillance, diagnosis and management of HCC in the control group was done according to the same recommendations and guidelines followed by the GEHEP-002 cohort [4,5,18].
Main variables of the study
Two main approaches were performed to evaluate the effectivity of surveillance. First, we assessed the proportion of HCC cases diagnosed by screening soon after a previous normal ultrasound. For this purpose, we considered as ‘ultrasound lack of detection’ cases of HCC diagnosed within the first 3 months after a previous surveillance ultrasound examination not showing suspicious hepatic nodules. The characteristics of these cases were compared with the remaining cases diagnosed by screening. Second, we assessed the proportion of HCC cases in which surveillance failed to detect HCC at early stage, which we defined as ‘surveillance failure’. Namely, a ‘surveillance failure’ occurred when a HCC diagnosed within the screening program was classified at BCLC stage B, C or D at initial presentation. In addition, we assessed the proportion of HCC cases diagnosed by screening beyond Milan criteria. Milan criteria is defined as the presence of a single nodule less or equal to 5 cm or up to three nodules less or equal to 3 cm . These definitions of ‘surveillance failure’ are in the line of the current recommendations of the European Association for the Study of the Liver  and have been also proposed in previous studies assessing the effectiveness of HCC surveillance [16,19–22].
Continuous variables and survival times are expressed as median (Q1–Q3), whereas categorical variables are presented as numbers (percentages). Continuous variables were compared by means of the Student t test or the Mann–Whitney U, depending on the normality tests. The chi-square and the Fisher tests were used for comparisons between categorical variables. For survival analyses, the date of HCC diagnosis was considered as the baseline time-point, defined as the date when the patient firstly met the American Association for the Study of Liver Diseases (AASLD) criteria for HCC. The time-to-event was computed as the time elapsed from baseline to the emergence of death of any cause or the censoring date, 31 December 2017. Kaplan–Meier estimates were calculated for survival end points and curves were compared using the log-rank test. Associations with a P less than 0.05 were considered significant. The statistical analysis was carried out by means of the SPSS statistical software package release 24.0 (IBM Corporation, Armonk, New York, USA).
The study was designed and conducted following the Helsinki declaration. The Ethics committee of the Hospital Universitario de Valme approved the study.
Features of the study population
Three hundred and forty-six cases of HCC in HIV-infected patients have been included in the GEHEP-002 cohort before 31 December 2017. Of them, 186 (54%) were diagnosed while being on an ultrasound surveillance program. One hundred and twelve (32.4%) patients received a curative therapy against HCC (liver transplantation n = 23, ablative therapies n = 53, hepatic resection n = 36), 111 (32%) received noncurative therapies and 123 (35.6%) did not receive treatment for HCC. After a median follow-up of 9 (2–28) months, 254 (73%) patients died. Of them, 199 (78%) died due to HCC, 30 (12%) due to non-HCC liver-related causes and 25 (10%) due to nonliver relates causes. The main characteristics of the study population are summarized in Table 1.
Performance of surveillance in the GEHEP-002 cohort
Ultrasound lack of detection rate
Sixteen (8.6%) out of 186 HCC cases diagnosed within an ultrasound surveillance program were diagnosed in the first 3 months after a previous ultrasound not showing suspicious hepatic nodules (Fig. 1a). The characteristics of these 16 patients were compared with the remaining 170 HCC cases diagnosed by screening (Table 2). The median (Q1–Q3) of alpha-fetoprotein was 31.6 (6.8–331) ng/dl in cases after ultrasound lack of detection, whereas it was 55 (7.4–372.5) ng/dl in the remaining cases (P = 0.8). HCC cases after ultrasound lack of detection were more frequently at Child–Pugh stage C and had an advanced stage at diagnosis (Table 2).
Ultrasound surveillance failure rates
Among 186 HCC cases diagnosed while being on screening, 107 (57%) cases were diagnosed at BCLC stage equal or greater than B and 104 (56%) were outside Milan criteria. Thus, ultrasound surveillance failure rates were 57 or 56%, depending of the approach which was used (Fig. 1b). The characteristic of HCC cases with ultrasound surveillance failure were compared with the remaining cases diagnosed by screening (Table 3). HCC cases after ultrasound surveillance failure showed a lower frequency of undetectable HIV viral load at diagnosis (Table 3). The median (Q1–Q3) nadir of CD4+ cell count prior to HCC diagnosis was 132 (81–209) and 154 (79–245) cells/μl in those with and without ultrasound surveillance failure (P = 0.5).
Performance of ultrasound surveillance according to the period of hepatocellular carcinoma diagnosis
The proportion of HCC cases diagnosed within a screening program in the GEHEP-002 cohort increased during the study period. Fifty-nine (45.4%) out of 130 cases diagnosed before 2010 and 127 (58.8%) out of 216 diagnosed after 2010 were diagnosed by screening (P = 0.015). However, the performance of ultrasound surveillance did not vary between both periods. Seven (12%) out of 49 cases diagnosed by screening before 2010 had a normal ultrasound in the previous 3 months, whereas this occurred in nine (7%) among 127 diagnosed by screening in the second period (P = 0.3). Finally, ultrasound surveillance failed to detect HCC at early stage in 36 (61%) and 71 (56%) in each period, respectively (P = 0.5).
Impact of ultrasound surveillance on survival
The median (Q1–Q3) survival time after HCC diagnosis was 13 (4–36) months for cases diagnosed by surveillance and 4 (2–17) months in the remaining cases (P < 0.0001) (Fig. 2a). When only cases diagnosed by screening were considered, the median survival in the 79 patients in which ultrasound surveillance failed to detect HCC at early stage was 7.5 months whereas it was 82 months in the remaining patients (P < 0.0001) (Fig. 2b).
Impact of HIV infection on the performance of ultrasound surveillance
Sixty-two (60%) out of 104 cases from the control group were diagnosed while being on a screening program. Five (8.6%) patients out of 58 HCC cases diagnosed by screening in which the exact date of the last ultrasound prior to HCC diagnosis could be ascertained, had undergone a previous ultrasound examination in the preceding 3 months that did not revealed suspicious hepatic nodules (Fig. 1a). This figure was similar to what had been found in the HIV-infected population (Fig. 1a). Table 2 summarizes the characteristics of these five cases. Forty-four (71%) patients out of 62 diagnosed by surveillance were at early stage at presentation (Fig. 1c). The performance to ultrasound surveillance to achieve an early diagnosis of HCC was significantly lower for HIV-infected patients than that found in non-HIV infected counterparts (Fig. 1b and c). As an example, ultrasound surveillance failed to detect HCC within Milan criteria in 104 (56%) out of 186 cases diagnosed by screening in HIV-infected patients whereas this occurred in 18 (29%) out of 62 HCV-monoinfected patients (P < 0.0001) (Fig. 1b). The worst performance of ultrasound surveillance in HIV-infected patients translated into lower survival rates when compared with HIV-uninfected patients. Thus, the probability of 1-year and 2-year survival after HCC diagnosis among those diagnosed by screening was 56 and 45% in HIV-infected patients, whereas it was 79 and 64% in HIV-negative patients (P = 0.038) (Fig. 2c).
The current study has found that ultrasound surveillance has a low performance for the diagnosis of HCC in HIV-infected patients. According to our findings, in spite of similar rates of HCC diagnosis by ultrasound surveillance and similar rates of ultrasound lack of detection, HIV-infected patients are more prone to a late diagnosis of HCC. Of note, ultrasound surveillance does not seem to translate into an earlier HCC diagnosis in HIV-infected patients. Importantly, survival of HCC diagnosed by screening was lower for HIV-infected patients than that of non-HIV-infected counterparts in the current study. Taken these findings together, we conclude that an HCC surveillance policy based on ultrasound examinations every 6 months might be insufficient in HIV-infected patients with cirrhosis.
Ultrasound is the preferred procedure for the surveillance of HCC in high-risk patients [2,4,5,17,18]. A previous meta-analysis had reported a pooled sensitivity of ultrasound to detect HCC at any stage of 94% . Drawbacks of ultrasound surveillance are its operator dependency and low uptake rates in clinical practice [20,21]. According to our data, the rate of lack of ultrasound detection in HIV-infected patients is similar to what has been reported out of HIV infection, as 8.6% of HIV-infected patients in our cohort were diagnosed of HCC soon after a previous normal surveillance ultrasound. In addition, our control group of HIV-uninfected patients showed a similar ultrasound sensitivity to detect HCC at any stage. Surprisingly, cases after ultrasound lack of detection tended to show a more advanced presentation at diagnosis, including higher frequencies of portal thrombosis. Notably, the aggressiveness of these cases tended to be more pronounced in HIV-infected patients. Ultrasound lack of detection in HIV-infected patients was more frequent in Child–Pugh stage C cirrhosis, which is in line with a recent retrospective study that found Child–Pugh C cirrhosis to be one of the main predictors of inadequate ultrasound quality for HCC surveillance .
Screening programs are aimed to reduce mortality by means of an anticipated cancer diagnosis. In the specific case of HCC, the ultimate goal of surveillance is to detect HCC at stages in which potentially curative therapies can be applied. On the contrary, the effectiveness of ultrasound as a surveillance tool is limited by its lower efficacy to detect HCC at early stage. A previous meta-analysis of studies conducted in non-HIV-infected patients reported a sensitivity of 63% for the detection of HCC at an early stage . Similarly, 71% of HCC cases diagnosed by ultrasound screening were at early stage in a multicenter Spanish study conducted in patients without HIV infection , similar to what was found in our control group of HCV-monoinfected patients. This suboptimal diagnostic yield of ultrasound to detect early stage HCC was even lower in HIV-infected patients in our study. Ultrasound surveillance was able to detect only 42% of cases at early stage in HIV-infected patients and, importantly, the proportion of early diagnosis was not significantly higher than that seen in the group without previous screening. In addition, when surveillance failure was defined as HCC diagnosis out of Milan criteria, the same findings were observed. As a consequence of the overall low performance of surveillance in HIV-infected patients, survival after HCC diagnosis by screening was very poor and significantly lower to what was observed in the HCV-monoinfected control group. On the other hand, despite the lower yield of ultrasound screening in HIV-infected patients, previous surveillance was associated with longer survival.
The reasons for the unexpectedly low performance of ultrasound surveillance in our HIV population are unclear. Previous reports have suggested that HCC might have an accelerated and aggressive pattern of presentation in HIV-infected patients [13–15]. In line with this hypothesis, despite rates of HCC diagnosis within screening program being similar, HIV-infected patients were more prone to a late diagnosis of HCC in our study. Of note, HCC cases in which ultrasound surveillance failed showed lower rates of controlled HIV viral replication and a nonsignificant trend for lower CD4+ cell counts. A possible explanation for these findings could be a poorer adherence to antiretroviral therapy and the surveillance protocol in a subset of patients. However, this is very unlikely as we only considered that a diagnosis was made within surveillance when all scheduled ultrasound had been performed at least within 1 year prior to HCC diagnosis. Previous studies have shown that the incidence of certain types of non-AIDS-defining neoplasms is increased in HIV-infected individuals [24,25], probably as a consequence of impairment of immune system due to HIV. In the case of HCC, it is tempting to speculate that HIV-associated chronic immunosuppression and inmunoactivation, neither of both are fully reversed by antiretroviral therapy, could affect tumor immunosurveillance, facilitating faster tumor growth and progression. In a context of accelerated progression, a substantial subset of patients can progress very fast between surveillance imaging techniques and fall into an advanced stage when diagnosis is firstly made.
The need for better tools for HCC surveillance have been claimed by most of experts [2,4,5]. Serum markers are an attractive alternative for this purpose. Although combined alpha-fetoprotein and ultrasound has been reported to increase sensitivity when compared with ultrasound alone , it also increases false-positive suspicious and costs [2,19]. On the contrary, other tumor markers that have been evaluated so far have not provided enough accuracy . Given the limited sensitivity of ultrasound, other imaging techniques, as MRI with liver-specific contrast, are being explored for surveillance purposes [26,27]. Finally, the potential role for early HCC diagnosis of assays evaluating molecular profiles of circulating free DNA and microRNA is also being investigated . Although ongoing surveillance studies are performed, alternative strategies should be considered in HIV-infected patients on an individual basis. One possible approach is to perform periodic ultrasound at a 3-month interval in very high-risk patients, as those harboring Child–Pugh stage B or C. In a previous randomized clinical trial, a 3-month interval increased the detection of small nodules but had no impact on survival . However, the efficacy of this strategy in HIV-infected patients cannot be completely ruled out, as patients with HIV infection were excluded in that trial . Given that a 6-month interval is clearly insufficient in HIV-infected patients, as indicated by our study, and the rationale of a faster tumor growth of HCC in the setting of HIV infection, the prospective assessment of ultrasound surveillance at a 3-month interval in this specific scenario should be evaluated.
The main limitation of this study is the retrospective design, as patients are included in the GEHEP-002 cohort once the diagnosis of HCC is made. Thus, the information of the whole cirrhotic population at risk attending at the participant centers was not available and precluded us to perform some additional relevant analyses. In addition, although ultrasound examinations were performed by an experienced radiologist in each center following a standard protocol, biases due to the operator dependent nature of ultrasound cannot be excluded. A further limitation was that non-HIV cases used herein as a control group came from a single center and was comprised only of HCV-related HCC. However, these cases showed the same features as those reported in a larger multicenter Spanish registry of non-HIV cases conducted in the same centers participating in the GEHEP-002 cohort . In addition, when the analysis was restricted to HCC cases coming from the Hospital de Valme, the similar findings were found, as surveillance failure occurred in 50% of those infected by HIV vs. 29% of those HIV uninfected. On the other hand, this is the first study that has assessed ultrasound surveillance in HIV-infected patients, using several approaches and including a control group of non-HIV-infected patients attended by the same protocol and during the same period of time, which allow us to perform direct comparisons. These are strengths of our study.
In summary, the performance of currently recommended policies for HCC surveillance in HIV-infected patients is very poor and worse than what is achieved outside HIV infection. Consequently, new effective options for HCC surveillance are urgently needed for these patients. Given the considerable burden of HCC in the setting of HIV infection and the lack of efficacy of ultrasound-based surveillance in this scenario, HIV-infected patients should not be systematically excluded from coming clinical trials assessing new tools for HCC screening.
The authors wish to thank the following members of the GEHEP-002 cohort for their contribution to this work: María R. Alemán Valls (Hospital Universitario de Canarias, San Cristóbal de la Laguna, Tenerife), Luis Metola (Hospital de San Pedro, Logroño) and Miguel Raffo (Complejo Hospitalario de Huelva, Huelva).
Author contributions: N.M. had full access and to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: N.M. and J.A.P.
Acquisition, analysis or interpretation of data: all authors.
Statistical analysis: N.M., M.R.-F., J.M. and J.A.P.
Drafting of the article: N.M.
Critical revision of the article for important intellectual content: all authors.
Obtained funding: N.M., J.M., J.A.P.
Study supervision: N.M.
The current study was supported by grants from the Consejería de Salud de la Junta de Andalucía (PI-0014/2014), the Servicio Andaluz de Salud (grant number SAS/111239) and the Fondo de Investigaciones Sanitarias ISCIII (grant number PI13/01621) and Project ‘PI16/01443’, funded by Instituto de Salud Carlos III, integrated in the national I+D+i 2013–2016 and co-funded by European Union (ERDF/ESF, ‘Investing in your future’). J.A.P. is the recipient of an intensification grant from the Instituto de Salud Carlos III (grant number Programa-I3SNS). Besides, this work has been partially funded by the Grupo para el Estudio de las Hepatitis Víricas (GEHEP) de la SEIMC (2017 grant to project GEHEP-002 and 2018 grant to project GEHEP-002), the SPANISH AIDS Research Network RD16/0025/0010 as part of the Plan Nacional R+D+I and cofinanced by ISCIII Subdirección General de Evaluación y el Fondo Europeo de Desarrollo Regional (FEDER).
Conflicts of interest
There are no conflicts of interest.
1. Akinyemiju T, Abera S, Ahmed M, Alam N, Alemayohu MA, Allen C, et al. Global Burden of Disease Liver Cancer Collaboration. The burden of primary liver cancer and underlying etiologies from 1990 to 2015 at the global, regional, and national level: results from the global burden of disease study 2015
. JAMA Oncol
2. Forner A, Reig M, Bruix J. Hepatocellular carcinoma
3. De Angelis R, Sant M, Coleman MP, Francisci S, Baili P, Pierannunzio D, et al. Cancer survival in Europe 1999–2007 by country and age: results of EUROCARE-5 a population based study
. Lancet Oncol
4. Bruix J, Sherman M. American Association for the Study of Liver Diseases. Management of hepatocellular carcinoma: an update
5. European Association for the Study of the Liver. EASL clinical practice guidelines: management of hepatocellular carcinoma
. J Hepatol
6. Zhang BH, Yang BH, Tang ZY. Randomized controlled trial of screening for hepatocellular carcinoma
. J Cancer Res Clin Oncol
7. Merchante N, Merino E, López-Aldeguer J, Jover F, Delgado-Fernádez M, Galindo MJ, et al. Increasing incidence of hepatocellular carcinoma in HIV-infected patients in Spain
. Clin Infect Dis
8. Rosenthal E, Roussillon C, Salmon-Céron D, Georget A, Hénard S, Huleux TF, et al. Liver-related deaths in HIV-infected patients between 1995 and 2010 in France: the Mortavic 2010 study in collaboration with the Agence Nationale de Recherche sur le SIDA (ANRS) en 20 Mortalité 2010 survey
. HIV Med
9. Sahasrabuddhe VV, Shiels MS, McGlynn KA, Engels EA. The risk of hepatocellular carcinoma among individuals with acquired immunodeficiency syndrome in the United States
10. Merchante N, Merino E, Rodríguez-Arrondo F, Tural C, Muñoz J, Delgado-Fernández M, et al. HIV/hepatitis C virus coinfected who achieved sustained virological response are still at risk of developing hepatocellular carcinoma
11. Merchante N, Rodríguez-Arrondo F, Revollo B, Merino E, Ibarra S, Galindo MJ, et al. Hepatocellular carcinoma after sustained virological response with interferon-free regimens in HIV/HCV-coinfected patients
12. Merchante N, Rivero-Juárez A, Téllez F, Merino D, Ríos-Villegas MJ, Villalobos M, et al. SVR with all-oral DAA regimens reduces the risk of hepatocellular carcinoma in HIV/HCV-coinfected patients with cirrhosis
. J Antimicrob Chemother
13. Garcia-Samaniego J, Rodriguez M, Berenguer J, Rodríguez-Rosado R, Carbó J, Asensi V, et al. Hepatocellular carcinoma in HIV-infected patients with chronic hepatitis C
. Am J Gastroenterol
14. Davila JA, Morgan RO, Shaib Y, McGlynn KA, El Serag HB. Hepatitis C infection and the increased incidence of hepatocellular carcinoma: a population-based study
15. Puoti M, Bruno R, Soriano V, Donato F, Gaeta GB, Quinzan GP, et al. Hepatocellular carcinoma in HIV-infected patients: epidemiological features, clinical presentation and outcome
16. Rodríguez de Lope C, Reig M, Matilla A, Ferrer MT, Dueñas E, Mínguez B, et al. Clinical characteristics of hepatocellular carcinoma in Spain. Comparison with the 2008–2009 period and analysis of the causes of diagnosis out of screening programs. Analysis of 686 cases in 73 centers
. Med Clin
17. Santos J, Valencia E. GeSIDA Expert Panel. Consensus statement on the clinical management of non-AIDS defining malignancies
. Enferm Infecc Microbiol Clin
18. Forner A, Reig M, Varela M, Burrel M, Feliu J, Briceño J, et al. Diagnosis and treatment of hepatocellular carcinoma. Update consensus document from the AEEH, SEOM, SERAM, SERVEI and SETH
. Med Clin
19. Singal A, Volk ML, Waljee A, Salgia R, Higgins P, Rogers MA, et al. Meta-analysis surveillance with ultrasound for early-stage hepatocellular carcinoma in patients with cirrhosis
. Aliment Pharmacol Ther
20. Singal AG, Conjeevaram HS, Volk ML, Fu S, Fontana RJ, Askari F, et al. Effectiveness of hepatocellular carcinoma surveillance in patients with cirrhosis
. Cancer Epidemiol Biomarkers Prev
21. Singal AG, Nehra M, Adams-Huet B, Yopp AC, Tiro JA, Marrero JA, et al. Detection of hepatocellular carcinoma at advanced stages among patients in the HALT-C trial: where did surveillance fail?
. Am J Gastroenterol
22. Chon YE, Jung KS, Kim MJ, Choi JY, An C, Park JY, et al. Predictors of failure to detect early hepatocellular carcinoma in patients with chronic hepatitis B who received regular surveillance
. Aliment Pharmacol Ther
23. Simmons O, Fetzer DT, Yokoo T, Marrero JA, Yopp A, Kono Y, et al. Predictors of adequate ultrasound quality for hepatocellular carcinoma surveillance in patients with cirrhosis
. Aliment Pharmacol Ther
24. Patel P, Hanson DL, Sullivan PS, Novak RM, Moorman AC, Tong TC, et al. Incidence of types of cancer among HIV-infected persons compared with the general population in the United States 1992–2003
. Ann Intern Med
25. Shiels MS, Cole SR, Kirk GD, Poole C. A meta-analysis of the incidence of non-AIDS cancers in HIV-infected individuals
. J Acquir Immune Defic Syndr
26. Kim HD, Lim YS, Han S, An J, Kim GA, Kim SY, et al. Evaluation of early-stage hepatocellular carcinoma by magnetic resonance imaging with gadoxetic acid detects additional lesions and increases overall survival
27. Kim SY, An J, Lim YS, Han S, Lee JY, Byun JH, et al. MRI with liver-specific contrast for surveillance of patients with cirrhosis at high risk of hepatocellular carcinoma
. JAMA Oncol
28. Pezzuto F, Buonaguro L, Buonaguro FM, Tornesello ML. The role of circulating free DNA and MicroRNA in noninvasive diagnosis of HBV- and HCV-related hepatocellular carcinoma
. Int J Mol Sci
29. Trinchet JC, Chaffaut C, Bourcier V, Degos F, Henrion J, Fontaine H, et al. Ultrasonographic surveillance of hepatocellular carcinoma in cirrhosis: a randomized trial comparing 3- and 6-month periodicities