NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
National Guideline Centre (UK). Cirrhosis in Over 16s: Assessment and Management. London: National Institute for Health and Care Excellence (NICE); 2016 Jul. (NICE Guideline, No. 50.)
8.1. Introduction
People with cirrhosis are at high risk of developing hepatocellular carcinoma (HCC) with an annual cancer risk of between 1% and 6% depending on the aetiology of the liver disease, for example HCV OR=13.4 (95% CI, 4.1 to 43.5), HBV OR=9.1 (2.1 to 39.5), and alcohol misuse OR=1.8 (0.7 to 4.3). Around 3,900 people are diagnosed with HCC each year in the UK, which accounts for about 1% of all cancers. However, HCC incidence rates are rising and over the next 20 years mortality is predicted to increase by 39% (from 4.2 to 5.9 per 100,000).56
The prognosis in people with HCC critically depends on tumour stage at the time of diagnosis. For example those with early HCC, defined as one nodule less than 5 cm or three nodules each less than 3 cm in diameter, can achieve 5-year survival rates near 70% with surgical resection or liver transplantation.102 In contrast, the survival rate in people who present with symptoms associated with a large tumour is poor, with median survival less than 6 months.123
Regular surveillance for people with cirrhosis, using liver ultrasound with or without serum alpha-fetoprotein (AFP) testing at 3- to 12-month intervals, endeavours to detect a tumour at an early stage when potentially curative treatment can be offered. Therefore, the GDG decided to compare the clinical and cost-effectiveness of ultrasound surveillance (with or without serum AFP testing) compared to no surveillance, and surveillance at difference frequencies of 3-, 6- and 12-month intervals for the detection of HCC in people with cirrhosis due to HCV, alcohol and non-alcohol-related fatty liver disease.
8.2. Review question: When and how frequently should surveillance testing be offered for the early detection of hepatocellular carcinoma (HCC) in people with cirrhosis?
For full details see review protocol in Appendix C.
Table 57
PICO characteristics of review question.
Randomised and observational studies comparing ultrasound surveillance (with or without serum AFP testing) with no surveillance for the early detection of HCC in people with cirrhosis were searched for. Randomised and observational studies comparing the effectiveness of different frequencies of ultrasound surveillance were also searched for. Implicit in the investigation of surveillance or surveillance frequency is that patients in whom HCC is detected earlier can be treated earlier, and have potentially better patient outcomes and a better chance of survival. The guideline does not, however, cover how HCC should be managed after diagnosis.
The review population was limited to people with confirmed cirrhosis. Due to the known link between HBV or HCV and HCC, some studies in the literature assess the effectiveness of surveillance in people with HCV or HBV regardless of the level of fibrosis or presence of cirrhosis. Studies including these populations were not included in this systematic review, which focuses on the effectiveness of surveillance in people with a confirmed diagnosis of cirrhosis. Studies including people with a mix of fibrosis stages were only included in the systematic review when the proportion of people with cirrhosis was >85%.
There is existing NICE guidance on surveillance for HCC in people with HBV (with fibrosis at any stage, including cirrhosis) and therefore, this guideline does not cover surveillance in people with cirrhosis due to HBV. There is evidence in the literature to suggest that people with HBV are at higher risk of developing HCC, even if they do not have cirrhosis. Therefore, the effectiveness of HCC surveillance may be different in this population compared with other disease aetiologies. For studies including people with cirrhosis and mixed aetiologies, studies were only included in the systematic review when the proportion of people with HBV was ≤15% due to the difference in prognosis in this population, and therefore an expected difference in the effectiveness of surveillance. A second approach agreed in the protocol was to use studies in people with cirrhosis and a proportion of people with HBV >15% if no evidence was identified using the above criteria. Evidence was identified for the comparison of surveillance versus no surveillance and for the comparison of different frequencies, therefore studies in people with cirrhosis and a proportion of people with HBV >15% were not included.
Observational non-randomised studies were included in the absence of evidence from RCTs. As pre-specified in the review protocol, only observational studies which performed a multivariate analysis to adjust for confounding factors were included in the review. Non-randomised studies reporting the characteristics of HCC (such as lesion size or the number of people with an advanced HCC stage) without adjusting for confounders were excluded from this review. All the observational studies identified in this review were retrospective cohort studies in people with a diagnosis of HCC, retrospectively analysed by their previous surveillance status prior to the diagnosis of HCC. Therefore, only people who developed HCC were analysed. No prospective observational studies were identified in people without HCC, followed up to see if HCC developed.
Studies of screened people who develop HCC compared to unscreened people who develop HCC are subject to lead-time bias. Lead-time bias is the apparent increase in survival that comes exclusively from diagnosis at an earlier stage of disease. The duration of survival from diagnosis to death is increased, even if no intervention is applied and some of the survival benefit could be ascribed to earlier diagnosis. Some of the studies included in this review attempted to correct for this bias by calculating the lead time and adjusting for it. This was taken into account when assessing the risk of bias of the included studies. Studies are also subject to length bias. This arises from the fact that surveillance is more likely to detect slow growing cancers than rapidly growing cancers, which might go from undetectable to death within the surveillance interval.
Following detection of lesions during ultrasound surveillance, all the included studies followed standardised protocols for the diagnosis and staging of HCC and treated HCC according to standard procedures. Details of the individual studies can be found in the clinical evidence tables in Appendix H.
8.3. Clinical evidence
8.3.1. Surveillance versus no surveillance
Six observational studies83,135,160,219,228,231 were identified comparing surveillance with no surveillance in people with cirrhosis. All were retrospective observational studies comparing outcomes in people with confirmed HCC depending on previous surveillance status. No RCTs were identified for this comparison. The outcomes of survival and HCC stage were reported from multivariate analyses. No evidence was identified from a multivariate analysis for the outcomes of quality of life, HCC occurrence, number of lesions, liver transplantation or lesion size less than 3 cm. Evidence for this comparison is summarised in the clinical evidence summary in Table 59.
Table 58
Summary of studies included in the review.
Table 59
Clinical evidence summary: Surveillance versus no surveillance.
8.3.2. Different surveillance frequencies
Two included studies compared surveillance at different frequencies: yearly versus 6-monthly surveillance in one observational study188 and 3-monthly versus 6-monthly surveillance in one RCT.233 Five other observational studies retrospectively compared yearly surveillance with 6-monthly surveillance, but were excluded from the current review due to the inclusion of people with cirrhosis of mixed aetiologies and the high proportion of people with HBV,187,229,230 or due to only reporting unadjusted results.45,59
8.3.2.1. Yearly versus 6-monthly surveillance
One observational study188 was identified comparing yearly surveillance with 6-monthly surveillance in people with cirrhosis. This was a retrospective observational study comparing outcomes in people with confirmed HCC depending on previous surveillance status. No RCTs were identified for this comparison. The outcomes of survival and HCC stage were reported from multivariate analyses. No evidence was identified from a multivariate analysis for the outcomes of quality of life, HCC occurrence, number of lesions, liver transplantation or lesion size less than 3 cm. Evidence for this comparison is summarised in the clinical evidence summary in Table 60.
Table 60
Clinical evidence summary: Yearly versus 6-monthly surveillance.
8.3.2.2. 3-monthly versus 6-monthly surveillance
One RCT233 was identified comparing 3-monthly surveillance with 6-monthly surveillance in people with cirrhosis. Evidence was available for all protocol outcomes with the exception of quality of life. Evidence for this comparison is summarised in the clinical evidence summary in Table 61.
Table 61
Clinical evidence summary: 3-monthly versus 6-monthly surveillance.
Narrative information for surveillance versus no surveillance
Trevisani 2004228 compared ultrasound surveillance every 6–12 months with both HCC detected by symptoms and HCC detected incidentally. For both comparisons, surveillance was not a statistically significant independent predictor of survival , so the adjusted relative risk for the effect of surveillance on survival from multivariate analysis was not reported. Trevisani 2007231 compared ultrasound surveillance every 6–12 months with HCC detected by symptoms or incidentally, with data stratified into Child-Pugh B and Child-Pugh C. Again, surveillance was not a statistically significant independent predictor of survival , so the adjusted relative risk for the effect of surveillance on survival from multivariate analysis was not reported.
The non-randomised studies included in this comparison also reported the characteristics of HCC including outcomes such as tumour size. However, these outcomes were not adjusted for confounding factors in a multivariate analysis, therefore results have not been extracted.
Narrative information for yearly versus 6-monthly surveillance
The non-randomised study included in this comparison also reported the characteristics of HCC including outcomes such as tumour size. However, these outcomes were not adjusted for confounding factors in a multivariate analysis, therefore results have not been extracted.
8.4. Economic evidence
8.4.1. Published literature
One economic evaluation was identified that compared annual surveillance and 6-monthly surveillance for HCC in people with cirrhosis of mixed aetiology.44
One economic evaluation was identified that compared no surveillance, annual AFP, annual ultrasound, annual AFP plus ultrasound, 6-monthly AFP, 6-monthly ultrasound, and 6-monthly AFP plus ultrasound for HCC in people with cirrhosis with either ALD or hepatitis C.226
These are summarised in the economic evidence profile below Table 62 and the economic evidence tables in Appendix I.
Table 62
Economic evidence profile: frequency of surveillance for HCC.
Table 63
Thompson Coon 2008 – Cost-effectiveness results for ALD patients.
See also the economic article selection flow chart in Appendix F.
8.4.2. Unit costs
See Table 64 in Appendix N.
Table 64
Thompson Coon 2008 – Cost-effectiveness results for HCV patients.
8.4.3. New cost-effectiveness analysis
Original cost-effectiveness modelling was undertaken for this question using the NGC liver disease pathway model developed for this guideline. A summary is included here. An evidence statement summarising the results of the analysis can be found below. The full analysis can be found in Appendix N.
8.4.3.1. Aim and structure
The aim of the health economic modelling for this question was to determine the optimal frequency of HCC surveillance. This was achieved by using the original lifetime diagnostic health state transition (Markov) model (see Section 6.4.3 and Appendix N) which followed the NICE reference case146 and by comparing overall cost and QALYs of the cirrhosis tests of preference in 2 different scenarios; annual and semi-annual HCC surveillance.
To apply the clinical benefit of HCC surveillance, figures from 2 different sources, identified by the clinical review (1 included in the review – Santi 2010), were combined. A study by Zhang 2004 with a 5-year follow-up on 18,816 people with hepatitis B reported that 6-monthly surveillance (using alpha-fetoprotein [AFP] blood test plus ultrasound) was associated with a 37% reduction in HCC mortality in comparison to a no monitor control group. This number was combined with an increased risk of death figure (HR 1.39) for patients under annual surveillance using ultrasound plus AFP compared to a 6-monthly surveillance strategy reported by Santi 2010 (649 patients of mixed disease aetiology).
To determine the most cost-effective surveillance frequency, incremental cost-effectiveness ratios (ICERs) were calculated to compare the options in each population. Base case results below were obtained through the probabilistic analysis to take combined parameter uncertainty into account.
8.4.3.2. Results
Table 65
ICERs comparing 6-monthly surveillance against annual surveillance.
At a cost-effectiveness threshold of £20,000 per QALY gained, 6-monthly surveillance was cost-effective compared to annual surveillance only for the HCV genotype 1 group. Although this group had the fewest liver-related deaths, risk of HCC progression was particularly high in this group compared to other model cohorts, making more frequent surveillance cost-effective at the specified threshold. However, at a cost-effectiveness threshold of £30,000 per QALY gained, 6-monthly surveillance was cost-effective compared to annual surveillance for all groups: ICERs £18,657–28,352. Variation in the ICERs was mainly due differences in cirrhosis prevalence, risk of progression to HCC, and competing risks of other complications in each aetiology.
In the deterministic sensitivity analyses, reducing the surveillance costs or increasing the 6-monthly surveillance effectiveness reduced the ICERs for each group by up to £2,000 per QALY gained.
8.5. Evidence statements
8.5.1. Clinical
Surveillance versus no surveillance
- Very Low quality evidence from 2 studies (n=351) indicated a clinical benefit of surveillance for survival when analysed using time-to-event data. Low quality evidence from 1 study (n=1729) indicated a clinical benefit of surveillance for the detection of HCC at a non-advanced stage.
Yearly versus 6-monthly surveillance
- Very Low quality evidence from 1 study (n=649) indicated a clinical benefit of 6-monthly surveillance for survival. Low quality evidence from the same study indicated a clinical benefit of 6-monthly surveillance for the detection of HCC beyond a very early stage.
6-monthly versus 3-monthly surveillance
- Moderate quality evidence from 1 RCT (n=1278) indicated a clinical benefit of 3-monthly surveillance for survival and HCC occurrence. Evidence ranging from Very low to Moderate quality from the same RCT indicated no clinical difference in HCC diameter >30 mm at detection, the number of HCC nodules detected or the HCC stage at detection.
8.5.2. Economic
- One cost-utility analysis that compared 6-monthly versus annual surveillance for HCC in people with cirrhosis (mixed aetiology) found that 6-monthly surveillance for was not cost-effective for either compensated cirrhosis or decompensated cirrhosis (ICERs: £21,230 and £40,540 per QALY gained). This analysis was assessed as partially applicable with potentially serious limitations.
- One cost-utility analysis that compared 7 relevant strategies for surveillance of HCC in people with cirrhosis (including use or ultrasound, alpha-fetoprotein, both or neither, at 6-monthly or annual intervals) in people with ALD or HCV found that:
- No other strategy was cost-effective compared to no surveillance at a cost-effectiveness threshold of £20,000 per QALY gained.
- Annual surveillance using alpha-fetoprotein only had an ICER of £25,490 per QALY gained compared to no surveillance.
This analysis was assessed as directly applicable with minor limitations. - One original cost-utility analysis that compared 6-monthly with annual surveillance for HCC in people with cirrhosis at a cost-effectiveness threshold of £20,000 per QALY gained found that:
- 6-monthly surveillance was cost-effective compared to annual surveillance for people with HCV genotype 1 (ICER: £18,657 per QALY gained).
- 6-monthly surveillance was not cost-effective compared to annual surveillance for people with NAFLD, ALD, HBV or HCV genotype 3 (ICERs: £20,128–28,352).
This analysis was assessed as directly applicable with minor limitations.
8.6. Recommendations and link to evidence
| Recommendations |
|
|---|---|
| Relative values of different outcomes | For the comparison of HCC surveillance with no surveillance, or different surveillance frequencies, the GDG chose transplant-free survival and quality of life as critically important outcomes. The GDG also considered HCC occurrence, the size of the HCC at the time of detection (≤3 cm, >3 cm), the number of lesions detected, the Barcelona Clinic Liver Cancer (BCLC) stage at the time of detection and the need for transplantation as important outcomes. All of the studies reported overall survival data rather than transplant-free survival. The GDG decided that overall survival was just as important and discussed this outcome when forming the recommendations. There is already existing NICE guidance on surveillance for HCC in people with HBV (with fibrosis at any stage, including cirrhosis) and therefore this guideline does not cover surveillance in people with cirrhosis due to HBV. |
| Trade-off between clinical benefits and harms | Surveillance versus no surveillance Six observational studies were identified comparing surveillance with no surveillance in people with cirrhosis. No RCTs were identified for this comparison. The outcomes of survival and HCC stage were reported from multivariate analyses. The GDG noted that there was a clinical benefit of surveillance on the outcome of survival from 2 studies reported as a hazard ratio, but a third study showed no clinical benefit of surveillance on survival, reported as an odds ratio. However, the GDG discussed that none of the studies adjusted for the lead-time bias. Two further studies reported in the narrative that surveillance was not found to be an independent predictor of survival, therefore the adjusted hazard ratios for the effect of surveillance on survival was not reported. One of these studies did adjust for possible lead-time bias. Evidence from 1 study showed an OR of 5.4 in favour of surveillance for detecting HCC at a very early stage (single nodule ≤2 cm). In addition, evidence from 2 studies showed a clinical benefit in favour of surveillance for detection of HCC at a non-advanced stage (according to the Milan criteria). No evidence was identified from a multivariate analysis for the outcomes of quality of life, HCC occurrence, number of lesions, liver transplant or lesion size less than 3 cm. Yearly versus 6-monthly surveillance One observational study was identified comparing yearly surveillance with 6-monthly surveillance in people with cirrhosis. No RCTs were identified for this comparison. The outcomes of survival and HCC stage were reported from multivariate analyses. The GDG noted that there was a clinical benefit of surveillance every 6 months compared to yearly surveillance on the outcome of survival, but the confidence intervals were wide. This study did adjust for the lead-time bias. There was also a clinical benefit of surveillance on the outcome of detection of HCC beyond a very early stage. No evidence was identified from a multivariate analysis for the outcomes of quality of life, HCC occurrence, number of lesions, liver transplant or lesion size less than or equal to 3 cm. 3-monthly versus 6-monthly surveillance One RCT was identified comparing 3-monthly surveillance with 6-monthly surveillance in people with cirrhosis. Evidence was available for all protocol outcomes with the exception of quality of life. There was a Hazard Ratio of 0.87 (95% CI 0.64 to 1.19) with 17 fewer patients per 1000 dying in the 3-monthly group versus the 6-monthly group. The GDG agreed this was a clinical benefit to the 3-monthly surveillance group. In the 3-monthly surveillance group the number of HCCs with a diameter ≤3 cm was less with a RR 0.85 (95% CI 0.57 to 1.27). There appeared to be no difference in the incidence of HCC >3 cm, the total number of nodules detected, or the number of infiltrative lesions in either the 3-monthly or the 6-monthly groups. The GDG noted that whilst there was no statistically significant difference between 3-monthly and 6-monthly surveillance, there was a trend towards a clinical benefit in favour of the shorter time interval. The GDG did note, however, that there were more patients who had HCC detected in the 6-monthly surveillance group so this may account for some of the variation in the stage of disease identified. Overall considerations of the trade-off between clinical benefits and harms The GDG felt that the fact that regular HCC surveillance is already common practice could not be overlooked. Existing guidance recommends that all people with hepatitis B and significant fibrosis should receive 6-monthly surveillance (NICE guideline CG165141) and this is established practice. Therefore, the GDG agreed that 6-monthly surveillance should be available to people with cirrhosis due to other underlying aetiologies. They also noted, however, that current practice is not supported by High quality evidence of improved survival. The GDG agreed that the decision to offer surveillance should be based on the ability to offer treatments for HCC. If the outcomes of surveillance are to be beneficial then effective treatments have to be available. The GDG discussed that new, more effective treatment options are available since the publication of some of the evidence included in this review and that this may enhance the benefit seen on patient survival. There are several effective treatment options available such as radiofrequency ablation, microwave ablation and chemoembolisation, as well as resection and liver transplantation. The GDG was concerned that there is some national variability in the availability of the treatment options offered to people with HCC. The management of HCC is not covered here as it is outside the scope of this guideline. The GDG's consensus was that it would be a disservice to patients not to recommend surveillance for HCC. Overall the GDG felt that there was a clinical benefit from surveillance in the detection of tumours at an earlier stage and that surveillance should be offered to give patients the opportunity to receive potentially curative treatment. The GDG considered this to outweigh the potential harms of surveillance in terms of the over-investigation of false positive results. It was discussed that the doubling times of liver tumours are likely to be between 50 and 200 days with small primary HCC generally within the shortest timescales. Survival is strongly linked to size and stage of tumour so more regular screening was the preferred option of the GDG. There was a strong consensus among the GDG that the recommendation should be for a high quality, inclusive surveillance programme looking at all aspects of management and care provision. There was a strong representation from the patient perspective that people with cirrhosis would want to be monitored, especially with the advent of modern potentially curative treatment. The GDG agreed that 6-monthly surveillance had clinical benefit over yearly surveillance. The GDG also discussed the potential economic and logistical considerations of recommending 3-monthly versus 6-monthly surveillance, and particularly the significant increased resource implication of 3-monthly surveillance in terms of radiology time. This would also increase the number of hospital visits required. Overall, the GDG agreed that 3-monthly surveillance was too frequent and that 6-monthly surveillance should be recommended. |
| Trade-off between net clinical effects and costs | The GDG noted that the costs of no surveillance take into account all of the treatment costs and the lifetime cost of care for patients who develop HCC. Two relevant economic evaluations were identified. The data from the paper by Cucchetti were based on 918 patients in 11 centres comparing surveillance with ultrasound and alpha-fetoprotein (with a CT scan arranged for all positive results). The results suggested that 6-monthly surveillance was not cost-effective compared to 12-monthly surveillance at a cost-effectiveness threshold of £20,000 per QALY gained (an ICER of only £21,230 for people with compensated cirrhosis but £40,540 per QALY for people with decompensated cirrhosis). A second paper by Thompson Coon assessed surveillance of patients with compensated cirrhosis under the age of 70 years. There were 7 different combinations of surveillance strategies proposed using ultrasound, alpha-fetoprotein or both, at intervals of 6 months or 12 months. They found that none of the surveillance strategies were cost-effective compared to no surveillance at a cost-effectiveness threshold of £20,000 per QALY gained, although the base case ICER for 12-monthly surveillance using alpha-fetoprotein only was £22,353 for people with hepatitis C. The GDG had a concern about the validity of the mortality rates associated with large tumours used as a basis of the economic calculations in this paper. The GDG acknowledged that the economic evidence presented did not support regular HCC surveillance at a threshold of £20,000 per QALY, however it highlighted that none of the presented models took into account the costs of non-HCC complications associated with cirrhosis (ascites, hepatic encephalopathy, variceal bleeding). In addition, in the case of Cucchetti, the GDG noted that differences in the healthcare system make the results less applicable to the UK. The GDG noted that the clinical evidence tended to be from more recent papers than the economic evidence. Although all the current interventions for HCC were available at the time the economic papers were published, the GDG agreed that these interventions may have been refined and the effectiveness of the interventions may have improved. The original economic modelling conducted for this guideline (see Appendix N) included 12-monthly surveillance using ultrasound with or without AFP in the base case, and investigated the cost-effectiveness of reducing this to every 6 months in a sensitivity analysis. Six-monthly surveillance was cost-effective at the threshold of £20,000 per QALY gained for the HCV genotype 1 cohort. For the other populations the ICERs for 6-monthly compared to annual surveillance varied between £20,128 and £28,352per QALY gained in each case. The second and third highest ICERs were for the HBV population (surveillance is already recommended for this group in the NICE Hepatitis B guideline CG165141). |
| Quality of evidence | Observational, non-randomised studies were included in the absence of evidence from RCTs. As pre-specified in the review protocol, only observational studies which performed a multivariate analysis to adjust for confounding factors were included. Non-randomised studies reporting the characteristics of HCC (such as lesion size or the number of people with an advanced HCC stage) without adjusting for confounders were excluded. All the observational studies identified were retrospective cohort studies in people with a diagnosis of HCC, analysed by their previous surveillance status prior to the diagnosis of HCC. Therefore, only people who developed HCC were analysed. No observational studies were identified in people without HCC, followed up to see if HCC developed. Studies of screened patients who develop HCC compared to unscreened patients who develop HCC are subject to lead-time bias. Lead-time bias is the apparent increase in survival that comes exclusively from diagnosis at an earlier stage of disease. The duration of survival from diagnosis to death is increased, even if no intervention is applied and some of the survival benefit could be ascribed to earlier diagnosis. Some of the included studies attempted to correct for this bias by calculating the lead time and adjusting for it. This was taken into account when assessing the risk of bias. Studies are also subject to length bias. This arises from the fact that surveillance is more likely to detect slow growing cancers than rapidly growing cancers, which might go from undetectable to death within the surveillance interval. As the studies only include patients with HCC detected with or without a surveillance program the magnitude and effect of false positives cannot be measured. Surveillance versus no surveillance Six observational studies were identified for this comparison. Evidence was available from a multivariate analysis for the outcomes of survival and HCC stage. However, all the evidence was graded Low or Very Low quality. The main reasons for downgrading the quality of the evidence for risk of bias and for imprecision. None of the studies adjusted for lead-time bias. Yearly versus 6-monthly surveillance One observational study was identified comparing yearly surveillance with 6-monthly surveillance in people with cirrhosis. No RCTs were identified for this comparison. Evidence was available from a multivariate analysis for the outcomes of survival and HCC stage. However, all the evidence was graded Low or Very Low quality. This study did adjust for the lead-time bias. 3-monthly versus 6-monthly surveillance One RCT was identified comparing 3-monthly surveillance with 6-monthly surveillance in people with cirrhosis. Evidence was available for all protocol outcomes with the exception of quality of life. Evidence for the critical outcome of survival was of Moderate quality. Evidence for all other outcomes ranged from Moderate to Very Low quality. Overall considerations on the quality of the evidence There was a lack of High quality evidence and RCTs identified in this area to support recommendations made by the GDG. The GDG noted that very few RCTs are available in this area due to the ethical considerations of randomising patients to an arm without surveillance. Those few RCTs that do exist were excluded as they were in patients with HBV and not all the population had cirrhosis. The GDG chose to only include in the protocol studies in which more than 85% of the population investigated had cirrhosis. The GDG reviewed the search strategy and discussed the excluded papers. The main reasons for exclusion were the absence of, or small proportion of, patients with cirrhosis and/or the presence of more than 15% of patients with HBV. A number of studies were excluded if they had not adjusted their outcomes for other confounding factors (such as the severity of liver disease or the frequency of decompensations). |
| Other considerations | The GDG discussed that the evidence for the benefit of HCC surveillance is dependent on the effectiveness of current surveillance strategies. For example, the accuracy of ultrasound may be reduced in patients with NASH or obesity and will also be dependent on the extent of the cirrhosis. The quality of the ultrasound scan is also operator-dependent. However, the GDG agreed that ultrasound is still the favoured option for surveillance. Biomarkers such as AFP can aid diagnosis of HCC, but it is thought that only around 60% of HCCs are AFP-secreting. The accuracy of AFP would also be reduced in certain aetiologies such as alcohol-related cirrhosis. The GDG also felt that an important clinical aspect of ultrasound surveillance was not only the detection of HCC, but also the assessment for other complication of cirrhosis, such as portal hypertension, portal vein thrombus and ascites. It was discussed that surveillance for HCC could have further benefit if used as part of an integrated package of surveillance for other complications of cirrhosis. The GDG noted that recent changes in HCV treatment may have some impact on the necessity of surveillance in the population with HCV who have achieved viral clearance. At the current time the level of impact this may have on surveillance programmes was unknown. The GDG noted that the current system is an ‘ad hoc’ clinician-initiated surveillance programme. There was a general consensus that if surveillance is thought to be effective then it should be run in an organised fashion at an institutional level to ensure all patients are offered the opportunity to take part. The GDG noted recent statements from the UK Royal College of Radiologists and the United States National Cancer Institute highlighting the lack of High quality evidence to support routine HCC surveillance. However, it is very unlikely that an RCT of surveillance versus no surveillance will be undertaken on which to base recommendations. Overall, the GDG agreed that there was evidence that surveillance can detect HCC at an earlier stage and would give patients the opportunity to receive potentially curative treatment. The risks and benefits of surveillance should be discussed with the patient. The GDG noted that there were certain groups that would not benefit from surveillance as they are not eligible for treatment or for transplantation, such as those on an end of life strategy. |
- Surveillance for the early detection of hepatocellular carcinoma (HCC) - Cirrhos...Surveillance for the early detection of hepatocellular carcinoma (HCC) - Cirrhosis in Over 16s
Your browsing activity is empty.
Activity recording is turned off.
See more...