SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

The purpose was to assess the cost-effectiveness of sorafenib in the treatment of hepatocellular carcinoma (HCC) patients incorporating current prices and the results of the recent published field practice SOraFenib Italian Assessment (SOFIA) study. We created a Markov Decision Model to evaluate, in a hypothetical cohort of Caucasian male patients, aged 67 years with Barcelona Clinic Liver Cancer (BCLC) C HCC, or BCLC B HCC who were unfit or failed to respond to locoregional therapies, well compensated cirrhosis, and with performance status 0-1 according to Eastern Cooperative Oncology Group (ECOG), the cost-effectiveness of the following strategies: (1) full or dose-adjusted sorafenib for BCLC B and C patients together; (2) full or dose-adjusted sorafenib for BCLC B patients; (3) full or dose-adjusted sorafenib for BCLC C patients. Outcomes include quality-adjusted life years (QALYs), costs, and incremental cost-effectiveness ratio (ICER). In the base-case analysis dose-adjusted sorafenib was the most effective of the evaluated strategies. For dose-adjusted sorafenib, QALY was 0.44 for BCLC B and C patients together, 0.44 for BCLC C patients, and 0.38 for BCLC B patients. The ICER of dose-adjusted sorafenib compared with BSC was €34,534 per QALY gained for BCLC B and C patients together, €27,916 per QALY gained for BCLC C patients, and €54,881 per QALY gained for BCLC B patients. Results were sensitive to BSC survival rate, and sorafenib treatment duration. Conclusion: In daily practice dose-adjusted, but not full-dose, sorafenib is a cost-effective treatment compared to BSC in intermediate and advanced HCC. (HEPATOLOGY 2013)

Hepatocellular carcinoma (HCC) is the fifth most common cancer worldwide with more than half a million new cases each year with an increasing incidence in the USA and Europe, and contributes substantially to healthcare spending. Most patients with HCC (∼50%) diagnosed at the Barcelona Clinic Liver Cancer (BCLC) are B (intermediate) and C (advanced) stages. The prognosis is grim for advanced HCC or intermediate HCC patients who are contraindicated for or do not respond to locoablative treatments like transarterial chemoembolization (TACE) or radiofrequency ablation (RFA). 1-3

Sorafenib (Nexavar, Bayer Healthcare Pharmaceuticals-Onyx Pharmaceuticals) is an oral multikinase inhibitor that restrains tumor blood vessel development and tumor cell proliferation. Clinical efficacy data came from the Sorafenib Hepatocellular Carcinoma Assessment Randomized Protocol (SHARP) study, which was a registration multicenter, double-blind, placebo-controlled randomized trial in patients with advanced HCC. 4 The evidence indicated that for patients with Child-Pugh class A and Eastern Cooperative Oncology Group (ECOG) performance status of 0-2, with BCLC stage C, or B who failed locoregional therapies, sorafenib started at 400 mg twice daily improved survival on average by 83 days relative to placebo. Because the SHARP trial was stopped early for benefit, the efficacy and risks of sorafenib could not be fully assessed. 5 Additional data on the safety and effectiveness of sorafenib in daily practice are also required.

To help inform clinical and policy decisions about effectiveness, safety, and cost-effectiveness of sorafenib, we conducted the SOFIA (SOraFenib Italian Assessment) study—a field practice prospective study in Italy. Overall, the SOFIA study 6 confirms the effectiveness of sorafenib with a lower safety profile than that of the SHARP trial, also showing that a significant proportion of patients needed adjustment in the dosage of sorafenib. Nevertheless, multivariate analysis demonstrated an increase in survival rates for those patients who, following adverse events or presence of comorbidities, received dose-adjusted sorafenib (400 mg daily) for ≥70% of the treatment period compared with those who were instead full-dosed (800 mg daily). Based on the SOFIA study, we did a cost-effectiveness analysis of full versus dose-adjusted regimens of sorafenib for intermediate/advanced HCC.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Summary Data from the SOFIA Study.

The SOFIA study 6 is a multicenter, prospective, observational, noninterventional study to assess the safety and effectiveness of sorafenib in patients with advanced HCC and patients with an intermediate HCC who were not eligible to or failed ablative therapies. In all, 296 patients were consecutively evaluated. A total of 260 (88%) patients were in the Child-Pugh A class. None had ascites, clinically overt jaundice, or hepatic encephalopathy. At baseline, 115 patients (39%) had macroscopic vascular invasion by the tumor, whereas 104 (35%) had extrahepatic spread of the tumor. Overall, 222 (75%) patients were in BCLC-C stage and 74 (25%) in BCLC-B stage, including 26 (35%) who were unfit for locoablative treatment. Sorafenib treatment was interrupted in 103 (44%) for disease progression, in 95 (40%) for an adverse event, and in 38 (16%) for liver decompensation. By Kaplan-Meier test, the median survival was 10.5 months in the overall cohort, 8.4 months in BCLC-C versus 20.6 months in BCLC-B patients (P < 0.0001), and 21.6 months in the 77 patients treated for ≥70% of the time with a half dose versus 9.6 months in the 219 patients treated for <70% of the time with a full dose (Fig. 1). ECOG performance status, macrovascular invasion, extrahepatic spread of the tumor, radiologic response at month 2, and full dose sorafenib were independent predictors of shortened survival (further data from the SOFIA study are given in Supporting Table 1).

thumbnail image

Figure 1. Kaplan-Meier (stair-step line) and estimated (smooth line) survival curves of patients treated with full-dose (solid line) or dose-adjusted (dashed line) sorafenib, according to BCLC stage: (A) entire population (BCLC B and C together); (B) BCLC B stage; (C) BCLC C stage.

Download figure to PowerPoint

Effectiveness of Different Sorafenib-Based Treatment Strategies.

In the SOFIA study all patients started with the recommended dosage of 800 mg daily and dose reduction (400 mg daily) was carried out and maintained according to the recommendations provided by the manufacturer. The average received dosage of sorafenib in the whole population of the SOFIA study was 696 mg/die that was strictly similar to the mean received dose of the SHARP trial (710 mg/die). 7

In the group of patients treated with dose-adjusted sorafenib for ≥70% of the treatment period the average received dose was 474 mg daily (469 mg daily for BCLC B and 476 mg daily for BCLC C patients). Instead, in the group of patients who maintained full dose of sorafenib for the entire treatment period or received a dose-adjusted for <70% of the whole treatment period the mean received dose was 748 mg daily (723 mg daily for BCLC B and 754 mg daily for BCLC C patients).

In the SOFIA study the average actual received doses of sorafenib were 474 mg daily in the dose-adjusted group and 748 mg daily in the full-dose group. These average doses were strictly similar to the theoretical doses of 400 mg and 800 mg daily. Therefore, we also performed analyses according to these theoretical doses.

Sorafenib-based treatment strategies were evaluated according to BCLC (B or C) stage and sorafenib dose (full dose: 800 mg daily; dose-adjusted: 400 mg daily). The strategies analyzed were: (1) full or dose-adjusted sorafenib for BCLC B and C patients together (Fig. 1A); (2) full or dose-adjusted sorafenib for BCLC B patients (Fig. 1B); (3) full or dose-adjusted sorafenib for BCLC C patients (Fig. 1C).

Given that there are no other agents besides sorafenib that have demonstrated significant survival benefit or have been approved for this patient population by the Food and Drug Administration (FDA), all sorafenib strategies were compared to best supportive care (BSC). BSC incorporated medical staff visits, hospitalizations, and laboratory and radiology tests. Survival of patients who underwent BSC was modeled by application of risk ratios from a recent meta-analysis of 30 randomized controlled trials (RCTs) of untreated HCC patients enrolled in trials of palliative treatments. 3, 4

Structure of the Model.

Treatment effectiveness was modeled by application of Kaplan-Meier survival curves from the recent field practice prospective SOFIA study (6). We used a Markov model to simulate the costs and effects associated with sorafenib treatment and BSC over a 5-year time horizon. The model was designed to simulate cohorts of Caucasian male patients, 67 years old, with BCLC C HCC (75%), or BCLC B HCC who failed locoregional therapies (25%), well-compensated cirrhosis, and with performance status of 0-1, as included in the SOFIA study. The model comprised three health states: BCLC B HCC, BCLC C HC,C and death (Fig. 2). In such a model patients suffering an acute event could die during that month or survive (at least for that month). The health states were mutually exclusive, i.e., a patient could experience a single health state at any given time. For each transition, we obtained the time-dependent transition rates by assuming a Weibull distribution, parameters of which were estimated using available data (6). We used the Weibull distribution because the hazard function of this distribution may remain constant or may increase, fitting better the real clinical situation. Face validity and verification were assessed during model construction, debugging, and testing for internal consistency.

thumbnail image

Figure 2. General structure of the Markov model.

Download figure to PowerPoint

We used quality-adjusted life year (QALY) as the main health outcome and life year gained (LYG) as a secondary measure of effectiveness. QALYs were calculated by multiplying the time a person remained in a certain health state by the utility associated with that particular health state and subsequent summing up over all health states. Utility weights for the health states before disease progression (0.76) and after disease progression (0.68) (Table 1) were derived from the National Institute for Health and Clinical Excellence (NICE) technology appraisal guidance 178. 7

Table 1. Base-Case Values and Ranges Used in Sensitivity Analyses (± 30%)
VariablesBase CaseLower LimitUpper LimitReference
  1. BCLC, Barcelona Clinic Liver Cancer; mo, months.

Natural history
 Untreated BCLC B and C median survival (mo)7.95.5-10.3 3, 4
 Untreated BCLC B median survival (mo)117.7-14.3 3, 4
 Untreated BCLC C median survival (mo)6.94.8-9.0 3, 4
Median survival (mo) in treated patients
 Full dose for BCLC B and C10.17.7-14.4 6
 Dose-adjusted for BCLC B and C1611.2-20.7 6
 Full dose for BCLC B16.311.4-21.2 6
 Dose-adjusted for BCLC B17.412.2-22.6 6
 Full dose for BCLC C107-13 6
 Dose-adjusted for BCLC C1510.5-19.5 6
Median duration of sorafenib treatment (mo)
 Full dose for BCLC B and C3.02.1-3.9 6
 Dose-adjusted for BCLC B and C6.84.8-8.8 6
 Full dose for BCLC B4.63.2-6.0 6
 Dose-adjusted for BCLC B10.67.4-13.8 6
 Full dose for BCLC C2.82.0-3.6 6
 Dose-adjusted for BCLC C5.13.6-6.6 6
Time to progression (mo)9.26.4-12.0 6
Cost of each sorafenib capsule (€)31.820-36 9
Annual costs for disease progression (2012 euros) for Hepatocellular carcinoma6,438.794,507-8,371 10, 11
Base-case value of quality of lifeQ values  
Before disease progression0.760.76-0.80 7
After disease progression0.680.60-0.68 

A panel of local experts (three hepatologists and one expert in economic evaluations) was consulted to ensure that assumptions taken into consideration in the model reflected routine clinical practice.

Model creation and analyses were performed using R (R Foundation for Statistical Computing, Vienna, Austria) 8 and Microsoft Excel 2007 (Microsoft, Redmond, WA).

Resource Consumption and Costs.

The analysis was conducted from the perspective of a third-party managed-care payer in Italy. Hence, only direct medical costs were included. Indirect costs, such as lost earnings due to poor health, were not estimated. We conducted a costing analysis of the treatment strategies, calculating all costs in 2012 euros. Total cost per strategy was the unit cost multiplied by the quantity used. In particular, the drug cost (sorafenib) and the costs associated with disease progression (e.g., diagnostic exams, visits, hospitalization) were considered. Sorafenib is administered orally as 200 mg tablets. The recommended dosage is 400 mg twice daily (a total daily dose of 800 mg). The dosage may be adjusted to two 200 mg tablets once daily if adverse drug reactions are suspected. The summary of product characteristics recommends that treatment should be continued as long as clinical benefit is observed or until unacceptable toxicity occurs. In Italy the price from the factory for a pack of 200 mg tablets (112 tablets per pack) is €3,562 excluding value-added tax (VAT). 9

Estimates of annual direct costs for each health state included the frequency and costs of inpatient and outpatient visits, diagnostic and laboratory testing, medications, and procedures. These costs were updated based on a previous study 10 in which the medical resource use associated with each disease state was estimated based on the DRG tariffs11 and national ambulatory fees. The drug costs and costs associated with disease progression are reported in Table 1. Future costs and life-years were discounted at 3% per year.

Cost-Effectiveness Analysis.

We calculated the incremental cost-effectiveness ratio (ICER) of the different sorafenib-based treatment strategies compared with BSC. The ICERs among the different strategies were calculated using difference in the drug and disease costs in 2012 euros divided by the difference in QALY or LYG effectiveness.

We performed a one-way sensitivity analysis to explore the impact of each variable on results. Analyses were done for survival of untreated patients, duration of sorafenib treatment, disease costs, discounting rate, and utilities. We also explored the impact of alternative survival distributions (lognormal, log-logistic, exponential) on the predicted survival probability. A Tornado diagram was used to represent and assess the relative weight of each variable on overall uncertainty in one-way sensitivity analyses.

Parameter uncertainty was dealt with by probabilistic sensitivity analysis using Monte Carlo simulation by randomly sampling a distribution of all variables 10,000 times and then simulating outcomes. Results from the probabilistic sensitivity analysis were presented as a cost-effectiveness acceptability curve.

To decide whether to perform an intervention it is necessary to choose a cost-effectiveness threshold: the amount of money that we are willing to spend to gain 1 year of life. There is no empiric evidence to support the choice of a particular threshold. However, the cutoff worldwide considered plausible in the developed world is $50,000 (which corresponds to about €38,000). 12

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Base-Case Analysis.

The results of our base-case analyses are shown in Tables 2 and 3, with the total costs versus LYG (Table 2) and QALY (Table 3) among the competing strategies.

Table 2. Results of Cost-Effectiveness Analyses: Life Year Gained (LYG)
Treatment Strategies According BCLC and DoseCosts in 2012 EurosLYGICER/LYG Base-Case Analysis (2012 Euros)
  1. BCLC, Barcelona Clinic Liver Cancer; ICER, incremental cost-effectiveness ratio.

Best supportive care4,142
Full dose for BCLC B and C16,0810.1863,197
Dose-adjusted for BCLC B and C19,9440.5925,874
Full dose for BCLC B24,2240.4044,794
Dose-adjusted for BCLC B26,9140.5041,782
Full dose for BCLC C14,8410.1859,922
Dose-adjusted for BCLC C16,6250.5920,896
Table 3. Results of Cost-Effectiveness Analyses: Quality-Adjusted Life Year (QALY)
Treatment Strategies According BCLC and DoseCosts in 2012 EurosQALYICER/QALY Base-Case Analysis (2012 Euros)
  1. BCLC, Barcelona Clinic Liver Cancer; ICER, incremental cost-effectiveness ratio.

Best supportive care4,142
Full dose for BCLC B and C16,0810.1669,344
Dose-adjusted for BCLC B and C19,9440.4434,534
Full dose for BCLC B24,2240.3257,385
Dose-adjusted for BCLC B26,9140.3854,881
Full dose for BCLC C14,8410.1665,551
Dose-adjusted for BCLC C16,6250.4427,916

The BSC strategy costs €4,142 on average for BCLC B and C patients considered together. It was, therefore, the least expensive, but also the least effective, of the competing strategies. The introduction of sorafenib in the entire population of the SOFIA study at the received mean dose of 696 mg/die increased the total cost significantly (€18,418), with a slight increase in effectiveness. Specifically, compared with BSC, the sorafenib treatment had an ICER of €47,796 for LYG and €58,456 for QALY. In the group of BCLC B HCC patients, the sorafenib treatment at the received mean dose of 705 mg daily had an ICER of €42,527 for LYG and of €55,242 for QALY. Instead, in the group of BCLC C HCC patients, the sorafenib treatment at the received mean dose of 682 mg daily had an ICER of €39,766 for LYG and of €48,009 for QALY.

In the group of patients treated with a dose-adjusted of sorafenib for ≥70% of the treatment period who received an average dosage of 474 mg daily, the sorafenib treatment had an ICER of €29,469 for LYG and of €39,332 for QALY (ICER for QALY of €62,889 for BCLC B and of €31,585 for BCLC C patients). In the group of patients who maintained full dose or received dose-adjusted sorafenib for <70% of the whole treatment period (an average dosage of 748 mg daily), the sorafenib treatment had an ICER of €59,508 for LYG and of €65,296 for QALY (ICER for QALY of €52,655 for BCLC B and of €62,186 for BCLC C patients).

In Tables 2 and 3, total costs, LYG, and QALY, according to BCLC stages, were reported using theoretical sorafenib doses of 400 mg daily for dose-adjusted strategies and of 800 mg daily for full-dose strategies.

In the whole population, the dose-adjusted strategy was more cost-effective than the full dose in terms of both LYG and QALY. Specifically, compared with BSC the full-dose strategy had an ICER of €63,197 for LYG and of €69,344 for QALY, while dose-adjusted strategy had an ICER of €25,874 for LYG and of €34,534 for QALY.

As in the entire SOFIA cohort, in the BCLC B patients the dose-adjusted strategy was more cost-effective than the full dose in terms of both LYG and QALY. Specifically, compared with BSC the full-dose strategy had an ICER of €44,794 for LYG and of €57,385 for QALY, while the dose-adjusted strategy had an ICER of €41,782 for LYG and of €54,881 for QALY.

Similarly, in BCLC C patients, considering both LYG and QALY, the dose-adjusted strategy was more cost-effective than the full dose. Specifically, compared with BSC the full-dose strategy had an ICER of €59,922 for LYG and of €65,551 for QALY, while the dose-adjusted strategy had an ICER of €20,896 for LYG and of €27,916 for QALY.

Performing analysis in the subgroup of the SOFIA cohort obtained after excluding patients with early radiologic progression, ICER per QALY in dose-adjusted sorafenib strategies marginally improved. Specifically in this subgroup of patients, dose-adjusted sorafenib strategy had an ICER per QALY of €25,569 for BCLC C and of €58,265 for BCLC B patients.

Sensitivity Analyses.

One-way sensitivity analysis was done for two dominant strategies: dose-adjusted sorafenib therapy for both BCLC B and C HCC patients. Figure 3 summarizes the results of one-way sensitivity analyses, using tornado diagrams.

thumbnail image

Figure 3. Tornado diagram for the one-way sensitivity analysis of variables that influenced the ICER per additional QALY of BCLC B (A) and BCLC C (B) at dose-adjusted strategies compared to BSC. The horizontal axis represents the incremental cost effectiveness ratio for QALY. The width of the bars illustrates the range of cost per additional QALY. Upper and lower limits of values evaluated in sensitivity analysis are indicated next to the bars. The bars are ordered from greatest width at the top to least width at the bottom.

Download figure to PowerPoint

Analyses showed that the results of the model were most sensitive to an assumption on survival rates of BSC patients, sorafenib treatment duration, and type of survival distribution.

Changes in survival rates in patients managed with BSC had a great effect on cost-effectiveness. In fact, sensitivity analysis with a hypothesized variation of survival of ±30% in BSC patients showed that ICER for QALY ranged significantly from €41,325 to €100,544 in BCLC B (Fig. 3A) and from €24,450 to €36,032 in BCLC C (Fig. 3B) patients treated with dose-adjusted sorafenib.

The cost effectiveness of dose-adjusted sorafenib was sensitive to change (±30%) in the treatment duration. With a longer time of therapy, the ICER for QALY impairs both the BCLC B and BCLC C patients. Instead, for the sensitivity analysis on the disease costs, a variation of ±30% was assessed, and the model had low sensitivity. With an increase in the disease costs, the ICER for LYG and for QALY marginally increased in both BCLC B (Fig. 3A) and BCLC C (Fig. 3B) dose-adjusted strategies. Lower variations were found for both strategies by applying a discount rate ranging from 0% to 5%.

The cost effectiveness of dose-adjusted sorafenib was also greatly sensitive to the type of survival distribution. Specifically, in both BCLC B and C patients the ICERs for QALY progressively increased from log-normal, to Weibull and further to exponential and log-logistic survival distributions.

Further sensitivity analysis was performed to evaluate whether the increase of the expected survival of HCC patients treated with full-dose sorafenib allowed achieving the willingness-to-pay threshold of €38,000 per QALY. Specifically, we observed that full-dose sorafenib reached cost-effectiveness if survival increased from 11.1 to 19.2 months (percentage increase of 74%) in BCLC B and C considered together, from 16.3 to 27 months (percentage increase of 67%) in BCLC B, and from 10 to 17.4 months (percentage increase of 74%) in BCLC C patients.

Probabilistic Sensitivity Analyses.

Varying all variables simultaneously in the Monte Carlo simulation, and using a willingness-to-pay threshold of €38,000 per QALY, dose-adjusted strategies were cost-effective in 68% and 9% of the simulations, in BCLC B and BCLC C HCC patients, respectively (Fig. 4).

thumbnail image

Figure 4. Cost-effectiveness acceptability curves representing the probability that each treatment strategy is cost-effective for a given maximum willingness-to-pay threshold per QALY gained. The vertical axis represents the cutoff of €38,000.

Download figure to PowerPoint

In a best-case scenario, the model generated an ICER of €16,464 per QALY for BCLC C and an ICER of €24,407 per QALY for BCLC B HCC. In a worst-case scenario, the model generated an ICER of €119,673 per QALY for BCLC C and an ICER of €1,295,276 per QALY for BCLC B HCC.

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

In the present study, we show that in patients with advanced/intermediate HCC included in the SOFIA study, 6 sorafenib at the recommended dose of 800 mg daily is not a cost-effective treatment compared with BSC. Our analysis, in agreement with the NICE recommendation,7 relies on data from the single available manufacturer-sponsored SHARP trial4 demonstrating that sorafenib, within its licensed dose and indication, is not recommended for the treatment of patients with advanced HCC. In addition, we showed that full-dose sorafenib was not cost-effective also when considering separately BCLC B and BCLC C HCC patients. However, we demonstrated that in field practice dose-adjusted sorafenib is cost-effective compared with BSC in the treatment of BCLC B and C HCC patients considered together, improving survival by about 0.59 LYG and 0.44 QALY. This gain came at an acceptable cost, resulting in an ICER of €34,000 per QALY gained, which was lower than the generally accepted societal threshold for willingness to pay.12

According to the SOFIA study, which differs from the SHARP trial 4 in terms of rates of poor tolerability, leading to dose reduction or discontinuation, and that provides indirect evidence of effectiveness of half-dose sorafenib, we hypothesized that sorafenib could be utilized in a more cost-effective manner in the setting of intermediate/advanced HCC. Specifically, the key issue is the identification for both BCLC B and C HCC of an effective but not toxic sorafenib dose. The availability of individual patients' data makes it possible to assess the cost-effectiveness of treatment according to sorafenib dose and BCLC stage.

Our base-case analysis suggests that dose-adjusted sorafenib was cost-effective in advanced but not in intermediate HCC. Our results are of further relevance considering that most patients with HCC are diagnosed at an advanced stage and that this group included the majority of patients who were candidates for sorafenib treatment. By contrast, BCLC B patients who failed or were not suitable for locoregional treatments are a heterogeneous group of patients, representing a small subgroup of HCC patients who are candidates for sorafenib treatment, and not fully representative of the entire BCLC B spectrum.

The robustness of these results was confirmed in the probabilistic sensitivity analysis, showing that the probability of dose-adjusted sorafenib providing a cost-effective alternative to BSC was about 70% for advanced and 10% for intermediate HCC at a willingness-to-pay threshold of €38,000 per QALY. The cost-effectiveness of dose-adjusted sorafenib in both advanced and intermediate HCC was sensitive to the BSC survival rate, sorafenib treatment duration, and the choice of parametric model used to predict survival gain. It is important to highlight that dose-adjusted sorafenib remains cost-effective in BCLC C HCC patients under a wide range of survival of untreated patients. This issue is very relevant, given the high heterogeneity and the unpredictability of clinical behavior of advanced HCC. 3

On the efficacy side of the cost equation, identification of robust predictive biomarkers would increase the overall efficacy of sorafenib for HCC by treating only those patients most likely to respond. This is not a trivial point, because sorafenib therapy is costly and still in search of optimization due to the lack of serum biomarkers of early response that are deemed necessary to generate response-guided therapeutic algorithms. Moreover, the identification of stopping rules based on predictors of mortality (such as early radiologic progression) could assist the clinician in the cost-effective management of patients with HCC. Unfortunately, stopping sorafenib in patients with early radiologic progression only marginally improves the cost-effectiveness ratio. Therefore, the identification of the optimal sorafenib dose, effective but not toxic, remains to date the only strategy to substantially improve the ICER. Modeling the indication for treatment of advanced/intermediate HCC with dose-adjusted sorafenib clearly improves its cost effectiveness. In this line, the role of dose-adjusted sorafenib should be taken into account also for the future perspectives in the adjuvant setting after resection/ablation or after transarterial chemoembolization and for the design of future comparative trials versus novel targeted therapies.

Although the proposed algorithms are useful tools for decision making, the treatment strategy should be carefully agreed upon with the patient, taking into account all the different factors, particularly the safety profile, that can interfere with treatment response. In particular, the choice of treatment should be targeted to obtain the best efficacy in the individual patient without any economic analysis affecting the clinical value and ethical impact of this decision.

The strength of this study was the access to individual patient data. The analysis relied on data from a field practice prospective study enrolling patients with broad eligibility criteria reflecting the complexity and diversity of clinical practice in HCC. Instead, individual patient data from the SHARP trial are not available and also not directly transferable to clinical practice due to the fact that trial patients are more adherent, and more intensively monitored. Furthermore, it would be interesting to validate this analysis on individual data of the Global Investigation Of Therapeutic Decision In HCC and Of Its Treatments With Sorafenib (GIDEON) study, 13 an ongoing, large noninterventional study in HCC patients receiving sorafenib.

This study has several caveats. First, the model was run for 5 years instead of the complete lifetime. However, survival data are associated with increasing uncertainty as the time axis extends and in these circumstances it will be appropriate to exercise caution by modeling the more robust data from the earlier part of the study.

Second, how extrapolations were generated had considerable impact on estimates of survival gain, of time under treatment (costly for the intervention arm) and each in turn impinged on the cost-effectiveness estimates. Therefore, we explored the impact of several alternative parametric functions on the predicted treatment-dependent survival gain. Logarithmic models, assuming a decreased hazard through time, delivered double the survival advantage that was derived from Weibull model, which assumes an increased hazard over time. In our base-case analysis we selected the Weibull distribution because it seems more biologically and clinically plausible than logarithmic models.

Third, the study's perspective was not societal. Therefore, our analysis was limited to direct medical costs. If indirect costs were included, sorafenib treatment would be expected to become more cost-effective. However, indirect costs such as lost productivity and caregiver salaries probably have a low impact in this clinical setting.

Fourth, we used utility values from NICE for the treatment of advanced renal carcinoma with sorafenib or BSC. 7 It is well known that utilities may vary widely across different patient populations and depend critically on quality-of-life assumptions. Thus, it may not be the most appropriate approach to estimate utility scores.

In conclusion, full-dose sorafenib was shown not to be cost-effective in intermediate and advanced HCC patients. Instead, we found that dose-adjusted sorafenib is cost-effective in patients with advanced HCC over a wide range of model assumptions, but not in those with intermediate HCC who were not eligible to or failed locoregional therapy.

Author contributions: C. Cammà, G. Cabibbo, S. Petta, M. Enea, M. Iavarone, A. Grieco, R. Bruno, A. Gasbarrini, E. Villa, C. Zavaglia, M. Colombo, A. Craxì had full control of the study design, data analysis and interpretation, and preparation of the article. All authors were involved in planning the analysis and drafting the article. All the authors approved the final draft article.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information
  • 1
    Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 2005; 55: 74-108.
  • 2
    Llovet JM, Di Bisceglie AM, Bruix J, Kramer BS, Lencioni R, Zhu AX, et al. Design and endpoints of clinical trials in hepatocellular carcinoma. J Natl Cancer Inst 2008; 100: 698-711.
  • 3
    Cabibbo G, Enea M, Attanasio M, Bruix J, Craxi A, Camma C. A meta-analysis of survival rates of untreated patients in randomized clinical trials of hepatocellular carcinoma. Hepatology 2010; 51: 1274-1283.
  • 4
    Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, Blanc JF, et al. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med 2008; 359: 378-390.
  • 5
    Brassler D, Briel M, Montori VM, Lane M, Glasziou P, Zhou Q, et al. Stopping randomized trials early for benefit and estimation of treatment effects: systematic review and meta-regression analysis. JAMA 2010; 303: 1180-1187.
  • 6
    Iavarone M, Cabibbo G, Piscaglia F, Zavaglia C, Grieco A, Villa E, et al. Field-practice study of sorafenib therapy for hepatocellular carcinoma: a prospective multicenter study in Italy. Hepatology 2011; 54: 2055-2063.
  • 7
    NICE technology appraisal guidance 189 — Sorafenib for the treatment of advanced hepatocellular carcinoma www.nice.org.uk/guidance/TA189
  • 8
    R Development Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2011. ISBN 3-900051-07-0, URL http://www.R-project.org/
  • 9
  • 10
    Munari L, Picciotto A. Recombinant interferon alfa-2b therapy for chronic hepatitis C in Italy: an economic analysis. FORUM Trends Exp Clin Med 1996; 6: 347-353.
  • 11
    Conferenza delle Regioni e Provincie Autonome. Tariffa Unica Concezionale per le prestazioni di assistenza ospedaliera. Regole e tariffe valide per il 2006. Rome, 15 December 2005.
  • 12
    Owens DK. Interpretation of cost-effectiveness analyses. J Gen Intern Med 1998; 13: 716-717.
    Direct Link:
  • 13
    Lencioni R, Marrero J, Venook A, Ye SL, Kudo M. Design and rationale for the non-interventional global investigation of therapeutic decisions in hepatocellular carcinoma and of its treatment with sorafenib (GIDEON) study. Int J Clin Pract 2010; 64: 1034-1041.

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
HEP_26221_sm_SuppTab1.doc31KSupporting Information Table 1.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.