Trans-arterial radioembolization for intermediate-advanced hepatocellular carcinoma: a budget impact analysis

Liver cancer is one of the most frequent cancers in the world, with a 5-year prevalence of 633,000 cases, 782,000 new diagnoses in 2012, causing more than 700,000 deaths globally per annum [1]. Hepatocellular carcinoma (HCC) is the most frequent type of liver cancer, accounting for 90% of all liver cancers [2].

In order to guide the therapeutic approach and to predict the prognosis of patients with HCC, different staging systems are used. The BCLC (Barcelona-Clinic Liver Cancer) classification is considered the standard system by the American Association of for the Study of Liver Disease (AASLD) [3] and European Association for the Study of the Liver [4]. The system identifies patients with early HCC (stage 0 and A), intermediate (stage B) or advanced stage (stage C) and those with very poor life expectancy (stage D).

Treatment schedules are recommended for each stage, ranging from curative therapies, such as resection or transplant for early stage patients, to best supportive care for terminal patients.

Intra-arterial transcatheter embolotherapies are recommended for non-surgical patients in the intermediate HCC stage, while sorafenib is the standard systemic herapy for patients with advanced HCC and well-preserved liver function and those with intermediate-stage HCC who progress following trans-arterial chemoembolization (TACE). However, in a sub-analysis of trials involving sorafenib, the tolerability of this treatment resulted suboptimal [5]. This situation opened the way to new therapies for the management of intermediate or advanced-stage HCC and, in this setting, transarterial radioembolization (TARE) showed to be a valuable therapeutic option [6‐10].

TARE is one type of intra-arterial brachytherapy used to treat HCC. TARE is performed using glass (TheraSphere®, MDS Nordion Inc.) or resin (SIR-Spheres®, Sirtex Medical Inc.) microspheres including β-emitter Y-90. The potential clinical benefits of TARE for the treatment of HCC are under investigation. Its organizational and economic impacts should also be carefully evaluated as TARE is a complex and expensive intervention.

One of the main issues in the evaluation of the cost-effectiveness of locoregional treatments for HCC is the lack of published randomized clinical trials’ results. Decision models may help filling this gap since they allow pooling information from different sources to perform cost-effectiveness analyses. In the literature few studies using this approach are reported. Chaplin and colleagues [11] performed a cost-effectiveness analysis of TARE compared to sorafenib for the treatment of HCC in the UK. The study showed that TARE yielded a total lifetime cost lower than sorafenib (£21,441 vs. £34,050) with a quality adjusted life year (QALY) gain of 0.27 (TARE dominant). Another study assessed the cost-effectiveness of TARE in comparison to conventional transarterial chemoembolization in the United States [12], showing lifetime costs of $31,000 and $48,000 for unilobar and bilobar radioembolization, respectively. For advanced stage patients, the incremental cost-effectiveness ratio of TARE versus TACE resulted 360$ per month (3120$ per year). A more recent study performed in Italy [13] reported for intermediate stage patients an incremental cost-utility ratio (ICUR) for TARE vs. sorafenib of 3302€/QALY, whilst for advanced stage patients, TARE seemed to be a dominant strategy (lower costs and greater health improvements) compared to sorafenib.

But what does this evidence bring to the financial prospects of the national healthcare budget? Although few studies assessed the cost-effectiveness of TARE in comparison to other locoregional or systemic treatments, the budget impact analysis (BIA) of the introduction of this technology at national level is still unexplored. The BIA is an essential component of a complete economic assessment of health technologies aiming at estimating the financial consequences, in the short-medium term, on the total healthcare national budget derived from the diffusion of a new therapeutic intervention, in combination with treatments already used for the management of a particular disease within a specific healthcare system [14].

The cost-effectiveness model developed by Rognoni et al. [13] has been applied to perform a BIA considering increasing utilization rates of TARE in place of sorafenib for the treatment of intermediate-advanced stage HCC patients in the Italian healthcare system over a 5-years horizon.

HCC is a life-threatening disease. Despite considerable improvements in the diagnoses and treatments, the available cure options are only partially effective [24] and the disease is very difficult to control when presenting in the advanced stage [3, 25]. Within a resource-limited healthcare system, this context highlights the need for resources to be used efficiently in order to guarantee the best outcomes to this population.

Real world clinical data of two cohorts of patients, treated with TARE or sorafenib, matched according to Child-Pugh score, presence or absence of PVT and number of nodules, were used to populate a model in order to estimate lifetime costs and health outcomes. In the final matched cohorts, intermediate stage patients undergoing TARE and sorafenib yielded mean survivals of 24.0 (median 18.5) and 18.4 months (median 13.0), respectively. In the group of advanced stage patients these values decreased respectively to 14.9 (median 11.2) and 16.1 (median 11.3) months. Patients’ survivals in the TARE group appear lower than other published figures data [9, 26, 27]. The propensity score matching selected pairs of patients with similar clinical characteristics, eligible indifferently to TARE or sorafenib treatments, favoring the selection of patients in worse clinical conditions and, indeed, with lower life expectancy. As regards the costs, the evaluation focused on first-line treatments, control visits and examinations, management of side effects (liver decompensation) and subsequent treatments (second line treatments after TARE or sorafenib).

The budget impact analysis showed that the Italian Healthcare Service could save about 7 million Euros in the hypothesis of an increased utilization of TARE, from 20 to 50%, in place of sorafenib in the next 5 years. The model robustness has already been tested performing a number of sensitivity analyses [13], however, two scenario budget impact analyses showed reduced savings (about 1,150,000€) in case of 1.5 TARE treatments performed per patient or an increase in the national healthcare budget (about 4,800,000€) in case of halved sorafenib cost. It should be noted that these scenarios are unlikely to be representative of the actual practice but can give information on the budget variations in extreme cases.

Budget impact analysis is an important component of the economic evaluation of healthcare interventions, which is gaining significant relevance in formal health technology assessment systems in place across several jurisdictions and certainly in Italy. This is the first study estimating the impact on the Italian healthcare budget of TARE and sorafenib strategies in intermediate-advanced stage HCC patients. A previous study tried to evaluate the patterns of treatment and costs related to HCC treatments, reporting an overall expenditure of 12,215€ for sorafenib and 26,106€ for TARE patients per year. However that analysis was performed from the hospital perspective and based on structured interviews with physicians in four Italian centers [23]. Another study [28], presented the results of a BIA from a hospital perspective in Canada. This study showed that, for a hospital managing 200 HCC patients annually, an increased use of TARE over TACE and sorafenib could incur savings of approximately $37,000, $55,000 and $75,000 in years 1, 2 and 3, respectively.

This study has some limitations. First of all, costs were estimated from the Lombardy Region perspective, and then considered as a proxy for the other Italian Regions. In a decentralized system such as the Italian NHS, one should consider separately each different Region to entirely capture all the specific features of the healthcare service provision and costs [29]. However, most of Italian Regions performing TARE have a special DRG reimbursement rate according to the use of Y-90 microspheres (DRG 409) and in these regions TARE reimbursement tariffs are quite similar (e.g. 8500€ for Emilia Romagna, 8568€ for Piedmont, 9510€ for Lombardy) meaning that the model could be considered reasonably conservative in this respect. Moreover, the healthcare resource consumption has been expressed in natural units (Table 1), as suggested by EunetHTA [30], to allow the model extension to other countries. Furthermore, the budget impact does not take into account the sunk costs of setting up the TARE procedure in a new organization. However, provided the angiographic room and imaging equipment are already available, these costs will refer mainly to a thorough training of the staff [31].

The consumption of healthcare resources has been retrieved from clinical data only for cancer related therapies (i.e. duration of treatment with sorafenib, mean number of TARE treatments per patient and subsequent treatments) while a predefined schedule, although validated by the focus group, for visits and examinations was applied for the follow-up period. This approach might have underestimated the real healthcare resource consumption. Analysis of data on best supportive care offered to patients after treatment failure resulted mainly in the use of off-label drugs or chemotherapy, with no reported indication of dose and duration, and an evaluation including this aspect was not possible. Even though liver decompensation was highlighted by the focus group as the main and most expensive adverse effect caused by TARE or sorafenib, other adverse events could have an impact on both costs and patient’s quality of life. More data on the occurrence of other side effects would allow for a more comprehensive evaluation.

As regards epidemiological data on prevalence and incidence of HCC in intermediate and advanced eligible for TARE and sorafenib therapies, as well as the current mix of treatment strategies in the eligible population, we relied on the largest source in Italy, the ITALICA registry, reporting data updated to 2014. Considering that TARE is an emerging treatment showing a promising efficacy in terms of disease control with a good safety profile [32], it is likely that its diffusion has been underestimated. In this regard, the results shown may have overestimated the potential savings due to the diffusion of this advanced medical device technology occurred since the last update of the ITALICA registry (2014).

Radioembolization can be considered a valid treatment option, giving an “additional” chance of survival for patients with intermediate-advanced hepatocellular carcinoma. The attitude towards this type of treatment is usually positive, while eventual side effects are considered tolerable. The present study adds evidence about the economic sustainability of TARE in comparison to standard systemic chemotherapy, sorafenib, at national level, showing that a decrease of the Italian healthcare budget is possible through an increase of the diffusion of this advanced medical device technology. Further prospective studies and increased awareness around the cost-effectiveness profile of healthcare technologies in this area will be able to provide additional data to confirm our conclusions.