In Europe, all member countries assess pricing and reimbursement of new drugs individually. Health-technology assessment and decision-making are based on efficacy and safety profiles from randomized controlled trials (RCTs) evaluating new drugs, and the allocation of scarce resources. Post-marketing studies help to assess differences between ‘efficacy’ and ‘effectiveness’. In clinical trials, the studied drug or treatment must adhere to strict operational conditions. For example, good clinical practice, patient’s inclusion and exclusion criteria, and efficacy or toxicity are well defined in the treatment groups. Studies of the same treatment in the general population may yield considerable differences from an RCT because the treatment conditions are less well defined, the number of participants is greater, and diagnoses and treatments are more varied.
Randomized-controlled trials involve a limited number of carefully selected participants, and treatment schedules, compliance of patient doses, and intervals are well monitored. Strict clinical diagnostic and pharmacological controls are used. Even an RCT, however, could fail to show relevance in modifying the natural history of a disease in clinical practice. Clinical outcomes in the general population tend to be more variable and treatment conditions less selected; therefore, clinical outcomes may differ. In clinical practice, the entire population is treated compared with specific target groups, such as children or elderly people. Consequently, the following may occur:
Less frequent detection of adverse drug reactions, i.e. one in 1,000 treated patients
More frequent side effects owing to age or associated disease
Differences in various ethnic groups
Frequent errors and failures associated with doses and schedules
Non-recording of patient compliance
Furthermore, many oncology treatments can be associated with multiple diseases and polypharmacy. Therefore, the real cost–benefit value should be revised once effectiveness data from clinical practice are available.
When a drug enters the market after approval, clinical trials may leave residual but important uncertainties from the point of view of the prescriber and purchaser. For example, the effectiveness may differ from that which was established in an experimental setting, and the frequency and nature of adverse events may differ. The Italian national health system (Servizio Sanitario Nazionale, SSN), therefore, needs new tools to determine the price of new drugs and health technologies.
The Italian and regional health system
The SSN covers all citizens and legal foreign residents. The SSN replaced a Bismarckian system of health insurance funds in 1978, and was modelled on the UK National Health Service (NHS) to provide uniform comprehensive care. Prescription drugs are divided into three tiers (A, H, and C) according to clinical effectiveness and, in part, cost-effectiveness; the SSN covers the first tier in all cases (A) and the second tier (H) only in hospitals, but does not cover the third tier (C). In Italy, reimbursement decisions for cancer drugs are taken by one regulatory body, the SSN, in consultation with the (Agenzia Italiana del Farmaco, AIFA). In Italy, almost all high-cost anticancer drugs are listed as Class H drugs (hospital-only prescribing drugs); therefore, significant savings can be achieved because drugs bought directly by hospitals are granted a minimum 50% discount by pharmaceutical companies , or 33% rebate (ex-factory price), when approved by EMA.
The financing and delivery of care in Italy are separate, with central government generally determining regional budgets and the regions deciding how to organize and deliver care . Although the central government determines the required minimum benefit package, and mostly controls the distribution of tax revenue, the 20 regions have responsibility for organizing and delivering health services.
Regions are financed by central government to assure the basic health benefits package (livelli essenziali di assistenza), and can choose to offer non-basic services, but must finance these themselves. Regions can raise their own revenue, but have limited capacity to do so. The division between financing and delivery creates tension, as the regions claim the government under-budgets and the government claims the regions need greater cost control. The regions have significant autonomy on the revenue side of the regional health budget, and are required to fund any deficit that might occur from their own resources, beginning with the 1992–1993 reforms. The regions organize services that are designed to meet the needs of their specific populations, define ways to allocate financial resources to all the local healthcare authority within their territories, monitor healthcare services and activities provided by the local healthcare authorities, and assess their performance .
The 1978 health reform set up the SSN according to the principles of universal coverage and a fully tax-based public healthcare system . During the late 1990s, Italy’s administrative and institutional settings became more federalized. The progressive move towards fiscal federalism started in 1997, with the abolition of social insurance contributions and the introduction of a regionally collected system of tax financing. General taxation was left to play a complementary role; its main role was to redistribute resources to regions with a narrower tax base to ensure that all residents receive adequate levels of care.
In Italy, as with most member countries of the organization for economic cooperation and development, healthcare expenditure has steadily increased over time, making its containment a major issue for successive governments. The existence of a large public deficit, and the need to reduce it drastically to comply with the requirements of the European Economic and Monetary Union, added further pressure to control healthcare expenditure. The central government’s response to the disconnection between funding responsibility and spending power created a situation of permanent financial crisis. The chronic regional deficits reflected two tendencies of central government policy: to underestimate the funding needs of the SSN systematically, although not overtly, and to overestimate the savings to be obtained from expenditure-containment strategies .
In recent years, deficits have been the subject of heavy inter-governmental negotiation. The central government attempted to limit these deficits by asking regions to underwrite yearly ‘pacts for health’ (accordi), which tied additional resources to the achievement of healthcare planning and expenditure goals. In addition, regional governors were requested annually to balance the books in healthcare expenditure; failure to do so might refer the administration of the region to an external commissioner nominated by the regional government. In recent years, expenditure has increased, largely on drugs. This increase in spending is attributed mainly to the loss of co-payment revenue, and partly to an increase in consumption (+ 17.4%). This resulted in reduced restrictions on prescribing and prescriptions of newer and more expensive drugs (+ 2.3%).
After the 1992–1993 reforms, the rate of growth in healthcare expenditure declined substantially, with northern regions more successful in containing real per capita expenditure growth. A main feature of Italy’s healthcare system is the presence of deep regional inequality in healthcare expenditure and in the supply and utilization of healthcare services.
Historically, health-budget deficits have been a major problem in most Italian regions. Since the early 2000s, however, the introduction of efficiency measures and tighter procedures on financial management have contributed to a significant decrease in the regional health budget deficit. A 2008 financial law established that regions will be financed through ‘standard costs’. Standard costs, although not yet operationally defined, would be the sectorial and overall costs of services provided in the regions that are considered the most efficient and effective.
Once the efficacy of a particular drug has been considered, the economic effect on the regional healthcare budget needs to be evaluated. As costs rise, countries have opted to control utilization by restricting off-label use or limiting reimbursement to subpopulations with the greatest benefits as determined by cost-effectiveness analysis .
People with cancer have an increased chance of survival if they are exposed to all available agents, but this benefit comes at a high cost, and may exceed regional health-service budgets. Therefore, all prescriptions for eligible patients need to be evaluated carefully, and a broader screening campaign might be implemented to contain costs of late tumour detection.
Regional control of appropriate drug use
Hospital pharmaceutical spending has also been under control of the SSN since its inception. According to national laws, regions have the power to set up their own regional hospital formularies based on the national therapeutic formulary, with the intention of monitoring and rationalizing the use and consumption of drugs. In order to establish its own formulary, each hospital within the region refers to the regional hospital formulary. Efficacy and safety criteria are applied for the selection, with priority being given to drugs for acute and chronic diseases. Pharmacoeconomic aspects are taken into account, especially when considering expensive drugs. A special class of drugs (class H) is used exclusively in hospitals; drugs included in class H are selected on the basis of certain types of pathologies and side effects in relation to benefits. Most antineoplastic agents are examples of class H drugs, which of course cannot be excluded from hospital formularies. Tight budgets and the need to restrain rising healthcare expenditures have led the SSN to undertake several cost-containment measures to encourage cost-conscious behaviour by consumers and providers.
Recently, the National Institute of Health and Clinical Excellence (NICE) in the UK has rejected four drugs, among those sunitinib and sorafenib, for use across the NHS for advanced renal cancer, claiming that they are clinically but not cost-effective .
Many studies have shown that our society can no longer sustain the traditional practices of the pharmaceutical industry [7, 8]. Companies invest in the drug market because of its financial attractiveness, mainly because patents and marketing authorization grants them a temporary monopoly that protects their markets and profits during a fixed period. Marketing authorization requires companies to demonstrate a favourable risk–benefit ratio, using a highly standardized process of drug development, with a central role given to RCTs. In this context, the risk-sharing principles are simple: companies know that they may face, at best, a restriction of indications, at worse, a withdrawal from the market and legal actions. In this early phase, no provision has been made to share the innovation cost. The payer organizes the solvency of the market, expects that the price will decline once the patent has expired, and may pressure the drug companies to lower prices earlier if competitive products are produced.
Studies have shown that public health policymakers are resistant to applying rigorous health-economic evaluations in the decision-making process. This is because the available healtheconomic information is often framed in a way that cannot be directly applied to a specific population . Economic evaluation studies may focus on cost-effectiveness ratios or qualityadjusted life years gained; however, it may be difficult to apply these in the ordinary healthcare domain because they rarely provide data localized to the specific area or population . The choice of well-specified points of view is an essential ingredient in the critical assessment of economic evaluation .
For decision makers, economic and clinical evaluation of a new intervention is more effective when information is presented in the general terms of a cost of illness analysis, or, in a disaggregated way, by means of a list of the costs, outcomes, or consequences of the intervention, namely as a cost-consequence analysis . Furthermore, some countries conduct economic evaluations at a national level, but health-economic data and evaluations are lacking at a regional level. This disconnects economic evaluations from the types of decisions that local institutions must normally face . The SSN costs in Italy are controlled by regional governments (which are financially accountable for healthcare expenditure, including hospital drug budgets), but agreements about price and discounts are decided by AIFA. For this reason, some regional governments ask the pharmaceutical companies to produce a ‘health technology assessment’ before deciding to introduce a drug in the regional hospital formulary.
Sharing the risks of rising costs
As healthcare costs continue to increase, it is anticipated that, at the current growth rate, the total Italian healthcare expenditure will account for growing proportions of the gross domestic product by 2020. Chemotherapy agents and, in particular, targeted treatments such as bevacizumab, have been closely evaluated to determine which specific populations should receive the drugs to minimize societal use and cost. Many studies evaluating colorectal cancer and bevacizumab, for example, have addressed the specific and societal costs of these drugs. Most studies have concluded that more information is needed to assess the true implication of high-cost drugs with their proven benefit. In this scenario, advances in oncology treatment have presented significant financial implications for healthcare systems in economically developed countries at both national and regional levels [1, 5]. The current challenge for oncologists and pharmacists lies in the optimal integration of high-cost targeted treatments.
In oncology, expenditure on pharmaceuticals is rising more rapidly than on any other healthcare component, with costs of cancer products growing at 21% per annum in recent years. European governments and health authorities have introduced successive reforms and initiatives to address these challenging resource issues, including funding innovative drugs or using methods to share financial risk .
Such measures include initiatives to ensure lower prices through the introduction of generics and biosimilars, as well as allowing interchangeability of brands within a class. Reasons for rising costs and solutions for cost reduction by healthcare purchasers are shown in Figure 1. In this scenario, risk sharing is a contract between two parties who agree to engage in a transaction in which one party is ready to accept a reward or a penalty depending on observed performance . The obstacles include attempts by payers to postpone introduction of new technologies of unclear effectiveness, e.g. use of quality-adjusted life year by NICE in the UK. Private contribution refers to healthcare providers asking patients to contribute to the payment of the treatment to reduce the direct costs of interventions.
*Figure 1 pending to upload.
The traditional pre-2000 risk model described earlier has lost ground in favour of new methods designed to deal with limited healthcare resources. Financial risk sharing is a contract between two parties who agree to engage in a transaction either to control the drug budget or to accept a reward or penalty depending on observed performance .
Risk sharing seems to be a stimulating method for dealing with post-marketing uncertainties of new drugs. It allows companies to base their value arguments on preregistration clinical efficacy and also on post-registration effectiveness. Risk sharing proposes a controlled framework to deal with uncertainties while offering companies predictability. Moreover, risk sharing implies that companies will spend more resources on producing good knowledge on the expected value of their products, before or at the time of market entry, and those purchasers will be forced towards transparency in decision making over time. Purchasers see risk sharing as a way of reducing uncertainty around cost-effectiveness estimations of new treatments and also uncertainty relating to drug performance in such a way as to control the available budget. Use of this strategy allows companies to gain or accelerate access to market, even when a treatment has an unclear result in practice. Often, such agreements are confidential and the conditions may not be disclosed.
The situation in Europe is far too complex for a detailed analysis, and is beyond the scope of this discussion. What can be said, however, is that a number of systems are in existence, see Table 1, of which the price-volume agreement is currently the most popular. The other main systems are the patient-access scheme and the performance-based model, and these have been implemented in various forms across the board, see Table 2.
*Tables 1 and 2 pending to upload.
In the UK, the NHS claims that it has a breakthrough method for determining prices, which it will begin implementing by 2014, and has already launched public consultations which closed on 17 March 2012. The purpose of value-based pricing is to improve NHS patient access to effective and innovative drugs by ensuring that drugs are made available at a price that reflects their real value. It will give patients and clinicians greater access to medicines, which are based on assessments of outcomes they are capable of achieving. The threshold should be based on quality of life. A basic threshold would reflect the benefits displaced elsewhere in the NHS when funds are allocated to new medicines. Higher thresholds for greater ‘burden of illness’; for greater therapeutic innovation and improvements compared with other products; and for medicines that can demonstrate wider societal benefits.
The Italian registry
In Italy, a national registry has been active since 2006. It records use of new drugs that have a high socioeconomic impact . The registry, named Onco-AIFA, initially started with oncology drugs. Now, it is a subset of a registry that monitors many different therapeutic groups. The aim of the registry is to check for appropriateness, safety, and the correct application of risk sharing. For this, an initial estimate is made on cost-effectiveness based on preregistration trials. It is like a large post-marketing study and acts in a similar to ‘conditional drug approval’, with the potential of identifying patients who experience major benefits and less toxicity .
The registry requires that doctors and hospital pharmacists register to obtain a username and personal password. According to the system, a physician must record vital statistics, disease, and treatment plan for each patient who receives drugs via the registry. If the patient is deemed eligible, the doctor can prescribe the medicine and the pharmacist can deliver the drug. The clinical data from this database allow prospective investigations, and the registry has become a data source for effectiveness and appropriate evaluations. Progression-free survival and overall survival in clinical practice can be calculated, and may be used as a valuable indicator in efficacy and effectiveness assessment. On the basis of any difference in benefit expressed in progression-free survival or overall survival, a new price adjusted for effectiveness should be proposed.
The efficacy–effectiveness assessment consist of analysis and discussion of main survival estimates—overall survival, time to progression, end treatment time, reason of stopping treatment—and predictive factors of the institutional sample as an example of the results achieved in the clinical environment. These data are compared with RCT outcomes of the approval trial, which is characterized by a sample with selected patients and ‘ideal’ conditions of treatment. The efficacy–effectiveness evaluation focuses on the hard endpoints, such as time to progression and overall survival as valuable indicators of effectiveness, providing detailed analysis of baseline characteristics in two groups of patients. The comparison of clinical outcomes is feasible only if the baseline characteristics are equally distributed between these groups.
Total expenditure for drugs recorded in the Istituto Oncologico Veneto and the cost per patient is shown in Table 3. Each drug has been marketed at different times, and the number of patients is proportional to the length of the introduction time as well as to the frequency of the disease. The highest cost per patient is for sunitinib, and the absolute highest cost is for bevacizumab, which is mainly used in colon rectal cancer. The average cost in all cases is high.
*Table 3 pending to upload.
The AIFA Oncologic Working Group has suggested two risk-sharing arrangements, plus price-capping, for new anticancer medicines to enhance their reimbursement potential based on the following:
epidemiological data for the disease
the possibility defines clearly a subset of the population responsive to the treatment
results from the submitted clinical trials.
To date, only 24 out of the 30 drugs included in the Onco-AIFA registry are subjected to conditional reimbursement. Of those 24, only 16 are listed in the registry section devoted to the calculation of the price discount. Reimbursement mechanisms are specific for each drug, see Table 4; no generalized reimbursement method is available. With cost sharing, a fixed discount is applied but it varies from brand to brand; with pay by result, on the other hand, a partial or complete reimbursement of the first cycles of treatment for non-responder patients is applied, using various schemes. Finally, in the case of bevacizumab and catumaxomab, a type of capping is made available.
*Table 4 pending to upload.
A cost-sharing scenario is presented in Figure 2. In this scenario, a discount is offered on some drugs, usually 50% of the ex-factory price for the first two or three treatments. The aim is to dilute the discount to patients who are responders and thus proceed with the treatment.
*Figure 2 pending to upload.
With the ‘pay-by-result’ system, a partial or complete reimbursement for medicines will be offered if it can be shown that the patient is a non-responder, after an established period of time.
With capping, responders who continuously use a drug eventually arrive at the capping limit, at which point the drug is provided for, see Figure 2. Different possibilities can be applied as seen fit; examples of how different schemes are applied are detailed below and in Table 4.
With risk sharing, a hospital initially purchases the drug, and then a physician, using the system, registers his patient if eligible and prescribes medicines for each cycle. When the treatment is completed or interrupted owing to progression or toxicity, the physician is required to close the ticket (record card). Subsequently, at the pharmacist’s request, the system evaluates or validates each prescription and automatically applies to the manufacturer for ‘pay-back’.
This process has been criticized because it is based on active documentation of non-responders. Many agree that it would be much better if drugs were purchased only if treatment were successful .
Evaluation of risk sharing with bevacizumab
In 2010, we, at the Istituto Oncologico Veneto, calculated the theoretical discount of bevacizumab with the complex risk-sharing method proposed, and obtained a 21% discount. At first glance, it seemed a reasonable pricing method for this drug; however, it is flawed because, when patients respond positively to bevacizumab, physicians eventually stop the treatment but tend not to close the ticket because of the possibility of a relapse. Statistically speaking, a non-closed ticket equates to a non-treated patient, but what is worse is that the discount is never applied. The result, as can be seen in Figure 3, is an actual discount of 7% with a constant increase in total costs over time, see Figure 4. The envy of any industry!
*Figures 3 and 4 pending to upload.
Use of the Onco-AIFA registry to obtain effectiveness results
How could efficacy and effectiveness be expressed in oncology trials? Outcome endpoints could be expressed as tumour mass shrinkage, or variations in biochemical markers, but these are usually considered ‘soft endpoints’; the scientific community suggests the use of hard endpoints, such as time to progression, progression-free survival, disease-free survival and overall survival, see Table 5. The most commonly used method to show the results of a treatment in a group of patients is a Kaplan–Meier curve or ‘survival curve’ chart.
*Table 5 pending to upload.
The progression-free survival calculation, obtained from the registry at the Istituto Oncologico Veneto for bevacizumab in colon rectal metastatic cancer, showed a slightly shorter median time (9.9 months) than that obtained in the efficacy study by Hurwitz et al.  (10.6 months) In reality, many tickets will remain open and thus discounts will be low.
At the Istituto Oncologico Veneto, we analyzed the use of sorafenib in cases of hepatocarcinoma. Use of data from the Istituto Oncologico Veneto (n = 81; overall survival eight months; time to progression three months) proved to be less effective than data published in the preregistration clinical trials (n = 299; overall survival 10.7 months; time to progression 5.5 months) .
But why do we have so little benefits, especially in people with metastatic cancer? Probably, the subgroup of patients who have the potential to obtain the maximum benefit should be identified more precisely . The registry allowed researchers at the Istituto Oncologico Veneto to conduct a study that showed no less effectiveness compared with efficacy, but also helped to identify responder patients. In a first report presented at the 2011 European Society for Medical Oncology Congress, in a subgroup of 130 patients treated with erlotinib for non-small cell lung cancer, a strong correlation was found between post-Erlotinib chemotherapy and prolonged overall survival . Progression-free survival and overall survival may be used as an indicator for efficacy–effectiveness assessment. On the basis of the difference in benefit expressed in progression-free survival and overall survival, a new price adjusted for effectiveness can be calculated. If effectiveness in real clinical practice is lower than expected on the basis of approval RCT results, then adjusted price should be applied reducing it to percentage value of ‘ineffectiveness’.
Methods to assess cost-effectiveness of lapatinib
The effectiveness of lapatinib, a tyrosine kinase inhibitor used for in advanced breast cancer, has been questioned . The lifetime cost of metastatic breast cancer is a key component for the economic evaluations of these innovative products : indeed, the addition of targeted drugs for cancer treatment to standard chemotherapy entails a substantial growth of the costs related to cancer care determined by the high-costs of new anticancer drugs .
In the UK, NICE issued a statement in 2009 that lapatinib, in combination with capecitabine, within its licensed indication, is not recommended for the routine treatment of women with previously treated advanced or metastatic breast cancer whose tumours overexpress human epidermal growth factor receptor 2 (HER2+), except in the context of clinical trials. The analysis showed that capecitabine monotherapy is likely to be the most cost-effective treatment option up to a threshold of about GBP 80,000 per quality-adjusted life years gained [24, 25].
So the question is how much are governments willing to pay per unit of efficacy gained? If efficacy is represented by overall survival, then life years can be used. NICE, for instance, is willing to pay a value up to Euros 60,000 (Euros 5,000 per month); according to NICE, more is not sustainable for the NHS. Assuming that the 1.9 months gained are quality-adjusted time without symptoms of progression or toxicity, the value commonly accepted of Euros 2,500 per month is used to calculate the maximum price attributable by the NHS (Euros 4,845).
If Euros 4,845 is a threshold value, and the median of treatment is 189 days, the daily cost of oral treatment should be Euros 25.6. In reality the ex-factory price in Italy is 3.4 times higher, at Euros 87.5. GlaxoSmithKline has agreed with AIFA a pay by result for lapatinib, which probably will never achieve the 70% discount corresponding to the acceptable price of NICE, as shown in Figure 5.
*Figure 5 pending to upload.
In Italy, the SSN costs are controlled by the regional governments which are financially accountable for healthcare expenditure, including hospital drug budgets; but agreements about price and discounts are decided by AIFA. For this reason, some regional governments ask the pharmaceutical companies to produce a health technology assessment before deciding to introduce a drug in the regional hospital formulary. The algorithm for lapatinib shows a typical budget impact calculation. Assuming the threshold described previously, the foreseen expenses would save more than Euros 1,200,000 per year, see Figure 6.
*Figure 6 pending to upload.
At the time of market entrance, a great deal of uncertainty surrounds the real performance of a drug, which usually leads to overpricing, the ‘traditional drug market situation’. In the wider population and in routine use, however, RCTs may fail to show relevant clinical benefit, as certain patient subtypes are under-represented, notably those with poorer prognostic factors and those who have received previous treatment. The Italian Onco-AIFA registry allows prospective investigations and has become a data source for drug effectiveness and appropriateness analysis, followed by a pharmacoeconomic analysis. This method could be a powerful tool in negotiations with pharmaceutical companies on a balanced risk-sharing scheme. By decreasing the uncertainty of the cost-effectiveness ratio of a drug during the post-marketing phase, a price can be established that reflects the true value of the drug. The more variables that are known, the more confident we can be with risk-sharing schemes. Hospital pharmacists can play a pivotal role in creating the conditions for a sustainable drug treatment culture.
Marta P Trojniak, PharmD
Angelo C Palozzo, PharmD
Director of Pharmacy Department
Venetian Oncology Institute (Istituto Oncologico Veneto IRCCS)
64 Via Gattamelata
IT-35128 Padova, Italy
1. Lo Scalzo A, Donatini A, Orzella L, et al. Italy: health system review. Health systems in transition. 2009;11(6). Available from: http://www.euro.who.int/__data/assets/pdf_file/0006/87225/E93666.pdf
2. Maio V, Manzoli L. The Italian health care system: WHO ranking versus public perception. Pharm Ther. 2002;27(6):301-8.
3. Drummond MF, Sculpher MJ, Torrance WG, et al. Methods for the economic evaluation of health care programmes. 3rd ed. Oxford; New York: Oxford University Press; 2005.
4. France G, Taroni F. The evolution of health-policy making in Italy. J Health Polit Policy Law. 2005;30(1–2):169-87.
5. Cheema PK, Gavura S, Migus M, et al. International variability in the reimbursement of cancer drugs by publically funded drug programs. Curr Oncol. 2012 Jun;19(3):e165-76.
6. O’Dowd A. Watchdog set to reject four drugs for kidney cancer on the NHS. BMJ. 2008;337:a1262.
7. Gordon C. Innovative approaches to pricing IMS pharma pricing and reimbursement. IMS Pharma Pricing and Reimbursement, August 2007;224-7.
8. De Pouvourville G. Risk-sharing agreements for innovative drugs: a new solution to old problems? Eur J Health Econ. 2006 Sep;7(3):155-7.
9. Smith N, Mitton C, Peacock S. Qualitative methodologies in health-care priority setting research. Health Econ. 2009; 18(10):1163-75.
10. Hoffmann C, Graf von der Schulenburg JM. The influence of economic evaluation studies on decision making: a European survey. The EUROMET group. Health Policy. 2000;52(3):179-92.
11. Mauskopf JA, Paul EJ, Grant. MD, Stergachis A. The role of cost-consequence analysis in healthcare decision-making. Pharmacoeconomics. 1998;13:277-88.
12. Eddama O, Coast J. Use of economic evaluation in local health care decision-making in England: a qualitative investigation. Health Policy. 2009;89(3):261-70.
13. AdamskiJ,GodmanB,Ofierska-SujkowskaG,etal.Risksharingarrange-mentsforpharmaceuticals:potentialconsiderationsandrecommendations for European payers. BMC Health Serv Res. 2010;10:153. Available from: www.biomedcentral.com/1472–6963/10/153
14. Agenzia Italiana del Farmaco. Available from: http://antineoplastici.agenziafarmaco.it/
15. Strom BL. How the US drug safety system should be changed. JAMA. 2006;295(17):2072-5.
16. Garattini L, Casadei G. Risk sharing agreements: what lessons from Italy? Int J Technol Assess Health Care. 2011;27(2):169-72.
17. Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer. N Engl J Med. 2004;350(23):2335-42.
18. Trojniak MP, Palozzo AC, Mazurek M, et al. Sorafenib in hepato-cellular carcinoma—a post marketing evaluation. Immunopharmacol Immunotoxicol. 2012;34(3-4):419-22.
19. Llovet JM, Ricci S, Mazzaferro V, et al. SHARP Investigators study group. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med. 2008;359(4):378-90.
20. Trojniak MP, Jirillo A. Does chemotherapy have a role after erlotinib treatment in NSCLC? Lung Cancer. 2012 Jan;75(1):136.
21. Crivellari D, Spazzapan S, Lombardi D, et al. Lapatinib-based therapy in heavily pretreated HER2-positive metastatic breast cancer: a single institution experience. Tumori. 2012; 98(1):33-8.
22. Bonastre J, Jan P, Barthe Y, et al. Metastatic breast cancer: we do need primary cost data. Breast. 2012;21(3):384-8.
23. Lidgren M, Wilking N, Jönsson B. Cost of breast cancer in Sweden in 2002. Eur J Health Econ. 2007;8(1):5-15.
24. National Institute for Health and Clinical Excellence. Final appraisal determination—advanced or metastatic breast cancer—lapatinib. Issue date: February 2009.
25. Cameron D, Casey M, Press M, et al. A phase III randomized comparison of lapatinib plus capecitabine versus capecitabine alone in women with advanced breast cancer that has progressed on trastu-zumab: updated efficacy and biomarker analyses. Breast Cancer Res Treat. 2008;112(3):533-43.
26. Konecny G, Pauletti G, Pegram M, et al. Quantitative association betweenHER-2/neuandsteroidhormonereceptorsinhormonereceptor-positive primary brest cancer. J Natl Cancer Inst 2003;95(2):142-53.