Abstract
Introduction
The main purpose of this analysis was to quantify the time elapsed between the validation date of European Medicines Agency (EMA) centralized procedure and the first purchase of a product by at least 1 Italian health care structure, evaluating different variables that affect the process, the number of products approved by the Committee for Medicinal Products for Human Use (CHMP) that are available on the Italian market (July 2016), and the impact of the Cnn class for oncology drugs in Italy.
Methods
A panel of oncology products has been defined, which considered drugs approved by the EMA between January 2013 and December 2015, and authorized for the treatment of oncology diseases, excluding generics. Data were obtained via the EMA website by the Agenzia Italiana del Farmaco (AIFA; the Italian Medicine Agency) meeting reports, by official administrative acts of marketing authorization, and the date of the first purchase (first day of the first handling month).
Results
The mean time of EMA evaluation for the considered panel of medicines was about 441 days (standard deviation (SD) 108; range 266-770); the average approval time for AIFA was about 248 days (SD 131; range 85-688). Interestingly, the mean AIFA evaluation time decreased significantly from 264 days for products submitted to AIFA assessment in 2013-2014 to 219 days for products evaluated in 2015-2016. Focusing on the regional access, both the timing and the number of drugs available for patients were widely different from region to region.
Discussion
A reduction in the approval time in the last 2 years has been observed in Italy. However, several variables influence the efficiency of the process and need to be addressed to make the access to drugs timely and efficient.
Introduction
New oncology drugs in Europe are assessed by the European Medicines Agency (EMA) and are authorized with a centralized procedure, which came in force in 1995 and allows applicants to obtain a marketing authorization that is valid throughout the entire European Union (EU) (1). However, even if a European marketing authorization has been granted, this does not imply that the product will be immediately available to patients everywhere in the EU (2).
The Italian model is characterized by a centralized National Health Service (NHS), responsible for drugs assessment, pricing and reimbursement (P&R). The regions are the budget holders and, therefore, are responsible for the local provision of healthcare (3), and they represent a second real hurdle for the market access of new treatments (4).
Once authorized by the EMA, Italy, as a member state of the European Community, automatically accepts drug marketing authorizations by defining the supply regimen and starting with the pricing and reimbursement process.
In Italy, the price setting of medicines reimbursed by the NHS is regulated at the central level by the Agenzia Italiana del Farmaco (AIFA), the Italian national regulatory authority (5). AIFA provides, with its scientific authority and autonomy, the clinical, scientific, and economic evaluation regarding medicines. AIFA activities are supported by 2 technical scientific committees consisting of experts of well-established experience.
Technical Scientific Committee (CTS, Commissione Tecnico Scientifica): it assesses the national and European marketing authorization applications, and delivers a consultative opinion on them and provides classification for reimbursement; it takes over the tasks formerly attributed to the National Drug Evaluation Board at the Ministry of Health.
Prices and Reimbursement Committee (CPR, Comitato Prezzi e Rimborso): it carries out the activity of negotiation with pharmaceutical companies for the setting of prices of medicinal products reimbursed by the NHS according to transparent methods, timelines, and procedures established by the Interministerial Committee for Economic Planning (Comitato Interministeriale per la Programmazione Economia – CIPE) deliberation of February 1, 2001 (6).
For a nonorphan drug, companies are allowed to submit a Price and Reimbursement Dossier for AIFA, CTS, and CPR evaluation in accordance with CIPE decree provisions after the Commission Decision publication in the Official Journal of the European Union, in order to start the pricing and reimbursement procedure. Instead, when the reimbursement process involves an orphan drug, the related dossier could be submitted at the national level at the Committee for Medicinal Products for Human Use (CHMP) positive opinion stage.
When new pharmaceutical treatments are considered very innovative after EMA evaluation, this evaluation needs to be translated and acquired by AIFA in its appraisal. The first definition of therapeutic innovation was obtained from a meeting of the International Society of Drug Bulletins in 2001. The innovation concept was declined in various meanings (commercial, technological, and real therapeutic innovation -when it offered real benefits compared to the therapeutic options already available) (7). In order to be considered as innovative, a medicinal product has to demonstrate its added therapeutic value. This value could be considered as superior compared to treatments already available, or it could be deputed to treat diseases without clinical options. A positive evaluation could account even as a sustainable expenditure for the NHS (8). The attribution of the innovative status ensures the immediate availability of medicines reimbursed by the NHS. This advantage permits the immediate prescription at a regional level, independent of inclusion in the regional formularies.
The prices of pharmaceuticals reimbursed by the NHS is set through negotiation between the Italian medicines agency (AIFA) and the pharmaceutical companies, in accordance with Law no. 326 of November 24, 2003 and CIPE (Resolution of February 1, 2001). The aim of these legislative measures is to regulate the temporary redefinition of the company margin and the new positive reimbursement list based on cost and efficacy in order to ensure the full respect of pharmaceutical price regulation criteria.
In Italy, a product can be assigned to (i) Class A, which includes essential products and those intended for chronic diseases, and these are fully reimbursed by the NHS; (ii) Class H, which includes products that are only fully reimbursed in the hospital; and (iii) Class C, which includes other products that do not have the characteristics of Class A and are not reimbursed. In this analysis, we have focused on oncology drugs and investigated mainly Class H products. (In rare cases, we found Class C products, when an agreement between AIFA and companies was not reached.)
When specific products obtain the reimbursed status and a monitoring system is considered as necessary (in order to monitor the appropriateness of use and economic parameters), AlFA's committees discuss with companies the related monitoring registry, often linked to a Managed Entry Agreement (MEA) (9). At the same time, this system ensures careful monitoring of the therapeutic benefits and the safety/efficacy ratio, and provides support for the decision-making process leading to price determination and to reimbursement conditions for innovative medicines (10).
Once the national phase is completed - by a decision published in the Gazzetta Ufficiale (GU), Italy's official journal – an oncology drug is usually not immediately available in every hospital pharmacy. The drug approval process needs to undergo further steps that may be quite different from one region to the other. Even within a single region these steps may vary among the different health districts and hospitals.
The point is to understand how these regulations interact with each other, fostering or disfavoring the reimbursement process. Moreover, in this analysis, how various factors can influence the final time to gain reimbursement has been assessed, such as orphan designation, route of administration, application of a monitoring registry, and the typology of MEA applied. Considering all these complexities, the main purpose of this analysis was to quantify, for each single step of the approval process, the time elapsed between the drug's validation date of EMA through a centralized procedure, and the first purchase of a product by at least one Italian healthcare structure. This analysis identified the variables that affect these times, by calculating the number of products approved by the CHMP that are available on the Italian market, and by estimating the impact of the Balduzzi Decree (by the introduction of the Cnn class, where “C” stands for “not reimbursed” and “nn” stands for “not negotiated”), for oncology drugs.
Methods
The patient access time to new oncology drugs was analyzed, assessing the time elapsed along each step from the European context to the availability of the product at the regional level. The evaluated panel of oncology products has been defined considering brand name drugs approved by the EMA between January 2013 and December 2015 authorized for the treatment of oncology diseases, excluding generics.
The analysis considered just the first application, excluding all possible further dossiers for the same product for any extension of therapeutic indication or other different procedures (new formulation, new pack size). Data were obtained by EMA website (EPAR) (11), through a systematic review of all the reports of each AIFA, CTS, and CPR meeting (12), by the analysis of the official administrative acts of marketing authorization published in the GU (13). The date of the first purchase has been considered as the first day of the first handling month (data provided by IMS Health).
The steps detailed in our analysis are expressed in days, as follows:
EU time: defined as the difference between the date of EU Commission decision and the application dossier date;
Submission period time: this step includes the time attributable both to pharmaceutical companies for P&R dossier submission and to administrative workloads, which may delay the “CTS opening”: defined as the difference between AIFA “CTS opening” and Commission decision;
AIFA assessment and health technology assessment (HTA) time: defined as the difference between “CTS opening” and “CPR closure”;
AIFA administrative and GU time: defined as the difference between “CPR closure” and publication in the GU;
Best regional time: defined as the difference between publication in the GU (the one that includes the Cnn classification when the product is commercialized in the Cnn class, and the one with price and reimbursement for the other products) and the date of the first purchase of the product by at least 1 of the public healthcare structures (first day of the month);
Last regional time: defined as the difference between publication in the GU (the one that includes the Cnn classification when the product is commercialized in the Cnn class, and the one with price and reimbursement for the other products) and the date of the first purchase of the product by the last Italian region (first day of the month, considering the data available).
We used descriptive statistics to describe patterns of approval time in Italian regions, and performed an analysis of variance (ANOVA) test to assess whether time for approval differs for orphan and nonorphan drugs.
Results
In the time elapsed between January 2013 and December 2015, the EMA granted 34 positive opinions for branded oncology products, 14 of them with a European orphan drug designation (14).
In July 2016, 17 of 34 products were commercially available. Of them, 8 completed the national assessment and 9 have been marketed using the Cnn class. Seven of the remaining products are still waiting for P&R dossier submission to AIFA; 4 are still under AIFA evaluation; 1 just completed the AIFA P&R process (waiting for publication in the GU); 4 are waiting for inclusion into regional formularies; and 1 is still in the Cnn class (from Oct 2014).
European and national approval pathway
The mean time of EMA evaluation for the considered panel of medicines was 441 days (including “clock-stop” days not attributable to the agency). Our analysis highlighted that AIFA assessment and HTA evaluation required a lower time of a mean of 248 days (median 217 days, considering the high variability of procedures (i.e., when an agreement on price between AIFA and the pharmaceutical company has been difficult to reach). Interestingly, the mean AIFA evaluation time decreased significantly: from 264 days for products that have undergone AIFA assessment in 2013-2014, to 219 days for products evaluated in 2015-2016 (median from 224 days in 2013-2014 to 215 days in 2015-2016). Table I shows that the average approval time for EMA is about 441 ± 108 (range 266-770) days and that the average approval time for AIFA is approximately 248 ± 131 (range 85-688) days.
Approval time (no. of days)
AIFA = Agenzia Italians del Farmaco; EMA = European Medicines Agency; Max = maximum value; Min = minimum value; Num. of drugs = number of drugs analyzed; SD = standard deviation.
Table I also shows further details related to the EMA and AIFA time in the distinction between orphan and nonorphan drugs. The data indicate that, for nonorphan drugs, the average approval time for EMA is approximately 450 ± 107 (range 290-770) days and that the average approval time for AIFA is approximately 196 ± 74 (range 85-385) days. Instead, for orphan drugs, the average approval time for EMA is approximately 428 ± 113 (range 266-617) days, and the average approval time for AIFA is approximately 335 ± 164 (range 202-688) days.
These figures, overall, are well above the expected approval time for EMA and AIFA, which are 210 and 180 days, respectively. In addition, the AIFA approval time for orphan drugs is above the 100 days timespan fixed by the recently introduced Balduzzi Legislative Decree for the approval of orphan and innovative drugs in the Italian NHS. This can be explained by considering various “clock stops” (i.e., the EMA “clock-stop” at day 120 in order to enable applicants to have enough time to respond to the CHMP List of Questions, or the AIFA CTS requests addressed to the companies in order to collect more information that is useful for the evaluation). In these cases, the increase in the procedure time cannot be attributed to EMA or AIFA, but depends on how quickly the company responds.
The relation between EMA time and AIFA time is negative and significant (r = −0.2095; p = 0.002 [r = linear correlation coefficient]). A t-test assessed whether the means between AIFA time (247.63 days) and EMA time (397.16) are statistically different. The result shows that the 2 means are statistically different at the significant level of p<0.05.
Table I also shows the average EMA time and AIFA time for both drug categories. The third column of Table I shows that, for EMA, the average time for approval of the orphan drugs is about 427 days, whereas for nonorphan drugs it is 449 days. For AIFA, the average time for approval of the orphan drugs is about 335 days, which is substantially higher than the average time reported for nonorphan drugs (196 days). This could be easily explained by the long negotiation process required for a single drug, which took 688 days.
Table I also documents that the AIFA time is lower than the EMA time for both orphan and nonorphan drugs. The results of the ANOVA test show that the difference is not statistically significant for EMA time (F = 0.33; p = 0.5668 [F = variation between sample means/variation within the samples]), while it is significant for AIFA time (F = 6.53; p = 0.0205).
Our analysis found that when products are authorized and reimbursed with a monitoring registry, the resulting timeline for the reimbursement approval is strictly reduced. Products listed in our analysis and reimbursed with a monitoring registry, (n = 13), were associated with a mean shortening of the timing of approval (232 vs. 298 days for products not subjected to a registry).
The final steps are related to AIFA administrative activities (after CPR closure, defined as decree's drafting) and the GU publication timeline, they required a further 114 days.
Regional approval pathway
Focusing on the regional access, both the timing and the number of drugs available for patients were widely different from region to region, and this was mainly due to organizational factors, such as the use of regional/local/hospital formularies periodically updated by regional commissions. The mean best regional time (defined as the average number of days after AIFA market authorization as published in the GU and the first purchase date in the first Italian region) was 29 days. The last regional time (the number of days between GU and the first purchase in the last region for which data are available to date (July 2016)) was 293 days. This last data are underestimated, considering that not all products were available in all the Italian regions. Out of the 9 products commercialized in C or Cnn class (not reimbursed), an average of 8 were available on the Italian regions and out of the 8 commercialized after AIFA price and reimbursement achievement, the average was 6. On average, 10 Italian regions started to treat patients with drugs in the C or Cnn class before the pricing and reimbursement process was concluded (with a wide variability among the products, range 1-18). Italian regions made available a mean of 4 products classified in the C or Cnn class (range 2-8). We found that the regions with the higher number of products commercialized in the C or Cnn class (Puglia and Lombardia) were also the regions where time to access was the shortest for most of the drugs under evaluation.
Table II shows the time elapsing for the purchase in all Italian regions. Particularly, it shows the elapsing time for the first purchase not even reimbursed (min) and the last purchase not even reimbursed (max) (when a product is commercialized using Cnn classification) and the elapsing time for the first purchase reimbursed (min) and the last purchase reimbursed (max) (when a product is classified in a reimbursed (Class A or H)).
Regional access time
Max = maximum value; Min = minimum value; SD = standard deviation.
We noticed that some regions are progressively abandoning their own formularies (e.g., Veneto) and that others, such as Lombardia, do not have a formulary. The figures reported assess that Valle D'Aosta and Piemonte are the regions with the fastest process and that Basilicata and Molise are those which, in contrast, show a longer delay in dispensing approved drugs. In some regions, the timing of access differs substantially for selected drugs, such as in Abruzzo (standard deviation (SD) = 269.2), Basilicata (SD = 279.7), and Sardegna (SD = 268.4). The figures show an overall worsening in the approval timing when compared to data previously reported in the literature (15). This might be due an overall decrease in the allocation of funding by the central government to regions. Figures 1 and 2 summarize the data shown in Table II, illustrating the average time and the range of timespans for each single region.
Timespan from Gazzetta Ufficiale publication to regional access.
Best and worst time for each Italian region.
Table III shows the average access timing for selected drugs, which were calculated as time elapsing from the date of publication in the GU (Cnn classification date by the official journal for Cnn commercialized products and the official journal reimbursement classification for Class A and H drugs). The distinction was made between products commercialized in the C and Cnn class (not reimbursed) and the first/last purchasing date in those regions where the drug was included in the formulary. The figures show that the relevant differences characterize the average access timing of the drugs being considered here.
Access timing for selected drugs
SD = standard deviation.
In our analysis, we found that the average elapsing time was widely different. For one product, in fact, the average elapsing time was almost 30 days, whereas the timespan was substantially longer for another drug, which reached 2 years from GU publication in 1 instance.
An additional key point to consider is related to the assessment that is conducted at the regional level. Even if the product is considered efficacious and safe by both the European and the national authorities, regional committees conduct a deep appraisal about products and their place in therapy at local level.
Discussion
The current analysis showed that before patient access, the drug assessment process is complex, multifaceted, and time consuming. The largest proportion of time is spent by the EMA for the official assessment about the safety, efficacy, and clinical value of the product. A recent IMS report (16) highlighted the comparison between the average times to access drugs in the US, the EU5 (France, Germany, Italy, Spain, UK), and in the rest of Europe. According to the presented data referring to year 2014, Italy ranked 18 (out of 22 countries) in average time from regulatory approval to market access (14.5 months). However, when considering the average time for market authorization and for final decisions on price, Italy is the country with the lowest average time amongst the EU5.
In the last few years, a number of steps have been taken to improve time to market through legislation. The most important was the creation of a new class (Cnn), which grants marketing authorization within 60 days after EMA approval, even before price negotiations begin. Class Cnn was introduced by the legislative decree no. 158/2012, the so-called Balduzzi Decree, converted by law no. 189/2012 (17). Initially, if companies decided to start the commercialization, these drugs were not reimbursed and were listed in the Cnn class. This classification means that the cost, if sustainable, is covered by the health care structure where the drug is prescribed, or by patients (18). After EMA approval, thanks to the introduction of the Cnn class, the drugs are potentially immediately available to Italian hospitals even while they are waiting for price negotiation. In fact, by using the Cnn class, pharmaceutical companies are allowed to commercialize their products (by communicating prices and the first day of commercialization), even if they are not yet reimbursed. The novel classification establishes a predefined timeline (about 60 days) so the drug is classified in Cnn class and the resulting timing to patients' access is substantially reduced.
By our analysis, AIFA requires an average of 248 days for its scientific and HTA evaluations, but this time has significantly decreased over recent years. AIFA, in its evaluations, considers specific criteria for price negotiation of a new chemical entity, which are:
Cost/benefit ratio positive: when the product is considered useful for the treatment of diseases for which there is no effective treatment, or to ensure a more appropriate response than drugs already available for the same therapeutic indication;
Cost/benefit ratio more favorable than other drugs already available and used for the same indications;
Economic burden evaluation, in the perspective of the NHS;
Better treatment cost/day ratio, compared to other products with the same efficacy;
Estimation of the potential market share acquirable in the next 24 months in the specific market segment;
Comparison with the prices and consumption data of other European countries.
A positive impact is attributable to the monitoring registries and the related MEAs, which grant a balance between a rapid access to market and appropriateness.
Since oncology drugs are often associated with an intrinsic uncertainty related to their efficacy for the specific patient, monitoring registries are in force in Italy. In 2006, the AIFA published the first monitoring registry related to oncology drugs. The main purpose of registries and shared schemes arose from the necessity to identify and grant a trade-off between innovation and economic sustainability, considering the change of perspective in terms of economic burden and real clinical effectiveness.
In order to ensure the access to medicines and balance the economic expenditure, the AIFA has defined that an oncology drug is to be reimbursed only when it is effective for the specific patient (considering that the health care system cannot support treatment failures associated with high-cost medicines). In addition, the use of oncology drugs is to be monitored in order to pay only the effective treatment in responding patients, charging pharmaceutical companies with the costs of failure, as per contract agreed (19). There are two main types of registries:
Performance-based registries: these were set up in order to assess the real effectiveness and other parameters related to clinical aspects. Associated with this type of monitoring registry, AIFA usually defines performance-based risk-sharing agreements in order to share the intrinsic risk associated with the drug use. When a drug is ineffective for a specific patient, the treatment is stopped and the economic burden associated with the treatment is shared between the AIFA and the company.
Finance-based registries: these are associated with financially based agreements.
The regional situation is widely variable: patient access is not simultaneously achieved for all oncology products in each Italian region and, in some cases, drugs are still not available, even after years from the initial EMA approval.
At regional and hospital levels in Italy, therapeutic formularies are in force that define which drug can be used, in which situations (according to disease stage or patient characteristics) and how long the treatment is to be continued.
Local formularies are usually defined as PTOR (Prontuario Terapeutico Ospedaliero Regionale) and are relevant to the entire regional level. Also, at the local level (intended as district area and hospital levels) additional formularies can be defined as a cascade, including drugs only included in the major formulary.
A regional formulary can be defined as “binding” when the minor formularies (hospital level) can be constituted as just selecting drugs from the regional formulary, and physicians are allowed to prescribe only the drugs listed in the PTOR.
A regional formulary can be defined, instead, as a “not binding” formulary, when the minor formularies (hospital level) can be constituted with drugs considered useful for patients at a local level and can be freely prescribed by physicians.
Actually, in Italy, 18 regional formularies (considering Bolzano and Trento provinces as 2 different regional entities) are in force. Only 3 regions do not have local formularies and drugs can be prescribed, after AIFA approval and publication in the official journal, without additional delays and restrictions. The administrative decisions to include or exclude a product into the formulary at local and regional levels, are defined by several technical and scientific committees. In analogy with national and European evaluation, even if simplified, this step undergoes a structured process that could significantly delay the first drug utilization. An additional step of evaluation could cause disparities in the availability of drugs for patients resident in different regions (20).
In 2009, the Italian Society of Medical Oncology (AIOM) conducted a survey to describe the impact of regional pharmaceutical formularies on the disparity of access to 8 new drugs among cancer patients treated in Italian regions (4) by analyzing the availability of 17 new anticancer drugs registered by AIFA from 2005 to 2009.
The analysis considered the differences between the regions with and without pharmaceutical formularies and the characteristics of technical-scientific committees for the evaluation of the inclusion of hospital drugs in these formularies. By this study, the median time from the EMA to the AIFA was 11.2 months (range 2.9-17.1), and the median time from the AIFA to patient availability was 1.4 months (range 0.0-50.5) in regions with drug formularies versus 0.0 months in regions without drug formularies. The median total time from the EMA to patient availability was longer in regions with formularies (13.3 months; range 2.9-65.3) than in regions without formularies (11.2 months; range 2.9-24.0), where drugs were immediately available after AIFA marketing authorization. However, in this analysis, the availability of the drug at the regional level has been considered as the inclusion into the regional/provincial pharmaceutical formularies (when existent) and not as the effective first use of the drug. Therefore, in 4 Italian regions (Piemonte, Lombardia, Friuli-Venezia Giulia, and Marche) and in the autonomous province of Bolzano (which did not have regional pharmaceutical formularies), national marketing authorizations and drug availability for the patients have been considered the same. A successive study (15) analyzed a panel of oncology products approved by the EMA in the period 2006-2008 evaluating the mean time for the sequential phases of the time to market for the EMA, for the pharmaceutical companies, for the AIFA, and for the Italian regions, as final providers of healthcare. The overall mean time required before patient access was 2.3 years. The EMA accounted for the greater proportion of time (31.8%), followed by the AIFA (28.2%). However, the duration for both pharmaceutical companies and Italian regions was associated with the highest variability. Considering that both the methodology and the sources used for these analyses are different from those followed in our study, we cannot compare the results collected.
A more recent study conducted by the observatory on oncology patients' health conditions (2015), which took into consideration 16 oncology drugs approved in the time period 2012-2014 in 10 Italian regions, the average approval time totaled 400 days for the EMA and 530 days for the AIFA (of which 380 days related to the technical analyses by the CTS and the CPR of drug dossiers). Interestingly, this study investigated the different steps that comprise the overall approval process that allows identifying the specific stages in which a delay of the approval procedure is observed. Even in this case, a detailed comparison between this study and our research, unfortunately, is not feasible given the different data available for the calculation of the various intermediate steps. However, a summary assessment of the approval times reported in both studies seems to suggest that while the overall approval time for the EMA is quite stable, the AIFA timing of the CTS and the CPR for technical analyses decreased over time, passing from 380 days for drugs considered in 2012-2014 to 248 days for drugs approved in 2013-2015.
Further research and monitoring are needed to examine which factor most affects the time to patient access, and to assess causal relationships between drug characteristics and time to market.
Methodological limits
Not all of the information is public and available. For example, when we considered the “EU time” we did not have any detail about the “clock-stop” days (generally connected to additional requests for technical documentation). Similarly, when the product is classified in the Cnn class, the company is allowed to start the commercialization by just communicating the price; nevertheless, this information is not public, and we were not able to estimate the patient access time for the 9 drugs launched on the market in the Cnn class. Regarding the handling data (data expressed in months), we considered the first day of the handling month as collected by IMS Health.
Footnotes
Abbreviations
Financial support: No grants or funding have been received for this study.
Conflict of interest: None of the authors has financial interest related to this study to disclose.