Advertisement

Early Dialogue Between the Developers of New Technologies and Pricing and Reimbursement Agencies: A Pilot Study

      Abstract

      It is common practice for developers of new health care technologies to engage in early dialogue with the major regulatory agencies; such discussions frequently center around the proposed clinical trial designs to support the registration of new interventions and suggestions on their improvement. Pricing and reimbursement agencies are increasingly using the results from health technology assessments to inform their decision making for new technologies. Such assessments are invariably underpinned by the phase 3 clinical trial evidence which may not provide answers to the key questions. Technology developers are beginning to realize that direct, early dialogue on the evidence requirements of the major pricing and reimbursement agencies, before phase 3 clinical trial designs for their key development compounds have been finalized, may be beneficial. This article reports on the pioneering efforts of one technology developer in seeking early dialogue with seven pricing and reimbursement agencies in five countries globally in 2007–2008 on their likely evidence requirements for a new oral treatment for patients with chronic plaque psoriasis. The pilot project demonstrated that a feasible process of early dialogue could be established, through a face-to-face meeting with prior circulation of a briefing book. Although there was some variation in the advice the similarities far outweighed the differences. More experience of early dialogue needs to be accumulated, involving a wider range of pricing and reimbursement agencies and compounds. The conclusion of this study, however, was that early dialogue can be a worthwhile process for all parties and can lead to a common understanding about evidence development for market access.

      Keywords

      Introduction

      It has been common practice for several years for the developers of new health care technologies (primarily new pharmaceuticals) to engage in early dialogue with regulators, such as the US Food and Drug Administration. The reasons for this are self-evident. Given the large and costly investment in clinical research and development to establish the efficacy and safety of a new product, it is important for the company to be reasonably certain that, if research of an adequate quality is conducted, approval to market will be granted if the net clinical benefit is considered to be positive. Therefore, technology developers often present their proposed clinical trial designs to regulators in order to obtain feedback and to receive suggestions on their improvement.
      A more recent development is that in several jurisdictions a health technology assessment, including an economic evaluation, is now required as part of the processes for determining the price and reimbursement status for new technologies. In these jurisdictions the process is that the sponsor of a new technology prepares a submission containing such an evaluation, according to an agreed set of guidelines [
      International Society of Pharmacoeconomics and Outcomes Research (ISPOR)
      Pharmacoeconomic guidelines around the world.
      ]. Although the published guidelines give some clues as to the scientific quality of clinical and economic evidence that may be required, they do not provide sufficient detail to answer important strategic questions such as “to obtain reimbursement for first line use in a given category of patients, precisely what evidence should be submitted?” This is understandable, given they are not written for this purpose; they are written to help potential applicants prepare a submission from the evidence that has already been generated rather than specifying, before the event, what evidence should be generated.
      Although the requirements for the submission of evidence to pricing and reimbursement agencies do not have the same legal status as submissions to regulators, from the manufacturer's perspective they are gaining increasing importance, because they are critical for market access for a new product. Market access, not product licensing, is the ultimate goal, and technology developers are increasingly viewing the clinical development program as providing data for both purposes.
      A potential complication is that the evidence required for a pricing and reimbursement submission might involve the production of clinical data that go beyond that required for a regulatory submission. For example, in order to obtain reimbursement for first line therapy, it may well be necessary to demonstrate the product is more efficacious than an existing first line therapy, or it may be necessary to collect data on an outcome that is of interest to a pricing and reimbursement agency, although not required by a regulator. For example, some new anticancer drugs have obtained regulatory approval (in some jurisdictions) based on demonstration of improvements in progression-free survival, whereas some pricing and reimbursement agencies would likely require evidence on overall survival, unless they can be convinced that the relationship between the two outcomes has been well established [
      • Drummond M.F.
      • Evans B.
      • LeLorier J.
      • et al.
      Evidence and values: requirements for public reimbursement of drugs for rare diseases – a case study in oncology.
      ,
      • Trotta F.
      • Apolone G.
      • Garattini S.
      • Tafuri G.
      Stopping a trial early in oncology: for patients or for industry?.
      ,
      • Fleming T.R.
      • Rothmann M.D.
      • Lu H.L.
      Issues in using progression-free survival when evaluating oncology products.
      ].
      Therefore, a growing number of technology developers have realized that a dialogue on evidence requirements may be beneficial with pricing and reimbursement agencies before phase 3 trial designs are finalized with regulators [
      • Backhouse M.E.
      Raising the standards of trial-based economic evaluation: the devil is in the detail.
      ]. This article reports on the pioneering efforts of one technology developer in seeking early dialogue with several pricing and reimbursement agencies. It reflects on the main lessons learned and considers future directions and challenges.

      Methods

      General approach

      An approach was made to seven pricing and reimbursement agencies in five countries globally in order to establish whether they would be open to early dialogue. The agencies concerned varied in their remit and here the terms “pricing and reimbursement agency,” “health technology assessment agency” or “payer” are being used interchangeably. Some of the agencies were directly involved in determining reimbursement status, such as listing of drugs on the national or regional formulary. Others merely issued general guidance on the use of products within the country's health care system, which could be mandatory or not. Some had a role in price negotiations, others not. However, the common feature of all the agencies approached was that they were known to be interested in the role of health technology assessment in helping determine the appropriate use of treatments to obtain a comparative health gain within their health care system. They were selected either because they were agencies in large pharmaceutical markets, or because they had an established track record in using economic evaluation to inform reimbursement decisions.
      The agencies were asked questions about the likely evidence requirements for a development compound that might be the subject of a pricing and reimbursement submission sometime in the future. A briefing book was produced, giving some background on the compound, its planned positioning in current and anticipated future treatment pathways and outlining several key questions on which the company required responses.
      The questions related to various aspects of the proposed phase 3 clinical trials, such as;
      • What is the most relevant target population for this therapy;
      • With which alternative therapies should this drug be compared;
      • What outcomes (e.g., clinical, quality of life) should be measured;
      • Over what time period should measurement be made (i.e., length of patient follow-up);
      • Should the issue of non-responders be examined and how should “non-response” or “treatment failure” be defined;
      • Are there relevant subgroups of the patient population that should be considered?
      In each case the agency concerned determined the precise nature of the provision of early dialogue (e.g., informal conversation or formal meeting) and the nature of the feedback given (e.g., written advice or not). The similarities and differences of the various approaches to early dialogue and the variations in the advice offered were noted.

      The development compound

      The compound chosen for the study was a new orally active treatment for patients with chronic plaque psoriasis. This choice was based on several considerations: 1) (plaque) psoriasis was an important disease area for the company at the time, with four compounds in development; 2) the compound was thought to have considerable commercial potential; 3) (plaque) psoriasis was considered to be a major disease with a high level of unmet need; 4) the compound was in the right (i.e., early) stage of clinical development, with the major phase 2 and phase 3 clinical trials still at the design stage; 5) new products for patients with plaque psoriasis had recently been launched, some of which had been subjected to rigorous health technology assessments; 6) internal documentation already existed that could form the basis of the briefing book and there was general support in the company for the study; and 7) there were several important unresolved issues about the positioning of the compound and the nature of the clinical development program that would be required.

      Briefing book

      A similar briefing book was used for all the meetings. This covered the overall meeting rationale and objectives (from the company's viewpoint) and background on current treatment options for patients with chronic plaque psoriasis (see Fig. 1 for an example of presenting clinical management pathways for the purpose of pinpointing new treatment positioning options). The briefing book also discussed issues arising from the technology assessments on the recently launched products (e.g., the anti-tumor necrosis factors), the data currently available on the development compound, and the company's draft plans for future evidence generation.
      Figure thumbnail gr1
      Fig. 1Pinpointing new treatment options within a clinical management pathway.
      The continual use of one or more topical agents, such as the emollients, is warranted in all patients. The need for systemic therapy indicates the failure of the topical agents (and phototherapy if available and warranted). Phototherapy may not be available in all countries. The failure of the use of at least two agents in a class, alone or in combination, is often required before the use of an agent in the next class is warranted. The use of balneotherapy has not been considered. It may be an option for some patients, which may require patients having to travel to specific locations.
      The latter items were the most important because they represented expensive decisions that were largely irreversible once the trial designs had been set. Therefore, a substantial amount of detail was given on the design of the planned phase 2 (dose finding) trials and the guidance that the company had received from regulatory agencies on the design of the planned phase 3 clinical trials. Finally, a few proposed phase 3 clinical trial designs were presented, including diagrammatic representations.
      This information formed the backdrop for a discussion with the agencies included in the pilot study. In the case of those agencies that use the results of economic evaluations to make pricing or reimbursement decisions, discussions also took place on the company's outline plans for economic evaluation and a series of specific questions. Some of these dealt with specific scientific issues (e.g., the method for determining utility estimates), but the most important ones addressed strategic issues like ‘if the company's aspiration is for the compound to be used first line for both the short-term treatment and long-term control of adults with chronic plaque psoriasis who are candidates for systemic therapy, what types of comparative clinical and economic evidence would best support it?’

      Results

      Agencies surveyed

      All the seven agencies approached agreed to participate in the exercise, and the meetings took place between November 2007 and June 2008. In this article the names of the agencies are not given (at the agencies' request), but they are all well-known organizations, either based in the largest pharmaceutical markets, or in smaller markets with a long history of using economic evaluation in pricing and reimbursement decisions and whose decisions often have an impact on decisions in other countries.

      Advice received

      Some of the advice received from the agencies is summarized in Table 1, Table 2, Table 3, Table 4.
      Table 1Summary of payers' advice (included population and prespecified population subgroups of interest).
      AttributeAdvice
      Agency 1Agency 2Agency 3Agency 4Agency 5Agency 6Agency 7
      Included population
       Active moderate or severe disease by PASI only✓✓
       Active moderate or severe disease by patient impact only✓✓✓✓
       Active moderate or severe disease by PASI and patient impact✓✓✓✓
       Failed topical and phototherapy✓✓✓✓✓✓✓✓✓✓
       Include subset who have failed at least 1 systemic oral drug
       Include subset who have failed 1 anti-TNF drug✓✓
       Include people over age 65 years and with comorbidity✓✓
       Similar to other trials to allow indirect comparisons
       Prespecified and adequately powered subgroups✓✓✓✓✓✓✓✓✓✓
      Prespecified population subgroups of interest
       Disease severity✓✓
       Treatment history✓✓
       Baseline QOL✓✓
       Site of disease✓✓
       Age and ethnicity✓✓✓✓
       Geographic region
       Metabolic status
       Comorbid depression and other conditions
       Alcohol use and smoking
      ✓ – Suggested or nice to have; ✓✓ – strongly recommended or required; no tick – not discussed or no information provided but does not imply disagreement with the statement.
      Multiple items with the same number of ticks in the same subsection imply the same level of importance to the agency.
      PASI, psoriasis area and severity index; QOL, quality of life; TNF, tumor necrosis factor.
      Table 2Summary of payers' advice (comparators and other trial design features).
      AttributeAdvice
      Agency 1Agency 2Agency 3Agency 4Agency 5Agency 6Agency 7
      Comparator(s)
       Methotrexate - continuous✓✓✓✓✓✓
       Cyclosporine - induction or intermittent✓✓
       Injectable biologic - intermittent or continuous✓✓✓✓✓✓✓✓
       Placebo
       OK not to compare directly to all relevant drugs
      Other trial design features
       Allow preselected topical agents during the trial
       Switching allowed using likely treatment pathway✓✓
       Intermittent treatment would be useful
       Reconsider rationale for 2 maintenance doses in trial
       Perform separate induction and maintenance trials
       PASI 100/90✓✓✓✓
      Maintenance
       Time to relapse - 50% reduction in PASI gain
       Measure impact of withdrawal from treatment
       Measures of productivity and other activity losses
       Measure adherence and impact on outcomes
      ✓ – Suggested or nice to have; ✓✓ – strongly recommended or required; no tick – not discussed or no information provided but does not imply disagreement with the statement.
      Multiple items with the same number of ticks in the same subsection imply the same level of importance to the agency.
      PASI, psoriasis area and severity index.
      Table 3Summary of payers' advice (trial duration).
      AttributeAdvice
      Agency 1Agency 2Agency 3Agency 4Agency 5Agency 6Agency 7
      Trial duration
       Induction
        16 weeks if methotrexate comparator✓✓✓✓
        12 weeks otherwise✓✓✓✓
        Develop early stopping rules✓✓
       Maintenance (responders)
        Long enough to assess relapse and rebound✓✓✓✓
        1 year is sufficient✓✓✓✓
        1 year is the minimum✓✓
        3 years✓✓
       Safety (all)
        Long-term follow-up for safety because new drug class✓✓✓✓
       Non-responders
        More than 4 weeks
        12 weeks✓✓
        12–16 weeks after switch therapy
        Trial duration or at least 1 year✓✓
      ✓ – Suggested or nice to have; ✓✓ – strongly recommended or required; no tick – not discussed or no information provided but does not imply disagreement with the statement.
      Multiple items with the same number of ticks in the same subsection imply the same level of importance to the agency.
      Table 4Summary of payers' advice (trials' clinical outcomes and QOL/utility outcomes).
      AttributeAdvice
      Agency 1Agency 2Agency 3Agency 4Agency 5Agency 6Agency 7
      Trials QOL/utility outcomes
       DLQI✓✓
       PSORIQOL
       EQ-5D✓✓✓✓✓✓
       HUI✓✓
       SF-36✓✓✓✓
       Depression measure
       Address payer concern about EQ-5D ceiling and floor effects✓✓
       Measures of productivity and other activity losses
       Measure adherence and impact on outcomes
      ✓ – Suggested or nice to have; ✓✓ – strongly recommended or required; no tick – not discussed or no information provided but does not imply disagreement with the statement.
      Multiple items with the same number of ticks in the same subsection imply the same level of importance to the agency.
      DLQI, dermatology life quality index; EQ-5D, EuroQoL; HUI, health utilities index; PSORIQOL, psoriasis quality of life instrument; QOL, quality of life; SF-36, Short Form-36.

      Patient population

      Table 1 illustrates that most of the agencies felt that the target patient population for the compound needed to be carefully defined and that its use as a first line systemic treatment for patients with moderate to severe disease would be difficult to substantiate within the confines of the studies currently envisaged, at least in the short run. Therefore most of the advice focussed on defining specific patient populations, either by severity of disease, or by previous treatment history (e.g., failure of first line systemic treatment). This reflects a common concern of pricing and reimbursement agencies to target new, potentially more expensive therapies to those patients who will benefit the most. The main message for the company was that a clear understanding of the likely attainable market positioning of a new product in the pathways of disease management is critical, since this forms the foundation for a clearly targeted, consistent, and coherent evidence generation strategy.

      Treatment comparators and strategies

      A major difference between the needs of regulators and payers is that, on occasions, regulators will consider data from one or more placebo-controlled trials to be satisfactory, whereas payers prefer to see data from head-to head active controlled trials. This is clearly demonstrated in Table 2, where only one of the pricing and reimbursement agencies thought that a comparison with placebo would be of interest. Most recommended an active comparison with either oral methotrexate (in the case of positioning as first line systemic use) and/or an injectable biologic (in the case of positioning as second line systemic use).
      Potentially, this may pose a dilemma for technology developers, particularly in situations where the regulator is willing to accept a placebo-controlled trial. Trials against an active comparator pose more risk (of failure) for the company, particularly if the intention is to demonstrate clinical superiority over a current treatment for the target patient population. Such comparisons can, and often are, synthesized using the methodology of indirect or mixed-treatment comparisons [
      • Ades A.
      • Sculpher M.
      • Sutton A.
      • et al.
      Bayesian methods for evidence synthesis in cost-effectiveness analysis.
      ], but a head-to-head clinical trial is most likely to be preferable to most payers. Nevertheless, where the requirements of payers indicate the need for a wide range of comparisons, it is inevitable that some of these will have to be made indirectly. An important objective of early dialogue would be to identify those comparisons which particular payers feel should be made directly, in head-to-head trials.
      The other clinical trial design features discussed in Table 2 illustrate that pricing and reimbursement agencies are also interested in clinical trials that are more “pragmatic,” or more closely reflecting actual clinical practice. Again, these needs may be difficult for the company to satisfy in its pivotal phase 3 clinical trials, depending on the trial designs required by regulators. Therefore, it may be necessary to conduct other clinical trials, alongside the main regulatory trials, to satisfy the needs of payers. The other possible option may be to include a third arm, reflecting current care, in a pivotal trial. However, the need for the company to satisfy both the information needs of regulators and payers is emerging as one of the key issues in seeking early dialogue. This will be considered in the discussion section below.

      Trial duration

      Within most companies there appears to be strong pressure, even rewards, to complete the clinical development program as quickly as possible, with the desired outcome being regulatory approval for minimal investment. Therefore, the objective is often to collect the minimum amount of data to satisfy the regulatory agencies. However, a short-term clinical trial to demonstrate the efficacy of a new treatment may not demonstrate its true clinical value, particularly if this can only be apparent in the longer term. Because pricing and reimbursement agencies are likely only to pay for demonstrated clinical (and economic) value, it is important that the company's evidence generation strategy should address this need. Depending on the jurisdiction, this can either mean that the relevant outcomes need to be measured within the clinical trials themselves, or can be reliably modeled, using a combination of the relative treatment effect from the trials and observational studies.
      Table 3 illustrates that the pricing and reimbursement agencies distinguished between the trial duration necessary to demonstrate clinical benefit in induction, as opposed to maintenance therapy. In maintenance therapy, the proposed trial's duration did not, in most cases, conflict with the evidence requirements of the regulators. In this case both parties required a follow-up period, in the maintenance phase, of 3 to 6 months after stopping treatment and at least 1 year's worth of safety data. However, it is often the case that payers like to see data from longer term clinical trials than those required by a regulator.
      The other issue explored in Table 3 is that payers increasingly like to see defined stopping rules for expensive new treatments, in situations where the patient does not respond to therapy. This raises the question of how long is required to determine non-response and whether this would require an increase in the duration of the clinical trials. Again, it may be possible for companies to accommodate the need for these assessments in their pivotal phase 3 clinical trials, as long as this does not conflict with the requirements of the regulators.

      Trial outcomes

      It is self-evident that pricing and reimbursement agencies may require different, or additional, trial outcomes from those traditionally used to demonstrate efficacy to regulators. Typically, these include various patient-reported outcomes and measures of quality of life. Very occasionally, they may include various ‘economic’ outcomes, such as reductions in health care resource use (e.g., shortening of length of hospital stay) or measures of productivity gain. It is also possible that the clinical measures favored by pricing and reimbursement agencies may differ from those required to demonstrate efficacy for regulators. For example, in evaluating some new anticancer drugs, payers want to see assessments of overall survival, whereas the regulators were satisfied with assessments of progression-free survival [
      • Drummond M.F.
      • Evans B.
      • LeLorier J.
      • et al.
      Evidence and values: requirements for public reimbursement of drugs for rare diseases – a case study in oncology.
      ].
      In this case, the efficacy (i.e., response) outcomes suggested by regulators included a reduction of the psoriasis area and severity index (PASI) score ≥ 50%, or ≥ 50% reduction in body surface area (BSA) if the PASI is not applicable. In addition, the trials were being powered to detect a 5-point or greater improvement in dermatology life quality index (DLQI). The feedback from payers (Table 4) was that the majority wanted to see a greater change in baseline PASI score, or the investigator's global assessment (IGA) of response (0/1) as co-primary outcomes. On the other hand, IGA was only requested by one of the regulators. In addition, most of the payers expressed some interest in quality of life or utility outcomes, although there were differences among them as to the most relevant measures. Three of the seven payers expressed an interest in one of the generic health utility measures, such as the EuroQol (EQ-5D) or health utilities index (HUI).

      Differences between regulators' and payers' needs

      In this particular example there were relatively few differences between regulators' and payers' requirements. Payers always wanted to see an active comparator, whereas regulators would accept comparison with a placebo. Payers were generally happy with the clinical outcomes specified by regulators, although they also wanted to see data on quality of life and utilities. However, in other clinical fields we might expect more divergence of opinion and this is one aspect of the pilot study that should be explored further in future studies.

      Feedback on the process and impact within the company

      Although informal feedback was obtained from all the pricing and reimbursement agencies, only one agency provided formal (written) feedback. The reasons for this were varied. In some circumstances the legal framework governing the operation of the agency did not permit written feedback. In others, general uncertainties about the process may have led the agency to be somewhat guarded. To a limited extent, the absence of written feedback reduced the value of the exercise but in all cases the company received fairly specific answers to its questions. Variability of the format for documenting early scientific advice is also observed among regulators.
      The general feeling within the company was that the advice given was both reasonable and helpful. It is difficult to establish that particular changes in the clinical development program were made as a direct result of the advice obtained. The feedback from payers, however, did add weight to the arguments of those within the company that felt that the original aspirations for market positioning would require clinical evidence beyond that which was likely to be provided by the existing clinical development program.

      Discussion

      Is early dialogue worth the effort?

      The experience from the company's viewpoint was extremely positive, in that specific advice was obtained on major aspects of the design of key clinical trials in order to help inform the downstream decisions of pricing and reimbursement agencies. Potentially, this could save the company considerable time and resources by focussing effort on the main requirements for an evidence package to support the compound in the desired indication/target patient population. There was no evidence that payers had used the opportunity of providing advice to enter into negotiations about price or reimbursement status of the product. Rather, the advice was given and it was left up to the company as to how best to respond. Another way of obtaining scientific evidence would be to assemble a group of expert advisors in the area of pricing and market access. This approach had also been followed by the company for the compound concerned. Early dialogue with the agencies, however, was judged to provide additional value, mainly because direct advice from the payers themselves enhances the awareness of the project team to the evidence issues that will need to be addressed.
      The assessment of value is more difficult from the perspective of the agencies. It could be argued that time spent on offering advice could divert the agencies from other important tasks. On the other hand, if companies are given advice on the appropriateness of different types of evidence, the process of making ‘correct’ (i.e., more informed) reimbursement decisions would be enhanced; submissions may be shorter and more focused, explicitly addressing evidence known to be relevant to payer decisions. Compounds discussed at this early stage, however, may not reach the market, for one reason or another.
      From a societal viewpoint the value of early dialogue would depend on whether it leads to a fairer and more efficient (i.e., timely) reimbursement process. This can only be assessed once more experience is accumulated. One important point to assess is whether the advice is ‘wasted,’ either because it is ignored by the companies concerned or because some of the compounds discussed at this early stage may not reach the market. Advice, however, once developed, is likely to be partially transferable to other developers with products for treatment of the same disease in the future.
      Another reason why early dialogue might not be useful is that, owing to changes in circumstances, advice could change over time. For example, another alternative therapy could become available after the time that the advice was given. However, there was recognition from all sides that advice can only be given based on existing knowledge and the potential for changes in the future should not undermine the process.
      One societal concern is the possibility of ‘regulatory capture’ as a result of agencies having more dialogue with manufacturers (that is, agencies may become overly sensitive to the concerns of manufacturers, to the detriment of the interests of consumers). Again, this would need to be assessed in the longer term, but the possibility of this is unlikely, given the rigorous procedures for technology assessment that have been established by most of the agencies. In addition, stakeholder engagement is recognized as a desirable characteristic of the pricing and reimbursement process by most agencies [
      • Neumann P.J.
      • Drummond M.F.
      • Jönsson B.
      • et al.
      Are key principles for improved health technology assessment supported and used by health technology assessment organizations?.
      ]. Indeed, it might be argued that the early dialogue initiative could lead to a paradigm shift; a change from an (at times) adversarial relationship between the developer of a new technology and the pricing and reimbursement agency, to that of a partnership in evidence development for market access. Certainly, the developer involved in this pilot study was keen to give feedback to the agencies concerned about the value and practical implications of the advice offered.
      If, on balance, early dialogue is judged to be socially beneficial, it would make sense for the budgets of the agencies to be increased to enable them to engage in such activities. In particular, it would be important that those individuals offering advice on behalf of agencies are adequately trained and also have personal experience of the agencies' decision-making processes (e.g., through having been an expert committee member in the past). Since the time of this pilot study, one agency has initiated a consultancy service, whereby manufacturers can pay for scientific advice on clinical trial design (and other matters). National Institute for Health and Clinical Excellence (NICE) scientific advice consultancy service. Details available from www.nice.org.uk. (Accessed October 1, 2010.)
      One concern is that, given the large number of pricing and reimbursement agencies, having early dialogue with all of them could be very resource-intensive both for companies and the agencies themselves. To this end, the finding that the key elements of the advice did not differ greatly across agencies is encouraging, although more investigation is required before we could be confident to suggest that companies need only consult a subset of the agencies.
      Where resource constraints prevent an otherwise willing agency to provide advice, it may be possible for pan-national consortia to evolve, in order to spread the load of offering early advice. However, the scope for this is likely to be limited to considerations of comparative clinical evidence due to the fact that views about economic value and decision-making criteria vary from jurisdiction to jurisdiction. Also, the differing remits and objectives of the various agencies might further limit the scope for joint advice.

      How could the process be improved?

      The company received feedback that some elements of the process worked well, such as the production of a briefing book. However, it was also observed that, in most cases, the time allocated for the meetings was too short. One minor disappointment from the company's viewpoint was that only one of the agencies was willing, or able, to provide written feedback. This was regarded as important by the company—it removed any doubt about precisely what advice was being given. In the future, it might be possible to produce a summary of the advice and send it to the agency for comment or even approval.
      A related issue is whether the advice offered should be binding on either party. Although this would further reduce any uncertainty on the part of the company, it is probably premature to take such a step. Indeed, it is probably unrealistic to expect such advice to be binding because the assessment of a given product is comparative to any other relevant alternatives existing at the time. Therefore, it will depend not only on the evidence of the company's compound, but also on that of the current and future alternatives. In any case, as with the advice given by regulatory agencies, the advice offered relates only to desirable evidential standards and does not include a discussion of what the agency would be likely to recommend if presented with a given set of data about the compound.
      Another question is whether the advice, and the discussions that lead to it, should remain confidential to the parties concerned, or be made public. This echoes a similar debate about whether company submissions to pricing and reimbursement agencies should be in the public domain [
      • Drummond M.F.
      Should commercial-in-confidence data be used by decision makers when making assessments of cost-effectiveness?.
      ]. The main argument against publication is that some of the material discussed is likely to be commercially sensitive (e.g., the company's aspirations for market positioning, or the results of the phase 2 trials).
      In some instances, however, there may be arguments to make public some of the elements of the advice being given. For example, if several companies are developing compounds in the same clinical area there may be benefits from standardizing the design or data collection within clinical trials. This would assist payers in making comparisons among the compounds concerned and would be analogous to the process, followed by some regulatory agencies, of issuing disease-specific guidance notes [
      • Backhouse M.E.
      Raising the standards of trial-based economic evaluation: the devil is in the detail.
      ].
      Therefore, the need to preserve the confidentiality of advice is not an easy issue to resolve, although in situations where the developer is asked to pay for the advice, it must surely remain proprietary. Should the provision of early advice be considered beneficial from a societal perspective, there would need to be more discussion of the terms on which it is offered; it clearly has both ‘private good’ and ‘public good’ components.

      Should tripartite advice be considered?

      It was mentioned earlier that sometimes the scientific evidence requirements of regulatory agencies and payers differ. This raises the question of whether early dialogue should be sought with regulators and payers in parallel (i.e., tripartite advice). It could be argued that, through the growing interest of regulators in “relative efficacy,” some convergence is already taking place [
      • Eichler H.-G.
      • Bloechl-Daum B.
      • Abadie E.
      • et al.
      Relative efficacy of drugs: an emerging issue between regulators and third-party payers.
      ]. If tripartite meetings could facilitate this process they may be beneficial. For example, the pricing and reimbursement agency might suggest a non-trivial change to proposed phase 3 clinical trial design; in a tripartite meeting, the developer of a new technology would be able to get a response on the proposed change from the regulator in real time.
      A pilot study is currently being conducted in Sweden, where companies can have a joint meeting with the Medical Products Agency (Lakemedelsverket) and the Dental and Pharmaceutical Benefits Agency (TLV) in order to obtain parallel advice [

      Ljungberg B. Presentation at the Health Technology Assessment International (HTAi) Annual Conference. Dublin, Ireland; June 2010.

      ]. In the future it will be possible to assess whether these meetings have proved successful.

      Conclusions

      This pilot project demonstrated that a feasible process of early dialogue could be established. Although there was some variation in the advice obtained from the various agencies, the similarities far outweighed the differences. Several important conclusions were reached about the viability of the company's aspirations for product positioning, the desirability of particular comparisons, the outcomes to be used in trials and the length of trial follow-up.
      More experience of early dialogue needs to be accumulated, involving a wider range of agencies and compounds in different clinical areas. If the process can be shown to be beneficial from a societal perspective, it would be desirable for adequate funding to be made available, either by expanding the budgets of the relevant agencies, or through self-financing arrangements like those recently established by one of the agencies.

      References

        • International Society of Pharmacoeconomics and Outcomes Research (ISPOR)
        Pharmacoeconomic guidelines around the world.
        www.ispor.org
        Date: 2010
        (Accessed February 1, 2010)
        • Drummond M.F.
        • Evans B.
        • LeLorier J.
        • et al.
        Evidence and values: requirements for public reimbursement of drugs for rare diseases – a case study in oncology.
        Can J Clin Pharmacol. 2009; 16: e273-e281
        • Trotta F.
        • Apolone G.
        • Garattini S.
        • Tafuri G.
        Stopping a trial early in oncology: for patients or for industry?.
        Ann Oncol. 2008; 19: 1347-1353
        • Fleming T.R.
        • Rothmann M.D.
        • Lu H.L.
        Issues in using progression-free survival when evaluating oncology products.
        J Clin Oncol. 2009; 27: 2874-2880
        • Backhouse M.E.
        Raising the standards of trial-based economic evaluation: the devil is in the detail.
        Value Health. 2005; 8: 519-520
        • Ades A.
        • Sculpher M.
        • Sutton A.
        • et al.
        Bayesian methods for evidence synthesis in cost-effectiveness analysis.
        Pharmacoeconomics. 2006; 24: 1-19
        • Neumann P.J.
        • Drummond M.F.
        • Jönsson B.
        • et al.
        Are key principles for improved health technology assessment supported and used by health technology assessment organizations?.
        Int J Technol Assess Health Care. 2010; 26: 71-78
        • Drummond M.F.
        Should commercial-in-confidence data be used by decision makers when making assessments of cost-effectiveness?.
        Applied Health Econ Health Policy. 2002; 1: 53-54
        • Eichler H.-G.
        • Bloechl-Daum B.
        • Abadie E.
        • et al.
        Relative efficacy of drugs: an emerging issue between regulators and third-party payers.
        Nature Rev Drug Discovery. 2010; 9: 277-291
      1. Ljungberg B. Presentation at the Health Technology Assessment International (HTAi) Annual Conference. Dublin, Ireland; June 2010.