Section 2.1 of this Chapter describes the Canadian generic drug manufacturing sector. Section 2.2 outlines the considerations manufacturers take into account in determining whether to supply a particular generic drug. Section 2.3 discusses the barriers to entry into the supply of a generic drug. Section 2.4 examines the dimensions for competition among generic manufacturers. Finally, section 2.5 considers the state of manufacturing competition in Canada.
There are over 15 suppliers of generic drugs in the country with 13 companies having manufacturing facilities in Canada. The largest Canadian manufacturer, Apotex, is domestically owned and controlled. 13 Of the next nine largest suppliers, seven have a parent company or group that is foreign-based.
The larger manufacturers tend to offer a large portfolio of drugs across multiple therapeutic classes and in a variety of forms, while others are less diversified or more specialized. For example, Taro Pharmaceuticals, an Israeli pharmaceutical company entered the Canadian market in 1984 and specializes in topical products. Hospira, a 2005 entrant, specializes in products used in hospitals including critical care products and specialty injectable pharmaceuticals. Sandoz acquired Sabex in 2004, and it specializes in injectable and ophthalmic generic pharmaceutical products.
Table 1 shows the ranking of generic manufacturers based on the value of their sales to hospitals and retail pharmacies in Canada.
| 2006 Rank | Manufacturer | Year 2006 $(000s) | Year 2006 (%) | Year 2006 Cumulative (%) |
|---|---|---|---|---|
| 1 | Apotex |
1,100.8 |
34.16 |
34.16 |
| 2 | Novopharm |
483.0 |
14.99 |
49.15 |
| 3 | Genpharm 14 |
365.3 |
11.34 |
60.48 |
| 4 | Ratiopharm |
359.5 |
11.16 |
71.64 |
| 5 | Pharmascience |
280.5 |
8.70 |
80.34 |
| 6 | Sandoz Canada |
190.1 |
5.90 |
86.24 |
| 7 | Cobalt Pharma |
77.4 |
2.40 |
88.65 |
| 8 | Mayne Pharma Canada 15 |
54.8 |
1.70 |
90.35 |
| 9 | Taro Pharmaceuticals16 |
37.3 |
1.16 |
91.50 |
| 10 | Ranbaxy Pharmaceuticals Canada |
34.2 |
1.06 |
92.56 |
| 11 | Laboratoires Riva |
28.2 |
0.88 |
93.44 |
| 12 | Nu-Pharm |
14.8 |
0.46 |
93.90 |
| 13 | Hospira |
14.3 |
0.44 |
94.34 |
| 14 | Dominion Pharmacal |
12.5 |
0.39 |
94.73 |
| 15 | ProDoc |
11.6 |
0.36 |
95.09 |
| Others |
158.2 |
4.91 |
100.00 | |
| All Manufacturers |
3,222.5 |
100.0 |
||
Source: IMS Health.
Generic manufacturers provide their products through three main supply routes: Independent pharmacy distrubutors (IPD), pharmacy chain self distributors, and direct to pharmacy shipments. IPD, discussed in the next chapter, are the principal supply route followed by self distribution. Some direct sales continue to occur but are a declining means for providing supply.
Manufacturers consider several factors when determining whether or not to develop and introduce an independent generic (IG) product. Key considerations include the following:
Once all factors and risks are considered, the manufacturer is then in a position to calculate its projected sales versus costs. If the expected return on investment is favourable, then the decision to develop the product may go forward. There is no unique entry threshold for molecules coming off patent. It varies among manufacturers and depends on the characteristics of the molecule, the manufacturer and the barriers to entry.
Generics may be classified into IGs, developed and supplied without authorization by the brand drug manufacturer, and authorized generics (AG) that are supplied under licenses granted by the relevant brand drug company. 19 In bringing an IG to the market, a manufacturer encounters various barriers to entry. Key barriers to entry relate to sunk costs associated with drug development, regulatory approval and provincial formulary listings. 20
The development of IGs normally involves three key steps:
The development costs of an IG may not be specific to the sale of the product in any particular country. Generic products developed and manufactured in one country can be supplied to other countries, provided they meet the other countries' specific regulatory requirements for approval.
Those contacted for this study indicated that development costs for a generic product can vary greatly from one to the next. Even in simple cases, costs may be around $1.5 million. However, they can be several times higher for more complicated products, such as biologics.
In order to market an IG in Canada, a manufacturer must obtain approval from Health Canada under the Patented Medicines (Notice of Compliance) Regulations (NOC Regulations) . The NOC Regulations , as explained in detail in Appendix 1, address two issues, first, whether the IG is bio-equivalent to the Canadian brand reference product, and, second, whether the IG infringes any valid patents.
To market an IG, the manufacturer must file an Abbreviated New Drug Submission (ANDS) with the Therapeutic Products Directorate (TPD) of Health Canada, containing data that demonstrate the drug's bio-equivalence with a Canadian reference brand product.
The ANDS must contain sufficient information for Health Canada to assess the bio-equivalence of the generic to the brand-name product, as well as evidence of tests conducted on potency, purity and stability of the new drug. 21
Standard bio-equivalence studies measure the rate and extent of absorption - or bio-availability - of a generic drug. This is then compared to the same characteristics of the reference drug product. The bio-availability of the generic drug must fall within an acceptable range of the bio-availability of the reference product. According to those contacted for this study, typical costs for conducting bio-equivalency studies are in the range of $1-1.5 million per product.
In the case of generic drugs, clinical trials are generally required for:
For example, topical products do not enter the blood stream so they are tested through clinical trials.
Clinical trials are research programs conducted to evaluate a new medical treatment, drug or device. These studies involve patients in the testing of treatments and therapies. Clinical trials, measure a drug's safety, effectiveness, dosage requirements and side effects. They are normally much more costly and time-consuming than bio-equivalence studies.
In doing its assessment of the bio-equivalence of a generic product (or an ANDS), Health Canada relies on data provided by the brand-name firm at the time it applied for a Notice of Compliance (NOC) for its product. These data are subject to a minimum period of protection from the date the reference product received its approval from Health Canada to be marketed. This period of protection, originally five years, was lengthened to eight years under amendments to the NOC Regulations in 2006. Where it extends beyond the life of the patent, the extended period of data protection may create an additional delay in bringing the generic drug to the market. The new regulations also allow six added months of data protection for drugs that have been the subject of clinical trials in children.
Once the ANDS is filed and, when applicable, the period of data protection ends, Health Canada typically takes between 12 and 18 months to complete its review. 22
After filing an ANDS with the Minister, generic manufacturers are required under the NOC Regulations to serve a Notice of Allegation (NOA) on the patentee that the generic product will not infringe any patent rights. The patentee may then apply to the court for an order prohibiting the Minister of Health from issuing an NOC on the basis that one of its patents is being infringed. In such cases, the Minister cannot issue an NOC until 24 months have passed or the application has been dismissed. Therefore, the patentee can prevent a generic product from entering the market for up to 24 months, simply by alleging that its patents have been infringed.
Prior to 2006, generics were required to address all patents added by the patentee to the Patent Register with respect to the reference drug product. In 2006, the NOC Regulations were amended to restrict the ability of a drug innovator to prevent a generic from getting an NOC by adding patents to the patent register after the generic manufacturer files an ANDS. 23 The generic now only has to address patents that were listed on the register in respect of the reference drug prior to the filing date of the ANDS. 24
If a patentee obtains a stay preventing the Minister from issuing an NOC, but the patents relied upon are later found to be invalid or not infringed, the generic firm that was kept off the market may seek damages for its losses. Under s. 8 of the NOC Regulations , the court may ?make any order for relief by way of damages that the circumstances require?. 25
In addition to the NOC Regulations , in some cases, the patentee may rely on a patent lawsuit to prevent entry of a generic drug or to recover damages. In such cases, a generic might succeed under the NOC Regulations , market the drug and then be sued by the brand-name manufacturer for patent infringement. In this case, if the brand-name manufacturer is successful, the generic would likely be required to pay damages to the patentee. Conversely, a generic manufacturer may challenge the validity of a patent under the Patent Act if it is preventing the company from receiving a NOC.
Success in the NOC proceedings by a particular firm does not automatically create free entry for all generic firms. Other generic firms still have to obtain an NOC, and address any patents on the Patent Register. Subsequent generic firms may, however, make the same arguments in litigation as the first successful generic. In some cases, the patentee may stop contesting these NOC cases.
Those interviewed for this study, while not providing related data, indicated that patent challenges under the NOC Regulations are commonly encountered and are a normal part of bringing an IG to market. Legal costs for the first generic to challenge were said to be commonly in excess of $1 million and potentially much higher in complicated cases. However, the costs for subsequent generic manufacturers, for the same reference product, can be as low as a few thousand dollars when NOAs are no longer being challenged.
Once an NOC is issued, a product can be sold anywhere in Canada. However, in order to be reimbursed under provincial drug programs and obtain significant sales volumes the generic product must be listed on provincial formularies. For an IG, the formulary listing process can take several months from the time an NOC is issued.
In sum, from the time a decision is made to produce a generic drug, manufacturers typically require between three to six years to bring the product to market. While costs can vary widely from case to case, they can be in the range of $3.5 million (including costs for bio-equivalence studies, development and regulatory approval) even for a relatively non-complex product.
These costs may be lower where, for example, patent challenges are not encountered or product development costs can be spread across sales in countries other than Canada. On the other hand, they can be much higher when product development is more complicated, clinical trials are required, or relatively high patent challenge costs are encountered. For example, the costs for the development of bio-generics can be as high as $25 to $50 million. Industry sources have indicated that it may take as long as three years after a generic product is introduced to market before it will break even, recouping its sunk developmental and approval costs.
Competition between generic manufacturers takes place in a number of dimensions. The key ones are: timing to market, patent challenges, pricing, AG, and breadth of product line.
Those contacted for this study cited timing to market as being a key dimension of generic competition. Pharmacies are less likely to switch to a new generic product if they already have one or two versions in stock. Stocking multiple manufacturers of the same molecule is cumbersome and inefficient. For this reason, ?timing is of the essence? in the generic drug industry. Product development and approval is carefully planned to maximize the likelihood of having a generic version ready as soon as a brand-name product loses patent protection.
The advantage of being first to market is supported by analysis performed on molecules that lost patent protection and encountered generic entry between January 1998 and December 2006. As shown in Table 2, for about two thirds of the molecules, the first entrant was able to maintain the leader's position at the end of 2006.
Table 2. Status Of The First Generic Entrant
| Number of Molecules | Percentage | ||
|---|---|---|---|
|
First generic entrant stayed first |
49 |
65.3 | |
|
First generic dropped to 2 nd position |
14 |
18.6 | |
|
First generic dropped to 3 rd position |
6 |
8.0 | |
|
First generic dropped to 4 th position or lower |
6 |
8.0 | |
|
Total |
75 |
100.0 |
Data source: IMS Health.
A competitive dimension related to timing to market is companies' patent challenge strategies. A generic company may file its ANDS to market a generic because the brand-name drug's main patent has expired or is about to expire. By marketing the generic, the generic company is not infringing on any of the other patents that are held by the brand-name company. 26 However, sources contacted for the study indicated that generic companies commonly enter the market prior to the expiry of all listed patents based on the belief that any remaining brand company patents are invalid or would not be infringed.
Companies that are the first to file a challenge may gain an advantage over others by getting their product into the supply chain earlier. However, not all generic manufacturers aggressively pursue legal challenges. According to industry sources, some generic manufacturers challenge only those patents where there is a perceived certainty of a positive outcome, such as where a brand company is no longer challenging NOAs. They may avoid the costs of legal proceedings altogether by timing their entry to the market in line with the brand's patent expiration.
While a generic that first successfully challenges brand patents may have the advantage of being first to market, this can be a costly process. The generic manufacturer has to evaluate whether costs sunk into a patent challenge can be recouped after the product launches.
In cases where the brand manufacturer fights the first generic challenger but gives up further challenges, thereby opening the market to all generics, the first generic challenger may not obtain a major first mover advantage. The generic may be in a situation where it is out of pocket for legal costs and has to compete against other generics, IGs or AG, which did not incur the same costs. 27
In the case of sales to retail pharmacies, pricing decisions by manufacturers consist of two elements: the establishment of the product's invoice price and the net pharmacy price. The net pharmacy price is the price paid by the pharmacy net of any off invoice rebates and discounts. Invoice prices are the amounts typically reimbursed by public and private drug plans. As developed further in section 5.A., limited competition appears to take place in invoice prices. Until recently, invoice prices have tended to reflect maximum generic prices allowed under Ontario legislation. Price competition among manufacturers has tended to take place at the pharmacy level in the form of lower net pharmacy prices. Once generic versions of brand-name products are placed on provincial formularies and are designated as interchangeable, they essentially become commodity products. 28
This situation results in pharmacies being the most important and influential customers of generic manufacturers. Traditionally, the most important factor in competing for pharmacies' business, where there are multiple generics available, has been generic manufacturers providing rebates off invoice prices. 29 Rebates on generic drugs are not recorded on invoices, but are provided to pharmacies and hospitals in a separate transaction often as a lump sum for drugs purchased in a given period.
It has not been possible to obtain information about the precise size and nature of rebates from manufacturers to retail pharmacies and hospitals. Average rebates have been estimated to be 40%, although sources indicated they may have been higher. 30 Sources further indicated that rebates have been as high as 80% for individual generic products.
The traditional role of rebates as a competitive dimension is being altered by the Ontario Transparent Drug System for Patients Act, 2006 , discussed further in Section 4.A.2. The legislation prohibits the granting of rebates to pharmacies. While it allows professional allowances to be provided as a possible alternative to rebates, these are capped at 20% of pharmacies' costs for drugs dispensed under Ontario Drug Benefit (ODB) programs. In addition, the legislation, with certain exceptions, reduces the maximum amount that can be reimbursed for generics, under ODB plans, to 50% of the brand drug price. These generic drug price or professional allowance caps do not apply to drugs dispensed under private drug plans. The legislation makes Ontario the second province in Canada to prohibit rebates. Such rebates have been prohibited for several years in Quebec and have been recently the subject of a number of legal actions. 31
While the full effects of the Ontario legislation are to be determined, the capping of generic drug professional allowances limits a key dimension of competition among generic drug manufacturers. The altered competitive framework may be particularly problematic for generic drug manufacturers with limited product portfolios. The ability to grant higher rebates or allowances can provide them a means to enter and expand market share in competition against rivals with broader product lines. With rebates and allowances being restricted or prohibited, it can be anticipated that competition in other areas, such as breadth of product line, will assume greater importance.
While the full effects of the Ontario legislation are to be determined, the capping of generic drug professional allowances limits a key dimension of competition among generic drug manufacturers. The altered competitive framework may be particularly problematic for generic drug manufacturers with limited product portfolios. The ability to grant higher rebates or allowances can provide them a means to enter and expand market share in competition against rivals with broader product lines. With rebates and allowances being restricted or prohibited, it can be anticipated that competition in other areas, such as breadth of product line, will assume greater importance.
AG are ?the actual brand-name drug product manufactured by the brand company, but sold as a generic by a licensee or subsidiary of the brand, competing with independent generics.? 32 Because they are identical to the branded drugs and approved by the patent holder, AG do not encounter the product development and federal regulatory approval barriers to entry that apply to IGs. Although in some provinces listing of AG on provincial drug formularies can be faster, under the streamlined formulary listing process employed by most provinces there is no advantage for AG.
Introducing an AG prior to the expiration of a brand-name product's period of patent protection runs counter to the business interests of a brand-name manufacturer. The lower-price AG will simply erode the market share of its higher priced brand-name counterpart diminishing the brand company's revenues. However, licensing the supply of an AG after the end of patent protection potentially provides the brand company a means to make some returns on a portion of generic drug sales.
A brand-name manufacturer may decide to license the manufacturing and distribution of the AG to an IG manufacturer. The decision of an IG manufacturer to partner with a brand-name manufacturer for the release of an AG is based on several factors. These may include their ability to source APIs to produce their own generic version and the expected return on supply of the AG versus developing and marketing its own IG. IG manufacturers differ on their AG strategies. While some engage in little if any supply of AG, others incorporate them as a component of their business strategy. According to industry sources, the number of AG available in the Canadian market has been trending downwards. In 2006, AG accounted for only about 7% of the generic sales, compared to about 15% in the early 90s.
An issue about introducing an AG is that it may affect the incentive for a generic manufacturer to develop an IG. 33 This is unlikely to be an issue for drugs having high sales relative to entry costs. However, it has the potential to affect the entry of IGs for drugs having relatively smaller valued sales. This may be particularly significant when the AG is able to obtain a first mover advantage. This matter is considered in Table 3.
Statistical analysis was performed on a set of molecules that lost patent protection between 2001 and 2006 and where the first generic competitor entered within the period. An AG entered 26 (36%) of the 75 drug markets in the sample. 34 No clear pattern was found of AG entering first. Of the 26 markets in which both an AG and an IG entered, the IG entered first in 12, the AG entered first in 11. They both entered in the same month in three markets. Note that in about half of the cases, the AG entered the market after an IG. However, in only two of the cases where it entered first, was the AG able to maintain the highest share. Table 3 shows the status of the AG in January 2007 and the timing of AG entry.
Table 3. Status Of The Authorized Generic After Independent Generic Entry
| Number of molecules | |
|---|---|
|
AG entered before the IG |
11 |
|
AG entered 1 st and retained highest share |
2 |
|
AG entered at the same time as the IG |
3 |
|
AG entered after the IG |
12 |
|
Total |
26 |
Data source: IMS Health.
The sample does not show a clear and consistent pattern of AG entering before IGs. Moreover, where they do enter first, AG, while they may obtain high market share for an initial period, retain leadership over time in only a small number of cases. 35
As discussed further in section 4.A, given the commodity nature of generic drugs, other things equal, pharmacies can reduce their costs by dealing with as few manufacturers as possible. This provides more diversified manufacturing firms with a competitive advantage over competitors with smaller product lines as they are able to bundle a portfolio of products across multiple therapeutic classes. 36 As indicated above, one means by which less diversified manufacturers have been able to overcome this disadvantage has been by offering lower net pharmacy prices.
The current competitive structure of the Canadian generic drug manufacturing sector is significantly different from that of the early 1990's. At that time, Apotex and Novopharm accounted for the majority of sales in the domestic market (72.8%). 37 In 2006, although the two largest firms remained Apotex and Novopharm, with approximately 50% of sales, the top four firms accounted for under 72% of sales.
The dynamics of the generic drug manufacturing sector is also being altered by increasing globalization. In 2000, Teva, a large Israeli generic drug manufacturer, entered the Canadian sector by purchasing Novopharm. This was followed by the expansion into Canada of Ratiopharm, a German generic drug company and one of the leading international generic producers. The third Canadian largest supplier, Genpharm, was recently acquired by a U.S. generic company, Mylan Laboratories from Merck, based in Germany. Indian generic manufacturers have also entered the Canadian sector through the entry of Ranbaxy in 2005, and the acquisition of Taro by Sun Pharmaceuticals in 2007.
An in depth analysis of the competition across the sector could not be done as the information on such matters as the net pharmacy prices and manufacturing costs for individual drugs was unavailable. 38 However, it appears that supply for many generic products is highly competitive. The expiration of brand-name pharmaceutical patents can be met by the introduction of multiple generic products. The number of competitive suppliers is more likely to be large in markets for popular molecules, the so-called ?blockbuster drugs.? Chart 1 shows the number of generic entrants per molecule and the sales of the brand in the year prior to generic entry. As the chart indicates, molecules with large sales tend to attract a large number of generic competitors. 39
Chart 1. Generic Entry

Data source: IMS Health.
The effects of the competition among manufacturers have traditionally not been reflected in invoice prices for generic drugs. Rather, with price competition focused on pharmacies, its effects are reflected in net pharmacy prices. As indicated above, these prices have been estimated to be on average at least 40% below the invoice prices used by the PMPRB and other pricing studies.
This suggests that other elements of the Canadian generic sector competitive framework must be taken into consideration to explain the differences between invoice prices in Canada and other countries. As noted above, work done by the PMPRB indicates that although Canada ranks in the middle of six countries studied in terms of the average number of generic suppliers for each non-patented product, the country has substantially higher invoice prices for generic drugs than 10 of 11 countries covered in its 2006 generic prices study. 40
13 For the purpose of this analysis, we use the term ?manufacturer? , even though a company did not manufacture but just distributes the product in Canada. According the Food and Drug Regulations, C.R.C., c. 870 , a ?manufacturer? of a drug is not necessarily the company that makes the product, but the company to which the product is registered at the time of approval.
14 Recently bought by Mylan Laboratories Inc. as part of its acquisition of Merck KGaA's generic business, Genpharm's parent company.
15 Recently bought by Hospira Inc. as part of its acquisition of Mayne Pharma Limited, Mayne Pharma Canada's parent company.
16 Recently bought by Sun Pharmaceutical Industries Limited, an Indian pharmaceuticals company.
17 While NOC Regulations prevent a firm from using the process to delay a generic version of the original formulation when the brand-name drug loses patent protection, it does not prevent a brand-name firm from marketing ?new and improved? formulations.
18 The approval process is described in more detail in the next section.
19 Licensing may also take place between two generics manufacturers.
20 Sunk costs are costs that are non-recoverable once spent.
21 The generic firm may undertake its own clinical trials instead of conducting bio-equivalence studies. In practice, however, showing bio-equivalence is much less expensive and generic firms almost always choose this path. See Bristol-Myers Squibb Co. v. Canada (Attorney General) , 2005 SCC 26.
22 In the case of topical products, the NOC application cannot be submitted until after the clinical trial results are available. Once the NOC application has been submitted, approval of topical prescription products takes from six to eight months.
23 In a subsequent 2006 decision, the Supreme Court of Canada held that a generic manufacturer is only required to address patents on the Patent Register that are relevant to the actual comparator drug. In addition, the generic manufacturer is not required to address patents issued after the NOA was made (since the generic manufacturer could have received no benefit from those patents). See AstraZeneca Canada Inc. v. Canada (Minister of Health), [2006] S.C.J. No. 49.
26 In addition to patents related to the active ingredient(s), formulation and process patents are listed by brand-name companies on the Patent Register. Typically, the patents on active ingredients expire first, thus giving generic manufacturers the possibility to enter the market by challenging the remaining patents prior to their expiration.
27It has been suggested that this could result in there being limited incentive to challenge patents. While this may be unlikely to be the case for popular drugs, it could affect the supply of generics for drugs with limited use and/or smaller sales. Examining this matter is an empirical issue beyond the scope of this study.
28 As developed in section 5.A., there may be limited exceptions for medical reasons.
29 Effective supply chain management is another key consideration. Pharmacies want to be sure that a drug is available to be dispensed to patients when needed.
30 In 2004, the province took four different legal actions before the Superior court of Quebec against four manufacturers of generic drugs (Apotex, Novopharm, Pharmascience and Ratiopharm) alleging that they had, between 2000 and 2003 given approximately 37% of illegal rebates and discounts. See for example the decision of the Superior Court of Quebec dated July 27, 2004, with respect to Quebec (Régie de l'Assurance-maladie) c. Pharmascience Inc., 2004 CanLII 4667 (QC C.S.) . See also respective files of the Superior court of Quebec no 500-17-015571-030, no 500-17-015460-036 and no 500-17-015406-039. In Quebec, Bill 130 adopted in 2005 and the Quebec Drug Policy published in February 2007 have set the stage for future ?professional allowances? similar to Ontario's to be provided. However, they are not yet included in regulations.
31Public sources that put the average rebate at 40% include: i) CIBC World Markets, ?2003 Investors' Guide To The Canadian Drugstore Industry?, May 26, 2003 and ii) Ontario Ministry of Health and Long-term Care, ?Challenges Facing Ontario's Drug System And How We Are Responding To Them?, available at: www.health.gov.on.ca/english/media/news_releases/archives/nr_06/apr/bg_041306_a.pdf. The implications of rebates for pharmacies are discussed in section 4.A.2.
32 Aidan Hollis and Bryan Liang, ?Assessing the effects of authorized generics on consumer prices? Journal of Biolaw and Business , forthcoming.
33 The issue of authorized generics and their role in providing competition to independent generics is being considered by the US Federal Trade Commission, which is conducting a related market study.
34 A drug market was defined for the purpose of the study as a unique combination of molecule and dosage form.
35 These results are partial, based on a limited set of drugs. More information (e.g. a broader sample size, information on terms of contract and market size) is needed to assess fully the impact of AG on the competitive framework for generic drugs.
36 While such bundling is not inherently anti-competitive, bundling can have anti-competitive effects in certain circumstances, for example, where it is used by a dominant firm to exclude competitors from the market resulting in a substantially lessening of competition.
37 Source: Canadian Generic Pharmaceuticals Association ( Canadian Generic Pharmaceuticals Association ).
38 Further, such an analysis would require detailed information regarding which products should be included in the relevant markets and related barriers to entry. For example, the mere finding that a non-patented product has one or a small number of suppliers, is not adequate to conclude that is not subject to competition.
39 A set of 32 molecules for which the first generic entered between January 2002 and July 2006 was analyzed. Brand sales in the year prior to the first generic entrant are considered.