International Journal of the Economics of Business
`
`ISSN: 1357-1516 (Print) 1466-1829 (Online) Journal homepage: http://www.tandfonline.com/loi/cijb20
`
`Similar Products at Different Prices: Can
`Biopharmaceutical Companies Segment Markets?
`
`Richard Manning, Christopher Stomberg, Benjamin Scher, Kathleen Twigg &
`Andrew Huson
`
`To cite this article: Richard Manning, Christopher Stomberg, Benjamin Scher, Kathleen Twigg &
`Andrew Huson (2015) Similar Products at Different Prices: Can Biopharmaceutical Companies
`Segment Markets?, International Journal of the Economics of Business, 22:2, 231-243, DOI:
`10.1080/13571516.2015.1045739
`
`To link to this article: http://dx.doi.org/10.1080/13571516.2015.1045739
`
`Published online: 26 Jun 2015.
`
`Submit your article to this journal
`
`Article views: 65
`
`View related articles
`
`View Crossmark data
`
`Citing articles: 1 View citing articles
`
`Full Terms & Conditions of access and use can be found at
`http://www.tandfonline.com/action/journalInformation
`
`Download by: [Ms Julia Livingston]
`
`Date: 06 July 2016, At: 11:36
`
`Samsung et al. v. Regeneron IPR2023-00884
`Regeneron Pharmaceuticals, Inc. Exhibit 2158 Page 1
`
`

`

`Int. J. of the Economics of Business, 2015
`Vol. 22, No. 2, 231–243, http://dx.doi.org/10.1080/13571516.2015.1045739
`
`Similar Products at Different Prices: Can
`Biopharmaceutical Companies Segment Markets?
`
`RICHARD MANNING, CHRISTOPHER STOMBERG,
`BENJAMIN SCHER, KATHLEEN TWIGG and ANDREW HUSON
`
`ABSTRACT New data on Medicare Part B payments show provider-level utilization of
`two similar but distinct biologic medications used to treat wet age-related macular
`degeneration. Their interaction provides insight into the ability of manufacturers to
`effectively segment markets and suggests some analogies for the future interplay
`between originator biologics and biosimilars. In particular, most ophthalmologists
`administer a mix of Avastin (used off-label), Lucentis, and other drugs; only a small
`share exclusively choose Lucentis. This is inconsistent with the hypotheses that (1) the
`manufacturer is able to effectively segment the market, and (2) the major factor
`driving physicians’ product choice is the financial motivation. The data are consistent
`with the notion that physicians typically exercise medical
`judgment on a patient-
`by-patient basis and that product choice is driven largely by factors other than simple
`financial interests. This raises important dynamic efficiency considerations regarding
`incentives for future biologic competition.
`
`Key Words: Pricing; Biotechnology; Physician; Prescription; Firm Behavior.
`
`JEL classifications: L110, D4, L650, I11, L2.
`
`Background
`
`The biopharmaceutical industry has been studied extensively by scholars in
`industrial organization and public policy because its structure, regulatory
`environment, and pricing are unique among modern industries. Early articles
`exploring the dynamics of this industry include Comanor (1964, 1986).
`Among the most fundamental aspects of the modern biopharmaceutical
`industry are the discovery and development of new medicines and the
`interplay between alternative treatments. In the most common case, when a
`
`The following product names referenced throughout the article are registered trademarks of the
`indicated companies: Lucentis, Genentech,
`Inc.; Avastin, Genentech,
`Inc.; Eylea, Regeneron
`Pharmaceuticals, Inc.; Visudyne, Novartis, AG; Macugen, Eyetech, Inc.
`
`Richard Manning, Christopher Stomberg, Benjamin Scher, Kathleen Twigg and Andrew Huson, Bates White,
`LLC, 1300 I Street NW, Suite 600, Washington, DC 20005, USA. All correspondence should be directed to
`Richard Manning, 202 216 115 7; e mail: richard.manning@bateswhite.com.
`
`Ó 2015 International Journal of the Economics of Business
`
`Downloaded by [Ms Julia Livingston] at 11:36 06 July 2016
`
`Samsung et al. v. Regeneron IPR2023-00884
`Regeneron Pharmaceuticals, Inc. Exhibit 2158 Page 2
`
`

`

`232 R. Manning et al.
`
`new therapy is introduced, that therapy typically enjoys patent protection (or
`in some cases another form of regulatory exclusivity) from immediate copying
`for some period, during which it may compete with previously developed
`alternative (nonchemically identical)
`therapies. After patent protection (or
`regulatory exclusivity) ends, small-molecule branded drugs are subject
`to
`intense competition. This competition comes from generic entrants who face
`low manufacturing costs, typically charge far lower prices, and take advantage
`of
`institutional pressures that encourage the use of generics. Key early
`contributions to the literature on the effects of patent expiration and the impact
`of generic entry are Caves et al. (1991) and Grabowski and Vernon (1992).
`An interesting development currently unfolding in the United States – one
`that has been unfolding in Europe for nearly a decade – is the competitive
`interplay between innovative large-molecule biologic medications and their
`“generic” counterparts, or “biosimilars.” A regulatory pathway for the review
`of biosimilars was authorized under the Biologics Price Competition and
`Innovation Act of 2009 (BPCIA; part of the Affordable Care Act), but the exact
`details of what
`it
`takes to gain Food and Drug Administration (FDA)
`marketing authorization are still evolving. An important step in that process,
`however, is an FDA review committee’s recommendation for approval of the
`first biosimilar in the United States (see Novartis 2015). Going forward, the
`interplay between competing similar but distinct biologic medications will be
`an issue of intense interest for policy makers, biopharmaceutical company
`management, patients, and academics.
`There is a substantial literature that attempts to forecast how competition
`will unfold between originator biologics and the biosimilars that reference
`them. The primary view is that
`the competition confronting originator
`biologics from biosimilars will bear only faint similarities to the brand-generic
`experience. A partial
`list of articles discussing the potential outcome of
`biosimilar entry includes the following: Grabowski, Guha, and Salgado (2014);
`Trusheim, Aitken, and Berndt (2010); US Federal Trade Commission (2009);
`Grabowski and Kyle (2007); Grabowski, Ridley, and Schulman (2007); and
`Trusheim, Berndt, and Douglas (2007).
`We may not see the equivalent of the patent cliff that occurs in the small-
`molecule sector for three reasons: (1) the manufacturing of biologics is far more
`complex and costly than the manufacturing of small-molecule drugs; (2) the
`regulatory approval process is and will
`likely remain far more involved,
`keeping the number of potential competitors down; and (3) physicians and
`patients may not be quick to adopt unproven biosimilars. Meaningful utilization
`of originator molecules is likely to persist beyond patent (or data exclusivity)
`expiration, and prices may not fall dramatically over short periods of time. Over
`the coming years, as biologic therapies come to play an increasingly important
`role in medical care,
`the industrial organization of the biopharmaceutical
`industry is likely to change as well. It is useful to look into current market
`dynamics to get a sense of the direction and degree of those changes.
`As suggested above, an important question will be what happens when two
`similar but distinct medications can be substituted for each other in a high-value
`medical application. Although, as indicated, biosimilar entry in the United States
`is only on the verge of becoming a reality, there are examples of competition
`between different branded versions of biologics employing similar mechanisms
`of action, and these can shed light on the important dynamics that will unfold.
`
`Downloaded by [Ms Julia Livingston] at 11:36 06 July 2016
`
`Samsung et al. v. Regeneron IPR2023-00884
`Regeneron Pharmaceuticals, Inc. Exhibit 2158 Page 3
`
`

`

`Similar Products at Different Prices
`
`233
`
`Here, we focus on one particular example of the interplay between similar
`biologic therapies that are used for the prevention of blindness as a result of
`wet age-related macular degeneration (AMD). These products are Genentech’s
`Lucentis and Avastin. The chemical names of Avastin and Lucentis are,
`respectively, bevacizumab and ranibizumab. The brand names will be used for
`ease of discussion.
`Avastin was approved by the FDA in February 2004 with an indication for
`the treatment, in combination with certain chemotherapy agents, of metastatic
`colon and rectal cancer. It has since been approved for treatment of a wider
`range of cancers. Lucentis received FDA approval in June 2006 to treat wet
`AMD and has also received FDA approval for two additional indications, but
`our focus here will be on wet AMD treatment.
`Following the early emergence of evidence that Avastin injections might be
`beneficial in the treatment of wet AMD, ophthalmologists began using Avastin
`for this unapproved (off-label) use. There is clinical trial evidence suggesting
`that patients receiving Lucentis and Avastin for AMD have comparable
`outcomes (US National Institutes of Health 2012), but the FDA has also issued
`an alert to healthcare professionals indicating concern about potential safety
`risk from using inappropriately repackaged use of Avastin in its off-label use
`in treating wet AMD (US Food and Drug Administration 2011a).
`While the price per treatment of Avastin in its approved use (cancer) is not
`terribly different from the price of Lucentis in its approved use (wet AMD),
`the quantities needed in each use differ dramatically. For example, the dose of
`Avastin used in clinical trials for cancer is 500 times the dose used to treat wet
`AMD (US National Institutes of Health 2012). As a result of the dramatically
`different demand structures for these two drugs, the off-label use of Avastin to
`treat wet AMD results in much lower expenditures (see, e.g., Pershing et al.
`2015). Lucentis sells for about $2,000 per injection, while the dose of Avastin
`typically administered by ophthalmologists for the treatment of wet AMD sells
`for approximately $50 per injection (see Whoriskey and Keating 2013). Hence,
`it may be in the manufacturer’s economic interest to attempt to separate the
`markets for Lucentis and Avastin in their respective approved uses.
`A key question in the industrial organization of the biopharmaceutical
`industry is whether the owner of two apparently substitutable products with
`two distinct uses is able to segregate markets and sustain different prices that
`correspond to the value of the products in those distinct uses. While it might
`seem obvious that in an open market, price sensitive buyers would select the
`lower-priced product that suits their needs, there are questions about how
`patients and physicians perceive and respond to products that may or may not
`seem identical. There are also issues of dynamic efficiency and the incentive to
`develop new therapies. If a prospective innovator is unable to earn financial
`rewards for developing products that provide valuable treatment for serious
`conditions, less than optimal investment will be called into the search for such
`innovations. The appropriate balance between the static efficiency gains
`associated with allowing physicians and patients to choose Avastin for wet AMD
`treatment and the dynamic efficiency gains associated with the development of
`new treatments is an important question that is beyond the scope of this article,
`but it deserves careful attention as the market for biosimilars evolves.
`Another important question is the degree to which biopharmaceutical
`manufacturers are able to determine the conditions of use of their products.
`
`Downloaded by [Ms Julia Livingston] at 11:36 06 July 2016
`
`Samsung et al. v. Regeneron IPR2023-00884
`Regeneron Pharmaceuticals, Inc. Exhibit 2158 Page 4
`
`

`

`234 R. Manning et al.
`
`if manufacturers attempted and were able to exert substantial
`Clearly,
`influence in cases such as this, one would expect to observe substantial use of
`the higher-priced Lucentis and minimal use of the lower-priced alternative. A
`recent change in Medicare policy provides some insight into the utilization
`patterns of these two products for the treatment of wet AMD. What we find is
`that despite economic incentives to segment the market for these two products,
`the market is in fact not segregated very completely. As indicated in recently
`released Medicare Part B data, a large share of the potential demand for wet
`AMD treatment is filled with Avastin. In addition, the physician-level patterns
`of use between these two products are consistent with conditions that might
`reasonably be expected if physicians are acting in their patients’ interests.
`
`Insight from a New Data Source
`
`Lucentis versus Avastin
`
`Recently, the Centers for Medicare and Medicaid Services made available to the
`public for the first time detailed data regarding the Medicare Part B program.
`The data cover all of 2012 and contain a wealth of information – more than nine
`million records of data for more than 880,000 healthcare providers. For the first
`time, one can look up a participating physician by name in a publicly available
`data set and observe a great deal of
`information,
`including the services
`provided, the drugs administered, and the amounts paid by the Medicare Part
`B program for those services and drugs. Medicare Part B covers doctors’
`services and outpatient care (US Centers for Medicare and Medicaid Services
`2012). A brief article discussing the reaction to this data release and the issues
`discussed here was recently published in the online magazine Pharmaceutical
`Executive (2014) by three of our co-authors, Scher, Twigg and Huson.
`The choice physicians make between Lucentis and Avastin has been a
`common topic in the health policy literature (see, e.g., Hutton et al. 2014). The
`release of the Medicare data attracted a great deal of attention in the popular
`press, with prominent articles appearing in many of the country’s most widely
`read newspapers, websites, and other periodicals (see, e.g., Abelson and Cohen
`2014; Chen and Pearson 2014; Luhby 2014). Many of these articles focused on
`high-billing physicians, highlighting the amounts they have been paid by
`Medicare, and in some cases questioning the medical practices in which these
`physicians have engaged. Ophthalmologists specifically received scrutiny as a
`result of the data’s public release; several articles in the press identified these
`physicians as frequent high billers.
`Several articles focused on ophthalmologists’ prescribing behavior with
`respect to Avastin and Lucentis for the treatment of wet AMD. Medicare Part
`B payment to doctors when they inject these drugs is set at a 6% markup over
`the drug’s Average Sales Price (ASP), and hence doctors receive higher
`Medicare payments when they use a drug that costs more to acquire.
`The difference in the Medicare Part B payment between these two products
`has been cited in articles to suggest that financial incentives may be motivating
`ophthalmologists to use Lucentis instead of Avastin. Of course, one needs to
`be cautious about what inferences can and cannot be drawn based on these
`data alone – for example, the data do not provide information on the reasons
`for the physician’s product choice, the severity of patients’ medical conditions,
`
`Downloaded by [Ms Julia Livingston] at 11:36 06 July 2016
`
`Samsung et al. v. Regeneron IPR2023-00884
`Regeneron Pharmaceuticals, Inc. Exhibit 2158 Page 5
`
`

`

`Similar Products at Different Prices
`
`235
`
`the public data do not contain claim-level
`or patient outcomes. Also,
`information that would provide additional insight. Despite these limitations,
`these new data are useful in providing preliminary insight into two competing
`hypotheses regarding doctors’ use of these drugs.
`
`(cid:129) Do ophthalmologists who administer Lucentis appear to be doing so
`primarily because they earn higher Medicare payments? Specifically, if
`doctors view Lucentis and Avastin as very similar or identical, and if the
`only factor affecting the decision to administer one versus the other is
`whether the doctor is driven by financial incentives (and hence inclined
`to use Lucentis) or cost sensitivity (and hence inclined to use Avastin),
`then one would expect doctors to administer one or the other but not
`both.
`(cid:129) Alternatively, do ophthalmologists who administer Lucentis appear to be
`doing so because they are exercising medical judgment on a patient-by-
`patient basis? Specifically, if doctors view Lucentis and Avastin as at
`least somewhat differentiated, then they may believe that Lucentis is
`better for some patients and Avastin is better for others. In this case, one
`would expect doctors to administer a mix of the two drugs to their
`patients, depending on patient need.
`
`into these competing hypotheses, we focused on
`To gain insight
`ophthalmologists who used these drugs frequently, because these physicians
`had a greater incentive to choose the higher-priced product. Specifically, we
`limited the data to ophthalmologists with 200 or more patients who received
`injections of
`these drugs over
`the
`course of 2012. A total of 547
`ophthalmologists met these criteria and were included in our analysis. Among
`these high-volume doctors, 11% administered only Avastin to their patients,
`19% administered only Lucentis, and 70% administered a mix of the two drugs
`– Avastin to some patients and Lucentis to others. Note that we obtained
`similar results when limiting our analysis to ophthalmologists with 100 or
`more patients. In particular, 18% administered only Avastin to their patients,
`24% administered only Lucentis, and 59% administered a mix of the two drugs
`to their patients. In total, Avastin and Lucentis use was split quite evenly
`between the two products; in this sample, if all Avastin use had been switched
`to Lucentis, Lucentis use would have increased by 97%. This pattern of mixed
`utilization is consistent with the hypothesis that the vast majority of these
`high-prescribing doctors used Avastin on some patients and Lucentis on
`others, and it suggests that they were exercising medical judgment in their
`choice of treatment on a patient-by-patient basis, not simply responding to
`economic incentives.
`the
`if
`Moreover,
`it
`is obvious from even this simple analysis that
`manufacturer attempted to and was successful in separating the market for
`Lucentis from Avastin, utilization of Lucentis would be much more common,
`perhaps universal relative to Avastin in treating wet AMD. This is clearly not
`the case. In fact, slightly more patients treated by ophthalmologists are treated
`with Avastin than are treated with Lucentis. Given the price difference
`between the two products, if all physicians in this group who use Avastin
`switched entirely to Lucentis, revenue to the manufacturer from this sample of
`physicians would nearly double. Obviously, other elements are exerting
`substantial influence on product choice.
`
`Downloaded by [Ms Julia Livingston] at 11:36 06 July 2016
`
`Samsung et al. v. Regeneron IPR2023-00884
`Regeneron Pharmaceuticals, Inc. Exhibit 2158 Page 6
`
`

`

`236 R. Manning et al.
`
`For doctors who used these products in high volumes, the “mix” group
`often included a substantial number of patients being treated with each of the
`products. That is, the mix group was not driven by doctors who used one
`drug on nearly all of their patients and the other on a small fraction. In
`Figure 1, we illustrate the number of patients treated with Lucentis as a
`fraction of the total number of patients treated with either Lucentis or Avastin
`for each of the high-volume doctors. Specifically, each data point represents
`one of 547 high-volume doctors, and the graph measures each doctor’s
`Lucentis “patient share.”
`As
`shown in Figure 1, many high-volume doctors administered a
`substantial amount of both products. For instance, the shaded region shows
`that many of these doctors did not concentrate >80% of their patients on one
`drug but rather used both drugs in significant shares. This shaded region
`amounts to >40% of the high-volume ophthalmologists. We obtain similar
`results when analyzing the proportion of
`treatments as opposed to the
`proportion of patients.
`This analysis thus casts some doubt on the hypothesis that many high-
`volume doctors are administering Lucentis primarily because they are
`motivated by profits, and the analysis lends some support to the hypothesis
`that doctors are exercising medical judgment in their choice of treatment or
`other factors are influencing product choice.
`
`Figure 1.
`
`Breakdown of Lucentis patient share for high-volume
`ophthalmologists.
`
`Downloaded by [Ms Julia Livingston] at 11:36 06 July 2016
`
`Samsung et al. v. Regeneron IPR2023-00884
`Regeneron Pharmaceuticals, Inc. Exhibit 2158 Page 7
`
`

`

`Similar Products at Different Prices
`
`237
`
`Interaction with Other Treatments
`
`Although much attention has been paid to the interplay between Lucentis and
`Avastin, several other drugs are also used to treat wet AMD. These include
`Visudyne, Macugen, Eylea, and Triamcinolone (which is not approved to treat
`wet AMD but
`is sometimes used off
`label
`to do so). Lucentis, Avastin,
`Macugen, and Eylea are administered as stand-alone injections. Visudyne is
`administered in conjunction with photodynamic therapy. Triamcinolone can be
`administered as a stand-alone treatment or in combination with one of the
`aforementioned drugs (see Gopal and Sharma 2007; Mayo Clinic 2014).
`Accounting for these alternatives in the analyses described above offers several
`additional interesting insights.
`First, after accounting for these drugs, very few doctors remain in the
`Lucentis-only category. In particular, as shown in Figure 2, only five (or <1%)
`of the 547 high-volume doctors prescribe only Lucentis. This suggests that very
`few physicians prescribe in a manner consistent with a solely profit-motivated
`orientation and that any efforts by the manufacturer to encourage exclusive
`use of Lucentis, as the first product approved for the treatment of
`this
`important disease, are relatively ineffective. These data suggest a very
`competitive set of
`interactions between alternative therapies. Note that
`ophthalmologists may use some of these drugs for conditions other than wet
`AMD, but the available data do not allow us to distinguish particular uses.
`Figure 3 provides a more granular look at the prescribing patterns of these
`doctors. Figure 3 expands Figure 1 to include Eylea and Triamcinolone as well
`
`Lucentis only
`1%
`
`Lucentis and other
`non-Avastin
`18%
`
`Avastin only
`3%
`
`Avastin and other
`non-Lucentis
`8%
`
`Avastin and
`Lucentis only
`5%
`
`Lucentis, Avastin,
`and other
`65%
`
`Note: “Other” includes Visudyne, Macugen, Eylea, and Triamcinolone.
`
`Figure 2.
`Breakdown of wet age-related macular degeneration (AMD)
`treatments for ophthalmologists with 200 or more patients treated with
`Lucentis and/or Avastin.
`
`Downloaded by [Ms Julia Livingston] at 11:36 06 July 2016
`
`Samsung et al. v. Regeneron IPR2023-00884
`Regeneron Pharmaceuticals, Inc. Exhibit 2158 Page 8
`
`

`

`238 R. Manning et al.
`
`100%
`
`LUCENTIS
`AVASTIN
`TRIAMCINOLONE
`EYLEA
`LUCENTIS vs AVASTIN patient share
`
`90%
`
`80%
`
`70%
`
`60%
`
`50%
`
`40%
`
`30%
`
`20%
`
`10%
`
`0%
`
`Patient share
`
`547 opthalmologists with more than 200 patients treated with Lucentis and/or Avastin
`
`Note: For illustrative purposes, this chart does not show Visudyne and Macugen, which account
`for a very small patient share.
`
`Figure 3.
`
`Share of wet AMD drug treatments for high-volume
`ophthalmologists.
`
`as Lucentis and Avastin. Specifically, each data point represents one of 547
`high-volume doctors, and the graph measures each doctor’s “patient share” for
`each of the drugs.
`Figure 3 illustrates that many of the doctors who were characterized as
`Avastin-only in Figure 1 actually administered other wet AMD drugs to a
`substantial number of their patients. Most notably, the chart shows that a
`sizeable share of these doctors’ patients received Triamcinolone. As mentioned
`above, Triamcinolone is used off label to treat wet AMD and is also relatively
`inexpensive, costing less than $20 for a 2–4 mg dose.
`Figure 3 also reveals that the doctors who were characterized as Lucentis-
`only in Figure 1 administered other wet AMD drugs to a substantial number
`their patients.
`In particular,
`the graph shows
`that
`these doctors
`of
`administered Eylea to a sizeable share of their patients, though clearly to fewer
`than they administered Lucentis. Eylea was approved in November 2011 and
`has been shown to help when patients do not respond to Lucentis or Avastin
`(see Bakall et al. 2013; US Food and Drug Administration 2011b). The drug is
`relatively expensive (approximately $1,850 per dose), but it only needs to be
`injected half as frequently as Lucentis and Avastin (see American Macular
`Degeneration Foundation 2014).
`these observations have various potential explanations.
`Taken together,
`First, one could interpret them as suggesting that certain doctors are more
`
`Downloaded by [Ms Julia Livingston] at 11:36 06 July 2016
`
`Samsung et al. v. Regeneron IPR2023-00884
`Regeneron Pharmaceuticals, Inc. Exhibit 2158 Page 9
`
`

`

`Similar Products at Different Prices
`
`239
`
`treatments, and hence doctors who tend to use
`inclined to use off-label
`Avastin over Lucentis also tend to use Triamcinolone relatively frequently.
`Second, because Avastin and Triamcinolone are relatively inexpensive, these
`observations could suggest that certain doctors choose less expensive products
`for their patients while others choose more expensive products, such as
`Lucentis and Eylea. This explanation, however, could also be consistent with
`various motivations. For instance, it could be the case that the doctors who use
`Lucentis and Eylea frequently have more complex patients or believe that
`these on-label products are better for patients than the off-label alternatives
`(Avastin and Triamcinolone). Although the data do not allow a fine distinction
`among the competing hypotheses, two key observations arise:
`
`(cid:129) Very few doctors (<1% as shown in Figure 2) administered only what
`to them.
`If financial
`was likely to be the most profitable product
`incentives dominated among this group of physicians, one would expect
`a much larger share of them to have prescribed the most expensive
`product.
`(cid:129) Doctors who more often administered Lucentis also provided certain
`services that were used to diagnose and track the progression of wet
`AMD and other eye conditions more frequently. For example, in our
`sample,
`the number of claims a doctor submitted for each of
`the
`following services is positively correlated with the number of Lucentis
`claims the provider submitted: optical coherence tomography (CPT code
`92134), fluorescein angiography (CPT code 92235), ICG angiography
`(CPT code 92240), and fundus photography (CPT code 92250). Optical
`coherence tomography (CPT code 92134) uses special rays of light to
`scan and produce an image of the retina that shows whether the macula
`is thickened or abnormal and whether fluid has leaked into the retina.
`Fluorescein angiography (CPT code 92235) involves injecting fluorescein
`dye into a vein and then taking pictures by using a special camera to
`see if any of the dye has leaked from the blood vessels behind the
`macula. ICG angiography (CPT code 92240) is similar to fluorescein
`angiography but uses a different dye that highlights different problems
`in the eye. Fundus photography (CPT code 92250) uses a special camera
`to take three-dimensional color images of the macula in order to see
`damage to different layers of the retina. In 2012, the Medicare physician
`fee schedule rates for CPT codes 92134, 92235, 92240, and 92250 were
`$44.93, $137.17, $248.47, and $76.58,
`respectively (National Health
`Service 2014). The relevant correlation coefficients are 0.42, 0.26, 0.37,
`and 0.20, respectively, all of which are statistically significant at the 1%
`level. This suggests that doctors who administered Lucentis had a more
`complex patient population with more severe conditions. Alternatively,
`it could indicate that these doctors provided more in-depth patient care.
`
`One other observation regarding these doctors merits noting. Those who
`prescribed Lucentis more often relative to Avastin tended to be located in
`counties with higher per-capita income than those who prescribed Avastin
`more often. (Per capita income was taken from the 2012 American Community
`Survey administered by the US Census Bureau.) While this relationship is not
`particularly strong,
`the correlation coefficient between a doctor’s Lucentis
`patient share and per-capita income in the doctor’s county is 0.24, and is
`
`Downloaded by [Ms Julia Livingston] at 11:36 06 July 2016
`
`Samsung et al. v. Regeneron IPR2023-00884
`Regeneron Pharmaceuticals, Inc. Exhibit 2158 Page 10
`
`

`

`240 R. Manning et al.
`
`70,000
`
`60,000
`
`50,000
`
`40,000
`
`30,000
`
`20,000
`
`10,000
`
`Per capita income in provider's county
`
`0
`0%
`
`10%
`
`20%
`
`30%
`
`60%
`50%
`40%
`Lucentis patient share
`
`70%
`
`80%
`
`90%
`
`100%
`
`Figure 4. Relationship between Lucentis patient share and per capita income
`in provider’s county for high-volume ophthalmologists.
`
`statistically significant at the 1% level. The relationship between income and
`Lucentis share is illustrated in Figure 4.
`Obviously, this correlation is modest, and variation around the trend is
`substantial. Of course,
`it would be far preferable to estimate the effect of
`patient income on therapy choice in the context of a regression framework that
`corrected for other individual and prescriber characteristics, but the data on
`individual patient and physician characteristics necessary for
`such an
`examination are not readily available. Nevertheless, physicians practicing in
`higher-income areas are likely to have patient populations that would be
`somewhat less sensitive to differences in out-of-pocket cost or who may view
`any therapeutic differences between the products as more important when the
`outcome involves retaining or losing vision. It is important to note that the
`majority of Medicare Part B beneficiaries have supplemental insurance that
`covers some or all of their coinsurance, so in many cases the patients are not
`likely to foot much if any of the bill (see Jacobson, Huang, and Neuman 2014).
`
`Conclusion
`
`In summary, the newly released data on Medicare Part B payments suggest
`that many doctors chose Avastin for some patients and Lucentis for others,
`and the data show most clearly that only a very small share of doctors
`exclusively chose Lucentis. Accounting for a broader set of drugs that can be
`used to treat wet AMD – Lucentis and Avastin, as well as Visudyne, Macugen,
`
`Downloaded by [Ms Julia Livingston] at 11:36 06 July 2016
`
`Samsung et al. v. Regeneron IPR2023-00884
`Regeneron Pharmaceuticals, Inc. Exhibit 2158 Page 11
`
`

`

`Similar Products at Different Prices
`
`241
`
`the vast majority of doctors
`that
`Eylea, and Triamcinolone – reveals
`administered multiple drugs to treat wet AMD. This is inconsistent with the
`hypothesis that the major factor driving product choice between these two
`products is a simple financial motivation on the part of the physician or the
`manufacturer.
`Rather, it is consistent with the notion that physicians typically exercise
`medical judgment or are responsive to other factors on a patient-by-patient
`basis. Moreover, and perhaps most interestingly from the perspective of the
`industrial organization of
`the biopharmaceutical
`industry, and for public
`policy discussions,
`it
`seems quite clear
`that
`there are limitations
`to
`manufacturers’ ability to segment markets.
`Furthermore, accounting for a broader set of drugs that can be used to treat
`wet AMD – Lucentis and Avastin, as well as Visudyne, Macugen, Eylea, and
`Triamcinolone – reveals that the vast majority of doctors administered multiple
`drugs to treat wet AMD. The physicians who were most likely to use Lucentis
`also tended to be those who most likely used the newer biologic therapy,
`Eylea, which is primarily indicated for patients who do not have satisfactory
`results on initial therapy. This may suggest that patients treated by such
`doctors have more complicated medical conditions or are for some unobserved
`reason more likely to respond better to products specifically indicated for the
`condition. Finally, physicians who administered Lucentis more often also
`performed more eye-imaging procedures to diagnose and track the progression
`of wet AMD and other eye conditions. This suggests that such patients may
`indeed be more medically complex and that they may be patients for whom
`the FDA-approved therapy may be deemed to be more appropriate by the
`physician, the patient, or both. All these patterns lend support to the notion
`that
`therapy choice is driven by patient-spec