Vanda Pharmaceuticals, Inc. v. Centers for Medicare & Medicaid Services

U.S. Court of Appeals for the Fourth Circuit
Vanda Pharmaceuticals, Inc. v. Centers for Medicare & Medicaid Services, 98 F.4th 483 (4th Cir. 2024)

Vanda Pharmaceuticals, Inc. v. Centers for Medicare & Medicaid Services

Opinion

USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 1 of 29

PUBLISHED

UNITED STATES COURT OF APPEALS FOR THE FOURTH CIRCUIT

No. 23-1457

VANDA PHARMACEUTICALS, INC.

Plaintiff - Appellant

v.

CENTERS FOR MEDICARE & MEDICAID SERVICES; CHIQUITA BROOKS- LASURE, in her official capacity as Administrator of Centers for Medicare & Medicaid Services

Defendants - Appellees.

Appeal from the United States District Court for the District of Maryland at Baltimore. Matthew James Maddox, Magistrate Judge. (1:22−cv−00977−MJM)

Argued: January 24, 2024 Decided: April 10, 2024

Before DIAZ, Chief Judge, WILKINSON, Circuit Judge, and MOTZ, Senior Circuit Judge.

Affirmed by published opinion. Judge Wilkinson wrote the opinion, in which Chief Judge Diaz and Senior Circuit Judge Motz joined.

ARGUED: Paul Whitfield Hughes, III, MCDERMOTT, WILL & EMERY, LLP, Washington, D.C., for Appellant. David L. Peters, UNITED STATES DEPARTMENT OF JUSTICE, Washington, D.C., for Appellees. ON BRIEF: Andrew A. Lyons-Berg, Alex C. Boota, MCDERMOTT WILL & EMERY LLP, Washington, D.C., for Appellant. Brian M. Boynton, Principal Deputy Assistant Attorney General, Alisa B. Klein, Civil Division, UNITED STATES DEPARTMENT OF JUSTICE, Washington, D.C.; Samuel USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 2 of 29

R. Bagenstos, General Counsel, Janice L. Hoffman, Associate General Counsel, Susan Maxson Lyons, Deputy Associate General Counsel for Litigation, Kara Wilcox Mundy, UNITED STATES DEPARTMENT OF HEALTH AND HUMAN SERVICES, Washington, D.C.; Erek Barron, United States Attorney, OFFICE OF THE UNITED STATES ATTORNEY, Baltimore, Maryland, for Appellees.

2 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 3 of 29

WILKINSON, Circuit Judge:

Under the Medicaid Drug Rebate Program, a drug manufacturer that increases its

prices faster than inflation rises must reimburse Medicaid for the difference. These

reimbursements are paid via rebates. Each drug’s rebate amount is usually determined

based on its own original price and inflation clock. But not always. Congress has instructed

in the Medicaid statute that some “line extension” drugs can be on the hook not only for

their own price increases, but also for the price increases of the drugs they evolved from.

In 2020, the Centers for Medicare and Medicaid Services (CMS) promulgated a

regulation that set forth criteria for what constitutes a line-extension drug. Vanda

Pharmaceuticals challenged that regulation in federal district court, arguing that it

expanded the definition of a line extension beyond what the Medicaid statute permitted.

The district court disagreed and granted summary judgment to CMS. Because the agency’s

regulation lies within the bounds of the Medicaid statute, we affirm.

I.

A.

The Medicaid Drug Rebate Program was not built in a day. It was created in 1990,

but Congress has regularly revisited the rebate regime in the decades since its enactment.

The line-extension provision at issue here was added in one such revision. It is best

explained in the context of what came before, and so we begin with a discussion of the

program as originally enacted.

Congress created the Medicaid Drug Rebate Program as part of an effort to reduce

Medicaid spending. See Omnibus Budget Reconciliation Act of 1990,

Pub. L. No. 101-508,

3 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 4 of 29

tit. IV, subtitle B, pt. 1,

104 Stat. 1388

, 1388-141. Congress was concerned that the

government was being swindled on drug prices. Medicaid, “the means-tested entitlement

program that purchases basic health care for the poor,” had been paying significantly more

for drugs than other large purchasers. See H.R. Rep. No. 101-881, at 96 (1990). To make

matters worse, drug prices had been rising rapidly. Between 1981 and 1988 drug price

increases had “more than tripled the rate of inflation.” See Majority Staff of S. Special

Comm. on Aging, 101st Cong., Prescription Drug Prices: Are We Getting Our Money’s

Worth? 8 (Comm. Print 1989). Congress devised the new rebate program in response to

both these problems.

Under the program, drug manufacturers must pay rebates to Medicaid that offset

part of the cost of the manufacturers’ drugs, thus lowering the ultimate price Medicaid

pays. The rebate amounts are calculated by CMS, but the Medicaid statute meticulously

lays out the formula that CMS must use. It has two components. The first (which is not at

issue here) is the “basic rebate,” which ensures that Medicaid pays the lowest price for any

given drug. See 42 U.S.C. § 1396r-8(c)(1).

This case concerns the “additional rebate,” which forces drug manufacturers to

reimburse Medicaid for price increases greater than the rate of inflation. § 1396r-8(c)(2).

It is calculated by taking the difference between the current average price of the drug and

the price of the drug when it was first marketed, adjusted for inflation. See

§ 1396r-8(c)(2)(A). In other words, once a drug manufacturer sets an initial price for a

drug, Medicaid will not pay more than that price (plus inflation). Medicaid thus locks in

that original price.

4 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 5 of 29

But the original additional rebate formula had a loophole. The statute calls for a

separate rebate calculation for “each dosage form and strength” of a covered drug. Id. A

manufacturer could thus raise the price of a drug without paying the additional rebate by

releasing a new strength (or form) of the drug at a higher price. The additional rebate for

the new strength would be based on the new strength’s higher release price rather than the

original one. The manufacturer could then discontinue the original strength, forcing

purchasers to switch. As a result, the manufacturer could avoid reimbursing Medicaid for

the jump in price. See H.R. Rep. No. 111-299, pt. 1, at 635 (2009); S. Rep. No. 111-89, at

92 (2009).

Congress added the line-extension provision in 2010 in part to close this loophole.

It provides that so-called “line extension” drugs are on the hook not only for their own

price increases, but also for any price increases to the original drug on which they were

based. Here is how the provision works: For a line-extension drug, the additional rebate is

the greater of two amounts. Amount One depends on the price increases of the line-

extension drug and is calculated the same way as above—that is, by taking the difference

between the current average price of the line-extension drug and the price of that drug when

first marketed, adjusted for inflation. § 1396r-8(c)(2)(C)(ii). So far, so good. A line-

extension drug is treated just like any other drug.

Amount Two, however, depends on whether the price of the original drug has

increased. It is calculated by multiplying the highest additional rebate percentage owed on

any strength of the original drug by the price of the line-extension drug. See

§ 1396r-8(c)(2)(C)(iii). In other words, Medicaid will treat the line-extension drug as if its

5 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 6 of 29

own price had increased proportionally with the original drug’s. For example, if one of the

strengths of the original drug (taking inflation into account) had doubled in price since its

release, the line-extension drug would also be treated as if it had doubled in price and would

thus be subject to at least a 50% rebate, regardless of whether the price of the line-extension

drug had changed at all. Importantly, because the statute instructs that the additional rebate

for a line extension is the greater of Amount One and Amount Two, line-extension status

can only lead to higher rebates.

B.

But what is a line extension? Answering that question involves consulting two

sources: the Medicaid statute and the corresponding CMS regulation challenged in this

case.

1.

First, the guidance given by Congress: According to the Medicaid statute,

[T]he term “line extension” means, with respect to a drug, a new formulation of the drug, such as an extended release formulation, but does not include an abuse-deterrent formulation of the drug (as determined by the Secretary), regardless of whether such abuse-deterrent formulation is an extended release formulation.

§ 1396r-8(c)(2)(C)(iii).

To break that down, the statute instructs that a “line extension” is a “new

formulation” of an existing drug. There is also a carveout: abuse-deterrent formulations do

not count. (The carveout was added in 2016 as part of Congress’s response to the opioid

crisis. See Comprehensive Addiction and Recovery Act of 2016,

Pub. L. No. 114-198, § 705

(a),

130 Stat. 695

, 753.) As is clear from this statutory language, however, the

6 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 7 of 29

legislative definition of line extension is incomplete: Congress inserted the language “such

as” before providing the example of “an extended release formulation,” indicating a gap

left to be filled. As is common in administrative regulation, the agency charged with

carrying out the program (here, CMS) was left to fill in the blanks.

The Medicaid statute provides one more wrinkle important to the case at hand. The

line-extension provision (and the higher rebates it brings) does not apply to all line-

extension drugs. It applies only to “a line extension of a single source drug or an innovator

multiple source drug that is an oral solid dosage form.” § 1396r-8(c)(2)(C)(i). In simpler

terms: the provision applies only to line extensions of two specific types of drugs (defined

later in the statute), and only to a “line extension of [either of those types of drugs] that is

an oral solid dosage form.” Id. (emphasis added). This oral-solid-dosage-form requirement

presented a puzzle: which drug(s) must be in oral solid dosage form? The original drug?

The line-extension drug? Both? Here, too, it fell to the implementing agency to resolve the

ambiguity.

2.

CMS took up those twin calls in June 2020. The agency had received “requests to

provide more specific guidance on how to identify a line extension drug.”

85 Fed. Reg. 37,286

, 37,294 (June 19, 2020). The agency also had reasons to think that a more explicit

definition of line extension was necessary. “After several years of experience with

manufacturers self-reporting their line extensions,” it had “noted inconsistency among

manufacturers.”

Id.

The agency suspected that line extensions were being underreported

because “manufacturers may have a financial incentive to be underinclusive in their

7 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 8 of 29

identification.”

Id.

It therefore proposed a regulation that would “interpret the term ‘line

extension’ broadly.”

Id. at 37,295

. After notice and comment, the agency finalized its

proposal (with modest changes) in December 2020.

The final regulation started with a definition for line extension that closely tracks

the statute:

Line extension means, for a drug, a new formulation of the drug, but does not include an abuse-deterrent formulation of the drug (as determined by the Secretary).

85 Fed. Reg. 87,000

, 87,101 (Dec. 31, 2020) (codified at

42 C.F.R. § 447.502

). It also

defined the term “new formulation” and expanded on the example from the statute:

New formulation means, for a drug, a change to the drug, including, but not limited to: an extended release formulation or other change in release mechanism, a change in dosage form, strength, route of administration, or ingredients.

Id. at 87

,101–02 (codified at

42 C.F.R. § 447.502

).

The agency’s regulation also clarified its understanding of the oral-solid-dosage-

form requirement: “only the initial brand name listed drug must be an oral solid dosage

form.”

Id. at 87,034

. In other words, only the original drug needs to be in an oral solid

dosage (read: pill) form for the provision to apply. The line-extension drug can be in a

different form—liquid, say, or injectable—and still qualify for the new rebate formula.

On this point, however, the agency has wavered. In 2012, it had proposed that both

the line-extension drug and the original drug must be in oral solid dosage form. See

77 Fed. Reg. 5318

, 5338 (Feb. 2, 2012). But it declined to finalize that proposal. See

81 Fed. Reg. 5170

, 5265 (Feb. 1, 2016). In its 2020 rulemaking, CMS changed course, proposing in June

8 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 9 of 29

and finalizing in December after notice and comment that the requirement applied to only

the original drug.

C.

Appellant Vanda was not happy with the agency’s new moves. Vanda develops and

manufactures innovative drugs that treat rare disorders. Its two flagship drugs are Hetlioz

and Fanapt, each of which has a sister drug that Vanda worries will be swept up in the

agency’s new line-extension definition.

Hetlioz is a drug approved to treat (among other conditions) severe sleep

disturbances caused by the developmental disorder Smith-Magenis syndrome. This

disorder affects both adults and children, but Hetlioz is approved only for adults because it

comes in oral solid dosage form (a pill capsule), and the dosage cannot be calibrated to

children’s weights. So Vanda created a liquid version called Hetlioz LQ for use in treating

sleep disturbances in children with Smith-Magenis syndrome. Hetlioz LQ was approved

by the Food and Drug Administration (FDA) in 2020 and is available today for patients 3

to 15 years of age.

Vanda also manufactures Fanapt, an antipsychotic used to treat schizophrenia.

Fanapt is currently offered in pill form and must be taken twice a day. Like many

antipsychotics, it is safest and most effective if taken at regular intervals without missing a

dose. But people with schizophrenia often struggle to adhere to such treatment regimens.

To solve this problem, Vanda has been working to create a long-acting injectable version

of Fanapt called Fanapt LAI. Fanapt LAI is still in development, but Vanda hopes to offer

it in a format that would require patients to receive only a few injections each year.

9 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 10 of 29

According to Vanda, this change could “revolutionize schizophrenia treatment.”

Appellant’s Opening Br. 16.

Vanda alleges that, under the agency’s pre-regulation approach, neither Hetlioz LQ

nor Fanapt LAI would have been considered line extensions. Under the new regulation,

however, both will be.

Vanda has a strong interest in avoiding line-extension status for its drugs. Many

patients who use Vanda’s drugs are enrolled in Medicaid. Fanapt users are especially likely

to be Medicaid recipients, as 70% of people with schizophrenia are covered under the

program. See Kimberly H. Geissler et al., Differences in Insurance Coverage for

Individuals with Schizophrenia After Implementation of the Patient Protection and

Affordable Care Act, 80(3) JAMA Psychiatry 278, 278–79 (2023); see also Appellant’s

Opening Br. 17. Medicaid sales are thus an important source of revenue for Vanda, and

Medicaid rebates cut into profits. Line-extension status can trigger even higher rebates,

which Vanda would prefer to avoid. To that end, Vanda seeks to cabin the definition of

line extension so as to exclude drugs like Hetlioz LQ and Fanapt LAI.

D.

Vanda sued in U.S. District Court to challenge the agency’s 2020 regulation. It

argued that the regulation’s definitions of “line extension” and “new formulation” were

contrary to the text of the Medicaid statute. It also challenged the agency’s new

interpretation of the oral-solid-dosage-form requirement. Finally, Vanda alleged that the

agency had violated the Administrative Procedure Act (APA) because the agency failed to

10 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 11 of 29

adequately address important industry concerns aired during the notice-and-comment

period.

On cross-motions for summary judgment, the district court rejected Vanda’s claims.

It held that the regulation was consistent with the text of the Medicaid statute, and that

CMS had satisfied its obligations under the APA. See Vanda Pharms., Inc. v. Ctrs. for

Medicare & Medicaid Servs.,

2023 WL 2743364

, at *16, 18, 23–24 (D. Md. Mar. 31,

2023). Vanda timely appealed.

II.

Under the APA, we set aside an agency’s regulation only if it is “arbitrary,

capricious, an abuse of discretion, or otherwise not in accordance with law.”

5 U.S.C. § 706

(2)(A). We review the district court’s evaluation de novo, independently assessing

whether the agency action was unlawful. See Ren v. U.S. Citizenship & Immigr. Servs.,

60 F.4th 89

, 93 (4th Cir. 2023).

III.

Vanda first argues that the 2020 regulation is “not in accordance with law,”

§ 706(2)(A), because it broadened the sweep of the line-extension provision beyond what

the Medicaid statute permits. We are thus called upon to evaluate whether the 2020

regulation contravenes the Medicaid statute’s text. That task, however, presents a

conundrum. Courts have traditionally reviewed an agency’s statutory interpretations in

accordance with the two-step framework of Chevron U.S.A., Inc. v. Natural Resources

Defense Council, Inc.,

467 U.S. 837

(1984). See, e.g., Cela v. Garland,

75 F.4th 355

, 360–

61 (4th Cir. 2023). Under that framework, we would first “consider[] ‘whether Congress

11 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 12 of 29

has directly spoken to the precise question at issue.’”

Id.

at 361 (quoting Chevron,

467 U.S. at 842

). If so, our inquiry would stop there. If the statute was ambiguous, however, we

would defer to the agency’s interpretation so long as it was reasonable.

Id.

But CMS did not invoke Chevron in its briefing or at oral argument. Recent

Supreme Court precedent has made clear that if the “government is not invoking Chevron”

a court can “decline to consider whether any deference might be due.” HollyFrontier

Cheyenne Ref., LLC v. Renewable Fuels Ass’n,

141 S. Ct. 2172, 2180

(2021); see also

Guedes v. Bureau of Alcohol, Tobacco, Firearms, & Explosives,

140 S. Ct. 789

, 789–90

(2020) (Gorsuch, J., statement respecting the denial of certiorari) (explaining that the

Supreme Court “has often declined to apply Chevron deference when the government fails

to invoke it” and collecting cases).

We, of course, follow the Supreme Court’s lead here. Thus, if we encounter

statutory ambiguity, rather than deferring to the agency’s interpretation, we will “approach

[that] interpretive problem[] methodically, using traditional tools of statutory

interpretation.” Facebook, Inc. v. Duguid,

592 U.S. 395

, 404 n.5 (2021). In doing so,

however, we will “pay particular attention to [the] agency’s views in light of the agency’s

expertise in [the] given area” and “its knowledge gained through practical experience.”

County of Maui v. Haw. Wildlife Fund,

140 S. Ct. 1462, 1474

(2020). But we will adopt

those views as our own only if they have the “power to persuade.” Skidmore v. Swift &

Co.,

323 U.S. 134, 140

(1944).

With those principles in mind, we turn to Vanda’s claims. It challenges two parts of

the agency’s 2020 regulation: the definitions of line extension and new formulation, and

12 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 13 of 29

the agency’s understanding of the oral-solid-dosage-form requirement. We take each in

turn.

A.

We begin with Vanda’s claim that the 2020 regulation is contrary to law in that it

gives impermissible definitions to the terms “line extension” and “new formulation.”

Generally, Vanda argues that the agency’s broad definitions contravene the purportedly

narrow focus of the line-extension provision in the Medicaid statute. Vanda then offers up

two specific limitations that it reads into the statutory regime. We take up the general point

first.

1.

We disagree with Vanda’s contention that the agency’s definitions stray outside the

general statutory ambit. Start with the definition of line extension. We are struck at the

outset by the similarities in the statutory and regulatory definitions of this term. In contrast

to Vanda’s claims, the substance of the regulatory definition hews quite closely to the

statute. Both define “line extension” as a “new formulation” and then carve out an

exception for “abuse-deterrent formulation[s].” Compare § 1396r-8(c)(2)(C)(iii), with

42 C.F.R. § 447.502

.

The similarities are so striking as to make this part of the regulation unimpeachable.

Each departure of the regulatory text is patently superficial, with no discernable effect on

the term’s reach. The agency’s definition of line extension is clearly within the bounds of

the statute and we see no reason it should be set aside.

13 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 14 of 29

We turn next to the definition of new formulation. Vanda claims that here the

regulation strays further afield, but it is plainly not so far as to call “foul ball.” Recall that

the statute explains the term by reference to a single example: “an extended release

formulation.” § 1396r-8(c)(2)(C)(iii). That example, however, is introduced with the

inclusive phrase “such as,” making clear that Congress intended the term to sweep more

broadly than the illustrative example it chose.

The regulation thus starts with a broader definition: a “new formulation” is “a

change to the drug.”

42 C.F.R. § 447.502

. It then provides an expanded list of changes that

can count. The list starts with the statutory example of “an extended release formulation.”

But it adds “other change[s] in release mechanism” and changes in “dosage form, strength,

route of administration, or ingredients” as well.

Id.

And like the statute, the regulation

makes clear (here with the phrase “including, but not limited to”) that those examples are

meant to be illustrative, not exhaustive.

Id.

This strikes us as a perfectly sensible way to implement the regime set by the

Medicaid statute. Start with the term Congress chose: “a new formulation.” Something is

“new” if it is “different from [what] previously exist[ed].” New, Oxford English Dictionary

(2d ed. 1989). And a “formulation” is “a material or mixture prepared according to a

particular formula.” Formulation, Oxford English Dictionary (2d ed. 1989) (emphasis

added). Putting it all together, a new formulation is created whenever a drug is prepared

according to a formula different from the formula or formulas previously in use. A different

formula creates a new formulation.

14 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 15 of 29

The agency’s choice to define a “new formulation” as a “change” aligns quite

closely with the words selected by Congress. With respect to drugs, change is achieved

precisely by using a different formula. Drug manufacturers often create several variations

of a drug. A drug may come as a pill, a rapid release gel, a chewable, a dissolve pack, or in

a liquid suspension measured out with a little plastic cup. As each successive format is

developed, the drug must be changed by using a different formula to achieve each final

product. It makes perfect sense, then, to say that each new iteration achieved by such a

change might permissibly be termed a “new formulation.”

Vanda, though, takes aim at the breadth of the definition. It says the agency’s

definition sweeps in too much, pointing for support to the Congressional Budget Office

report in which a line-extension provision was first suggested. See CBO, Budget Options,

Volume I: Health Care 143 (Dec. 2008) (the “CBO Report”). The report stated that drug

manufacturers could “avoid incurring an additional rebate obligation by making a slight

alteration to an existing product.”

Id.

It recommended that Congress close this “loophole”

by treating “a certain type of new formulation—specifically, extended-release versions”

more like “the original product.”

Id.

This, Vanda says, was the problem Congress intended

to solve with the line-extension provision as well: drug manufacturers avoiding the

additional rebate by making merely incremental changes to their original drugs to reset the

inflation clock. And so, according to Vanda, Congress must have intended the term “new

formulation” only to encompass such slight alterations. Vanda insists that larger changes

in formula create altogether new drugs—not simply line extensions to original drugs—and

that therefore such changes should not fall into the provision’s reach.

15 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 16 of 29

We do not think that this is a fair reading of the statute or a fair condemnation of the

agency’s regulation. As an initial matter, “legislative history is not the law.” Azar v. Allina

Health Servs.,

139 S. Ct. 1804, 1814

(2019). Courts must “apply faithfully the law

Congress has written” and “cannot replace the actual text with speculation as to Congress’

intent.” Luna Perez v. Sturgis Pub. Sch.,

598 U.S. 142

, 150 (2023). And the law that

Congress passed differs in important respects from the Budget Office proposal cited by

Vanda. That proposal would have gone after extended-release formulations only, whereas

the statute makes clear that extended-release formulations are but one example of what will

be swept in. And the proposal targets “a certain type of new formulation.” CBO Report, at

143. In contrast, the statute speaks in terms of new formulations writ large.

The 2016 amendment that carved out abuse-deterrent formulations confirms that the

statutory definition is broader than Vanda would have it be. Abuse-deterrent formulations

are not slight, meaningless alterations. They require new innovations and costly research

and development. They also provide important therapeutic benefits and help prevent the

harmful consequences attendant to prescription drug misuse. The fact that they had to be

carved out from the provision shows that the statutory definition encompasses more than

just the meaningless changes Vanda advocates.

And Vanda’s argument overlooks the fact that Congress chose to define line

extension as a “new formulation.” Vanda says that an alteration that is more than slight

creates, not a new formulation, but an entirely new drug. But the term “new formulation”

does not by its plain meaning limit line extensions in the way Vanda might hope. We use

the term “new” to describe differences both big and small, and both types of changes can

16 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 17 of 29

create an updated version of an original without creating an entirely novel item. Take as an

example the new editions of well-loved books. Sometimes the differences are atmospheric

or slight. Oxford World’s Classics’ new edition of Willa Cather’s My Ántonia added a new

introduction and notes that provide historical context, but Cather’s evocative descriptions

of pioneer life on the plains of Nebraska remain intact. See Willa Cather, My Ántonia

(Oxford World’s Classics ed. 2009). In other words, despite peripheral changes, the

substance of the book remains the same. But sometimes the differences in new versions of

something can be quite large indeed. New editions of classic legal treatises can differ

dramatically from the old, especially when there have been major intervening changes in

doctrine. But that does not mean that the fourth edition of Williston on Contracts cannot

be said to be a “new edition” with respect to the third. Even significant changes, therefore,

can still flow from the original so long as there is a throughline between them. Congress’s

choice of the definition “new formulation” simply fails to support Vanda’s contention that

anything more than a “slight alteration” should count as a new drug entirely.

Moreover, the structure of the statutory line-extension provision confirms that a

broad definition of new formulation is appropriate. The statutory provision speaks in terms

of inclusion, not exclusion. It makes clear that a line extension is “a new formulation of the

drug, such as an extended release formulation.” § 1396r-8(c)(2)(C)(iii) (emphasis added).

Terms of inclusion like “such as” are congressional invitations for agencies to apply their

expertise to fill out the list with further examples. “Congress knows to speak in plain terms

when it wishes to circumscribe, and in capacious terms when it wishes to enlarge, agency

17 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 18 of 29

discretion.” City of Arlington v. FCC,

569 U.S. 290, 296

(2013). Here, Congress chose the

broader route and reinforced it with a structure evincing inclusive intent.

The agency heeded the congressional invitation. When proposing the regulation at

issue here, the agency acknowledged that it had chosen a “much broader definition of new

formulation” than in previous (never-finalized) proposals. See 85 Fed. Reg. at 37,295. It

also made clear that its broadened definition was designed to expand the universe of drugs

that qualified as line extensions. Id. The agency “believe[d] that the statute g[ave] [it]

discretion and authority to interpret the term ‘line extension’ broadly” and to “include a

broad range of drugs” therein. Id. In response to “requests to provide more specific

guidance,” id. at 37,294, the agency informed manufacturers that—in addition to extended-

release formulations—“other change[s] in release mechanism” and changes in “dosage

form, strength, route of administration, or ingredients” would count as line extensions too.

42 C.F.R. § 447.502

.

We find the agency’s additional examples here “reasonable” and “consistent with

the statutory framework.” Fed. Exp. Corp. v. Holowecki,

552 U.S. 389, 402

(2008). In fact,

the agency’s added examples strike us as fairly similar to extended-release formulations.

And we can think of no “clearer alternatives” that are “within our authority or expertise to

adopt.”

Id.

Moreover, Vanda has not identified any individual element of the list that poses

a particular problem. In short, then, we do not find merit in Vanda’s general challenge to

the breadth of the agency’s “new formulation” definition.

That is not to say, however, that the agency is subject to no limitations at all. Line-

extension status is relational. A drug product cannot be a line extension in the abstract; it

18 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 19 of 29

is a line extension of a particular original drug. See § 1396r-8(c)(2)(C)(iii) (“[T]he term

“line extension” means, with respect to a drug, a new formulation of the drug.” (emphasis

added)). The two must bear some relationship to each other. And not all relationships will

suffice. Tylenol (active ingredient: acetaminophen) is not a line extension of Advil (active

ingredient: ibuprofen), even though both are useful in treating aches and pains.

Luckily, we need not confront the problem of when changes to a drug are significant

enough to take it from a new formulation to a distinct drug entirely. The agency’s chosen

examples of changes—different release mechanisms, routes of administration, dosages,

ingredients, and strengths—are limited enough (and similar enough to extended-release

formulations) to fit well within the statutory bounds.

2.

In addition to taking aim at the breadth of the provision generally, Vanda proposes

two specific textual limitations on the term “line extension.” We find neither one

persuasive.

First, Vanda looks to the definitions of the words “line” and “extension” to argue

that a line extension must serve the same “purpose” as an existing drug. For support, it

states that the definition of line is “material serving a particular purpose.” Appellant’s

Opening Br. 36. But the full text of the cited definition is: “a length of cord, rope, wire, or

other material serving a particular purpose.” Line, New Oxford American Dictionary (3d

ed. 2010) (emphasis added). And the examples given relate to telephone lines or lines of

cord used for hanging out the wash to dry. Id. We are skeptical that this is the definition

that Congress had in mind. A more on-point definition is “merchandise or services of the

19 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 20 of 29

same general class for sale or regularly available.” Line, Merriam-Webster’s Collegiate

Dictionary (10th ed. 1999) (emphasis added). Tellingly, this definition is broader than what

Vanda proposes.

Second, Vanda points out that the agency’s broad definition of line extension would

sweep in drugs for which the FDA requires a new drug application. It urges us to hold that

those drugs are not new formulations, but new drugs themselves ineligible for line-

extension status. But extended-release formulations (which, recall, are explicitly listed in

the statute as new formulations) sometimes require just such new drug applications. That

alone shows that a new drug application is not the silver bullet Vanda imagines.

In making its new-drug-application point, Vanda fails to consider the phrase “new

formulation of the drug” as a whole. Instead, Vanda counsels that Congress’s insertion of

the words “the drug” provides an oblique cross-reference to the oral-solid-dosage-form

requirement. Only by looking there, according to Vanda, can we understand what “the

drug” must mean. In that requirement, Vanda continues, the statute mentions two types of

drugs: “single source drug” and “innovator multiple source drug.” From there, Vanda

instructs that we should look at the definitions portion of the Medicaid Drug Rebate

Program to find out what qualifies a drug as either of those types. And (in a final step)

Vanda points out that both these types of covered drugs must fall “under a new drug

application approved by the Food and Drug Administration.” §§ 1396r-8(k)(7)(A)(ii),

(iv).

According to Vanda, this lengthy chain of cross-references establishes some sort of

statutory instruction that no drug marketed under a new drug application can ever be a line 20 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 21 of 29

extension because such a drug is always an original drug instead. Line extensions, on

Vanda’s reading, can therefore only be drugs for which the FDA requires a supplemental

new drug application.

This argument is wholly unpersuasive. The FDA has a different focus from the

Medicaid Drug Rebate Program. The FDA is concerned chiefly with the safety and efficacy

of drugs, not with their cost. And the cross-references to the FDA approval process in the

definitions of the Medicaid Drug Rebate Program are meant chiefly to limit Medicaid drug

coverage to drugs that the FDA considers safe and efficacious—not to create a hidden

constraint on line extensions.

If Congress had meant to limit line extensions to drugs approved via supplemental

new drug applications, we think it would have done so explicitly within the definition of

the term “line extension.” See Jama v. Immigr. & Customs Enf’t,

543 U.S. 335, 341

(2005)

(“We do not lightly assume that Congress has omitted from its adopted text requirements

that it nonetheless intends to apply, and our reluctance is even greater when Congress has

shown elsewhere in the same statute that it knows how to make such a requirement

manifest.”). Other provisions of the Medicaid Drug Rebate Program do explicitly depend

on FDA determinations: Special reimbursement limits apply when the FDA has rated

several drugs as “therapeutically and pharmaceutically equivalent.” § 1396r-8(e)(4). And

drugs “approved by the [FDA] exclusively for pediatric indications” are eligible for a lower

minimum rebate. § 1396r-8(c)(1)(B)(iii)(II)(bb). No such FDA references are made in the

line-extension definition at issue here. That definition is at the core of the matter. In

contrast, Vanda’s argument invokes provisions that are wandering on the periphery. 21 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 22 of 29

***

Having addressed Vanda’s arguments to the contrary (both general and specific),

we find that the agency’s definitions of line extension and new formulation are

appropriately within the Medicaid statute’s ambit. We thus affirm the district court as to

this aspect of the case.

B.

We turn now to Vanda’s challenge to the portion of the agency regulation regarding

the oral-solid-dosage-form requirement. Recall that the line-extension provision applies

only to “a drug that is a line extension of a single source drug or an innovator multiple

source drug that is an oral solid dosage form.” § 1396r-8(c)(2)(C)(i). The agency and

Vanda disagree about whether the oral-solid-dosage-form requirement applies only to the

original drug (“a single source drug or an innovator multiple source drug”) or to the line-

extension drug as well.

The agency states that the modifier “that is an oral solid dosage form” applies only

to the immediately preceding phrase “a single source drug or an innovator multiple source

drug.” The regulation therefore instructs that “only the initial brand name listed drug must

be an oral solid dosage form.” 85 Fed. Reg. at 87,034. The line-extension drug, according

to the agency, can be in any form at all.

Vanda disagrees. It asserts that the agency’s reading is contrary to the statute and

proposes that both the original drug and the line-extension drug must be in oral solid dosage

form for the provision to apply.

22 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 23 of 29

The agency’s interpretation is the more persuasive. The statute makes clear that only

one thing need be “an oral solid dosage form” by using the singular verb “is.” Verbs in

modifying phrases must correspond to their intended antecedent: If the antecedent is

singular, the verb must be singular. If the antecedent is plural, the verb must be plural. See

Rodney Huddleston & Geoffrey K. Pullum, The Cambridge Grammar of the English

Language § 18.1, p. 500 (2002). Thus a visitor to a local café can order a single cappuccino

that is made with whole milk, but if she wants two she will need to ask the barista for

cappuccinos that are made with whole milk. Vanda’s reading, on which two drugs must

meet the requirement, violates this grammatical rule.

In contrast, the agency’s reading conforms with the rule. According to the agency,

only one thing—the original drug—must be in oral solid dosage form. And, grammatically,

it is of no moment that the original drug can be either “a single source drug or an innovator

multiple source drug.” When singular nouns are joined by words like “or,” they take a

singular verb. See id. § 18.4(b), p. 508.

The agency’s reading also comports with the last antecedent rule, which counsels

that a limiting clause “should ordinarily be read as modifying only the noun or phrase that

it immediately follows.” See Barnhart v. Thomas,

540 U.S. 20, 26

(2003). Here, the phrase

“that is an oral solid dosage form” immediately follows the phrase “a single source drug or

an innovator multiple source drug.” That phrase consists of a series of two specific terms

with “straightforward, parallel construction.” Antonin Scalia & Bryan A. Garner, Reading

Law: The Interpretation of Legal Texts 147 (2012). And so the modifier at the end of that

list should be read to “appl[y] to the entire series.” Id.; see also Facebook, 592 U.S. at 402– 23 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 24 of 29

03. The “most natural reading” is thus that the oral-solid-dosage-form requirement applies

to the two elements in the parallel series before it: single source and innovator multiple

source drugs. See Facebook,

592 U.S. at 403

. That suggests that the original drug (whether

a single source drug or an innovator multiple source one) is the thing that must be in oral

solid dosage form.

We think the agency persuasively posits that the last antecedent rule applies here.

Vanda protests, however, that the last antecedent rule does not apply to an “‘integrated’

clause” that “refer[s] to a single thing.” Cyan, Inc. v. Beaver Cnty. Emps. Ret. Fund,

583 U.S. 416, 440

(2018). The phrase “a line extension of a single source drug or an innovator

multiple source drug,” Vanda continues, is just such an integrated clause. But even so, this

does not support Vanda’s suggested reading. When faced with integrated clauses, the

Supreme Court has instructed that “the modifier goes back to the beginning of the

preceding clause.”

Id.

(emphasis added). And so, even if we were to read the phrase at issue

as an integrated clause, only the line extension (and not the original drug) would need to

be in oral solid dosage form. But Vanda’s position is that both drugs must be. Even Vanda’s

proposed exception to the last antecedent rule does not support its preferred result.

Vanda lastly tries to avoid the most obvious grammatical reading by contending that

the agency’s interpretation would produce an “absurd” result: that pill-to-liquid

transformations can lead to line extensions, but liquid-to-pill transformations cannot. It is

true that the Supreme Court will sometimes reject “the most grammatical reading” of a

statute when it would produce a result that is “positively absurd.” United States v.

X-Citement Video, Inc.,

513 U.S. 64

, 69–70 (1994). But only when the absurdity is quite

24 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 25 of 29

serious indeed. See, e.g.,

id.

(rejecting the most grammatical reading because it would raise

“substantial constitutional questions” and impose criminal penalties without any scienter

requirement). We see no such absurdity here. A great many drugs start off as pills. And

new formulations can work off that initial form as well. It was reasonable for Congress to

target this common path of transformation. Further, judicial rulings that Congress produced

an “absurdity” must be rare indeed, lest we open the gates for our own policy preferences

to trump those chosen by the legislature.

In short, and fundamentally, Vanda’s reading cannot be squared with the language

of the statute. It has therefore failed to demonstrate that the agency’s interpretation of the

oral-solid-dosage-form requirement was contrary to law. We thus affirm the district court

and uphold the agency’s regulation as to the oral-solid-dosage-form requirement as well.

IV.

Having concluded that the regulation is not contrary to law, we turn to Vanda’s

contention that the agency’s rulemaking here was arbitrary and capricious in violation of

the APA. Unlike contrary-to-law analysis, arbitrary and capricious review is “very

deferential.” Rural Cellular Ass’n v. FCC,

588 F.3d 1095, 1105

(D.C. Cir. 2009). It is

focused not on the substance of the regulation, but on “the reasonableness of the agency’s

decisionmaking processes.”

Id.

We look only to whether the agency has “reasonably

considered the relevant issues and reasonably explained the decision” it made. FCC v.

Prometheus Radio Project,

592 U.S. 414, 423

(2021).

Vanda lodges three specific complaints against the procedure undergirding the

agency’s 2020 regulation. First, Vanda contends that the agency failed to confront a

25 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 26 of 29

“practical inconsistency” that results from the agency’s construction of the oral-solid-

dosage-form requirement, reiterating its view that it would be “absurd” for the line-

extension provision to apply to transformations from pill to liquid but not from liquid to

pill. But, given our discussion above, see supra III.B, any such inconsistency must be laid

at the feet of Congress, not the agency. This, then, cannot be cited as a flaw in the agency’s

rulemaking procedure.

Second, Vanda accuses the agency of failing to account for reliance interests

engendered by its 2012 proposal. But the agency explicitly declined to finalize that

proposal and instructed manufacturers instead to “rely on the statutory definition.” 81 Fed.

Reg. at 5265. The notice-and-comment procedure is designed so that an agency can float a

potential rule to the public without committing itself to enacting the proposed rule’s

content. We are thus loath to impose on agencies an obligation to make allowances for

industry players who relied on proposals never implemented. Such an obligation would

only serve to dissuade agencies from making exploratory proposals in the first place. The

agency here acknowledged that it was changing its position, see 85 Fed. Reg. at 87,034,

and explained that the new interpretation was based on “10 years’ experience” and

“supported by the statute,” id. at 87,036. That was sufficient to satisfy the APA.

Finally, Vanda accuses the agency of failing to consider the “chilling effects on

pharmaceutical innovation.” Appellant’s Opening Br. 23. But in alleging such effects,

Vanda overlooks two things. First, line-extension status leads to higher rebates only if a

drug manufacturer chooses to raise the price of the original drug at a rate that outpaces

inflation. Drug manufacturers are thus in control of the rebate that they pay. Second, a drug

26 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 27 of 29

manufacturer can set the initial price for a line-extension drug as high as it needs to recoup

the cost of its development, and can increase that price as well so long as those increases

do not outpace inflation either. As the agency explained, a line-extension drug incurs a

higher rebate “not due to the innovations in the new formulation” but “because the original

drug increased in price faster than the rate of inflation.” 85 Fed. Reg. at 87,041 (emphasis

added). The agency also rightly noted that if Congress had intended for factors such as “the

extent of the improvements” or “the value of an innovation” to “limit the scope of drugs

that are line extensions, it would have provided as much in the statute.” Id. at 87,038. We

think this satisfies the agency’s burden under the APA to respond to innovation-focused

comments and concerns.

In sum, none of the challenges Vanda raises to the agency’s decision-making

process suffice to show that the agency acted arbitrarily or capriciously. We therefore hold

that the process conformed to the APA’s strictures and affirm the district court with respect

to Vanda’s claim here as well.

V.

It is not surprising that the government advocates a broad definition of line

extension, both in the regulation and in this litigation. Line-extension status makes a big

difference to the federal purse. It creates the prospect of larger inflation-based rebates,

which make up an increasingly large portion of the total amount paid under the Medicaid

Drug Rebate Program—more than half since 2012. See Off. of Inspector Gen., U.S. Dep’t

of Health & Hum. Servs., Medicaid Rebates for Brand-Name Drugs Exceeded Part D

Rebates by a Substantial Margin, OEI-03-13-00650 (Apr. 2015).

27 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 28 of 29

Inflation-based rebates account for so much of the total amount rebated not because

of choices the agency has made, but because prescription drug prices continue to rise.

Recall that when Congress was considering the Medicaid Drug Rebate Program, a Senate

report noted that drug prices had been increasing at three times the rate of inflation. This

problem has not abated. Between January 2022 and January 2023, drug prices increased an

average of 15.2%. See U.S. Dep’t of Health & Hum. Servs., Changes in the List Prices of

Prescription Drugs, 2017-2023 (Oct. 6, 2023). That is still nearly three times the 6.4% rate

of inflation measured for the same period. See U.S. Bureau of Lab. Stats., Consumer Prices

for Shelter Up 7.9 Percent from January 2022 to January 2023 (Feb. 16, 2023).

Congress reasonably sought to respond to these soaring costs via innovative

solutions, such as the line-extension provision at issue here. In recent years, Congress has

doubled down on this mission to control drug prices while preserving the ability of

pharmaceutical companies to grow their research and development. Indeed, post-regulation

initiatives underscore this continuing focus. The Inflation Reduction Act of 2022

established for Medicare a rebate program based on the Medicaid Drug Rebate Program at

issue in this litigation. See

Pub. L. No. 117-169, § 11102

,

136 Stat. 1818

, 1871 (codified at

42 U.S.C. § 1395w-114b). The new Medicare rebate provision, like its Medicaid

counterpart, includes a line-extension provision corresponding to the one before us. See

§ 1395w-114b(b)(5)(B)(ii). And it instructs the agency that the Medicare rebate formula

should match the one the agency uses in Medicaid. See § 1395w-114b(b)(5)(B)(i).

Post-regulation initiatives do not operate as any sort of ironclad proof of original

congressional intent to broaden the universe of drug-expenditure regulation. If the agency’s

28 USCA4 Appeal: 23-1457 Doc: 38 Filed: 04/10/2024 Pg: 29 of 29

regulation was at odds with the line-extension provision in the statute, post-enactment

events in other statutes could not salvage it. See Food Mktg. Inst. v. Argus Leader Media,

139 S. Ct. 2356, 2366

(2019). But the fact that Congress is pushing in the same direction

as the agency to carefully balance progress and prices strengthens our conviction that the

statute it enacted allows for the regulation the agency imposed.

Inflation-based rebates, of course, are not permissible as a matter of policy but only

insofar as they conform to law. But given that costs have continued to rise and that

Congress has continued to expand its cost-control measures, we think the balance struck

by the agency’s challenged regulation between cost control and room for pharmaceutical

innovation is appropriate. By entrusting the Medicaid Drug Rebate Program to CMS’s care,

Congress charged the agency with trying to provide some modest set of controls on

healthcare costs to the government at a time when great medical advances bring great

benefits but also impose great expense. Vanda is entitled to disagree with the balance

struck, but we do not think that the statute gives us the authority to upset that balance so

long as it operates within the margins Congress drew. If Vanda prefers a different balance

between these competing forces, it should direct its arguments to Congress, not to the

federal courts.

Our job is simply to ensure that the agency’s rule is not contrary to law. In this case,

we think that the agency’s rule is in accord with law for all the reasons set forth above. The

district court’s judgment is thus

AFFIRMED.

29

Reference

Cited By
4 cases
Status
Published