Medicare has selected 40 drugs for price negotiation since 2023. Ten of them have had a government-set price since January 1, 2026. Fifteen more follow on January 1, 2027, and fifteen more on January 1, 2028. [1] On May 18, 2026, the Supreme Court declined to hear six manufacturer challenges to the program. [2] The main constitutional fight is over. The commercial questions are not.
This article answers four of those questions. How deep are the cuts? Which drugs are exposed next? Do company results show the effect yet? And what happens when the Inflation Reduction Act (IRA), the 2025 orphan-drug amendment and the administration’s most-favored-nation (MFN) deals all run at the same time? I use CMS price tables, SEC filings, court records and law-firm analyses. I mark every figure that I calculated myself. The status date is October 11, 2026. CMS has not yet published the 2028 prices.
DrugPatentWatch tracks the patent, exclusivity and litigation records that decide which drugs enter this program and when they leave it. The section on IP teams, near the end, explains how to use that data.
The Short Answer: What Does the IRA Price Reset Do?
The IRA lets Medicare set prices by negotiation for a growing list of single-source drugs. The first 25 negotiated prices average 62.4% below list price (calculated from CMS tables). [3][4] CMS estimates net Medicare savings of $6 billion for the first cycle and $12 billion for the second. [5][3]
What it does: it sets a Medicare price for each selected drug, with annual CPI-U increases after the price takes effect. [3]
What it does not do: it does not set a commercial-insurance price. Merck warns that government price setting “may also impact pricing in the private market.” [6]
What the data shows so far: list-price discounts are large, but company revenue effects are hard to see. Eliquis sales rose 22% in Q2 2026, in the first year of its Medicare price. [7]
What comes next: 15 more drugs in 2028, up to 20 per year from the 2029 price year, and a separate set of MFN programs that overlap with the IRA. [8][9]
The Seven Findings That Matter
Each finding below can be read alone. Findings marked “calculated” are my own analysis of the cited data.
The first 25 negotiated Medicare prices average 62.4% below list price, with a median of 65%. Cycle 1 averages 63.1%. Cycle 2 averages 61.9%. (Calculated from CMS tables.) [4][3]
Discount depth tracks list price. The 13 drugs listed under $1,000 for a 30-day supply average a 71.6% discount. The 12 drugs listed at $1,000 or more average 52.4%. (Calculated.) [4][3]
CMS estimates the first cycle would have cut 2023 net Medicare spending on its 10 drugs by 22%, or about $6 billion. It estimates the second cycle would have cut 2024 net spending by 44%, or about $12 billion. [5][3]
The 15 drugs in the 2028 cycle had about $27.0 billion in Part B and Part D spending from November 2024 to October 2025. Five drugs account for 57.5% of that total. (The percentage is calculated.) [10]
The Supreme Court denied six manufacturer petitions on May 18, 2026, without comment. Challenges to the second and third cycles are still pending. [2][11]
Eliquis worldwide sales were $4,481 million in Q2 2026, up 22% from $3,680 million in Q2 2025, although Medicare’s 2026 price for Eliquis is $231 against a 2023 list price of $521. [7][4]
CMS cut its savings estimate for the GLOBE Part B model from $11.9 billion in the proposed rule to $440 million in the final rule. [9]
KFF reports that the 40 drugs selected so far accounted for $125 billion, or 36%, of 2024 Medicare Part B and Part D spending. [1]
How Does Medicare Price Negotiation Work?
CMS picks high-spend drugs that have no generic or biosimilar competition. It sends the manufacturer an offer. The manufacturer accepts or counters. The talks end by a fixed date, and the agreed price starts about two years after selection.
What Is the Maximum Fair Price (MFP)?
The maximum fair price (MFP) is the price that Medicare pays for a selected drug after negotiation. The law sets a ceiling on it. The ceiling has two parts. One part uses the net price that Part D plans negotiate. The other part is a percentage of the non-FAMP, which is the non-Federal average manufacturer price. [1]
The non-FAMP percentage falls as a drug ages. It is 75% for small molecules and vaccines that are more than 9 and less than 12 years past approval. It is 65% at 12 to 16 years. It is 40% beyond 16 years. [1] After an MFP takes effect, it rises each year by the CPI-U percentage increase. [3]
Which Drugs Can Medicare Select?
A drug must be a single-source drug. The eligible pool holds up to 50 qualifying drugs with the highest Part B spending and up to 50 with the highest Part D spending. Drugs already selected, and drugs covered by the Small Biotech Exception, are not in the pool. CMS ranks the pool by total spending and selects the top 15 for the 2028 cycle. [10]
The 7-Year and 11-Year Clocks
A small-molecule drug must be at least 7 years past FDA approval on the selection date. A biologic must be at least 11 years past licensure. For the 2028 cycle, a small molecule must have been approved on or before February 1, 2019. A biologic must have been licensed on or before February 1, 2015. [1]
Orphan, Small-Biotech and Plasma Exclusions
CMS excludes certain orphan drugs, low-spend drugs and plasma-derived products. The 2025 reconciliation law (P.L. 119-21) widened the orphan exclusion. [10] A drug with several rare-disease designations is now ineligible. The 7-year or 11-year clock also waits for orphan drugs that later win a non-orphan approval. CBO estimates this change will cost the federal government $8.8 billion over the coming decade. [1]
The Small Biotech Exception removed five drugs from the 2028 pool. [10] CMS does not plan to codify that exception for 2029 and later. [8]
How Does a Negotiation Round Run, Step by Step?
The table shows the 2028 cycle. All dates are statutory or CMS-announced. The last four rows are scheduled events, not completed events.
Date
Event
Status on Oct 11, 2026
Source
Jan 27, 2026
CMS announces 15 drugs, plus Tradjenta for renegotiation
From selection on January 27, 2026 to the effective date of January 1, 2028 is 704 days (calculated). The negotiation window itself, from January 27 to November 1, 2026, is 278 days (calculated).
What Happens If a Manufacturer Refuses to Negotiate?
A manufacturer that does not reach an agreement owes an excise tax under section 5000D of the tax code. The company can also lose Medicare and Medicaid participation. [2] The court challenges argued that this makes participation coerced. The courts found that participation stays voluntary. [2]
Excise Tax Rates by Period of Noncompliance
The tax is stated as a share of the post-tax sale price. The implied rate on the pre-tax price is the share divided by one minus the share. [13]
The noncompliance period ends when the company signs an agreement or when a generic version becomes available. [13] No cycle so far has ended in a refusal. All 10 cycle-1 drugs and all 15 cycle-2 drugs ended with agreed prices, and all manufacturers signed for the 2028 cycle. [5][3][12]
What Did the First 25 Negotiated Prices Show?
The first two cycles produced 25 prices. All are Part D prices. The tables below give CMS list prices and MFPs for a 30-day supply.
Cycle 1 Prices: Effective January 1, 2026
CMS sent initial offers on February 1, 2024. The talks ended on August 1, 2024, with agreed prices for all 10 drugs. [5] The 10 drugs had $56.2 billion in 2023 Part D gross covered costs, about 20% of the Part D total. About 8.8 million of 54 million Part D enrollees used them. [4]
CMS estimated that the negotiated prices would have cut 2023 net Medicare spending on these drugs by about $6 billion, or 22%. It projected about $1.5 billion in savings for enrollees in 2026 under the standard benefit design. [4]
Cycle 2 Prices: Effective January 1, 2027
CMS sent initial offers on June 1, 2025. The period ended on November 1, 2025, with agreement on all 15 drugs. Eight prices were agreed in meetings or price exchanges, and in seven of those CMS accepted a revised counteroffer. The other seven came from a written final offer that the company accepted. [3]
Drug
Company
List price, 2024
MFP, 2027
Discount
2024 Part D gross covered cost
Ozempic; Rybelsus; Wegovy
Novo Nordisk
$959
$274
71%
$15.16B
Trelegy Ellipta
GSK
$654
$175
73%
$5.30B
Xtandi
Astellas
$13,480
$7,004
48%
$3.40B
Pomalyst
Bristol Myers Squibb
$21,744
$8,650
60%
$2.15B
Ofev
Boehringer Ingelheim
$12,622
$6,350
50%
$2.09B
Ibrance
Pfizer
$15,741
$7,871
50%
$2.04B
Linzess
AbbVie
$539
$136
75%
$1.98B
Calquence
AstraZeneca
$14,228
$8,600
40%
$1.70B
Austedo; Austedo XR
Teva
$6,623
$4,093
38%
$1.68B
Breo Ellipta
GSK
$397
$67
83%
$1.43B
Xifaxan
Salix
$2,696
$1,000
63%
$1.16B
Vraylar
AbbVie
$1,376
$770
44%
$1.14B
Tradjenta
Boehringer Ingelheim
$488
$78
84%
$1.13B
Janumet; Janumet XR
Merck
$526
$80
85%
$1.07B
Otezla; Otezla XR
Amgen
$4,722
$1,650
65%
$1.05B
Source: CMS. [3] List prices are wholesale acquisition costs based on 2024 fills.
The 15 drugs had $42.5 billion in 2024 Part D gross covered costs, about 15% of the Part D total. CMS estimated net savings of about $12 billion, or 44%, against 2024 net spending. With coverage-gap discount spending included, the figure is about $8.5 billion, or 36%. Enrollees spent $1.7 billion out of pocket on these drugs in 2024. CMS projects about $685 million in enrollee savings in 2027. [3]
Original Analysis: Discount Depth by List-Price Band
I tested whether discount depth relates to list price. The result is a clear split.
Method
I used the 25 prices in the two tables above. I counted each CMS row as one price, so the semaglutide family counts once. I split the drugs at a list price of $1,000 per 30-day supply. I recomputed each discount from the list price and the MFP. All recomputed values match the CMS rounded figures within 0.5 percentage point. [4][3]
Result
List price band (30-day supply)
Drugs
Mean discount
Median discount
Range
Under $1,000
13
71.6%
73%
53% to 85%
$1,000 or more
12
52.4%
50%
38% to 67%
All 25 drugs
25
62.4%
65%
38% to 85%
Source: author’s calculation from CMS tables. [4][3]
The 19-point gap is large. The four deepest cuts are Janumet (85%), Tradjenta (84%), Breo Ellipta (83%) and Januvia (79%). All four list under $1,000. The four shallowest are Imbruvica and Austedo (38% each), Calquence (40%) and Vraylar (44%). Imbruvica, Austedo and Calquence list above $6,000.
Limits
The sample is small. The $1,000 cut-off is a round number that I picked. Other cut-offs move the averages. CMS compares the MFP with list price, not with the net price that the manufacturer received. The gap between list and net varies by drug, and I did not adjust for it. I also did not review the per-drug CMS explanations of the agreed prices. Those documents are where to look for the reasons behind each discount.
Why a 66% Discount From List Is Not a 66% Saving
A discount from list price is not a saving against net spending. CMS measures savings against net spending after rebates. The two numbers differ a lot.
Cycle 1 saved an estimated $6 billion on $56.2 billion of gross covered cost. That is 10.7% of gross cost (calculated). Cycle 2 saved an estimated $12 billion on $42.5 billion. That is 28.2% of gross cost (calculated). [4][3]
The simple average list discount barely moved between cycles. It was 63.1% in cycle 1 and 61.9% in cycle 2 (calculated). The net saving share rose from 22% to 44%. The fact sheets I reviewed give no per-drug net savings, so I cannot split that change into its causes. For modeling, the number to use is the net price change, not the discount from list.
Which 15 Drugs Are in Negotiation for 2028?
CMS announced the third cycle on January 27, 2026. It covers 15 new drugs, and it adds Tradjenta for renegotiation. [12] The CMS list says the drugs are payable under Part B and/or covered under Part D. It does not classify each drug by part. [10]
The 15 Drugs and Their Medicare Spending
Spending covers Part B and Part D combined, from November 1, 2024 to October 31, 2025. [10] The share and per-user columns are my calculations.
The group totals about $27.0 billion, or about 6% of Part B and Part D spending in the period. About 1.78 million people used the drugs. [10] Trulicity alone is 18.1% of the total (calculated). The top five drugs are 57.5% (calculated). The two Eli Lilly drugs, Trulicity and Verzenio, are 23.4% together (calculated).
Spend per user separates two kinds of drug. Botox, Anoro Ellipta and Trulicity have low spend per user and large user counts. Lenvima, Erleada, Verzenio and Kisqali each exceed $90,000 per user and serve under 20,000 people each (calculated). The first group is a volume story. The second is a price story.
Who Is Exposed? 40 Selections by Manufacturer Family
I grouped the 40 selected drugs by parent company. I assigned subsidiaries as follows: Pharmacyclics to AbbVie, Immunex to Amgen, the Janssen entities to Johnson & Johnson, PF PRISM CV to Pfizer and Genentech to Roche. I count Tradjenta once.
Source: author’s grouping of CMS selection lists. [4][3][12] Eliquis is co-marketed by Pfizer, and Lenvima by Merck. I counted each drug once, under the company that CMS lists.
The counts sum to 40 (calculated). AbbVie has the most selections, with four. Six companies have three each. Merck also notes that Lenvima is now selected, with a 2028 price. [6]
Tradjenta: The First Renegotiation
Tradjenta, a type 2 diabetes drug from Boehringer Ingelheim, was first negotiated for 2027. Its agreed MFP is $78 against a list price of $488. [3] CMS selected it again in the third cycle. [10] A renegotiated drug already has an agreement, so it needs no new one. [10]
Renegotiation Criteria
A prior-cycle drug is eligible if it has no approved and marketed generic or biosimilar and at least one of three things is true. Its status has changed to “long monopoly.” It added a new indication. Or a statutory negotiation factor has materially changed. CMS then picks drugs that moved to long-monopoly status, or drugs where renegotiation is likely to change the MFP significantly. [10] From the 2029 price year, CMS proposes to treat a drug’s agreed MFP as one of the renegotiation factors. [8]
Did the Law Survive Court?
Yes, at the appellate level. The Supreme Court has left every first-wave ruling in place.
What the Supreme Court Did on May 18, 2026
The Court denied certiorari petitions from AstraZeneca, Novo Nordisk, Novartis, Bristol Myers Squibb, Johnson & Johnson (through Janssen) and Boehringer Ingelheim. It gave no reason. [2][14] A denial is not a ruling on the merits. It leaves the lower-court decisions standing.
The appellate record runs as follows. The Third Circuit decided the AstraZeneca case in May 2025. It was the first appellate merits decision, and it rejected a procedural due process claim. The Second Circuit affirmed dismissal of the Boehringer Ingelheim takings, compelled-speech and unconstitutional-conditions claims. The Third Circuit rejected each theory in the Janssen, Bristol Myers Squibb, Novo Nordisk and Novartis cases. Some of those rulings drew a dissent. [11]
Claims That Failed
Takings and due process under the Fifth Amendment. The courts held that manufacturers have no property right to sell at a given price. [11]
Compelled speech under the First Amendment, because manufacturers must call the negotiated price “fair.” [2]
Nondelegation, which argued that Congress gave too much power to the executive branch. [2]
Excessive fines and Administrative Procedure Act claims, which were also pressed across the first wave. [11]
The courts reasoned that Medicare participation remains voluntary. [2]
Which Lawsuits Are Still Open?
One first-wave case is pending in the Fifth Circuit. One is pending in federal court in D.C. Challenges to the second and third waves are also open, and one of them is on appeal in the D.C. Circuit. [11] Merck states in its 10-K that it has sued the government over the program. [6] I could not confirm the current status of the Merck case from the sources I reviewed.
What Happens to Revenue? Eliquis, Januvia and Keytruda
Company filings give the first real-world test. Three drugs show different cases: a large drug with a price already in force, a small legacy drug, and a very large drug with a price still ahead.
Eliquis: MFP $231, Yet Sales Rose 22%
Eliquis has a Medicare price of $231 for a 30-day supply, against a 2023 list price of $521. That is 56% below list. [4] The price took effect on January 1, 2026. Bristol Myers Squibb states in its Q2 2026 10-Q that maximum fair prices in Medicare may materially affect its business and results. [7]
Eliquis worldwide revenue was $4,481 million in Q2 2026. In Q2 2025 it was $3,680 million. That is growth of 22% (calculated from the filing figures). [7] Reuters reports that Eliquis’s share of new U.S. prescriptions is nearing 80%, according to BMS Chief Commercialization Officer Adam Lenkowsky. BMS raised its 2026 Eliquis growth forecast to 20% to 25%, from 10% to 15%. [15] On the earnings call, CFO David Elkins said the second half benefits from the end of an accumulated CPI penalty in certain U.S. government channels, after a price reduction at the start of the year. [16]
The sources do not break out the net effect of the Medicare price on Eliquis revenue. They do not say why reported revenue rose in the first year of the price. Two facts limit any conclusion. The 56% cut is measured from the 2023 list price, not from the net price that BMS received. And a product revenue line mixes volume, channel and price effects.
Analysis: one product revenue line is a weak test of the policy in year one. Share gains can hide a lower price per unit. Analysts should look for disclosed Part D net price and volume data before they draw a conclusion.
Januvia and Janumet: The Small-Molecule Test Case
Merck’s 2025 10-K says HHS selected Januvia in 2023 for the first year of the program. The government set a price that became effective on January 1, 2026. The price for Janumet and Janumet XR becomes effective on January 1, 2027. [6] Januvia and Janumet had $2,544 million in 2025 sales. [6] The filing does not state the dollar effect of the price. [6]
The Januvia case differs from Eliquis. PharmaVoice reported in August 2024 that Januvia revenue had already fallen in the first half of 2024 because of lower demand and generic competition. [17] Analysis: for a legacy drug in decline, the Medicare price adds to a trend that already exists. It is hard to separate the two causes in the reported numbers.
Keytruda: The 2029 Clock
Keytruda and Keytruda Qlex had $31,680 million in 2025 sales. That is 12.5 times the Januvia and Janumet total (calculated). [6] Merck writes that it “expects that Keytruda will be selected in 2027 for government price setting,” with a price effective on January 1, 2029. Merck also writes that “U.S. sales of Keytruda will decline materially after that time.” [6] The filing gives no dollar estimate. This is an expectation by the company, not a CMS announcement.
What the Orphan-Drug Change Did
KFF reports that the 2025 reconciliation law delayed selection of Keytruda and Opdivo beyond 2026. CBO estimates that the orphan provisions will cost the federal government $8.8 billion over the coming decade. [1] The original CBO estimate for negotiation savings was $98.5 billion for 2022 to 2031. [18] The $8.8 billion is about 9% of that figure (calculated). The two estimates have different dates and time windows. Treat the comparison as a size check, not as a measure of lost savings.
How Does a Negotiated Price Reach the Pharmacy Counter?
A negotiated price needs a payment system. CMS built one, and its timing affects pharmacy cash flow.
How Does MFP Effectuation Work?
At the point of sale, the pharmacy charges the enrollee on the basis of the MFP. The plan pays the pharmacy no more than the MFP plus the dispensing fee. The pharmacy then gets its money in one of two ways. In a prospective purchase, the pharmacy buys the drug at the MFP. In a retrospective refund, the pharmacy buys at its usual acquisition cost, and the manufacturer refunds the difference. The default refund is the wholesale acquisition cost minus the MFP. [19]
CMS built the Medicare Transaction Facilitator to route claims data and refunds. Plans have 7 days to submit prescription drug event data. Manufacturers have 14 days to decide how to handle each claim. Pharmacies wait at least 21 days for a refund. That is longer than the 14-day prompt-payment rule for plan payments. [19]
Manufacturers raised three concerns: processing millions of claims in 14 days, spotting 340B claims to avoid double discounts, and civil penalties for late refunds. Pharmacies raised cash-flow risk, especially for small independent pharmacies. [19] That brief is from October 2025, before the first prices took effect. I did not review post-January operating data. CMS plans draft guidance on MFP effectuation for 2028 in summer 2026. It plans to codify its effectuation policies for 2029 and later in CY2027 rulemaking. [8]
What Does the Part D Redesign Change in 2026?
The 2026 out-of-pocket threshold is $2,100, and the annual deductible is $615. In the initial coverage phase, the manufacturer discount program typically covers 10% of the cost of applicable drugs. In the catastrophic phase, it typically covers 20%. [20]
Selected drugs follow separate rules. In initial coverage, sponsors typically pay 65% of the cost of a selected drug, and CMS pays a 10% subsidy during the price applicability period. In catastrophic coverage, CMS pays a 40% reinsurance subsidy for selected drugs and 20% for applicable drugs. The selected-drug subsidy ends when the enrollee reaches the $2,100 threshold. [20]
What Are Inflation Rebates?
Manufacturers of single-source Part B drugs, and nearly all Part D drugs, pay rebates to Medicare when prices rise faster than CPI-U. Part B uses average sales price. Part D uses average manufacturer price. The base year is 2021. A manufacturer that does not pay owes a penalty of at least 125% of the rebate. [18]
For Part B drugs with above-inflation increases, enrollee coinsurance is calculated on 20% of the drug’s lower, inflation-adjusted price. [18] CBO estimated in 2022 that inflation rebates would cut the federal deficit by $63.2 billion over 2022 to 2031. It estimated $237 billion for all the drug-pricing provisions together. [18]
Does the IRA Reduce Drug Innovation? What the Data Shows
The evidence is early and narrow. One published study gives a measurable signal in one trial type.
What Does the Post-Approval Oncology Trial Study Show?
Zheng, Patterson and Campbell studied industry-sponsored phase I to III post-approval oncology trials in the Citeline Trialtrove database. The sample has 4,367 trials started from July 2014 to August 2024. Of these, 2,519 (57.7%) tested small molecules and 1,848 (42.3%) tested biologics. [21]
The monthly average of all post-approval oncology trial starts fell 40.0% after the IRA. [21]
In a difference-in-differences model, small molecules lost an additional 4.5 trial starts per month against biologics (95% CI, -7.1 to -1.9). [21]
In a post-COVID sensitivity analysis, small-molecule starts fell 43.6% and biologic starts fell 8.4%. The biologic change was not statistically significant. [21]
All three authors work for the National Pharmaceutical Council. The authors say the design does not prove cause. The effect may have begun before the IRA passed, possibly as early as November 2021. Some post-marketing commitments may be in the sample. The study does not measure long-term effects on innovation, spending or patient access. [21] It also covers one trial type. It does not test early-stage pipelines or total approvals.
What Is the “Pill Penalty,” in Dates?
The selection clocks are 7 years for small molecules and 11 years for biologics. [1] The price takes effect about two years after selection. Critics call the four-year gap between the two classes the “pill penalty.” The sources I reviewed show no enacted change to the 7-year and 11-year clocks. KFF (March 2026) and CMS (2028 selection) both still apply them. [1][10]
Worked Example: The 2028 Cutoff Dates
Drug class
Approval cutoff for 2028 selection
Years required at selection
MFP effective date
Approval to price (calculated)
Small molecule
Feb 1, 2019
7
Jan 1, 2028
8.9 years
Biologic
Feb 1, 2015
11
Jan 1, 2028
12.9 years
Sources: KFF and CMS for the dates. [1][10] The last column is the author’s calculation.
Many sources write “9 years” and “13 years” as eligibility ages. The table shows why that is loose. The eligibility ages are 7 and 11. The 9 and 13 years are the time from approval to a price in force.
IRA vs. MFN, GLOBE and GENEROUS: Who Sets the Price?
The IRA is one of several federal price programs. They differ in legal basis, scope and size. Mixing them up leads to wrong forecasts.
How the Programs Compare
Program
Legal form
Scope
Status on Oct 11, 2026
Source
IRA negotiation
Statute
Medicare Part D, and Part B from 2028
10 prices in force. 15 start in 2027. 15 start in 2028.
GLOBE Final Rule: From $11.9 Billion to $440 Million
CMS proposed GLOBE for Part B and GUARD for Part D on December 23, 2025. Both would test international reference pricing in place of inflation-based benchmarks. [22] CMS announced the GLOBE final rule on September 30, 2026. The model begins on January 1, 2027. [9]
GLOBE covers separately payable Part B single-source drugs and biologics in defined drug classes. The product must have more than $100 million in Original Medicare Part B spending over 12 months. The benchmark uses 19 reference countries, adjusted for GDP per capita at purchasing-power parity. Method I uses the lowest adjusted country price from commercial data. Method II uses the manufacturer’s voluntarily submitted net prices. [9]
CMS cut its projected federal savings from $11.9 billion in the proposed rule to $440 million over seven years in the final rule. The impact analysis says as few as four manufacturers will be covered. [9] CMS intends to waive mandatory participation, quarter by quarter, for any manufacturer in GENEROUS. An independent analysis found that 32 of 36 otherwise-qualifying medicines come from GENEROUS or Medicaid-deal companies. That leaves about four drugs from three companies in scope. [9]
What GLOBE Excludes
GLOBE excludes orphan-only drugs, cell and gene therapies on FDA’s approved list, plasma-derived products, and drugs with an MFP in effect. It also excludes biosimilars and their reference products once a biosimilar is marketed. CMS declined to exclude oncology drugs. [9] The MFP exclusion matters here. A drug that the IRA prices does not also face GLOBE.
Ozempic: Two Federal Prices for One Drug
The IRA price for Ozempic, Rybelsus and Wegovy is $274 for a 30-day supply from January 1, 2027. The 2024 list price was $959. [3] In November 2025, the administration and Novo Nordisk agreed that Medicare would pay $245 per month for Ozempic and Wegovy, tied to international prices. [25]
Bloomberg Law reported that both prices were set to apply in 2027 and that CMS had not explained how they relate. It also reported a difference in obligation. The IRA price must appear on Part D formularies. The MFN price is not required to be covered by insurers. [25] Bloomberg Law also reported that dosage differences put the actual IRA figures at about $276 for Ozempic and Rybelsus and about $385 for Wegovy. [25] I did not find a later CMS clarification in the sources I reviewed.
How Far Do the MFN Deals Reach?
The deals trade price concessions for tariff relief. Bristol Myers Squibb, for example, agreed to provide Eliquis free to Medicaid. [23] Eliquis therefore has a Medicare MFP of $231 and a separate Medicaid arrangement. The two programs touch the same drug in different payer channels.
Forbes reports that a White House report projects $64 billion in Medicaid savings over 10 years. Forbes calls the figure dubious because the assumptions are unclear and the deal terms are hidden. It notes that Medicaid already pays net prices that reflect mandatory and supplemental rebates, so the extra saving from MFN rates is unclear. It also notes that MFN does not directly address manufacturers’ list-price power. [24]
The two sets of programs differ in durability. The IRA is a statute, and the Supreme Court left the appellate rulings that upheld it in place. [2] The MFN deals are confidential agreements with terms that, for many, end in under three years. [24]
What Happens Next? Timeline and Scenario Analysis
Dated Milestones, October 2026 to January 2029
Date
Event
Type
Source
Jan 1, 2026
Cycle 1 MFPs take effect. Part D out-of-pocket threshold is $2,100.
CMS issued proposed rule CMS-4215-P on June 16, 2026. Comments were due on August 17, 2026. The rule would codify the program for the 2029 price year and later. For 2026 to 2028, CMS used program instruction authority, which expires. [8] I did not find a final version in my sources.
CMS would select up to 20 drugs in each price year from 2029, and could select drugs for renegotiation.
Fixed-combination drugs: if a combination product shares an active ingredient with another product from the same NDA or BLA holder, and the added ingredient makes a new formulation with a new route of administration, CMS would treat the products as one qualifying single-source drug. The example in the fact sheet is hyaluronidase combined with an active ingredient.
Bona fide marketing: CMS would change when it checks whether a generic or biosimilar is marketed, both for ineligibility and for deselection. It would also monitor drugs that were excluded or deselected.
Orphan drugs: CMS would clarify how it treats withdrawn indications when it assesses former orphan drugs.
Small Biotech Exception: not codified. The definition still applies to a new temporary floor for small biotech drugs in the 2029 and 2030 price years.
Payment: the price used for payment could not exceed the MFP plus dispensing fees, and Part D plans would have to cover selected drugs with an MFP in effect.
Analysis: the fixed-combination proposal matters for lifecycle strategy. It targets reformulations that add a new route of administration. It may reach subcutaneous versions of intravenous biologics that use hyaluronidase. The fact sheet names only that example, so apply this reading with care.
How Many Drugs Can the Program Reach?
This is scenario analysis. It sets a ceiling, not a forecast.
Assumptions
CMS selects the maximum number each year. No selected drug leaves the program. Renegotiations do not add new drugs. These assumptions follow the selection limits in the proposed rule. [8][1]
Result
Price year
New selections allowed
Cumulative maximum (calculated)
2026
10
10
2027
15
25
2028
15
40
2029
Up to 20
60
2030
Up to 20
80
2031
Up to 20
100
A spending scenario for the 2029 group shows the range. If each of the 20 drugs had the lowest spending in the 2028 group ($787 million, Cimzia), the group would total $15.7 billion. If each had the 2028 average ($1.80 billion), it would total $36.0 billion. Both are calculated from the 2028 table. [10] The 2028 group of 15 totals $27.0 billion. These scenarios leave out Keytruda. I did not find a verified Medicare spending figure for Keytruda in my sources. If Keytruda is in the 2029 group, the total rises.
What This Means for Brand Manufacturers, Generic Developers and IP Teams
What This Means for Brand Manufacturers
Selection depends on two things that a company can read in advance: the approval date and the competition status. A small molecule approved on or before February 1, 2019 met the 2028 age test. A biologic licensed on or before February 1, 2015 met it. [1] The spending rank then decides who is picked. [10]
The list discount is not the planning number. The first 25 prices average 62.4% below list, but CMS estimates net savings of 22% and 44% for the first two cycles. [4][3][5] Eliquis shows that reported revenue can rise in year one. [7] A plan should model the Medicare net price and the volume separately.
Reformulation plans need a second look. The June 2026 proposal would group some fixed-combination products with their parent drug. [8] A new route of administration may not reset the clock for such a product under that proposal.
What This Means for Generic and Biosimilar Developers
Generic and biosimilar marketing status appears in four places in the program. A drug is single-source only while no generic or biosimilar is marketed. [10] A selected drug can be deselected after bona fide marketing begins. [8] The excise-tax period ends when a generic version is available. [13] And renegotiation requires that no generic or biosimilar is approved and marketed. [10]
Biologics have a separate rule. CMS excludes from selection some biologics that qualify for a delay because of likely biosimilar entry. [10] CMS now proposes to change when it reviews bona fide marketing, and to monitor excluded and deselected drugs. [8]
Analysis: the launch date matters for the program as well as for market share. An entry date that a settlement sets is a date that CMS checks may also reach. I did not find CMS guidance on settlement-dated entry in my sources. Ask counsel before you rely on any assumption here.
What This Means for Patent and IP Teams
Patent and exclusivity records are inputs to the program. DrugPatentWatch aggregates patent, exclusivity and litigation records for approved drugs. An IP team can use it to lay each drug’s expiry and entry dates beside the CMS dates in the tables above. It does not replace the primary records. Those are the FDA Orange Book and Purple Book, court dockets and CMS notices.
Check the approval date against the 7-year and 11-year cutoffs for the next selection round.
Check competition status: marketed generics or biosimilars, pending Paragraph IV cases and settlement entry dates.
Compare patent and exclusivity expiry dates with the date a price would take effect. For a 2029 price year, that date is January 1, 2029.
Check any reformulation or combination plan against the fixed-combination definition in the proposed rule.
What This Means for Investors and Analysts
Use net price, not list price. Use the dated milestones above to place each drug in a price year. Read the company’s own filing language. Merck, for example, names Januvia, Janumet, Lenvima and Keytruda with their price years but gives no dollar effect. [6] Treat the GLOBE and MFN programs as smaller factors for now. CMS estimates $440 million in GLOBE savings over seven years. [9]
Methodology and Limitations
This section lets a reader reproduce the calculations.
Data analyzed: CMS fact sheets for the three negotiation cycles, KFF briefs, a CMS proposed-rule fact sheet, SEC filings, legal-publisher summaries of court rulings, and trade-press reports.
Time period: negotiation cycles from 2023 to 2026. Revenue data cover 2025 and Q2 2026. The status date is October 11, 2026.
Records included: all 25 negotiated prices that CMS had published on the status date. I excluded the 2028 cycle because it has no published prices. I excluded renegotiation.
Duplicates: Tradjenta counts once. The semaglutide family counts once. Rows that CMS lists with an extended-release form (for example, Austedo and Austedo XR) count once.
Calculations: discount equals one minus MFP divided by list price. Averages are simple, not spend-weighted. Shares equal a drug’s spending divided by the sum of the 15 listed values ($27.016 billion). CMS rounds the total to $27.0 billion. Growth equals Q2 2026 divided by Q2 2025, minus one. Day counts are calendar days.
Definitions: list price means the wholesale acquisition cost that CMS reports. Net savings means CMS’s estimate against net spending after rebates.
Sources I read only through a secondary summary: the BMS Q2 2026 10-Q figures (a StockTitan summary), the Reuters report (via Yahoo Finance), the GLOBE final rule (a Skadden summary) and the Supreme Court denial (legal publishers). I did not read the court opinions or the Supreme Court docket directly.
Other limits: CMS does not classify each 2028 drug by Medicare part in the fact sheet that I read. The trial-activity study comes from authors who work for the National Pharmaceutical Council. Events after October 11, 2026 are not covered.
Key Takeaways
Medicare has selected 40 drugs across three cycles. Ten prices are in force, fifteen start on January 1, 2027, and fifteen start on January 1, 2028. [1]
The first 25 negotiated prices average 62.4% below list price (calculated). Drugs listed under $1,000 for 30 days average 71.6%. Drugs listed at $1,000 or more average 52.4% (calculated).
CMS estimates net savings of about $6 billion for cycle 1 and about $12 billion for cycle 2. [5][3]
The 2028 cycle covers about $27.0 billion of spending. Five drugs make up 57.5% of it (calculated). [10]
The Supreme Court denied six manufacturer petitions on May 18, 2026. Other cases are still pending. [2][11]
Eliquis worldwide revenue rose 22% in Q2 2026 in the first year of its Medicare price. The sources do not separate price and volume effects. [7]
Merck expects Keytruda to be selected in 2027, with a price from January 1, 2029. This is a company expectation. [6]
The program can reach at most 60 drugs by the 2029 price year and 100 by 2031 if CMS selects the maximum each year (calculated). [8]
GLOBE is a separate Part B program. CMS now estimates $440 million in savings, down from $11.9 billion. [9]
FAQ
When Do Ozempic and Wegovy Get a Negotiated Medicare Price?
The IRA price starts on January 1, 2027. It is $274 for a 30-day supply of Ozempic, Rybelsus and Wegovy, against a 2024 list price of $959. [3]
Do Negotiated Prices Apply to Commercial Insurance?
No. The MFP is a Medicare price. Merck writes that government price setting may also affect pricing in the private market. [6]
What Is the Small Biotech Exception?
It removed five drugs from the 2028 selection pool. [10] CMS proposes not to codify it for 2029 and later. A related temporary floor for small biotech drugs would apply in the 2029 and 2030 price years. [8]
Will Keytruda Be Negotiated?
Merck expects selection in 2027 and a price from January 1, 2029. [6] KFF reports that the 2025 orphan-drug changes delayed selection of Keytruda beyond 2026. [1] CMS has not announced a 2029 list.
Does a Generic or Biosimilar Launch End a Drug’s Negotiation Status?
It can. A generic or biosimilar launch affects eligibility, deselection and the excise-tax period. [10][13] CMS proposes new timing for its review of bona fide marketing. [8]
How Does a Pharmacy Get Paid for a Drug With an MFP?
The pharmacy either buys at the MFP or buys at its usual cost and gets a manufacturer refund. Pharmacies wait at least 21 days for a refund. [19]
What Is the Part D Out-of-Pocket Cap in 2026?
The 2026 threshold is $2,100. The annual deductible is $615. [20]
How Many Drugs Can CMS Select After 2028?
CMS proposes to select up to 20 drugs for the 2029 price year and each later year. [8] That ceiling would put the program at 60 drugs by 2029 and 100 by 2031 (calculated).
Is GLOBE Part of the IRA?
No. GLOBE is a CMS Innovation Center model for Part B. It excludes drugs that have an MFP in effect. [9][22]
Does the IRA Cap Insulin Costs?
Yes. Monthly cost sharing is capped at $35. The cap began for Part D on January 1, 2023 and for Part B on July 1, 2023. From 2026, Part D insulin copayments are capped at the lesser of $35, 25% of the MFP for a selected insulin, or 25% of the negotiated price. [18]
