Deconstructing Non-Steroidal Anti-Inflammatory Drugs (NSAIDs) AOASM OMED15 October 18, 2015 Orlando, FLA F.A. Fitzpatrick, Ph.D.

Abbreviations ASA – Acetylsalicylic Acid (Aspirin) COX – Cyclooxygenase MI – Myocardial Infarction tNSAID – traditional non-steroidal anti-inflammatory drug RCT – Randomized Clinical Trial PG – Prostaglandin PGI2 – Prostacyclin TxA2 – Thromboxane USPTFS – US Prevention Services Task Force

Evolution of NSAIDs A rheumatologist is a gastroenterologist’s best friend

∼1975

Cell Membrane Phospholipids Arachidonic Acid

Diclofenac Piroxicam Naproxen Ibuprofen Indomethacin

A-spirin Salicylate

1899

Joints Gastric ConnecMve Tissue mucosa COX

COX

PGE2

PGE2

InflammaMon

Normal Ulcers Gastric GI Bleeding Mucosa

Swelling Edema Pain ‘Normal’

Discovery ∼ 1976 •  Thromboxane (TxA2) •  Prostacyclin (PGI2) Platelet:endothelium interface Cardiologists enter the story: “Paradise Lost”

Arachidonic Acid

Blood Platelets COX Thromboxane Vasoconstric4on !AggregaMon

ThromboMc

Vascular Endothelium COX Prostacyclin Vasodila4on "AggregaMon

AnM-thromboMc

Vessel-blood interface

Joints

Gastric mucosa

COX

COX

PGE2

PGE2

Gastric Swelling ProtecMon Edema Pain InflammaMon

Mechanism of Action •  At anti-inflammatory doses, aspirin is a pro-drug # salicylate, a reversible COX inhibitor. have any Do these nuances •  Aspirin itself is an irreversible inhibitor clinical because itsignificance? acetylates COX. •  tNSAIDs are reversible COX inhibitors with distinctive chemical & pharmacokinetic traits.

ASA

A single, daily low-dose aspirin Platelet aggregation (81 mg) has a durable (days) antithrombotic effect due to inhibited" unique Antithrombotic chemical features of acetylsalicylic platelets acid, platelets & mesenteric portal circulation. COX acetylated=inactive

=Thromboxane"

Platelet Low dose minimizes gastric ulcer COX risk FuncMonal but confers a durable systemic anti-

thrombotic effect

Platelet Aspirin inhibits COX irreversiblyCOX Platelets have no nucleus Inactivated

•  •  1 requires 2 3 platelet 4 5 turnover •  Recovery

Days

Does aspirin ‘prevent’ heart attacks? Models, Clinical Trials, Recommendations

‘Imbalance” Model: Blood Flow in Coronary (or Endothelium Platelets Prostacyclin Thromboxane Cerebral) Vessels Reflects The (Anti-Thrombotic) (Thrombotic) ‘Balance’ of Thrombotic & Antithrombotic Mediators ‘Basal’ capacity Excess Risk for vasoconstriction (angina) or occlusion (MI)

Clinical Trial Placebo 1016

1000 ISIS-2 2nd International Study of Infarct Survival 804 800 Aspirin Cumulative 17,187600 cases of suspected acute MI Vascular randomized into 4 arms. 400 Deaths Primary endpoint: 5 week vascular mortality. 200 Aspirin Both Placebo 160 mg 1 2 Streptokinase 4 5 3 1988 Se 1 month Weeks after Randomization

FDA & USPSTF recently reached different conclusions about aspirin’s benefits/risk in primary prevention of patients with heightened baseline risk of heart attack.

Possible reasons?

Actuarial perspective Gain = 33-60 life years Life4me Events* per 1000 Men Taking Aspirin

Age Benefit 50-59 Number Prevented

Risk Number Caused

Serious CVD Stroke Colon MI GI Ischemic Cancer Risk Bleeding

10% 22.5 20% 28.6

8.4 9.2

13.9 12.2

28.4 24.8

Strokes

Hemorrhagic

2.3 2.1

55 year old men, ≥ 2 risk factors : Risk of CVD death over next 20 years is ∼ 20%

Life4me Events* per 1000 Men Taking Aspirin Risk Age Benefit 50-59 Number Prevented Number Caused Serious CVD Stroke Colon MI GI Ischemic Cancer Risk Bleeding

10% 22.5 20% 28.6

8.4 9.2

13.9 12.2

28.4 24.8

Strokes

Hemorrhagic

2.3 2.1

With the exception of Aspirin, since 2005 All other NSAIDs have had a Black Box warning for cardiovascular risks.

Why ?

Indole Ace4c Acids Indomethacin, sulindac

tNSAIDS

Heteroaryl Ace4c Acids Diclofenac, ketorolac Aryl Ibuprofen, Naproxen, Propionic Acids

Flurbiprofen, Oxaprozin

Anthranilic Mefenamic acid Acids

Enolic Acids Pirox-, tenox-, melox-icam Alkanones Nabumetone

COX-1 ‘constitutive’; COX-2 inducible Inflammation induces COX-2 expression; COX-2 converts AA into PGE2 , worsening symptoms worse Knee Erythema Induced:COX-2 Edema AA Constitutive:COX-1 PGE2 Pain

Connective Tissue

∼ 1996

Are Selective COX-2 Inhibitors (Coxibs) The ‘Elusive’ Gastric Sparing NSAIDs?

∼300-fold selec4ve Rofecoxib Valdecoxib Celecoxib Diclofenac ∼10-30-fold selec4ve

Selec4vity is a rela4ve trait

‘

Ibuprofen Naproxen Ketorolac (parenteral)

More COX-2 SelecMvity More COX-1 SelecMvity

Arachidonic Acid

Joints ConnecMve Tissue COX-1

COX-2 Selec4ve COX-2 Inhibitors Rofecoxib PGE2 Celecoxib Pain

‘Normal’ InflammaMon

Gastric mucosa COX-1

PGE2 COX-1 sparing = ‘Normal’ Gastric sparing

Clinical Trials of Coxibs as Gastric Sparing NSAIDs

Rofecoxib

Celecoxib

VIGOR Vioxx GastrointesMnal Outcomes

CLASS Celecoxib Long-term ArthriMs Safety

Research

Study

RA (8076 enrolled) OA & RA (7968 enrolled) Comparator: Naproxen Comparators: Ibuprofen & (2 x 500 mg) Diclofenac Endpoint: Fewer Gastrointes4nal Events Favors rofecoxib ‘Trend’ favors celecoxib CV safety (MI) ‘favors naproxen’ CV safety (MI, etc) – equivalent 50 mg = 2x label dose Concomitant aspirin: NO

800mg = 2-4x label dose Concomitant aspirin:YES

Do Coxibs Confer Risk of Cardiovascular Toxicity? Q:CV Toxicity?

A:Aspirin precedent

APTC endpoint: Composite endpoint used in aspirin studies to get enough statistical power. MI, ischemic stroke, death from CV or unexplained causes.

After VIGOR & CLASS, four ongoing clinical trials provided a glimpse at cardiovascular safety of COXIBS. These were randomized, placebo-controlled, long-term trials for prevention of colorectal polyp recurrence & Alzheimer’s progression.

What did these trials show?

n

Rofecoxib APPROVe

Adenomatous Polyp Preven4on on Vioxx 25 mg

CV risk! RR 1.5

Celecoxib

APC

Adenoma Preven4on Celecoxib 400 mg

800 mg

ADAPT

PreSAP

Preven4on of Colorectal Sporadic Adenomatous Polyps Celecoxib 400 mg

*CV risk! CV risk vs placebo ‘1.0’ RR 3.4

Alzheimers Disease An4-inflammatory Preven4on Trial

CV risk vs. placebo

‘1.0’

Celecoxib 400 Naproxen 440

Imbalance confersAcid risk for occlusion Arachidonic

Blood Blood Vessels ‘Imbalance” Model Re-Visited platelets Endothelium GA FitzGerald Model COX-1 COX-1 COX-2 Selec4ve COX-2 Inhibitors Thromboxane Blood Flow in Coronary (or Prostacyclin Coxibs Cerebral) Vessels Reflects The Thrombotic Anti-thrombotic ‘Balance’ of Thrombotic & Atherosclerosis COX-1 sparing InflammaMon Mediators Antithrombotic in platelets

“Imbalance” Model & Coxib CV Risk NONE of the clinical trials mentioned was designed to test a model, nor ‘powered’ to reach statistically reliable conclusions about cardiovascular risks of coxibs (or tNSAIDs). The ‘imbalance’ model about cardiovascular risk of coxibs (or tNSAIDs) applies ONLY to ischemic thrombotic events (MI, stroke, death from MI, stroke ) = APTC endpoint

For patient safety each type of clinically relevant CV event matters INDEPENDENT of ‘Imbalance’ Model, FitzGerald Hypothesis, etc…

Does cardiovascular risk observed with rofecoxib & celecoxib indicate a coxib ‘class’ effect ?

This is a conceptual scheme, not data. High

Valdecoxib

Rofecoxib 2005 FDA ‘Yes’:

CV Events Other

Celecoxib

coxibs and all NSAIDs may Rofecoxib confer a cardiovascular risk. Base CVD line Risk

Where are CLASS, ADAPT, Pre-SAP?

Does cardiovascular risk observed High Rofecoxib in trials with rofecoxib & celecoxib Valdecoxib Celecoxib & valdecoxib validate the CV Events Rofecoxib ‘Imbalance’ (FitzGerald) Hypothesis ? Base l CVD ine Risk

Furthermore CABG I trial has been Much has been made a numerical ‘misinterpreted’. The trial of tested valdecoxib excessefficacy of adverse CV events the – analgesic (opioid sparingineffect) arm of CABG withoutto Ittreatment was not designed, nor 1, powered qualifyingcardiovascular that this is partly attributable evaluate safety. The only to randomization of patients the ratio significant ‘serious adverseinevent’ in 2 drug: 1 placebo. CABG 1 was sternal wound healing NOT cardiovascular events.

CABG 2 Trial Design

Post-CABG Surgery 3 days i.v. Parecoxib

7 days p.o. Valdecoxib

PLACEBO

Valdecoxib

PLACEBO

PLACEBO

It is imprudent/ inadvisable to draw any Mechanistic conclusions from Intent To Treat Analysis .

“Three deaths occurred among patients given placebo & valdecoxib…. these deaths occurred in patients who had not yet begun treatment with valdecoxib.”

Were are we now?

FDA decisions & announcements emerged from..

July 09, 2015 FDA strengthens warning that non-aspirin antiinflammatory drugs (NSAIDs) can cause heart attacks or strokes Retail Ibuprofen Naproxen

Dosing t1/2 = 2 hrs, 2-3 tablets, tid, qid t1/2= 14 hrs, 1-2 tablets,bid

Current NSAID class labeling implies that CV thrombotic risk is not substantial with short treatment courses. Some epidemiological studies conducted since 2005 suggest that there …patients should take is no, or minimal, latency period prior to the the lowest dose for the onset of CV thrombotic risk. Does the weight of evidenceshortest support reconsideration of advice period of time regarding the latency of CV thrombotic risk?

possible.

Yes= 14 No=11 Abstain= 0

Do the available data support a This is an important vote conclusion that naproxen has a because it means the FDA lower risk of CV thrombotic events will allow the PRECISION as compared to the other NSAIDs? Trial to proceed to completion Yes= 9 No=16 Abstain= 0

Prospective Randomized Evaluation of Celecoxib Integrated Safety Vs Ibuprofen Or Naproxen

•

High risk CV patients studied for first time.

• Full GI protection using a proton pump inhibitor.

RESULTS COMING SOON FINISH ∼ SUMMER •>50,000 patient-years exposure >> than2016 the meta• ASA permitted as indicated.

analysis of all prior trials comparing celecoxib to ibuprofen or naproxen.

• All CV, GI, & renal endpoints prospectively adjudicated.

Grateful thanks to Dr. John Dougherty, DO, FACOFP, FAOASM, FAODME Senior Associate Dean for Clinical Affairs & GME, Professor Clinical Education/GME for inviting me to speak. W. Joshua Cox, DO, FACOFP Chairman, Primary Care Medicine for helpful comments.