section name header

Info


A. Metabolism [1]

  1. Prostaglandins (PG) produced from arachadonic acid (AA)
    1. AA is contained in membrane phosphlipids and released by phospholipases
    2. AA metabolized from leukotrienes through lipoxygenases
    3. AA gives rise to PG and thromboxanes (TX)
  2. Key enzymes are cyclooxygenases (COX, prostaglandin synthetases or PGS) [19,28,38]
    1. Cyclooxygenase 1 (COX-1, PGS1) is a constitutively expressed "housekeeping" protein
    2. COX-1 also has protects gastric mucosa from acid damage [4]
    3. Cyclooxygenase 2 (COX-2, PGS2) is primarily induced by inflammatory stimuli
    4. COX-2 is constitutively expressed in brain and kidney, but its role is variable [29]
    5. COX-2 inhibition is usually the pharmacologically desired effect
    6. High potency COX-2 inhibition appears to have deleterious cardiovascular effects
    7. There may be a COX-3 isoenzyme in humans [28]
    8. COX-3 may be involved in synthesis of anti-inflammatory PG
  3. Metabolic Pathways
    1. PGS1 and PGS2 convert AA to PGG2
    2. PGG2 converted to PGH2
    3. PGH2 convereted to PGI2 (prostacyclin) by prostacyclin synthetase
    4. PGH2 converted to PGE2, PGF2, and PGD2 by isomerases
    5. PGH2 converted to thromboxane A2 (TXA2) by thromboxane synthetase
  4. Various Classes of PG Exist [28]
    1. Most have vasodilator or vasoconstrictor activity
    2. Some PG are pro- or anti-inflammatory
    3. Vascular dilatation primarily mediated by prostacyclin (PGI-2) and PGE2
    4. PGE2 is also major mediator of fever
    5. Vasoconstriction mediated by other prostaglandins
  5. PGs, Colonic Polyps, Colon Cancer
    1. Colonic polyp growth is believed to be dependent on cyclo-oxygenase 2 (COX2)
    2. Hyperproliferative colonic epithelium expresses elevated levels of COX2
    3. Various COX inhibitors reduce growth of colonic polyps
    4. Non-cyclo-oxygenase pathways may also be involved in NSAID effects
    5. Thus, some NSAIDs may block PPAR-delta transcription and/or activation of NFkB
    6. Use of ASA (>14 pills/week) or NSAIDs for >10 years associated with ~50% reduced colon cancer risk in women [59]

B. Specific PG and TX [1,38]

  1. PGE2 [40]
    1. Proinflammatory - induced by IL1ß, IL-6, TNFa
    2. Major mediator of fever induction in hypothalamus
    3. Potent vasodilators in vascular beds
    4. Reduction in blood pressure
    5. Increase in cardiac output
    6. Relax bronchial and tracheal smooth muscle
    7. Uterine smooth muscle relaxation
    8. Increased GI motility; causes diarrhea
    9. Protect gastric mucosa from acid damage
    10. Increased renal blood flow
    11. Receptors are EP1, EP2, EP3, EP4
  2. PGI2 (Prostacyclin)
    1. Important vasodilator
    2. About 5 fold more potent than PGE2
    3. Inhibits platelet aggregation
    4. Antithrombotic on vascular endothelium
    5. Bronchodilation
    6. Increased renal blood flow
    7. Receptor is IP
  3. PGD2
    1. Causes both vasodilation and vasoconstriction
    2. Vasodilation usually at lower doses than vasoconstriction
    3. Anti-inflammatory (particularly on mast cells)
    4. Receptors are DP1 and DP2
  4. PGF2
    1. Bronchoconstriction in asthmatics (with leukotrienes)
    2. Contraction of uterine smooth muscle
    3. Constriction of circular smooth muscle in GI tract
    4. Anti-inflammatory (PGF2a)
    5. Receptors are FP-alpha, FP-beta
  5. TXA2
    1. Potent vasoconstrictor
    2. Platelet aggregation and granule release
    3. Receptors are TP-alpha, TP-beta
  6. PG Receptors
    1. DP: PGD2 binds, increase cAMP; increased risk of asthma with specific variants [54]
    2. EP1: PGE and PGF2a bind; PIP2 hydrolysis
    3. EP2: PGE binds, cAMP increase
    4. EP3: PGE binds, variable cAMP response
    5. FP: PGF2a binds, PIP2 hydrolysis
    6. IP: PGI2 (and some PGE) binds, cAMP increased
    7. TP (nonplatelet): TXA2, PGH2 bind, increase PIP2 hydrolysis
    8. TP (platelet): TXA2, PGH2 bind, increase PIP2 hydrolysis

C. Prostaglandin Replacement Therapy

  1. Misoprostol (Cytotec®)
    1. Prostaglandin E2 analog
    2. Approved for prophylaxis against NSAID induced gastric / peptic ulcer
    3. ~40% reduction in serious GI bleeding at doses of 200µg qid po in patients on NSAIDS
    4. No more effective (more expensive) than standard treatment for gastric ulcer
    5. Superior to ranitidine or sucralfate for prevention of NSAID induced gastric ulcers
    6. Standard dose is 200µg po qid for gastric and duodenal ulcer prevention
    7. Lower doses (200µg bid-tid) are nearly as effective and better tolerated
    8. Major side effect is diarrhea; reduced incidence at doses of 100-200µg bid
  2. Epoprostenol (Flolan®) [6,7]
    1. Prostacyclin PGI2 analog
    2. Continuous iv prostacyclin improves survival to ~63% at 3 years
    3. Twelve weeks of PGI2 improved exercise capacity, RV pressures, quality of life
    4. Aerosolized PGI2 or Iloprost is very effective at reducing pulmonary pressures
    5. Aerosolized Iloprost is active for 60-120 minutes and causes little reduction in BP
    6. Increase in exhaled nitric oxide following iloprost is marker for efficacy of therapy
    7. IV Iloprost given through implanted central venous catheter and is well tolerated
    8. Long-term benefits: improved cardiac output and symptoms, and reduced mortality
    9. Annual Cost (all inclusive) for IV therapy at10ng/kg/min is about $58,000 per year
  3. Treprostinil (Remodulin®) [46]
    1. Prostacyclin analog
    2. Approved for chronic subcutaneous treatment at initially 0.625 up to 20ng/kg/min
    3. Generally well tolerated but all patients experience infusion site reactions
    4. Improves hemodynamics, symptoms in patients with severe P-HTN
    5. Jaw pain, diarrhea, flushing, lower extremity edema, gastrointestinal (GI) hemorrhage

D. Non-Steroidal Anti-Inflammatory Drugs (NSAIDs) [4,38]

  1. COX1 and COX2 Classification and Efficacy [19]
    1. Most older agents show preference for COX-1 over COX-2
    2. Celecoxib and rofecoxib are COX-2 selective (>100X selectivity)
    3. Valdecoxib, etoricoxib, lumiracoxib are >250X more selective for COX2 over COX1
    4. COX-2 inhibitors reduce polyps in familial adenomatous polyposis (FAP) [32]
    5. Long term NSAID use reduced risk of oral cancers (including in smokers) by >50% [60]
    6. COX-1/2 nonspecific and COX-2 specific equally effective for pain and inflammation
    7. Nonspecific or COX2 selective NSAIDs showed no reduction in progression of mild to moderate Alzheimer's Disease [2]
    8. Non-aspirin NSAIDs do not reduce the risk of cardiovascular disease [43]
    9. Effects of some NSAIDs go beyond COX inhibition and may include growth regulation [3]
    10. Side effects are discussed in more detail below
  2. Profen Group
    1. Ibuprofen (Motrin®, Advil® and others): 200-800mg q4 hrs as needed, OTC
    2. Ketoprofen (Orudis®, Oruvail®): 200mg po qd (may begin bid; now OTC)
    3. Flurbiprofen (Ansaid®): 100mg po bid
    4. Ibuprofen generally has lowest risk of upper GI complications
    5. Ketoprofen generally has highest risk of upper GI complications
    6. May be associated with slight (5-10%) increase risk of cardiac events [10]
  3. Naproxen (Naprosyn® and Naprelan®) and Naproxen Sodium (Anaprox®)
    1. Naprosyn 357-500mg and Anaprox® 550mg bid-tid; Naprelan 750-100mg qd
    2. Excellent for joint pain, usually from inflammatory arthritis
    3. Avoided in children as it may cause a pseudo-porphyria
    4. Over the Counter (OTC): 220mg (Aleve®, others) po bid-tid
    5. May be associated with slight (5-10%) increase risk of cardiac events [10]
  4. Nabumetone (Relafen®)
    1. Minimal specificity for blocking inflammatory (COX2) rather than gastric PG's (COX1)
    2. Whether this translates to a clinical benefit is not clear
    3. Dose is 1000-1500mg po qd (pill sizes are 500mg and 750mg)
    4. Some studies have shown relative safety of this agent, but this is unclear
  5. Diclofenac (Voltaren®, Cataflam®)
    1. Extended release forms available for bid dosing
    2. Usual dosing: 50-75mg bid-tid
    3. Mainly for rheumatoid and osteoarthritis, and other painful arthritis conditions
    4. Diclofenac 50/75mg + misoprostal 200µg (Arthrotec®) now available
  6. Oxaprozin (Daypro®)
    1. Good pain control with long half life (once daily dosing)
    2. May be safer than some of other agents with respect to chronic ulcer risk
    3. Dose is1200-1800mg qd usually as single dose (600mg po pill size)
    4. No preference for COX-2 over COX-1
  7. Sulindac (Clinoril®)
    1. Dose 150-200mg po bid
    2. Side effects no different than other NSAIDs at equivalent pain doses
    3. No reduction of polyps in patients with FAP [44]
    4. Moderate risk of ulceration (see below)
  8. Etodolac (Lodine®)
    1. Relatively more specific for inflammatory than gastric PG synthetases
    2. Dose is 300-400mg po bid
    3. May have some preference for COX-2 over COX-1
    4. Safety profile appears to be better than other agents
  9. Ketorolac (Toradol®)
    1. Narcotic strength NSAID may be used only short term for pain control
    2. Available im (or iv) as 15 or 30mg injections
    3. Oral form as 10mg pills, q6 hours up to 40mg po per day
    4. Significantly increased risk of bleeding with this NSAID compared with others [23]
    5. Ketorolac should not be used long term (>4 days)
  10. Indomethacin (Indocin®)
    1. Rapid acting, excellent for crystal arthropathy
    2. Dose 25-50mg tid to qid
    3. GI distress common; avoid in elderly
    4. CNS side effects (light-headedness, wooziness) are fairly common
  11. Non-acetylated Salicylates
    1. Efficacy is probably lower than that of other agents
    2. Choline Magnesium Salicylate (Trilasate®):1000mg tid
    3. Salicylsalicylic Acid (Salsalate®, Disalcid®): 500-1000mg po bid-tid
    4. Reversible platelet inhibition and very low GI side effects
    5. Can probably be used safely in patients on warfarin
  12. Aspirin (ASA)
    1. Enteric coated and extended release forms of ASA available
    2. Excellent for cardiovascular disease and/or stroke prevention (81-325mg qd)
    3. Both primary and secondary prevention in men and women is beneficial [12,18]
    4. Safe and beneficial within 48 hours of coronary artery bypass grafting [48]
    5. May reduce risk of colonic polyps or frank colonic neoplasia
    6. High doses 1000-1600mg tid-qid often required for anti-inflammatory activity
    7. Risk of bleeding higher than most other agents (due to platelet inhibition)
    8. No longer recommended for chronic use at doses >325mg/d
    9. ASA+esomeprazole (Nexium®) 20mg qd is safe in patients with previous ASA-induced peptic ulcer bleeding; recurrent bleeding rate 0.7% over 1 year [57]
    10. ASA at least 300mg qd x 5 years associated with 25% reduced CRC risk at 10 years [64]
    11. Regular use of ASA associated with ~35% reduction in risk of CRC expressing high levels of COX-2; no effect on CRC with weak or absent COX-2 [13]
  13. Celecoxib (Celebrex®) [19,20,24,38]
    1. COX-2 specific inhibitor with >7X preference for COX-2 over COX-1 in vitro
    2. FDA approved for use in chronic rheumatoid arthritis and osteoarthritis
    3. FDA also approved for reduction in polyposis in FAP [32]
    4. Ulcerogenic potential (even at very high doses) is similar to placebo [26,35]
    5. Contains a sulfur moiety; contraindicated in patients allergic to sulfonomides [21]
    6. Celecoxib is metabolized by CYP2C9 and inhibits CYP2D6 activity
    7. CYP2C9 inhibitors such as fluconazole, fluvastatin, zafirlukast may increase levels
    8. Dose 100-200mg bid for rheumatoid arthritis, 200mg qd or divided for osteoarthritis
    9. As effective as non-selective COX-2 inhibitors [27]
    10. May reduce risk of cardiac events ~20% [10,11], have no effect [47], or showed 2.3-3.4X increased risk in a single study [15]
  14. Etoricoxib (Arcoxia®) [63]
    1. Highly selective COX-2 inhibitor
    2. Effective in rheumatoid and osteoarthritis
    3. Cardiovascular risk similar to diclofenac
    4. Reduced total and uncomplicated, but not complicated, GI events versus diclofenac [22]
    5. Dose is 60-90mg po qd
  15. Valdecoxib (Bextra®) [45]
    1. COX-2 speciific inhibitor with ~30X preference for COX-2 over COX-1 in vitro
    2. FDA approved for osteoarthritis, rheumatoid arthritis, primary dysmenorrhea
    3. Dose is 10mg qd to 20mg bid
    4. Has been withdrawn from market
  16. Rofecoxib (Vioxx®) [19,25,38]
    1. COX-2 specific inhibitor with >100X preference for COX-2 over COX-1 in vitro
    2. Withdrawn from market due to increased cardiovascular risks [55,62]
    3. Also associated with increased renal and arrhythmia risks [61]
    4. These risks have not been observed with celebrex [10,11,55,61,62]

F. NSAID Side Effects

  1. Main Target Organs
    1. Gastrointestinal (GI) - mainly stomach and duodenum
    2. Renal - due to vasoconstrictive effects
    3. Cardiovascular - due to inhibition of prostacyclin production
  2. GI Risks [4]
    1. Risk of GI bleeding overall with NSAIDS is ~5X above baseline
    2. Translates to ~1-2% GI ulceration annually
    3. Ketorolac and piroxicam have highest risk; ibuprofen low risk, naproxen modest risk [23]
    4. Meloxicam has lower GI side effects than piroxicam or diclofenac [30]
    5. COX-2 specific are no more ulcerogenic than placebo [4,26,27]
    6. COX-2 specific inhibitors are as effective for pain relief as nonselective agents with 50-75% fewer GI side effects [8,22,27]
    7. COX-2 inhibitors have the same effects on the kidney as nonselective NSAIDs [33]
  3. Prophylaxis against Upper GI Events [4]
    1. Age is a major risk factor for bleeding
    2. Switching to COX-2 selective NSAIDs will prevent most GI ulceration and reduces the need for adjunctive GI medications [49]
    3. Misoprostol, double-dose H2-blockers, and PPI are equally effective [37]
    4. Adding a proton pump inhibitor (PPI) to standard NSAID reduces risk of GI side effects to level with COX-2 specific inhibitors [42]
    5. Standard dose H2-blockers are not effective
    6. All of these agents add expense to therapy (up to $80-100 per month)
    7. Note that 7-30 patients need to be treated to prevent one endoscopic ulcer [37]
    8. Prophylaxis recommended in patients with ulcer history, cardiac disease, age >75 years
    9. Misoprostol dose is usually 200µg qid; lower doses may be effective (see above)
    10. Diclofenac + misoprostal (Arthrotec®) is now available for chronic use
    11. Consider H. pylori eradication prior to initiating chronic NSAID therapy
    12. Prophylaxis is not required with COX-2 specific inhibitors
    13. Use of nitrovasodilator (nitrates) drugs 40% reduced risk of GI bleeding [34]
  4. Treatment of GI Toxicity from NSAIDs
    1. Proton Pump Inhibitors (PPI) are most effective and/or best tolerated of all agents
    2. PPI include omeprazole (Prilosec®) and lansoprazole (Prevacid®)
    3. PPI heal NSAID induced ulcers and erosions within 4-8 weeks, even on NSAIDs
    4. PPI + diclofenac has similar PUD risk as celecoxib with 6 months' chronic use [42]
    5. Misoprosol (Cytotec®): PGE compound effectively reduces gastric acid secretion
    6. PPI better tolerated than misoprostal and equal for NSAID induced ulcers
    7. PPI more effective at healing ulcers than usual dose ranitidine (Zantac®)
    8. High dose (2X normal) H-2 blockers are required if used instead of PPI
    9. Avoid NSAIDs in all patients on warfarin
    10. Non-acetylated salicylates or COX2 specific inhibitors appear to be safe with warfarin
    11. Addition of nitric oxide moiety to standard NSAID molecule may reduce ulcers [4,19]
  5. NSAIDs and Helicobacter pylori [52]
    1. H. pylori infection nearly doubles the risk of bleeding with non-selective NSAIDs
    2. H. pylori eradication may reduce risk of NSAID induced PUD
    3. Consider H. pylori eradication before starting chronic NSAID therapy (in high risk)
    4. Effects of COX-2 specific inhibitors on H. pylori are not clear
  6. Risk Factors for NSAID-Associated Gastroduodenal Ulcers [4]
    1. Advanced age (linear increase)
    2. History of ulcer
    3. Concomitant use of glucocorticoids
    4. Higher doses of NSAIDs / use of >1 NSAID at a time
    5. Concomitant administration of warfarin or other anticoagulants
    6. Serious underlying systemic disorder, especially liver disease
    7. Concomitant infection with H. pylori may be a risk factor (see above)
    8. Cigarette smoking and alcohol consumption may be risk factors
    9. Ketorolac likely has the highest risk for serious GI bleeding
  7. Cardiovascular (CV) Risks
    1. Some concern with overall NSAIDs which have shown up to 2X increased CV risk with long term use [60]; meta-analysis suggests no overall risk with non-selective NSAIDs [5]
    2. Nonselective NSAIDs have shown 5-10% increased cardiovascular risks compared with no NSAID use or with celecoxib [10,11] but not in meta-analysis [5]
    3. Increased risk (1.8-3.6X) of cardiac events with >25mg/day rofecoxib (Vioxx®) [10,11,17,39,41,47]
    4. Modest increased risk (1.4-1.9X) of cardiac effects with 25mg/day rofecoxib [10,11,14]
    5. Rofecoxib or non-selective NSAID use increase risk of CHF admission 1.8X versus use of celecoxib or no NSAID use [53,55]
    6. Rofecoxib withdrawn from market due to 2-5X increased cardiovascular risks [17,55,56]
    7. Increased cardiovascular risks with rofecoxib (1.3-2.2X) and diclofenac (1.4X risk) [62]
    8. No increase in cardiovascular risk with naproxen, piroxicam or ibuprofen [5,62]
    9. No significant increased cardiac risk with lumiracoxib [8,9] or etoricoxib [63]
    10. Paracoxib and valdecoxib have shown increased adverse events after cardiac surgery [15]
    11. Celecoxib has generally shown protective or no cardiovascular risk [10,11,17,47] or
  8. 3-3.4X risk in a single study [15]
    1. Increased bleeding risk with aspirin is generally not seen with other agents
    2. Ibuprofen for 7 days has mild antiplatelet effects which disappear after 24 hours [58]
  9. NSAIDs and Blood Pressure
    1. NSAIDs may also increase blood pressure
    2. Can antagonize effects of HTN medications
  10. Renal Insufficiency
    1. Inhibition of prostaglandin synthesis at efferent arteriole leads to GFR
    2. Interstitial nephritis with with acute renal failure (ARF) can also occur [51]
    3. Oliguria, Acute Renal Failure, hypertension (HTN)
    4. Renal problems are especially prevalent in patients with underlying renal disease
    5. Avoid chronic NSAID use in patients on ACE inhibitors, especially elderly
    6. Relative contraindications: diabetes, dehydration, myeloma, renal artery stenosis
    7. COX-2 specific inhibitor renal side effects similar to nonselective agents

References

  1. Hardman JG and Limbird LE, editors. 1996. Pharmacologic Basis of Therapeutics. pp 601 ff
  2. Aisen PS, Schafer KA, Grundman M, et al. 2003. JAMA. 289(21):2819 abstract
  3. Huls G, Koornstra JJ, Kleibeuker JH. 2003. Lancet. 362(9379):230 abstract
  4. Wolfe MM, Lichtenstein DR, Singh G. 1999. NEJM. 340(24):1888 abstract
  5. Salpeter SR, Gregor P, Ormiston TM, et al. 2006. Am J Med. 119:552 abstract
  6. Epoprostenol. 1996. Med Let. 38(968):14 abstract
  7. McLaughlin VV, Genthner DE, Panella MM, Rich S. 1998. NEJM. 338(5):273 abstract
  8. Schnitzer TJ, Burnmester GR, Mysler E, et al. 2004. Lancet. 364(9435):665 abstract
  9. Farkouh ME, Kirshner H, Harrington RA, et al. 2004. Lancet. 364(9435):675 abstract
  10. Graham DJ, Compen D, Hui R, et al. 2005. Lancet. 365(9458):475 abstract
  11. Kimmel SE, Berlin JA, Reilly M, et al. 2005. Ann Intern Med. 142(3):157 abstract
  12. Berger JS, Rongcaglioni MC, Avanzini F, et al. 2006. JAMA. 295(3):306 abstract
  13. Chan AT, Ogino S, Fuchs CS. 2007. NEJM. 356(21):2131 abstract
  14. Bresalier RS, Sandler RS, Quan H, et al. 2005. NEJM. 352(11):1092 abstract
  15. Solomon SD, McMurray JJV, Pfeffer MA, et al. 2005. NEJM. 352(11):1071 abstract
  16. Nussmeier NA, Whelton AA, Brown MT, et al. 2005. NEJM. 352(11):1081 abstract
  17. Levesque LE, Brophy JM, Zhang B. 2005. Ann Intern Med. 142(7):481 abstract
  18. Ridker PM, Cook NR, Lee IM, et al. 2005. NEJM. 352(13):1293 abstract
  19. Hawkey CJ. 1999. Lancet. 353(9149):307 abstract
  20. Celecoxib. 1999. Med Let. 41(1045):11 abstract
  21. Celecoxib for Arthritis. 1999. Med Let. 41(1048):28
  22. Laine L, Curtis SP, Cryer B, et al. 2007. Lancet. 369(9560):465 abstract
  23. Garcia Rodriguez LA, Cattaruzzi C, Troncon MG, et al. 1998. Arch Intern Med. 158(1):33 abstract
  24. Simon LS, Weaver AL, Graham DY, et al. 1999. JAMA. 282(20):1921 abstract
  25. Langman MJ, Jensen DM, Watson DJ, et al. 1999. JAMA. 282(20):1929 abstract
  26. Emery P, Zeidler H, Kvien TK, et al. 1999. Lancet. 354(9196):2106 abstract
  27. Feldman M and McMahon AT. 2000. Ann Intern Med. 132(2):134 abstract
  28. Willoughby DA, Moore AR, Colville-Nash PR. 2000. Lancet. 355(9204):646 abstract
  29. Lipsky PE, et al. 2000. Arch Intern Med. 160:913 abstract
  30. Meloxicam. 2000. Med Let. 42(1079):47 abstract
  31. Janne PA and Mayer RJ. 2000. NEJM. 342(26):1960 abstract
  32. Steinbach G, Lynch PM, Phillips RKS, et al. 2000. NEJM. 342(26):1946 abstract
  33. Swan SK, Rudy DW, Lasseter KC, et al. 2000. Ann Intern Med. 133(1):1 abstract
  34. Lanas A, Bajador E, Serrano P, et al. 2000. NEJM. 343(12):834 abstract
  35. Silverstein FE, Faich G, Goldstein JL, et al. 2000. JAMA. 284(10):1247 abstract
  36. Bombardier C, Laine L, Reicin A, et al. 2000. NEJM. 343(21):1520 abstract
  37. Rostom A, Wells G, Tubwell P, et al. 2000. J Rheumatol. 27:2203 abstract
  38. FitzGerald GA and Patrono C. 2001. NEJM. 345(6):433 abstract
  39. Mukherjee D, Nissen SE, Topol EJ. 2001. JAMA. 286(8):954 abstract
  40. Aronoff DM and Neilson EG. 2001. Am J Med. 111(4):304 abstract
  41. Cardiovascular Safety of Cox-2 Inhibitors. 2001. Med Let. 43(1118):99 abstract
  42. Chan FKL, Hung LCT, Suen BY, et al. 2002. NEJM. 347(26):2104 abstract
  43. Ray WA, Stein CM, Hall K, et al. 2002. Lancet. 359(9301):118 abstract
  44. Giardiello FM, Yang VW, Hylind LM, et al. 2002. NEJM. 346(14):1055
  45. Valdecoxib. 2002. Med Let. 44(1129):39 abstract
  46. Treprostinil. 2002. Med Let. 44(1139):80 abstract
  47. Ray WA, Stein CM, Daugherty JR, et al. 2002. Lancet. 360(9339):1071 abstract
  48. Mangano DT. 2002. NEJM. 347(17):1309 abstract
  49. Lisse JR, Perlman M, Johansson G, et al. 2003. Ann Intern Med. 139(7):539 abstract
  50. Buvanendran A, Kroin JS, Tuman KJ, et al. 2003. JAMA. 290(18):2411 abstract
  51. Singh AK and Colvin RB. 2003. NEJM. 349(21):2055 (Case Record) abstract
  52. Papatheodoridis GV, Papadelli D, Cholongitas E, et al. 2004. Am J Med. 116(9):601 abstract
  53. Mamdani M, Juurlink DN, Lee DS, et al. 2004. Lancet. 363(9423):1751 abstract
  54. Oguma T, Palmer LJ, Birben E, et al. 2004. NEJM. 351(17):1752 abstract
  55. Celebrex and Vioxx. 2004. Med Let. 46(1194):87 abstract
  56. Juni P, Nartey L, Reichenbach S, et al. 2004. Lancet. 364(9450):2021 abstract
  57. Chan FKL, Ching JYL, Hung LCT, et al. 2005. NEJM. 352(3):238 abstract
  58. Goldenberg NA, Jacobson L, Manco-Johnson MJ. 2005. Ann Intern Med. 142(7):506 abstract
  59. Chan AT, Giovannucci EL, Meyerhardt JA, et al. 2005. JAMA. 294(8):914 abstract
  60. Sudba J, Lee JJ, Lippman SM, et al. 2005. Lancet. 366(9494):1359 abstract
  61. Zhang J, Ding EL, Song Y. 2006. JAMA. 296(13):1619 abstract
  62. McGettigan P and Henry D. 2006. JAMA. 296(13):1633 abstract
  63. Cannon CP, Curtis SP, FitzGerald GA, et al. 2006. Lancet. 368(9549):1171
  64. Flossmann E and Rothwell PM. 2007. Lancet. 369(9573):1603 abstract