Blood Pressure Destabilization on Non-steroidal Anti-Inflammatory Agents: Acetaminophen Exposed?

Non-steroidal anti-inflammatory drugs and blood pressure

It has been conventional wisdom for some time now that non-steroidal anti-inflammatory drugs (NSAIDs) - whether selective for inhibition of cycloxygenase 2 (COX-2) or non-selective - increase blood pressure (BP) or interfere with BP control while acetaminophen should not have this effect (1–4). Controlled clinical trials have shown that NSAIDs have heterogeneous effects on clinic and 24-hour BP in normotensive and hypertensive subjects (3). The inhibition of COX-2 by NSAIDs results in decreased actions of both vasodilatory and natriuretic prostaglandins (3). In most individuals, homeostasis of plasma volume is reestablished by small, nearly undetectable increases in BP (5) while in patients with impaired excretory function, more substantial volume retention occurs that is maybe associated with hypertension, edema, and congestive heart failure (3,5).

Prior to 2002, little was known about the effects of NSAIDs on 24-hour BP in patients with arthritis who also had hypertension and/or vascular diseases (6–8). Because ambulatory BP has improved reproducibility compared to clinic measurements, detection of small, but clinically meaningful differences in drug treatment groups (9,10) is more likely to occur. Ambulatory monitoring also allows for improved evaluation of pharmacodynamic effects of any drug, including the NSAIDs (6, 7). For example, in the Celecoxib Rofecoxib Efficacy and Safety in Comorbidities Evaluation Trial (CRESCENT), destabilization of systolic BP control occurred for approximately 18 of the 24 hours post-dosing of rofecoxib at standard osteoarthritis doses of 25 mg each morning (8). In contrast, there were no changes from baseline in 24-hour systolic BP observed with celecoxib nor naproxen at their respective, standard dosing regimens for osteoarthritis. Of the NSAIDs studied with 24-hour BP monitoring and widely used in clinical practice, ibuprofen, indomethacin and etoricoxib appear to cause the largest increases in BP, while celecoxib and naproxen seem to induce smaller increases in BP and destabilization of BP control (Table 1).

Table.

EFFECTS OF VARIOUS NON-STEROIDAL ANTI-INFLAMMATORY AGENTS ON AMBULATORY BLOOD PRESSURE

Study (ref)	Design and Trial Duration	Patient Population	N	Age (years)	Study drugs	Baseline 24-hour BP (mmHg)	Change in 24-hour BP (mmHg)
Izhar M (7)	Randomized, cross-over Active controlled, 2 weeks	Hypertensive non-arthritis	25	58	Celecoxib 200 mg BID	129/80	1.6/1.9
				58	Diclofenac 75 mg BID	129/79	4.2/3.0
MacDonald T (24)	Parallel, double-blind, active controlled, 6 weeks	Osteoarthritis hypertensive	787	65	Lumiracoxib 100 mg QD	127/74	−2.7/−1.5
				64	Ibuprofen 600 mg TID	127/74	2.2/ 0.5
Morgan TO (25)	Parallel, randomized, double-blind, cross-over, 6 weeks	Hypertensives non-arthritis	41	69	Amlodipine + indomethacin 50 mg BID	141/77	1.0/0.0
				72	Enalapril + indomethacin 50 mg BID	134/73	12.0/5.0
Polonia J (26)	Randomized, cross-over, single-Blind, 1 week of NSAID	Hypertensives non-arthritis	18	53	Nifedipine + indomethacin 75 mg	135/88	0.3/0.6
				53	Enalapril + indomethacin 75mg	135/87	6.8/1.3
Schwartz J (27)	Parallel, double-blind, placebo-, and active-controlled, 15 days	Elderly normal controlled diet	85	66	Etoricoxib 90 mg QD	NR	7.7/3.2
				65	Celecoxib 200 mg BID	NR	2.4/1.1
				67	Naproxen 500 mg BID	NR	3.6/1.4
				66	Placebo	NR	−2.4/−0.8
Sowers J & White WB (8)	Parallel, double-blind, active-Controlled, 6 and 12 weeks	Osteoarthritis, Hypertension, Type 2 DM	404	64	Rofecoxib 25 mg QD	132/76	4.2/1.5
				62	Celecoxib 200 mg QD	132/76	−0.1/−0.1
				64	Naproxen 500 mg BID	134/76	−0.8/−1.0
Sudano F & Flammer AJ (15)	Randomized, double-blind cross-over, placebo-controlled, 2 weeks	Coronary artery disease	33	61	Acetaminophen 1 g TID	122/73	2.9/2.2
				61	Placebo	123/74	−0.5/0.2
White WB (6)	Parallel, double-blind, placebo-Controlled, 4 weeks	Hypertensive on ACE inhibitor	178	55	Celecoxib 200 mg BID	135/84	2.6/1.5
				53	Placebo	131/82	1.0/0.3

Bolded typeface depicts values that are statistically greater than comparator; QD – once daily, BID – twice daily, TID – three times daily; NR – not reported

In addition to interfering with BP control, NSAIDs at high doses have been shown to increase cardiovascular event rates in placebo-controlled studies of patients with colonic polyps and following coronary artery bypass surgery (11–14). Due to concerns for these potentially negative cardiovascular effects of the NSAIDs, acetaminophen has typically been suggested as a safer alternative for initial therapy (2, 4) in patients with osteoarthritis and co-morbid cardiovascular disorders.

In this issue of Circulation, the findings of Sudano and Flammer and co-workers (15) cast some doubt on the cardiovascular safety of acetaminophen, at least from the perspective of BP. The investigators evaluated the effects of acetaminophen at a standard osteoarthritis dose (1 gram three times daily) on ambulatory BP, a variety of serum biomarkers and platelet and vascular function in 33 patients with known coronary artery disease using a randomized, double-blind, placebo-controlled cross-over design. Even with truncated treatment phases of 2 weeks (the investigators wished to minimize exposure time to acetaminophen since these subjects had no pain indication), acetaminophen induced statistically significant increases in both mean 24-hour systolic and diastolic BP from baseline compared with placebo (approximately 3/2 mmHg). This increase in 24-hour mean BP is not dissimilar to changes observed with many of the NSAIDs (Table 1). Additionally, a small but significant increase in 24-hour heart rate (2 beats/minute) occurred on acetaminophen relative to placebo.

A comprehensive attempt by the investigators to obtain mechanistic data associated with the effect of acetaminophen on BP was inconclusive. Small relative reductions from baseline in brachial artery flow mediated dilation (FMD) were noted in the acetaminophen arm (− 5% compared to placebo) but were not statistically different between acetaminophen and placebo; nor was there a significant difference in endothelial-independent FMD. In addition, the investigators found no effects of acetaminophen on platelet adhesion, the proportion of endothelial progenitor cells, concentrations of highly sensitive C-reactive protein, plasma or urinary concentrations of PGE₂ or thromboxane B₂. Safety serum chemistries showed a statistically significant increase in the hepatic enzyme, gamma-glutamyltransferase on acetaminophen treatment compared to placebo.

What do we know about acetaminophen’s mechanism that could lead to a blood pressure destabilizing effect?

While one of the most widely used analgesics world-wide, the mode of action for relieving pain by acetaminophen has been poorly understood for nearly a century but recently several interesting hypotheses have been reported (16–20). It is questionable whether these novel mechanisms of action that may convey pain relief by acetaminophen could theoretically translate to a pressor effect in susceptible individuals. Acetaminophen is an indirect COX-2 selective inhibitor that acts via a peroxidase site on prostaglandin H₂ synthetase – 2 (PGHS-2) and reduces the conversion of arachidonic acid to prostaglandin-H₂ (PGH₂) (16). It is not known whether this directly affects vascular resistance and BP. Another carefully described mechanism of action for analgesic benefits of acetaminophen is the indirect activation of the cannabinoid receptors. N-arachidonoyl phenylamine, a metabolite of acetaminophen, has been shown to indirectly stimulate endogenous cannabinoid release (17, 18). However, stimulation of cannabinoid release seems to modest lower BP, not increase it (19). Finally, acetaminophen and other NSAIDs have been shown to inhibit N-methyl-D-aspartic acid (NMDA) receptors, which play a role in pain neurotransmission (20) and in vasodilation (21). When stimulated, NMDA receptors release nitric oxide as a neurotransmitter in the spinal cord. The release of nitric oxide may modulate arachidonic acid metabolism by altering cyclooxygenase activity [21].

Where does acetaminophen fit in the management of arthritis and pain in patients with cardiovascular disease?

The data reported by Sudano and Flammer et al (15) imply that development of asymptomatic hypertension or destabilization of treated, controlled hypertension could occur in some patients with heart or vascular disease when treated with standard chronic doses of acetaminophen. However, the interpretation of most other findings of the study, with the exception of platelet adhesion, are limited by its brevity - with only 2 weeks of exposure to acetaminophen, results for the other biomarkers and functional assessments studied can not be considered conclusive. Additionally, while the study did demonstrate a significant increase in BP after only 2 weeks of therapy, it is possible that longer-term administration of acetaminophen could induce more substantial increases in BP then was observed here. In the future, it will be important to explore the effects of acetaminophen in a broader population that might include patients with higher levels of BP, arthritis and/or chronic pain, mild-to-moderate renal disease and those on various antihypertensive drug classes. The narrow patient population studied limits our ability to apply these results to the general population.

It is not likely that the clinical status of acetaminophen will be effectively changed in the short-term by the results of this new study (15). However, clinical cardiologists should be aware that while acetaminophen is relatively effective for some forms of minor pain, it is quite ineffective for inflammatory diseases, such as moderate to advanced osteoarthritis or rheumatoid arthritis (1). In most clinical trials of these latter patient populations, it has not been proven to be superior to placebo. In addition, the hepatic safety of acetaminophen, particularly in doses > 3 grams/day, has been under great scrutiny recently as the drug is implicated as one of the most common causes for liver transplantation in the United States (22).

In conclusion, despite broad recommendations to use acetaminophen as first-line therapy for pain and arthritis in patients with heart and vascular disease (2, 23), the agent is simply not that effective and as supported by new findings by Sudano and Flammer et al (15), a lot more is unknown about this drug from a cardiovascular safety perspective than we know about the conventional NSAIDs and selective COX-2 inhibitors (3,4).