#NephJC Chat
Tuesday March 8th 9 pm Eastern
Wednesday March 9th 9 pm IST 3:30 pm GMT
Circulation 2022 Feb 8;145(6):416-423.
doi: 10.1161/CIRCULATIONAHA.121.056015. Epub 2022 Feb 7.
Regular Acetaminophen Use and Blood Pressure in People With Hypertension: The PATH-BP Trial
Iain M MacIntyre, Emma J Turtle, Tariq E Farrah, Catriona Graham, James W Dear, David J Webb, PATH-BP (Paracetamol in Hypertension–Blood Pressure) Investigators*
PMID: 35130054
Introduction
Pain is the commonest reason to see a doctor. Even Celsus, just over 2 centuries ago, described pain (dolor) as one of the five pillars of inflammation. Hence, anti-inflammatory medications make a lot of sense to use for pain, chief among them non-steroidal anti-inflammatory drugs (NSAIDs). Of course NSAIDs are notorious in nephrology, having been known for the last few decades to cause various sorts of kidney trouble, and we have discussed them before on NephJC (PRECISION trial, Nelson et al; and the CRIC study). In all this discussion, acetaminophen (as it is known in North America, Colombia, Venezuela, Iran, and Japan) or paracetamol (rest of the world) has not been discussed much, since it is thought to be the safest of the lot - with little deleterious effects (outside of overdose causing hepatotoxicity). Interestingly enough, it is toxic to cats (who lack the glucuronidation enzyme to detoxify it), who die due to methemoglobinemia. A closely related predecessor of acetaminophen, acetanilide, also had a tendency to do this in humans, and is not used as a result. Also of interest, phenacetin, which was thought to be the major culprit behind analgesic nephropathy, is also closely related to acetaminophen. Though all three (acetanilide, phenacetin and acetaminophen) were discovered in the 1880s, it was only around 1950 when acetaminophen’s relative safety brought it to the forefront. It has been available over the counter since the late 1950s or so.
Acetaminophen is the first step of the WHO analgesic ladder for treatment of cancer pain, and now it is the most commonly used analgesic in the world. We have been prescribing it at therapeutic doses (up to 4 grams per day) without major concern for side effects for decades. A few initial trials (Saragiotto et al, Cochrane Database of Systematic Reviews 2016) were done to test the efficacy of acetaminophen, and they suggested that it actually had a limited benefit in treatment of acute or chronic pain. Despite its limited proven efficacy, the perceived safety has led to its use as a first line drug for pain, with sales exceeding $2 billion in 2016, and some estimates that a quarter of Americans take some acetaminophen every week.
Because of its widespread use, concerns have been slowly increasing about its safety with long-term use, especially its effect on blood pressure (BP). The uncertainty about its safety stems from the fact that initial evidence was based on cohort studies, observational data, and a few small RCTs with conflicting results (Dedier et al, Hypertension 2002, Curhan et al, Arch Intern Med 2002, Dawson et al, J Hypertens 2013). Because expectations of the drug’s benefit are decreasing, while concerns of it’s adverse effects are increasing, there has been a need for better quality data to address its safety in long-term use.
In order to address this need, this study was done to compare the effect of acetaminophen versus placebo on blood pressure in matched individuals with hypertension (HTN). It is a randomized, double-blind, cross over trial called Paracetamol Treatment in Hypertension–Blood Pressure (PATH-BP).
The Study
Design
PATH-BP was a randomized, double-blinded, placebo-controlled crossover study completed at a single center in the United Kingdom. It analyzed the effect of acetaminophen treatment on blood pressure in individuals with hypertension during a 2-week period.
Intervention
Participants were randomized to receive 2 weeks of acetaminophen 1 gram four times daily or a matched placebo, followed by a 2-week washout period. After which, the participants were crossed to the other treatment group for another 2 weeks. Both researchers and participants were blinded to treatment during the trial. Study drug shapes were identical, and importantly, the acetaminophen preparation had little sodium (0.04 g per tablet), unlike some effervescent formulations.
Study Population
Inclusion Criteria:
≥18 years of age and hypertensive, defined as either:
A) Treated for HTN with average of daytime ambulatory BP <150/95 mm Hg, on stable doses of ≥1 antihypertensive medication; or
B) untreated with average of daytime ambulatory BP ≥135/85 mm Hg but <150/95 mm Hg.
Exclusion Criteria:
History of ischemic heart disease or heart failure
History of cerebrovascular disease
Liver impairment (ALT >50 IU/L)
Chronic kidney disease Stage III to V
Suicidal ideation
Weight <55 Kg
Regular treatment with acetaminophen, NSAIDs, steroids, or oral anticoagulants
Recruitment: Local ambulatory BP clinics, general practices, and the Scottish health research register.
Measurements
During each of the 2 weeks of either intervention (2 weeks of treatment, 2 weeks of placebo) participants attended four follow up visits. Two long visits on days 0 and 14 where clinic BP was recorded (after 10 minutes of rest, 3 readings of non-dominant hand, the average of the 2nd and the 3rd read was recorded), a 24-hour ambulatory BP monitor (ABPM) was fitted, and blood samples (for urea, electrolytes, liver function test, and acetaminophen concentration) were taken. In addition, two short visits on days 4 and 7 where clinic BP and blood samples were taken.
Study Outcomes
Primary endpoint
The primary endpoint was a change in mean daytime systolic ambulatory BP after 2 weeks of treatment with acetaminophen vs placebo.
Secondary endpoints
Change in mean daytime diastolic BP
Change in systolic 24-hour or Diastolic 24-hour BP
Change in clinic BP
Sample Size and Statistical Analysis
The study was designed to detect a difference of 1.6 mmHg in systolic BP between acetaminophen and placebo groups. It was estimated that a sample of 110 patients would be needed to detect such difference using a 2-sided paired student t test with 90% power and 5% level of significance.
Analyses were based on a modified intention to treat population. The primary and secondary BP data were analyzed using a mixed model where treatment, period, and baseline BP were fitted as fixed effects and the participant as a random effect. Another analysis, a per-protocol analysis was performed according to adherence with treatment which was based on serum acetaminophen levels (<3mg/L in placebo group). Each of the comparisons in the trial was considered significant if p < 0.05.
Funding
The study was funded by a grant from the charity, British Heart Foundation.
Results
Study Population
The trial was conducted between September 2014 and June 2019. Out of 204 screened local participants, 110 participants were randomized into the study. Seven participants did not complete both arms of the study, so 103 participants were included in the modified intention to treat analysis and 90 participants (based on acetaminophen assay results) were included in the per-protocol analysis. See Figure 1 below.
Figure 1. Flow Diagram of the Study.
Patient Characteristics
The baseline characteristics are shown in table 1, according to which arm the participants were randomized to. Being a crossover study, all of them received acetaminophen and placebo, the randomization referring merely to the order. Overall, all of the participants were white, they were mostly men (80%) with a mean age around 60 years, of whom 70% were already on an antihypertensive drug.
Table 1. Baseline Characteristics.
Primary Endpoint
Table 2. Change in BP After Acetaminophen and Placebo: Modified Intention-to-Treat Analysis.
There was a statistically significant increase in mean daytime systolic ambulatory BP of 4.7 mm Hg (95% CI, 2.9-6.6; P<0.0001) with acetaminophen compared with placebo. (Table 2).
There was no difference between those who were on antihypertensive medications, or not.
Secondary Endpoints
Table 3. Change in BP After Acetaminophen and Placebo: Per-Protocol Analysis.
Similar to the changes in the primary outcome, even the mean daytime diastolic BP, and the 24 hour BPs were higher with acetaminophen compared to placebo.
The per-protocol analysis, which excluded the 13 patients with undetectable acetaminophen levels in the blood, also showed similar differences (Table 3).
Clinic BP
There was a statistically significant increase of 4.6 mm Hg in systolic BP with acetaminophen compared with placebo and an increase of 1.6 mm Hg in diastolic BP. Systolic BP increase was significant in the per protocol analysis (4.4 mm Hg) but there was no significant change in diastolic BP in the clinic (Table 3). From the supplementary data, this change was apparent as early as day 4 and had peaked by about day 7.
Biochemical Parameters
Table 4. Laboratory Values Before and After Acetaminophen and Placebo
There were no significant changes except for a small rise (still within normal range) in ALT during acetaminophen therapy, which normalized within 2 weeks of stopping the drug.
Adverse Events
Three adverse events reported:
Two events; Atrial Fibrillation and Myocardial infarction - not related to acetaminophen.
Exceeding the predefined safety stopping criteria for BP (clinic BP >180/110), happened on day 14 of acetaminophen treatment, after discontinuation of acetaminophen the participant’s BP normalized, because he did not complete 4 ABPM – excluded from the study.
Discussion
The PATH-BP trial provides clear evidence that acetaminophen raised BP during a 2-week period when compared to placebo in people with hypertension (treated or untreated) by about 5/2 mm Hg. Even such a small change in BP, if sustained over time, can have important effects on cardiovascular and cerebrovascular disease. In the BPLTTC meta-analysis (Ettehad et al, Lancet 2016), a 10 mm Hg lowering in systolic BP was associated with ~ 30% relative risk reduction (RRR) in stroke, ~ 20% RRR in heart failure, ~ 10% RRR in mortality (though no significant change in kidney failure). The impact of a 5 mm Hg lowering can be seen in the metaregression plot below.
Due to the widespread use of acetaminophen (3-5% of US adult population, up to 8% of patients with newly diagnosed HTN), the 4.7 mm Hg placebo-corrected rise could have considerable consequences on the population as a whole.
Other studies preceding this have raised this concern. The Nurses Health Study II (Curhan et al, Arch Int Med 2002; 80,030 Participants), found an association between regular acetaminophen use and HTN with a relative risk of developing HTN of 2 (near that of NSAIDS; RR of 1.9). A subsequent analysis of the same study suggested a possible dose-response relationship, with increasing dose of acetaminophen independently increasing the risk of HTN in women (Forman et al, Hypertension 2005). However, concerns about selection bias and residual confounding always remain since these individuals are taking acetaminophen for a reason (e.g. arthritis and back pain closely associated with age, obesity, and hypertension). Other prospective interventional studies have been generally limited by size and poor design. The largest and best designed prior study had only 33 participants with coronary artery disease, and reported that a 2-week treatment with 3 g daily acetaminophen significantly increased systolic BP 3-mm Hg compared with placebo (thus results keeping with the present study). The results of this study remarkably also are congruent with the effects of other NSAIDs on BP. In a meta-analysis of 38 RCTs, a 5 mm Hg higher systolic BP was reported (Johnson et al, Annals Int Med 1994), with a greater effect with piroxicam and indomethacin (6 to 7 mm Hg), and lower with ibuprofen and sulindac (~ 1 to 2 mm Hg).
Limitations
The study was done in participants with existing HTN - and without pain or fever and a clear indication for acetaminophen itself. Hence, it is not clear whether findings can be extrapolated to individuals who are not hypertensive at baseline. However, the rates of chronic pain and HTN increase with age, so it is expected that a considerable portion of individuals with chronic pain will have HTN as well.
This was a single center study, albeit well conducted and reported.
It is unclear whether the increase in BP with acetaminophen 2-week treatment will be sustained after the 2 weeks in people taking long term therapy.
The study was performed in a white population only, therefore it’s unknown whether these differences would be observed, and at the same magnitude in other populations also.
What could be the possible mechanism of acetaminophen raising BP? One possibility is that it is indeed similar to classical NSAIDs. Acetaminophen also inhibits cyclooxygenase (COX)-2 enzyme, but is mostly thought to do that in the central nervous system and not elsewhere. Therefore it has an analgesic, but lacks an anti-inflammatory, effect. The COX enzymes are responsible for prostaglandin synthesis, among which prostaglandins E2 and F2 promote natriuresis, and prostacyclin has vasodilatory effect, which could explain the BP raising effect of NSAIDs. However our presumption about the difference between acetaminophen action and NSAIDs may have been incorrect from these results. The present study was not a mechanistic study, and we do need to understand the science better. From the NSAID meta-analysis, it seemed that vasodilators and diuretics were more effective than beta-blockers in treating hypertension in this setting (Johnson et al, Annals Int Med 1994).
Some other questions still linger. Is this a dose dependent effect, and will this effect be attenuated at lower doses? Presumably short term exposures, especially at lower doses, are without much downside. What are the long term implications - surely if acetaminophen was causing cardiovascular badness, some large database study would have picked up the signal? Unfortunately, being over-the-counter makes pain relief easier, but also makes ascertainment of exposure in studies harder.
Lastly, what do we do for pain relief now? Less acetaminophen might mean more other NSAIDs (which probably have the same effect on BP, and might be worse for kidneys), or more opioids (probably not really safer than NSAIDs for kidneys, and potentially worse in many other ways). It does mean non-pharmacological means of pain relief should receive more than lip service.
Conclusion
The findings of the PATH-BP trial question the current guidelines suggesting that acetaminophen is a safe alternative to NSAIDs, and it suggests that caution should be used when prescribing regular use of acetaminophen especially in those with HTN or those at risk of ischemic heart disease and stroke.
Summary Prepared By
Momen Abbasi
Nephrology Fellow
Hadassah Medical Center
Jerusalem, Israel
NSMC Intern 2022
Reviewed by