Chronic pain and analgesic medication use are common among patients who receive ongoing dialysis.[1–3] According to recommendations in the World Health Organization pain ladder, opioids are generally avoided until nonopioid medications are proven ineffective. Regularly scheduled acetaminophen is used, however, many patients report little benefit. In the dialysis population, many physicians recommend ibuprofen and other nonsteroidal anti-inflammatory drugs (NSAIDs), reassured that once a person's kidneys fail the risk of nephrotoxicity is far less relevant.
Beyond nephrotoxicity, NSAID-related major adverse cardiac events are of concern. This was first identified in the general population with rofecoxib, which showed a 1.92-fold higher risk of myocardial infarction (MI) and ischemic cerebrovascular events over placebo.[5,6] This resulted in its removal from the market in 2004. While the risk was initially perceived as limited to cyclooxygenase-2 (COX-2)-specific NSAIDs, subsequent randomized controlled trials in the general population confirmed this risk also extends to nonselective NSAIDs.[7–9]
Patients who receive dialysis have a high prevalence and incidence of cardiovascular disease, yet NSAID-related risks of cardiac disease remain understudied. A population-based case–crossover study of 1190 nontraumatic stroke patients receiving dialysis in Taiwan found a 1.31-fold higher risk of stroke associated with NSAID use (versus no use) in the prior 30 days [95% confidence interval (CI) 1.03–1.66]. The association remained significant in a subgroup analysis of nonselective NSAIDs [odds ratio (OR) 1.27 (95% CI 1.00–1.61)] but not COX-2-specific NSAIDs [OR 1.18 (95% CI 0.67–2.09)], where the latter estimate was imprecise due because of the small sample size.
A similar analysis using the same set of administrative databases followed patients who had initiated chronic dialysis for a mean of 4.0 years. It found 78% of patients had at least one prescription for NSAIDs during the follow-up period and after adjustment for a number of demographic and comorbid factors, NSAID use (versus no use) was associated with a higher risk of all-cause mortality [hazard ratio (HR) 1.39 (95% CI 1.211–1.60)].
In the January edition of Nephrology Dialysis and Transplant, Jo et al. examined the association between NSAID use and cardiovascular events in patients receiving dialysis . Using a case–crossover design, they studied a cohort of patients receiving dialysis in Korea who suffered an in-hospital major adverse cardiac or cerebrovascular event (MACCE). The composite outcome of MACCE consisted of nonfatal MI, coronary reperfusion or a cerebrovascular event. Secondary outcomes were mortality or major bleeding. From a cohort of 68 771 incident dialysis patients, excluding those with a history of a MACCE or bleeding in the prior 5 years, 3433 suffered a MACCE, 3073 patients had major bleeding and 8524 patients died; all took NSAIDs in the 120 days preceding their event. The authors assigned the 30 days preceding the event as the case period and the 60–90 and 90–120 days preceding the event as the control period. The analysis was adjusted for time-varying confounding variables. The study used pharmacy dispensing records to determine the dose, timing and type of NSAID taken. It is unknown how many of the study prescriptions were prescribed as as needed rather than regularly, and there was no information on over-the-counter NSAID use. Data were available for a large number of prescription NSAIDs, but notably, naproxen was not used and celecoxib was the only COX-2 selective NSAID.
The average ages in the MACCE, mortality and major bleeding cohorts were 66, 69 and 63 years, respectively. Most (85%) were receiving hemodialysis with the remainder on peritoneal dialysis. Most (60%) were male with substantial comorbidities, including pre-existing congestive heart failure (15–20%), diabetes (50–54%), peptic ulcer disease (31–37%) and prior cerebrovascular event (27–32%). The presence of conditions associated with musculoskeletal pain was common; 20% of patients had a history of gout and 25% a history of osteoarthritis.
NSAID use in the prior 30 days when compared with remote use was associated with a higher risk of MACCE [OR 1.37 (95% CI 1.26–1.50)], mortality [OR 1.29 (95% CI 1.22–2.36)] and major bleeding [OR 1.81 (95% CI 1.65–1.98)]. In subgroup analysis, nonselective NSAID use was significantly associated with a higher risk of MACCE, mortality and major bleeding (OR 1.39, 1.32 and 1.80, respectively) but COX-2 selective NSAID use was not [OR 1.23 (95% CI 0.90–1.62) and OR 0.90 (95% CI 0.74–1.09) for MACCE and mortality, respectively].
Higher prescribed NSAID doses were associated with a higher risk of bleeding. However, there was no higher risk of MACCE or mortality with a higher cumulative daily dose of NSAIDs. Indeed, higher doses of COX-2 selective NSAIDs were associated with a lower risk of both mortality and MACCE.
Of those taking NSAIDs in the 30 days prior to an event, the risk of all three outcomes was higher closer to the date of NSAID prescription; the association became attenuated as the time from the last NSAID prescription date became longer to the adverse event.
Overall, the findings of this study are consistent with those in the general population.[7–9,13–17] Randomized, placebo-controlled trials demonstrate a higher relative risk of cardiovascular events or death, with most HRs between 1.3 and 2.2.[8,9,17]
Many prior observational cohort and case–control studies have also examined this issue; however, the choice of a suitable comparator group remains a challenge. Indications for NSAIDs such as inflammatory conditions or acute pain may, in themselves, increase cardiac risk either transiently or permanently, raising concerns about unmeasured confounding when NSAID use is compared with nonuse. Similarly, a comparison to other NSAIDs is problematic, as no NSAID has been proven to be without cardiac risk. In the study by Jo et al., a case–control crossover design was used to reduce the risk of time-invariable confounding, but as a result, the control patients still had remote NSAID exposure. This also meant the authors did not report the absolute risk of MACCE in this study population—important information when counseling patients about risk.
NSAID-induced cardiovascular disease is attributed to several mechanisms. In the short term, COX-2 inhibition lowers vessel prostacyclin production and thromboxane A2 inhibition, which may promote thrombosis. This provides a mechanism for the higher short-term risk observed with NSAIDs in studies by Jo et al. and others.[12–14] Long-term renal prostaglandin synthesis inhibition results in sodium and fluid retention, which may lead to hypertension and vascular changes, increasing cardiac risk even in those with remote exposure. A higher cardiac risk was seen in the Adenomatous Polyp Prevention on Vioxx placebo-controlled trial of rofecoxib after 18 months of use, suggesting a cumulative, time-dependent mechanism, which the observational trial by Jo et al. was not designed to detect.
Unlike the study by Jo et al., other observational studies report a higher dose of naproxen, ibuprofen, celecoxib or diclofenac is associated with a higher risk of cardiovascular disease.[8,13,15,16] In pharmacokinetic studies, higher doses of NSAIDs increase the degree of COX inhibition.[19,20] The point estimates for the increased MACCE risk with nonselective and COX-2 selective NSAIDs were similar, although the precision of the latter was limited by the small sample size. Although it would be premature to make a conclusion about the risk of COX-2 selective compared with nonselective NSAIDs from this study alone, the finding of increased cardiovascular risk, even with nonselective NSAIDs, is consistent with the established literature.
The Prospective Randomized Evaluation of Celecoxib Integrated Safety versus Ibuprofen or Naproxen trial randomized patients to celecoxib, ibuprofen or diclofenac and found no difference in the risk of MACCE between drug classes. Additionally, a meta-analysis of 639 trials comparing COX-2 selective NSAIDs to placebo or nonselective NSAIDs found that while COX-2 selective NSAIDs did have a higher risk of cardiac events over placebo, this risk was similar to high-dose diclofenac and ibuprofen. Some suggest traditional NSAIDs, particularly diclofenac, still exhibit strong COX-2 inhibition, in some cases in a more potent manner than the COX-2 inhibitors.
Unfortunately, pain often goes untreated in dialysis care. In the Dialysis Outcomes and Practice Patterns Study (DOPPS), three-fourths of patients who reported moderate to severe pain were not prescribed analgesics. Similar findings were described in an Italian population. Uncontrolled pain can lead to depression, insomnia and a choice to discontinue dialysis. In the general population, NSAIDs are equivalent to opioids in the treatment of osteoarthritic pain, however, pain management guidelines in end-stage renal disease make little comment about NSAID use while giving robust guidance on the use of opioids.[22–24] In a DOPPS cohort, 2% of patients took NSAIDs while 15% took opioids. Although opioids do effectively control pain, they have a significant side-effect profile.[22,25,26] In the dialysis population, opioid use is associated with a 28–67% higher risk of altered mental status, 28–45% higher risk of falls and 44–65% higher risk of fracture compared with nonuse. In another study, even short-term opioid use was associated with a higher risk of hospitalization and death compared with nonuse. Antiepileptics such as gabapentin and pregabalin can also be effective in the treatment of neuropathic pain but have significant side effects limiting their tolerability. About 10–20% of patients receiving dialysis discontinue use of these medications because of dizziness or fatigue.[27,28]
In conclusion, in a large study in dialysis, Jo et al. show NSAID use is associated with a modest increase in MACCE and mortality, similar to the increase observed in the general population. NSAID use also increases the risk of bleeding. However, there are few good alternatives for pain control and pain often goes untreated in dialysis care. When NSAIDs are used, it remains prudent to start at a low dose and for a short duration.
(See related article by Jo et al. Cardiovascular risk of nonsteroidal anti-inflammatory drugs in dialysis patients: a nationwide population-based study. Nephrol Dial Transplant 2021; 36: 909–917)
Nephrol Dial Transplant. 2021;36(5):749-751. © 2021 Oxford University Press