Vascular events such as myocardial infarction, stroke, or fatal cardiovascular complications occur about three times more often in patients with NSAIDs than in patients without NSAID medication. This significant increase was statistically proven in large clinical trials and meta-analyses for coxibs and diclofenac. An increase of the coronary risk could be shown for higher-dosed ibuprofen (1,200 mg/day) (RR 2.22; p = 0.0253). A similar high cardiovascular (CV) risk can be assumed for other less-investigated NSAIDs. High-dosed naproxen as an exception behaves neutral in terms of CV risk (RR 0.93). Low-dose ibuprofen does not increase the thromboembolic CV event rates. With all NSAIDs the risk of severe heart failure is doubled [6]. Increase in blood pressure is more common in patients taking etoricoxib than others taking naproxen or ibuprofen. In several observational studies, rofecoxib, celecoxib, and diclofenac dose dependently increased the risk of vascular adverse drug reactions by 20–30 % [6, 9]. One year of NSAID treatment with 1,000 patients at moderate CV risk results in three additional clinically relevant CV events including one fatal event [6, 7].

The common CV risk factors (e.g., hypertension, diabetes mellitus, hyperlipidemia, smoking, a history of CV disease such as coronary heart disease or myocardial infarction, positive family history with CV disease in ≤50-year-old patients) also apply to NSAID patients as predictors of increased CV risk; also the average risk increases with the NSAID dose and the duration of treatment [10]. But even without a corresponding risk factor, the rate of NSAID-associated CV side effects is slightly increased [6, 11–13]. Especially the risk of severe heart failure under any NSAID therapy is doubled independently of the platelet function [6]. Other studies have found an increased risk of stroke. However, the data in this area are not sufficient for a final assessment. Even after many years of NSAID treatment, an adaption of risk cannot be observed. After cessation of NSAIDs, the individual risk quickly disappears [10].

The FDA alerts of an increased CV risk in any NSAID therapy. In patients with coronary heart disease or congestive heart failure (NYHA grade ≥2) or history of stroke, the EMA declares the coxibs and diclofenac as contraindicated. With any NSAID therapy in patients at CV risk, the EMA urges caution [14].

CV and GI risks often are increased at the same time, especially in elderly patients. In patients with moderate combined CV and GI risks, vascular or gastrointestinal complications are expected to occur in 4–19 % [7] after a 10-year high-dose NSAID therapy.

Several physiological products of the COX enzymes are significantly involved in the regulation of renal function. NSAIDs interfere with the COX-1- and COX-2-dependent renal tissue level of prostaglandins. During the first 30 days of NSAID treatment, elderly patients (>65 years) have a significantly increased risk of acute renal failure (RR: 2.05. [15]), and a NSAID therapy lowers the glomerular filtration rate in patients with chronic kidney disease [16]. The risk of hypertension is increased [17–20]. In damaged kidneys water retention and edema are common events.

Other manifestations of diseases and serious side effects (liver, skin, etc.) are possible but occur rarely.

All NSAIDs inhibit the enzyme cyclooxygenase (COX) which interferes with the paracrine synthesis of prostaglandins (PGs) [21]. Prostaglandins regulate specific physiological processes via two different COX enzymes, dependent on the tissue location. COX-2 is induced in injured tissues. This isoenzyme synthesizes PGs, which locally induce inflammatory processes and mediate pain.

In other tissues, prostaglandins are synthesized continuously. This is especially important in the GI tract whose mucosa constantly expresses COX-1 and produces PGs to protect the tissue against low pH levels. Traditional NSAIDs (tNSAIDs) inhibit both isoenzymes of cyclooxygenases. Desired effects (pain relief through inhibition of COX-2-specific PG synthesis), as well as adverse effects (mucosa lesions due to elimination of the COX-1-mediated protective PGs), are attributed to the same principle, the inhibition of the synthesis of PGs. Individual effect and side-effect differences result from differences in organ tissue and metabolic reactions.

In therapeutic doses, selective COX-2 inhibitors (coxibs) virtually do not react with the COX-1, and therefore the gastrointestinal toxicity of coxibs is less than that of tNSAIDs [22].

NSAIDs usually are administered orally and are absorbed in the intestine. Thus, their effect is systemically mediated. This is true for the inflamed articular tissue as well as for the gastrointestinal tract. Some NSAIDs undergo an enterohepatic cycle and pass multiple times through the small intestine. In this case an additional local potential of harm could be added to the systemic toxicity.

Parenteral routes of administration have no advantage related to the gastrointestinal potential danger but bear greater local risks. Thus, this kind of application should be avoided. Suppositories have a longer residence time in the rectum, and therefore, local toxic effects cannot be ruled out. On the other hand, they are frequently eliminated in advance via naturalis, or they are insufficiently absorbed. That is why suppositories do not provide benefits in clinical trials, and therefore they are not recommended for pain therapy [23, 24].

Topical NSAIDs – containing ointments, gels, solutions, or patches, applied on the intact skin – alleviate musculoskeletal pain better than placebo [25] and are not inferior compared to oral NSAIDs therapy [25–27]. Their mode of action is unclear. Thus, topical NSAIDs can be used tentatively as low-risk NSAID preparation.

The risk of GI complications can be reduced by 50 % if a coxib instead of a tNSAID is used. An equally effective approach is to use a proton pump inhibitor (PPI) or misoprostol accompanying the intake of a NSAID. The PPI protects the gastroduodenal part of the GI tract but not the lower GI tract; misoprostol in turn has a higher potential for adverse GI symptoms. In high-risk patients where NSAID therapy is planned to be resumed despite an occurred GI complication, a treatment strategy with a coxib in combination with a PPI is recommended, thus avoiding recurrence of ulcers.