Is inhibition of cyclooxygenase required for the anti-tumorigenic effects of nonsteroidal, anti-inflammatory drugs (NSAIDs)? In vitro versus in vivo results and the relevance for the prevention and treatment of cancer

Biochem Pharmacol. 2002 Feb 1;63(3):343-7. doi: 10.1016/s0006-2952(01)00857-7.


Active research is being conducted to unravel the cellular mechanisms mediating the anti-tumorigenic effects of nonsteroidal anti-inflammatory drugs (NSAIDs) and their association with cyclooxygenase (COX) inhibition. The majority of NSAIDs inhibit either COX-1, COX-2, or both and exert their anti-COX, anti-inflammatory, and anti-tumorigenic effects in vivo in a parallel dose-dependent manner. The effects are seen at NSAID blood plasma concentrations of 0.1-5 microM. Significantly, the same compounds tested at the same concentrations in incubations with cultured tumor cells in vitro similarly inhibit COX activities but are devoid of anti-proliferative activity. Yet, at much higher concentrations (100-20,000 microM), these same NSAIDs do exert anti-proliferative effects in vitro due to apparent non-specific toxic effects, as evidenced by disruption of ion transport and mitochondrial oxidation in some cells. A small group of NSAIDs (e.g. sulindac) do not inhibit COX enzymes significantly but can reduce the synthesis of prostanoids by alternate mechanisms. One such mechanism is inhibition of agonist-stimulated phospholipase-mediated release of arachidonic acid from phospholipids leading to depressed synthesis of prostanoids, especially prostaglandin E(2) (PGE(2)). Another group of non-COX inhibitors are the R-isomers of NSAIDs, based on the structure of 2-arylpropionic acid. These compounds exert anti-proliferative effects in vivo, acting by an as yet undetermined mechanism. A possible caveat in these data is an R to S chiral transformation in vivo that would render the R-isomer effect as being due to the S-isomer generated in vivo from it. Demonstration of minimal or no R to S inversion under the experimental in vivo conditions employed is, therefore, a necessary control in these studies. The overall body of data supports the conclusion that, for COX-inhibiting NSAIDs, their anti-tumorigenic effect in vivo is due to, and depends upon, inhibition of tumor COX enzymes, primarily COX-2. The cellular effects seen when adding high concentrations of NSAIDs to tumor cells cultured in vitro and the mechanisms proposed to mediate these effects may not have substantial relevance to the mechanisms that mediate the effects of NSAIDs in vivo.

Publication types

  • Review

MeSH terms

  • Animals
  • Anti-Inflammatory Agents, Non-Steroidal / classification
  • Anti-Inflammatory Agents, Non-Steroidal / therapeutic use*
  • Antineoplastic Agents / therapeutic use*
  • Cell Cycle / drug effects
  • Cyclooxygenase Inhibitors / therapeutic use*
  • Dinoprostone / metabolism
  • Humans
  • Neoplasms / drug therapy
  • Neoplasms / metabolism
  • Neoplasms / prevention & control*
  • Prostaglandin-Endoperoxide Synthases / metabolism*


  • Anti-Inflammatory Agents, Non-Steroidal
  • Antineoplastic Agents
  • Cyclooxygenase Inhibitors
  • Prostaglandin-Endoperoxide Synthases
  • Dinoprostone