The thiophene ring of DuP 697 was replaced by a variety of heterocycles and the products were tested for their ability to inhibit human Cox-2 and Cox-1, the isozymes of cyclooxygenase.A highly selective inhibitor of the newly discovered cyclooxygenase (Cox) isozyme, 1 Cox-2, that would be bioavailable and metabolically stable, is a strongly pursued objective. This is because such an inhibitor has the potential to produce a new generation of nonsteroidal antiinflammatory drags (NSAIDs) with improved therapeutic properties and fewer side effects such as gastropathy. 2 This potential was explored in the "tricyclic" class of NSAIDs, best represented by DuP 697, 3 a series that exhibited excellent potency for the inhibition of Cox-2 and moderate selectivity over Cox-1.4 In an earlier letter, 5 we established that the structure-activity relationship (SAR) for the methylsulfonylphenyl ring of this series is particularly tight. Only the relatively small and polar primary sulfonamide group is tolerated at this position without appreciable loss of potency on Cox-2. Unfortunately, despite the promises of improved bioavailability with such a derivative, the concomitant loss of selectivity required that we investigate the SAR in other areas of the molecule. Removal of the brolno substituent at the C5 position of the thiophene ring yielded a potent and more selective Cox-2 inhibitor. In fact, this compound is inactive in our Cox-I assay at concentrations up to 100 laM. 5 Furthermore, the 2-bromothiophene moiety of DuP 697 has been reported to be metabolized to the 2-methylsulfonyl derivative, a long lasting metabolite. 6 Consequently, a simple, preferably unsubstituted ring to replace the thiophene nucleus was sought. In this letter, our SAR investigation focuses on ring replacements for the thiophene template. 7