Penicillin V, loracarbef and clindamycin in tonsillar surface fluid during acute group A streptococcal pharyngotonsillitis.

Patients with acute group A- strepotococcal pharyngotonsillitis were randomly assigned to treatment for 10 d with either phenoxymethylpenicillin (PcV), loracarbef or clindamycin. The concentrations of the drugs, respectively, were determined in tonsillar surface fluid (TSF), serum and the saliva in each patient on altogether 5 occasions; before, during and 4 d after end of therapy. On the same occasions blood was drawn for analysis of C-reactive protein (CRP) and orosomucoid. On the last d of treatment PcV could be detected in TSF in 1 of 6 patients only. Loracarbef had a slower decrease in TSF during therapy and measurable levels did occur 2 d after end of therapy corresponding to MIC 100 for GAS. This may be related to the somewhat better clinical results of the cephalosporins than of PcV, and possibly indicates that an extended therapy with these drugs in primary GAS pharyngotonsillitis for more than the arbitrarily chosen 10 d could reduce the number of recurrent episodes. PcV and loracarbef were not detected in serum after the end of treatment. The concentration of clindamycin in both TSF and the saliva was fairly longstanding during therapy and reached levels exceeding MIC 100 for GAS, in both TSF and serum 2 d after the end of treatment. Several investigations have shown that GAS, especially in the stationary phase may invade respiratory epithelial cells and are present intracellularly in patients with acute pharyngotonsillitis as well as in asymptomatic carriers. The same T-type, identical DNA fingerprints and arbitrarily primed patterns are found in GAS before and after treatment failure indicating that the primary episode and the failures are caused by the same strain. The longstanding concentrations of clindamycin in TSF, roughly independent of the degree of the local inflammation combined with its intracellular accumulation and activity against resting GAS seem to explain the efficiency of the drug in recurrent GAS pharyngotonsillitis. CRP and orosomucoid were of limited value in differing between bacterial and viral pharyngtonsillitis and a correlation between antibiotic concentration and CRP/orosomucoid levels was not found.