Randomized trial of adenotonsillectomy versus expectant treatment in PFAPA (periodic fever, aphthous stomatitis, pharyngitis, cervical adenitis) syndrome: Is the impasse over?

To the Editor:

The results of an interesting study by Garavello et al comparing adenotonsillectomy in children who have PFAPA (periodic fever, aphthous stomatitis, pharyngitis, cervical adenitis) syndrome with expectant treatment can be aptly described as an absolute risk reduction of 58% (95% CI, 29%-76%) in PFAPA episodes on 18-month follow-up and a number needed to treat of 2 (95% CI, 1-3).1 This very large effect size, however, needs to be triangulated in the light of some perspectives on the study. First, a surgical failure rate of 37% (95% CI, 19%-59%; 7 of 19 had PFAPA recurrences) in the first year may not be clinically acceptable for a benign disorder characterized by self-resolving episodes responding strikingly to single dose of oral corticosteroid, no short- and long-term physical consequences, and also, the propensity to outgrow the problem by the age of 10 years.2, 3, 4 Second, because the primary outcome was resolution or recurrence of episodes on follow-up, the choice of control group patients with only expectant treatment (ie, oral corticosteroid) seems methodologically less efficient. A more active control group with the use of a prophylactic drug like cimetidine (with 29% event resolution) was warranted because the use of episode-ameliorating corticosteroid can escalate recurrences in some patients.2, 5 Third, the duration of the episode-free period was compared with the t test, completely ignoring the skewed nature of such time-related data and presence of right censoring when patients had complete resolution. Ideally, this important outcome should have been compared more reliably by using Kaplan Meier survival analysis. Finally, the study fails to address the superiority of combined adenotonsillectomy compared with tonsillectomy alone (64%-100% efficacy),6 appropriate age of surgical intervention, and incremental cost-effectiveness in terms of costs per episode averted. To conclude, potentially modest benefits of adenotonsillectomy in PFAPA need to be weighed against the benign nature and heterogeneity of episode recurrence, comparative cost-effectiveness with medical therapy, and risk of surgery in centers with less experience.