Predicting Surgical Outcomes in Pediatric Cervicofacial Nontuberculous Mycobacterial Lymphadenitis

Abstract

We examined surgical outcomes in children with cervicofacial nontuberculous mycobacterial lymphadenitis and attempted to identify predictors of complications. A retrospective chart review from 2 tertiary pediatric centers was used to identify .11 presentation or operative variables (age at surgery, gender, symptom duration, pain, violaceous skin changes, skin breakdown, fluctuance, purified protein derivative positivity, operative procedure, use of nerve integrity monitoring, and use of skin flap advancement) and to compare these to 5 postoperative complications (facial nerve dysfunction [paresis or paralysis], poor scarring, recurrence, wound infection, and wound dehiscence without infection). The 45 patients analyzed for presentation or operative variables (28 female, 17 male; average age, 31.2 months) typically presented with painless masses averaging 8.2 weeks in duration, along with violaceous skin changes in 29 of the 45 cases (64%) and skin breakdown in 9 cases (20%). The surgical procedures included parotidectomy with or without selective lymphadenectomy in 38 of the 45 cases (84%) and lymphadenectomy alone in 7 cases (16%). Skin resection and cervicofacial advancement flap reconstruction was performed in 20 cases (44%). Nerve integrity monitoring was utilized in 32 cases (71%). In the 44 patients analyzed for postoperative complications, we found facial nerve paresis in 14 (31.8%), poor scarring in 9 (20.5%), wound infection in 6 (13.6%), recurrence in 4 (9.1%), and facial nerve paralysis in 2 (4.5%). Nine of the 14 cases (64.3%) of initial facial nerve paresis resolved. At final follow-up, facial nerve paresis persisted in 5 of the 14 children (35.7%) with initial postoperative paresis and in 1 of the 2 children (50.0%) with initial postoperative paralysis. Facial nerve paralysis persisted in the other child with initial postoperative paralysis. Overall, 6 of these 7 patients (85.7%) with persistent facial nerve dysfunction had follow-up of less than 1 month. All transient and permanent facial nerve dysfunction was in the distribution of the marginal mandibular nerve only. No statistically significant predictors of complications were identified. We report acceptable but not insignificant rates of marginal mandibular distribution facial nerve injury, poor scarring, wound infection, and recurrence following resection of cervicofacial nontuberculous mycobacterial lymphadenitis in children that must be discussed with patients and parents before operation. No presentation or operative variables predicted the complications.