Discussions
There are no diagnostic criteria for ANCA-associated vasculitis (AAV).2 Investigative accuracy has been limited in AAV; 77% of head and neck biopsies are non-diagnostic and 10% of AAV patients being ANCA negative. The latter is accentuated in localised GPA disease such as SGS.2
This is echoed in our cohort with 9 ANCA-negative cases and only 23% of biopsies confirming GPA. Considering the low yield of diagnostic tests, we recommend regular clinical assessments for signs of systemic involvement and early MDT input where systemic features of GPA are recognised.
Our medical strategy for SGS-GPA has shown that induction IS for active SGS-disease improves the PFI and decreasing need for more invasive surgery.7 Comparing with other studies utilizing a surgically centered approach (table 2),3,8-10 our PFI was notably better, with 19% of the cohort not requiring surgery. This demonstrates a 65% increase when comparing against the mean PFI of the quoted studies at 19.0 months (this was calculated with individual mean PFIs given different weighting based on the study’s cohort size).3,8-10
Four patients required a tracheostomy for emergency airway management. Ideally all SGS-GPA patients should be managed with an emergency airway dilatation where possible rather than an emergency tracheostomy. However, with the joint ENT vasculitis being a regional service, emergency airway dilatation procedures are not always available in the peripheral hospitals. Nevertheless, the tracheostomy rate in this study (19%) was lower than that quoted in the literature (41-52%),3 with all patients decannulated promptly.
The benefit of IS is further stated when assessing the four patients excluded from our study due to extensive SGS-GPA and late referral for MDT management: undergoing a combined 54 dilatational procedures, translating to a 6-fold increase in surgeries required in the absence of induction IS. This enforces the message of early MDT management with induction IS.
The regression model estimated an improved mean PFI of 48 months, reflecting a more accurate representation of the mean PFI obtained our patients’ individual rates at 42.8 months. Unfortunately, the impact of the other covariables could not be adequately analysed, due to small sample size obtained.