Development of a novel prognostic assessment tool for recurrent respiratory papillomatosis

BACKGROUND: Recurrent respiratory papillomatosis (RRP), caused by the human papillomavirus (HPV), is associated with an unpredictable clinical course. Although the Derkay Score is widely used to determine clinical severity, its prognostic value has rarely been evaluated. We developed a pathological severity score, the Hamamatsu Recurrent Respiratory Papillomatosis Pathological (HARRP) Score and combined it with the Derkay Score to develop a novel clinicopathological system-the Derkay-HARRP (D-H) Classification. We aimed to validate its prognostic value. METHODS: We retrospectively analyzed 125 patients who were clinically diagnosed with RRP from 16 Japanese institutions, randomly divided into validation (n = 38) and test (n = 87) cohorts. HPV-typing and immunohistochemistry for HPV-L1, HPV-E4, Ki-67, and p16 were performed. HPV particles were confirmed using the NanoSuit-correlative light and electron microscopy (CLEM) method. Receiver operating characteristic (ROC) curve analysis evaluated marker performance and defined recurrence cut-offs. HARRP and Derkay Scores were further assessed by ROC analysis and Cox proportional hazards models. We stratified patients using the D-H Classification and analyzed disease progression over time. RESULTS: No significant demographic differences were observed between the two cohorts. Positivity of HPV-L1, HPV-E4, and Ki-67 in the upper third of the epithelium was associated with recurrence. NanoSuit-CLEM confirmed HPV particles in HPV-L1-positive areas, supporting pathological relevance. The HARRP Score was calculated by assigning 1 point each for positivity of HPV-L1, HPV-E4, and ≥ 5% Ki-67-positive cells in the upper third of the epithelium. ROC analysis of the HARRP Score showed areas under the curve (AUCs) of 0.675 (validation) and 0.754 (test), whereas the Derkay Score showed AUCs of 0.709 and 0.834, respectively. The cut-off values were 1 and 4, respectively. Both scores were significant in the Cox analysis (p < 0.001). The D-H Classification stratified patients as Severe (both positive scores), Moderate (either positive), or Mild (both negative), with significant differences in relapse-free survival (p < 0.001). Severity tended to decrease with repeated surgeries and recurrence was rare in the Mild group. Findings remained consistent in HPV-positive cases only. CONCLUSIONS: Combination of the Derkay and HARRP Scores-the D-H Classification-provides a practical tool for risk stratification and personalized follow-up planning of patients with RRP.