A clinical classification system for grading platinum hypersensitivity reactions

Current grading systems for platinum hypersensitivities (pHSR) rely on subjective features rather than objective clinical signs leading to inconsistencies in grading. To standardize classification of pHSR, a clinical grading system was developed at our institution. We report the clinical outcomes our classification system and evaluate its correlation with the classification systems currently published and used in practice.

This was a retrospective review of patients with pHSR from 2011 to 2017. Demographics, chemotherapeutic histories (CT), and details of their initial HSR were collected. Mild reactions were defined as local skin manifestations only. Moderate-low reactions included widespread skin, respiratory or GI findings. Moderate-standard reactions were defined as transient cardiovascular compromise (CVC), hypoxia or neurologic changes whereas sustained changes (>10 min) were used to define severe reaction. Fischer Exact Tests (p < .05) and binary logistic regression analyses were performed. Spearman correlation were used to assess relationships between our grading system and the NCCN and CTCAEv4.0 criteria.

87 patients were identified with most having ovarian cancer (n = 55, 63.2%), receiving carboplatin (n = 62, 71.3%), and on second-line CT (n = 34, 42.5%). Chest pain was associated with transient CVC (OR 10.0, 95% CI 1.148-87.133) while nausea/vomiting (OR 8.420, 95% CI 1.263-55.275) was associated with transient hypoxia albeit less closely than transient hypotension (OR 17.010, 95% CI 2.026-142.825). Only presyncope/syncope remained associated with sustained CVC (OR 38.0, 95% CI 2.815-512.912) on logistic regression. The classification system was most strongly correlated with the NCCN grading system (ρ 0.761, p < .001).

This classification system offers an objective means of grading pHSR severity and correlates with currently-used grading systems.