A 63-year-old man with a history of superficial spreading melanoma of the left scapular region (Breslow 3.5 mm, ulcerated) diagnosed in 2019 was treated with wide local excision and sentinel lymph node biopsy, initially disease-free. In 2020, metastatic relapse involving skin, lymph nodes and lungs was confirmed, and a BRAF V600 mutation was identified. The patient received monthly nivolumab (480 mg) for 54 cycles over 5 years, achieving a sustained complete metabolic remission. Treatment was discontinued in November 2024. Two months later, he developed progressive bilateral temporal headaches and pulsatile scalp pain, followed by jaw claudication without fever, weight loss, neurological or visual symptoms. Within days, violaceous livedoid macules appeared over both temples, rapidly evolving into extensive necrotic plaques of the temporal scalp bilaterally (Figure 1 a left side, b right side). A temporal artery biopsy was performed (Figure 2). Giant cell arteritis (GCA) with bilateral scalp necrsis following anti–PD-1 therapy cessation. Laboratory tests revealed mild inflammation (C-reactive protein 17 mg/L; erythrocyte sedimentation rate 24 mm/h). Doppler ultrasound showed bilateral frontal artery occlusion without the typical halo sign of the temporal arteries. PET-CT excluded large-vessel involvement. Right temporal artery biopsy demonstrated granulomatous arteritis with lymphocytes, macrophages, epithelioid histiocytes and multinucleated giant cells with internal elastic lamina disruption, consistent with active GCA (Figure 2a,b). High-dose systemic corticosteroids (0.7 mg/kg/day) were initiated, resulting in rapid relief of headaches and jaw claudication. Over the following weeks, progressive re-epithelialization of scalp ulcers occurred, leaving cicatricial alopecia (Figure 3a,b). GCA is a granulomatous vasculitis of medium- and large-sized arteries, typically affecting individuals over 50 years old. While cranial symptoms such as temporal headache, scalp tenderness and jaw claudication are classical features, scalp necrosis is an uncommon but severe ischaemic manifestation, usually indicating extensive arterial involvement. Bilateral scalp necrosis is exceptional. Immune checkpoint inhibitors (ICIs), particularly PD-1 inhibitors, have transformed melanoma management but may induce a variety of immune-related adverse events (irAEs). Although cutaneous irAEs are frequent, vasculitis remains rare and GCA represents one of its least common presentations 1. PD-1/PD-L1 signalling plays a key role in maintaining peripheral immune tolerance. In GCA, PD-L1 expression is markedly reduced on vascular dendritic cells, while PD-1 is highly expressed on infiltrating CD4⁺ T cells, leading to unchecked Th1 and Th17 proinflammatory activity within the vessel wall. Experimental models have shown that PD-1 blockade further aggravates this dysregulation, enhancing vascular inflammation and remodelling 2-4. These findings provide a mechanistic rationale linking PD-1 inhibition to vasculitis processes. Only a few cases of ICI-associated GCA have been reported, involving agents such as nivolumab 5, 6 and pembrolizumab 7. Kreuter et al. described a woman who developed unilateral scalp necrosis 6 weeks after starting nivolumab 5; Betrains et al. reported GCA after 30 cycles of nivolumab 6; and Micaily et al. observed GCA within 1 week of pembrolizumab initiation 7. These reports suggest that GCA may occur at various stages of ICI exposure, from early onset to late during prolonged treatment. A global pharmacovigilance analysis using the WHO VigiBase database identified a significant association between ICIs and vasculitis, particularly temporal arteritis (reporting odds ratio ROR, 12.99; 95% CI, 8.12–20.77; 18 cases). Reported cases involved both men and women, with a mean age of 72.9 years (range, 60–83). Most cases were associated with ipilimumab (n = 10), followed by pembrolizumab (n = 4) and nivolumab (n = 3). The median time to onset of vasculitic irAEs was 21 days (range, 21−131) 8. Although most events occurred early, delayed presentations cannot be excluded. In our patient, biopsy-proven GCA manifested 2 months after discontinuation of long-term nivolumab, following 54 treatment cycles. While this extended delay challenges a direct causal relationship, late-onset irAEs—sometimes arising months or even years after therapy initiation or cessation—are increasingly recognized 1. In this case, the French Regional Pharmacovigilance Center deemed the imputability of nivolumab ‘probable’, based on chronology, mechanism and literature consistency. To our knowledge, this is the first report of bilateral scalp necrosis as the presenting manifestation of delayed GCA following nivolumab discontinuation, possibly related to ICIs. Management relies on prompt initiation of systemic corticosteroids to prevent irreversible ischaemic complications. The rapid improvement in symptoms and progressive healing of necrotic ulcers in our patient underscores the efficacy of early treatment. Rechallenge with ICIs after vasculitic irAEs is generally discouraged due to the potential for recurrence; in our case, permanent discontinuation was facilitated by the patient's sustained melanoma remission. In most published cases, immunotherapy was permanently discontinued. This case highlights a rare and striking presentation of possible ICI-related vasculitis and emphasizes the need for ongoing vigilance—even after immunotherapy withdrawal. Dermatologic findings such as scalp necrosis should alert clinicians to possible severe large-vessel vascular inflammation requiring urgent evaluation and treatment. Zoé Dumesnil: data curation, writing – original draft preparation. Manuel Croix: conceptualization, methodology. Anne Dompmartin, Jean-Matthieu L'Orphelin, Mohamed-Taha Tagmouti: supervision. Diane Kottler, Anael Dumont, and Pharmacovigilance Team: help for initial diagnosis. The authors received no specific funding for this work. The authors have nothing to report. All patients in this manuscript have given written informed consent for participation in the study and the use of their de-identified, anonymized, aggregated data and their case details (including photographs) for publication. Ethical Approval: not applicable. The authors declare no conflicts of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request.
Dumesnil et al. (Tue,) studied this question.