Late‐Onset Chimeric Antigen Receptor <i>T</i> ‐Cell Therapy‐Associated Pneumonitis: A Case Report and Literature Review

To the editor, A 4-year-old girl presented to our hospital with 1-week history of tachypnea and mild nonproductive cough. Chest computed tomography (CT) demonstrated bilateral lower lobe ground-glass opacities (GGO) and patchy consolidations (Figure 1A), prompting empirical antimicrobial therapy. She subsequently developed pyrexia and dyspnea 1 day prior to admission. Ten months ago she was diagnosed with Burkitt's lymphoma through a biopsy of ileocecal mass and underwent eight courses of chemotherapies with complete remission. Four months ago the positron emission tomography and computed tomography (PET-CT) revealed a new osteolytic destruction of the right mandible and a thickening of the rectal wall with increased 18F-FDG metabolism. Immunohistochemistry examination of the right mandibular bone biopsy specimen showed positivity for Kappa, CD19, CD22, and CD20. She was recruited into a clinical trial of chimeric antigen receptor (CAR)-T cell therapy with a costimulatory 4-1BB endodomain. Thirteen weeks ago, a total of 0.95*10^8 bicistronic CD19/22 CAR-T cells was infused and she developed grade 1 cytokine release syndrome (CRS) and grade 4 neutropenia. One week ago, peripheral blood B cells were reconstituted with proportions of 0.76%. On admission, a physical examination revealed the following: BP, 98/50 mmHg; heart rate, 105 beats/min; respiratory rate, 29 breaths/min; oxygen saturation, 95% on air, and temperature, 37.3°C. She was alert but in moderate respiratory distress. Her lungs sounded clear, with no crackles or wheezes. There was no palpable lymphadenopathy or relevant extrapulmonary findings. Laboratory findings showed anemia (Hb 8.8 g/dL), leukopenia (WBC 2.35 × 10⁹/L), mild elevated C-reactive protein (27.2 mg/dL) on admission, and high ferritin (388.9–449.4 μg/mL) at Days 5, 8 and 10 after admission. Serologic screening for autoimmune diseases yielded negative results. Infectious workup [blood/sputum cultures, respiratory polymerase chain reaction assays, metagenomic next-generation sequencing analysis of bronchoalveolar lavage fluid (BALF)] was negative during the first 48 h of hospitalization. CT on admission revealed the improved bilateral lower lobe consolidation with new left upper lobe opacities (Figure 1B,C). Empiric antibiotics were started. On hospital Day 7, she developed progressive hypoxemia requiring mechanical ventilation, with CT showing new bilaterally consolidation with air bronchogram, especially in the right lung and the improved consolidation in the left upper lung (Figure 1D–F). On Day 8, bronchoscopy and video-assisted thoracoscopic biopsy excluded infectious etiologies and malignancy recurrence. Histopathology demonstrated acute alveolitis and interstitial pneumonitis with alveolar hemorrhage, histiocytic infiltration, and CD8 + T-cell predominance (CD4/CD8 ratio 0.25) by flow cytometric immunophenotyping (Figure 1J–L and Table 1). No CD19 + B cells were detected in pulmonary tissue and blood specimens. Cytokine profiling revealed transient systemic IFN-γ/IL-6 elevation early in pneumonitis, while BALF showed progressive increases (IFN-γ: 7.24 → 96.62 pg/mL; IL-6: 68.21 → 471.98 pg/mL), peaking during respiratory failure (Table 1). The findings supported immunotherapy-associated pneumonitis secondary to CAR-T cell therapy. Corticosteroids were deliberately withheld to avoid potential suppression of CAR-T cell efficacy. No tocilizumab was given. Intravenous immunoglobulin was administered. She improved with supportive care alone—extubated within 48 h and discharged by Day 20. During 24-month follow-up period, B-cell aplasia recurred intermittently at 6, 18, and 24 months following transient reconstitution at 12 months. A series of follow-up imaging showed near-complete resolution by 24 months (Figure 1G–I), with sustained lymphoma remission. The advent of CAR-T cell therapy has transformed the management of relapsed/refractory hematologic malignancies. While CRS and immune effector cell-associated neurotoxicity syndrome are well-recognized acute toxicities, late-onset immune-related adverse events remain poorly characterized. We present a novel case of histopathologically confirmed pneumonitis occurring 13 weeks post-CAR-T infusion, featuring compartmentalized pulmonary inflammation without systemic CRS. To date, six cases of late immune-related pneumonitis post-CAR-T therapy have been documented, including the present case [1-3]. These encompass heterogeneous presentations: lymphocytic alveolitis (elevated ferritin, prior autologous HCT 18 months pre-CAR-T), eosinophilic pneumonia (concurrent lymphoma relapse), unspecified pneumonitis (post-HCT and ibrutinib therapy [2], or during lymphoma progression/pembrolizumab treatment [1]), interstitial pneumonitis (coinciding with CAR-T re-amplification in circulation [3]), and alveolitis (localized CRS with transient B-cell reconstitution, current study). Notably, three cases—including ours—occurred in patients maintaining complete remission (CR). But until now, no direct association could be made between CAR-T cells and the presumed immune-related events. Sun et al. described interstitial pneumonitis at 16 weeks post-CAR-T, outside the typical CRS window, with systemic IFN-γ elevation preceding pneumonitis and IL-6 surge at diagnosis [3]. Our findings diverge: peripheral IFN-γ/IL-6 exhibited only transient early increases, normalizing during pneumonitis progression, while BALF revealed escalating cytokine levels. Concurrently, CD8 + T-cell predominance re-emerged in pulmonary and bone marrow compartments, mirroring CRS dynamics after fluctuating B-cell reconstitution. Our findings suggest localized CAR-T cells reactivation, distinct from Sun et al. systemic IFN-γ/IL-6 surge [3]. Recently, localized immune cell-mediated toxicities associated with CAR-T therapy have also been reported in the treatment of autoimmune diseases [4]. Onset intervals vary markedly, ranging from 13 weeks (current study) to 19 months post-CAR-T-cell infusion [2]. Presenting features (dyspnea, hypoxia ± fever/cough) lack specificity, necessitating rigorous exclusion of alternatives. Bronchoscopy proves indispensable for excluding occult infections or neoplastic processes. Current evidence on CAR-T therapy related pneumonitis imaging remains sparse, with only three documented cases [1, 3]. Analogous to checkpoint inhibitor pneumonitis (CIP), hallmark CT features include GGO, consolidation, interlobular septal thickening, distraction branch expansion, nodules, and reticular shadow. Imaging of our case revealed two hallmark features: overlapping diffuse alveolar damage (DAD) and interstitial pneumonitis patterns, along with migratory consolidations. The histologic landscape of late CAR-T cell therapy related pneumonitis remains incompletely defined. While CIP classically demonstrates lymphocytic infiltration, granulomatous inflammation, or organizing pneumonia, our case exhibited acute alveolitis and interstitial pneumonitis with CD8+ predominance. Transbronchial biopsies in PD-1 inhibitor associated interstitial lung diseases revealed lymphocyte-predominant cellular alveolitis [5], mirroring our flow cytometry findings of pulmonary lymphocytic infiltration. Current ESMO-SITC guidelines recommend medium to high-dose glucocorticoids or even immunosuppressants for Grade 4 pneumonitis. However, strategic avoidance of corticosteroids and other immunosuppressive agents in our patient—prioritizing CAR-T persistence—yielded favorable outcomes. This underscores the need to recognize delayed immune toxicity of CAR-T therapy, characterized by compartmentalized inflammation and dynamic B-cell recovery. Our findings inform understanding of CAR-T's rare late pulmonary sequelae. Juan Qian: clinical team caring for the patient, investigation, writing – original draft, writing – review and editing, final approval of the manuscript. Yanjing Tang: clinical team caring for the patient, investigation, final approval of the manuscript. Jing Ma: pathological diagnosis, investigation, writing – original draft, final approval of the manuscript. Benshang Li: clinical team caring for the patient, investigation, final approval of the manuscript. The authors received no specific funding for this work. This study was reviewed and approved by the Institutional Review Board of Shanghai Children's Medical Center Affiliated to Shanghai Jiao Tong University School of Medicine (Reference Number: SCMCIRB-K2025317-1). The committee granted a waiver of informed consent for the publication of this anonymized case report. Informed consent was obtained from the parents, who were permitted to publish clinical images. The authors declare no conflicts of interest. The individual de-identified patient data presented in this case report are not publicly available due to privacy and ethical restrictions. Further inquiries regarding the data can be directed to the corresponding author upon reasonable request and with appropriate institutional approvals.