Radiological and Pathological Analysis of Pembrolizumab-Associated Lung Lesions: Diagnostic Challenges and Management

Introduction

Advanced-stage malignant melanoma presents significant therapeutic challenges and often carries a poor prognosis. Pembrolizumab, a humanized monoclonal antibody targeting the programmed cell-death protein 1 (PD-1) immune checkpoint, has shown efficacy in prolonging progression-free and overall survival in these patients [1]. Despite its benefits, pembrolizumab is associated with adverse effects, including drug-induced lung injury. Studies suggest that pembrolizumab monotherapy for malignant melanoma leads to drug-induced lung injury in approximately 1–3% of cases [2–5]. However, due to the rarity of this complication, its mechanisms are not fully understood, and definitive treatment protocols have yet to be established. Additionally, imaging results can vary significantly, and instances of pathologically confirmed diagnoses are uncommon.

In this report, we detail a case of pembrolizumab-induced lung injury distinguished by unique radiographic features. This case underscores the diagnostic complexities and the critical need for meticulous clinical evaluation to optimize patient outcomes. The discussion also aims to contribute to a deeper understanding of the pathological processes involved and to encourage the development of more effective management strategies for this serious complication.

Case Report

A 40-year-old woman with a previous diagnosis of left gluteal malignant melanoma was referred to our clinical department due to abnormal findings on chest radiography. The patient exhibited neither cough, fever, nor breathlessness at the time when the pathological pulmonary findings were identified. She had previously undergone surgery and received immune checkpoint inhibitors (nivolumab and ipilimumab) at the Department of Dermatology in our institution. During the course of nivolumab and ipilimumab therapy, she experienced severe hepatic injury (Grade 4) and developed type I diabetes mellitus (Grade 4). The diagnosis and treatment of liver injury were conducted by a hepatologist, whereas an endocrinologist from our medical institution managed the diagnosis and treatment of diabetes mellitus. The diagnoses of liver injury and diabetes mellitus were established based on criteria from international guidelines [6,7].

Subsequently, her treatment was switched to pembrolizumab, which she received for 31 courses. She had been diagnosed with dyslipidemia and had allergies to metals and adhesive bandages. The patient had no history of abnormal computed tomography (CT) findings or other diseases including interstitial lung disease or lung tumors. At the time of presentation to our clinic, she was being treated with atorvastatin, sitagliptin, and insulin. On presentation, her vital signs were stable, with a body temperature of 37.2°C, blood pressure of 114/72 mmHg, pulse of 108 beats/minute, oxygen saturation of 97% on room air, and a respiratory rate of 12 breaths/min. Physical examination was largely unremarkable, with no evidence of ocular or thyroid abnormalities, thoracic deformities, or hepatosplenomegaly. She had edema in the left lower leg secondary to lymph node dissection. Laboratory tests revealed elevated levels of surfactant protein A (SP-A), SP-D, and KL-6 (Table 1). The plain radiographs demonstrate bilateral patchy opacities, with a subtle reticulonodular pattern, more apparent in the lower lung zones (Figure 1). CT scans revealed multiple bilateral nodular opacities of varying sizes, predominantly peripheral in distribution. In addition to the nodules, there were areas of ground-glass opacity that were particularly noticeable in the lower lobes. These findings were accompanied by mild interlobular septal thickening (Figure 2). The bronchoscopic examination provided detailed visualization of the tracheal carina, as well as the bifurcation of the right upper and lower lobe bronchi, the right middle bronchus, and the segmental bronchi in the right lower lobe (B9 and B10) (Figure 3A–3D). Furthermore, an endobronchial ultrasound (EBUS) was utilized to facilitate accurate localization and biopsy of the nodule identified in the right lower lobe, ensuring precise sampling of the lesion (Figure 3E).

We considered several potential differential diagnoses, including drug-induced lung injury, metastatic lung tumor, multiple pulmonary infarctions, and infection. The total cell count of the bronchoalveolar lavage fluid was 2.3×105 cells/mL, with histiocytes comprising 32%, lymphocytes 58%, neutrophils 6%, and eosinophils 4%, indicating an elevated total cell count with lymphocytes predominating (Figure 4A). Histopathological examination of the lung biopsy specimens revealed chronic and acute inflammatory changes, with histiocytes being the predominant inflammatory cells and enlargement or swelling of the alveolar epithelial cells (Figure 4B, 4C). Cytological assessments, along with cultures for bacteria, fungi, and acid-fast bacilli, all yielded negative results. Furthermore, Pneumocystis jirovecii DNA polymerase chain reaction (PCR) testing on bronchoalveolar lavage fluid (BALF) was also negative. However, due to institutional protocols, routine DNA sequencing for pathogenic microorganisms in BALF was not performed, which limited our ability to conduct this specific analysis. Based on the bronchoalveolar lavage fluid and biopsy histological findings, the diagnosis of drug-induced lung injury (Common Terminology Criteria for Adverse Events, Grade 1) was suggested.

Pembrolizumab was implicated as the probable etiological agent. Pembrolizumab administration was ceased, and another chest CT scan was conducted 1 month later. The follow-up imaging revealed an expansion of the infiltrative shadow along the bronchovascular bundle, with certain areas demonstrating traction bronchiectasis (Figure 5A). Upon this finding, the patient was admitted, and steroid pulse therapy was initiated. The treatment course following steroid pulse administration is outlined in Figure 5B. Following steroid pulse therapy, the pulmonary opacities showed improvement (Figure 5A). Subsequently, a gradual tapering of the steroid dosage was commenced, culminating in the complete discontinuation of steroid therapy (Figure 5B). Since then, the patient has been under close observation without recurrence of either drug-induced lung injury or malignant melanoma.

Discussion

Malignant melanoma, arising from cancerous melanocytes, presents the highest mortality rate among skin cancers, posing a significant clinical challenge [8]. The global incidence of malignant melanoma exhibits substantial geographical variability, measured as new cases per 100 000 population [8]. The most affected regions include Australia/New Zealand, with Auckland, New Zealand, demonstrating the highest incidence rates for both men and women [9]. Conversely, Africa and Asia report considerably lower rates, with Japan averaging only 1 case per 100 000 individuals [10].

While the precise etiology of malignant melanoma remains elusive, ultraviolet (UV) radiation emerges as a key contributor to observed geographical disparities [11]. Sunlight emits 2 carcinogenic UV rays: UVA (315–400 nm) and UVB (280–315 nm). UVA, constituting 95% of sunlight’s UV rays, penetrates deeper into the dermis, whereas UVB induces more significant gene damage [12]. Interestingly, some melanomas develop independently of UV exposure, with germline mutations in CDKN2A implicated [13]. Additionally, BRAF V600E and NRAS mutations, present in 50% and 15% of melanomas, respectively, correlate with a more aggressive disease course, leading to the development of BRAF V600E-targeted therapies [13,14].

Immune checkpoint inhibitors have revolutionized oncology by empowering T cells to combat cancer cells through the inhibition of suppressive signals. Among these inhibitors, pembrolizumab, a humanized IgG monoclonal antibody targeting the PD-1 checkpoint, has shown efficacy in treating malignant melanoma [1,8]. Despite its success, treatment with anti-PD-1 antibodies, notably pembrolizumab, is linked to serious adverse effects, such as pneumonitis [15,16]. The incidence of this drug-induced condition often correlates with the presence of risk factors like chronic obstructive pulmonary disease (COPD), bronchial asthma, or prior chest radiotherapy [16–18]. Interestingly, our case showed none of these common predisposing factors. A meta-analysis has found that pneumonitis rates vary by the type of malignancy treated [15]. Notably, it occurs less frequently in malignant melanoma than in non-small cell lung cancer (NSCLC) or renal cell carcinoma [1,18,19]. The higher incidence in NSCLC patients can be attributed to underlying issues such as smoking, COPD, and interstitial lung disease [16,17,20]. However, the reasons for its high occurrence in renal cell carcinoma patients remain elusive.

Our current case developed pembrolizumab-associated pneumonitis. The incidence of pneumonitis induced by pembrolizumab is 5% in NSCLC patients and less than 1% in patients with malignant melanoma [1,18,19]. Due to its rarity, there is limited literature on this pathological condition, and the underlying mechanisms are not fully understood. Furthermore, standardized protocols for the management of pembrolizumab-induced lung injury are lacking. Cases of pneumonitis in patients with malignant melanoma treated with anti-PD-1 antibodies often show ground-glass opacities and infiltrative shadows in the lungs [21–26]. However, our present case exhibited previously unreported radiological features characterized by multiple nodular shadows in both lungs, making the diagnosis of lung abnormalities challenging. Initially, we considered metastatic lung tumors, pulmonary embolism, pulmonary mycosis, and pulmonary mycobacterial infection as differential diagnoses rather than drug-induced lung injury. A previous case reported by Arays et al also showed atypical lung radiological images for pneumonitis, necessitating a surgical lung biopsy to diagnose organizing pneumonia [27]. Among the 11 reported cases of pneumonitis induced by anti-PD-1/PD-L1 antibodies in malignant melanoma patients, only 4, including our present case, were pathologically diagnosed through transbronchial lung biopsy (TBLB) [21,23,27]. Therefore, while some cases omit bronchoscopy and surgical lung biopsy, tissue sampling remains essential for accurate diagnosis given the diverse potential causes. Our case underscores the importance of recognizing that atypical radiological findings of pneumonitis can occur in daily clinical practice and highlights the diagnostic value of TBLB in confirming drug-induced lung injury.

Conclusions

In summary, this case underscores the potential for pembrolizumab to cause lung injury in cancer patients. Despite its effectiveness in prolonging survival, clinicians must remain vigilant for respiratory complications. The patient’s positive response to steroid therapy and subsequent improvement emphasize the significance of early detection and management of pembrolizumab-related pulmonary adverse effects. Continuous monitoring is essential to ensure the patient’s sustained recovery without reliance on steroid therapy. Additionally, this case highlights the challenge of diagnosing lung lesions based solely on imaging, necessitating bronchoscopy with tissue sampling as a critical diagnostic tool.