Early Postoperative Recurrence and Systemic Metastases Following Curative Resection of Primary Melanoma of the Lung

ABSTRACT

INTRODUCTION

Primary melanoma of the lung (PML) is exceedingly rare, accounting for approximately 0.01% of lung cancers. Because cutaneous melanoma may undergo spontaneous regression, a solitary pulmonary melanoma may represent either a true primary lesion or a metastasis from an occult regressed cutaneous primary. Therefore, rigorous exclusion of extrapulmonary primary sites is essential. PML is generally associated with a poor prognosis, and its biology and optimal management remain incompletely defined. Here, we report a case of PML with early systemic relapse after curative resection and provide a brief review of the literature.

CASE PRESENTATION

A 57-year-old woman was referred to our department (Department of General Thoracic Surgery, Osaka International Cancer Institute) for surgical management after an abnormal chest shadow was detected on routine health screening. Chest CT revealed a 2.0-cm irregular nodule in the left lower lobe with ipsilateral hilar lymphadenopathy. ¹⁸F-fluorodeoxyglucose (FDG)-PET/CT showed uptake in both lesions without evidence of disease elsewhere. Initial bronchoscopic biopsy at the referring hospital suggested a neuroendocrine tumor, whereas repeat bronchoscopic sampling at our institution raised suspicion for melanoma. Detailed dermatologic examination found no suspicious cutaneous lesions. The patient underwent open left pneumonectomy with regional lymphadenectomy; final pathology confirmed PML. Adjuvant pembrolizumab was initiated 1 month after surgery. At 4 months postoperatively, FDG-PET/CT demonstrated recurrence at the bronchial stump, mediastinal lymph nodes, pleural dissemination, and multiple systemic metastases. Combination therapy with nivolumab plus ipilimumab was administered; however, the disease progressed, and the patient died of disease 8 months after surgery.

CONCLUSIONS

This case highlights the diagnostic challenges and aggressive clinical course of PML. Thorough exclusion of extrapulmonary primary sites and comprehensive histopathologic and molecular assessment are required. Given the high risk of early systemic relapse, further investigation of multidisciplinary strategies, including perioperative systemic therapy, is warranted.

Abbreviations

ATF1

activating transcription factor 1

BRAF

v-raf murine sarcoma viral oncogene homolog B1

CCS

clear cell sarcoma

CREB1

cAMP response element-binding protein 1

EWSR1

EWS RNA-binding protein 1

FDG

¹⁸F-fluorodeoxyglucose

FISH

fluorescence in situ hybridization

HMB-45

human melanoma black-45

ICI

Immune checkpoint inhibitor

MITF

microphthalmia-associated transcription factor

NRAS

neuroblastoma RAS viral oncogene homolog

PD-L1

programmed death-ligand 1

PML

primary melanoma of the lung

PRAME

preferentially expressed antigen in melanoma

SOX10

SRY-box transcription factor 10

SUVmax

maximum standardized uptake value

INTRODUCTION

PML is an exceedingly rare malignancy, accounting for approximately 0.01% of all lung cancers and about 0.4% of all melanomas.¹⁾ Since melanoma can undergo spontaneous regression,²⁾ a solitary pulmonary lesion histologically identified as melanoma must be carefully evaluated to distinguish a true primary tumor from a metastasis originating from an occult, regressed extrapulmonary site. This distinction is critical for both treatment planning and prognostic assessment.

Furthermore, PML is associated with a poorer prognosis compared with cutaneous melanoma, and its responses to radiotherapy and chemotherapy are limited; consequently, established treatment guidelines have yet to be defined.³⁾ Although surgical resection is commonly performed for resectable cases, early postoperative recurrence and distant metastasis have been reported.⁴⁾

Here, we report a case of PML treated with curative resection that nonetheless developed early recurrence and systemic metastases, followed by a brief review of the relevant literature.

CASE PRESENTATION

Case: A 57-year-old woman.

Chief complaint: Abnormal chest radiographic finding.

Past medical history: Ovarian endometriotic (chocolate) cyst, dyslipidemia, and atopic dermatitis.

Smoking history: Never smoker.

History of present illness: An abnormal chest shadow was detected during a screening exam, and the patient visited a referring hospital. Chest CT revealed a nodule in the left lower lobe. A bronchoscopic biopsy was suggestive of a neuroendocrine tumor, and the patient was subsequently referred to our department (Department of General Thoracic Surgery, Osaka International Cancer Institute) for surgical treatment.

Laboratory tests at initial presentation: Complete blood count and serum biochemistry were within normal limits. Tumor markers, including carcinoembryonic antigen, squamous cell carcinoma antigen, cytokeratin 19 fragment, and progastrin-releasing peptide, were not elevated.

Chest radiography: A nodular shadow was observed at the left hilum and in the left lower lung field.

Contrast-enhanced chest and abdominal CT: A 2.0-cm nodule was identified in the left lower lobe (Fig. 1A and 1B), along with enlargement of the hilar lymph nodes (#11 and #12l) (Fig. 1C and 1D).

Fig. 1. Contrast-enhanced CT findings of the thorax and abdomen. (A) Axial image showing a 2.0-cm nodule in the left lower lobe. (B) Coronal image of the primary tumor. (C) Axial image showing ipsilateral hilar lymphadenopathy. (D) Coronal image of the hilar lymphadenopathy.

FDG-PET/CT: FDG uptake corresponding to the left lower-lobe nodule was observed (Fig. 2B), with an SUVmax of 8.45. FDG uptake was also present in the enlarged hilar lymph nodes, with an SUVmax of 21.21 (Fig. 2C). No abnormal uptake was detected at other sites.

Fig. 2. FDG-PET scans showed abnormal uptake (A) in the nodular region of the lung (B, SUVmax = 8.45) and in the hilar lymph nodes (C, SUVmax = 21.21).

FDG, ¹⁸F-fluorodeoxyglucose; SUVmax, maximum standardized uptake value

Brain MRI: No findings suggestive of intracranial metastasis were identified.

Pulmonary perfusion scintigraphy: Perfusion defects were observed at the primary tumor and the hilar lymphadenopathy, whereas perfusion distribution in the remaining lung fields was homogeneous.

Pulmonary ventilation scintigraphy: Ventilation distribution was homogeneous in both lungs, with no localized ventilatory defects.

Pathological diagnosis: Due to the inadequate sample volume from the initial biopsy at the referring hospital for comprehensive subtyping and molecular testing, repeat bronchoscopy was conducted. The hilar lymph node specimen demonstrated a tumor with low mitotic activity and subtle nucleoli (Fig. 3A). The tumor cells expressed melanocytic markers—HMB-45, Melan-A, MITF, and SOX10—confirming a diagnosis of metastatic melanoma (Fig. 3B–3E).

Fig. 3. Bronchoscopic biopsy. The specimen from the bronchoscopic biopsy shows that the nodule of the left lung seems to be a melanoma metastasis to the lymph node (A, hematoxylin and eosin stain, ×400 original magnification), and the cells were positive for HMB-45 (B), Melan-A (C), MITF (D), and SOX10 (E) (×400).

HMB-45, human melanoma black-45; MITF, microphthalmia-associated transcription factor; SOX10, SRY-box transcription factor 10

Considering the possibility of pulmonary and lymph node metastases from a cutaneous primary melanoma, the patient underwent a comprehensive dermatologic evaluation by the dermatologic oncologist. No suspicious cutaneous lesions or post-regression changes were identified, supporting a working diagnosis of PML. The oral cavity was inspected by the oral care team as part of perioperative oral management, and no suspicious mucosal lesions were identified. After multidisciplinary discussion with the Department of Dermatologic Oncology, surgical treatment was selected because the disease was considered confined to the lung and deemed to be resectable. In addition, evidence for nonsurgical modalities in PML is limited, and responses to systemic therapy may be less favorable in non-cutaneous melanoma than in cutaneous melanoma, which also supported the decision for upfront resection. Preoperatively, left lower lobectomy was planned if complete resection appeared feasible; however, pneumonectomy was also considered likely given the suspected hilar nodal involvement and its proximity to the left pulmonary artery. Based on the preoperative assessments, including ventilation–perfusion scintigraphy, the patient was considered able to tolerate pneumonectomy if required.

Operative findings

Surgery was performed approximately 2 months after the initial presentation and 12 days after completion of the preoperative evaluation; this interval largely reflected the time required for pathological re-evaluation and assessment of operability in anticipation of possible pneumonectomy. Although preservation of the left upper lobe was initially attempted, intraoperative findings revealed enlargement of lymph node station #12u, and a frozen-section examination confirmed metastatic involvement. Furthermore, the hilar lymph nodes were firmly adherent to the main trunk of the left pulmonary artery; therefore, the procedure was converted to a left pneumonectomy. Mediastinal lymph node evaluation included stations #5 and #7, and hilar/intrapulmonary lymph nodes (stations #10, #11, #12u, and #12l) were also dissected. The operative time was 266 min and the estimated blood loss was 241 mL. Intraoperative pleural lavage cytology was negative. Intraoperative frozen-section examination confirmed a negative bronchial resection margin.

Histopathological findings

The resected tumor showed focal black pigmentation on gross examination. Atypical cells with marked nuclear pleomorphism proliferated in nests with focal melanin pigmentation (Fig. 4A and 4B). Immunohistochemical analysis showed positivity for HMB-45, Melan-A, and SOX10 (Fig. 5A–5C), whereas mutation-specific staining for BRAF V600E and NRAS Q61R was negative (Fig. 5D and 5E). PRAME was also negative (Fig. 5F). Although CCS was considered in the differential diagnosis, FISH showed no evidence of the EWSR1–ATF1 fusion gene or the EWSR1–CREB1 fusion gene, which are characteristic of CCS. Consequently, these findings were consistent with a diagnosis of melanoma. PD-L1 tumor proportion score (clone 28-8) was 1.2%. Junctional change, characterized by the presence of melanoma cells within the bronchial epithelium, was identified (Fig. 6). Venous and lymphatic invasion were present, and metastases were observed in hilar/intrapulmonary lymph nodes (stations #10, #11, #12u, and #12l), whereas mediastinal lymph nodes (stations #5 and #7) were negative. The bronchial resection margin was negative on permanent pathology. The closest distance from the bronchial resection margin to the nearest tumor component was 78 mm on permanent pathology. The final diagnosis was PML with hilar lymph node metastases.

Fig. 4. Histological findings of the surgical specimen (hematoxylin and eosin stain). The nodule in the left lower lobe showed nests of atypical cells with marked nuclear pleomorphism and focal melanin pigmentation (A, hematoxylin and eosin stain, ×40 original magnification; B, hematoxylin and eosin stain, ×400 original magnification).

Fig. 5. Immunohistochemical findings. Tumor cells were positive for HMB-45 (A), Melan-A (B), and MITF (C), and negative for BRAF V600E (D), NRAS Q61R (E), and PRAME (F) (×400 original magnification).

BRAF, v-raf murine sarcoma viral oncogene homolog B1; HMB-45, human melanoma black-45; MITF, microphthalmia-associated transcription factor; NRAS, neuroblastoma RAS viral oncogene homolog; PRAME, preferentially expressed antigen in melanoma

Fig. 6. Junctional change, defined as the presence of melanoma cells within the bronchial epithelium, was identified (hematoxylin and eosin stain, ×400 original magnification).

Postoperative clinical course

The postoperative course was uneventful.

Postoperative treatment and recurrence

Adjuvant pembrolizumab was initiated 1 month after surgery. Four months postoperatively, after 2 cycles of pembrolizumab, a new subcutaneous nodule was detected, prompting an FDG-PET/CT. The imaging revealed multifocal recurrence, including at the bronchial stump, left mediastinal lymph nodes, and the pleura, along with systemic metastases, including myocardial metastases (Fig. 7). Although treatment was escalated to combined nivolumab and ipilimumab for metastatic PML, the tumor proved refractory to therapy. The patient died of the disease 8 months after the initial surgery.

Fig. 7. FDG-PET scans showed abnormal uptake suggestive of bronchial stump recurrence, left mediastinal lymph node recurrence, left pleural dissemination recurrence, and multiple systemic metastases.

FDG, ¹⁸F-fluorodeoxyglucose

DISCUSSION

PML is exceedingly rare, and confirming a primary pulmonary origin requires rigorous exclusion of extrapulmonary primaries. A long-standing hypothesis suggests that PML originates from ectoderm-derived melanocytes that aberrantly migrate to, and persist within, the airway mucosa during embryogenesis.⁵⁾ Accordingly, junctional change, defined as melanoma cells within the bronchial epithelium, has been reported as a finding supportive of a primary pulmonary origin.⁶⁾ Diagnostic criteria incorporating junctional change and the absence of extrapulmonary melanoma at diagnosis have been proposed.⁷⁾ In the present case, junctional change was identified on histologic examination of the resected specimen, and all proposed diagnostic criteria were satisfied. However, because intraepithelial melanoma cells may also represent endobronchial extension of metastatic disease, PML should not be diagnosed on the basis of junctional change alone.¹⁾ Therefore, our diagnosis was based on an integrated clinicopathologic assessment rather than reliance on this finding in isolation.

Spontaneous regression can occur in melanoma.²⁾ Therefore, even when melanoma presents as a solitary pulmonary lesion, pulmonary metastasis from an occult regressed extrapulmonary primary should be considered. Because post-regression changes such as hyperpigmentation, depigmentation, or scar-like alterations may persist,^2,8,9) careful specialist examination and review of prior medical records are essential when excluding extrapulmonary primary sites. In this case, dermatologic oncologists identified neither a suspicious primary cutaneous lesion nor findings suggestive of spontaneous regression, supporting management under a working diagnosis of a primary pulmonary lesion.

Immunohistochemical findings can also provide supportive information when considering the primary site of melanoma, although interpretation requires caution. BRAF V600E and NRAS Q61R have been reported more frequently in cutaneous melanoma than in melanomas arising from other primary sites.¹⁰⁾ In this case, gene panel testing was not performed, but mutation-specific immunohistochemistry for BRAF V600E and NRAS Q61R was negative, which was not strongly supportive of a cutaneous primary. Nevertheless, these markers are not definitive for site assignment, and their absence does not exclude cutaneous melanoma; therefore, they should not be used in isolation to determine the primary site.

Although no single finding is definitive, a multifaceted diagnostic approach that integrates clinical, radiological, and pathological evidence—as demonstrated in this case—is essential to substantiate a diagnosis of PML. However, ophthalmologic examination including fundoscopy and gastrointestinal endoscopy/colonoscopy, as well as formal evaluation of the nasal cavity to exclude mucosal melanoma, were not systematically performed in this case; therefore, an occult extrapulmonary primary cannot be excluded with absolute certainty. In future similar cases, these additional evaluations should be considered to further strengthen diagnostic confidence.

Distinguishing melanoma from CCS, a malignant soft-tissue tumor with melanocytic differentiation, was important in this case. While PRAME immunohistochemistry can be a helpful adjunct, it lacks absolute specificity. Although melanoma typically exhibits diffuse, strong nuclear PRAME expression and CCS is more frequently negative or only focally positive, reported exceptions preclude its use as a standalone diagnostic tool.¹¹⁾ In this lesion, PRAME was negative and did not allow definitive exclusion of CCS. Ultimately, the diagnosis of melanoma was supported by marked pleomorphism and severe nuclear atypia, together with FISH showing no evidence of EWSR1–ATF1 or EWSR1–CREB1 fusions, which are characteristic of CCS.¹²⁾ These findings highlight the value of integrating morphology, immunohistochemistry, and targeted molecular testing in diagnostically challenging melanocytic tumors.

Due to the absence of established therapeutic standards for PML, management is often extrapolated from strategies used for cutaneous melanoma. Nevertheless, prognosis remains poor; a retrospective study of 76 patients reported a mean survival of 14.3 months.³⁾ Although surgical resection is frequently pursued when complete resection is feasible, early postoperative recurrence and distant metastasis have been reported, suggesting that surgery alone may be insufficient in some patients.⁴⁾ Consistent with this, our patient developed rapid multiorgan metastases and died shortly after curative-intent resection, indicating that PML can carry a high risk of recurrence that may not be adequately controlled by local therapy alone. Vessel-first approaches, including early division of the drainage vein, have been discussed in lung cancer surgery as a potential measure to reduce intraoperative tumor cell dissemination.¹³⁾ In this case, because we attempted left lower lobectomy until the final stage, we prioritized hilar dissection to determine resectability and the extent of resection and did not perform early pulmonary vein division, which can cause congestion and compromise exposure. While the role of a vessel-first strategy in PML remains uncertain, it may merit evaluation in biologically aggressive cases.

ICIs are central to systemic therapy for cutaneous melanoma. Although neoadjuvant ICI therapy has demonstrated clinical benefit in resectable stage III cutaneous melanoma, this approach is not approved in Japan.¹⁴⁾ Evidence regarding ICIs for PML is exceptionally sparse; to our knowledge, only a single case has been reported where ICI-induced tumor regression facilitated salvage surgery.¹⁵⁾ Nevertheless, a future neoadjuvant strategy may merit consideration in selected patients to assess treatment responsiveness and detect early systemic progression before definitive surgery, particularly in cases with hilar lymph node involvement or those in which pneumonectomy may be required. However, several caveats should be noted. Neoadjuvant immunotherapy for melanoma is not an approved indication in Japan, and much of the available evidence is derived from Western cohorts; some reports suggest relatively lower efficacy of ICIs in Asian patients than in White patients.¹⁶⁾ Additionally, non-cutaneous melanomas often show limited responsiveness to ICIs relative to their cutaneous counterparts.¹⁷⁾ Accordingly, extrapolation of international data to Japanese patients with PML should be made cautiously, and the role of neoadjuvant approaches in PML requires further evaluation.

This case highlights the diagnostic and therapeutic challenges of PML. Distinguishing PML from pulmonary metastasis of an occult extrapulmonary primary, potentially after spontaneous regression, remains difficult and requires integrated clinicopathologic assessment, including careful systemic evaluation, imaging, and histopathology with ancillary testing when appropriate. Despite curative-intent resection and adjuvant ICI therapy, the patient developed rapid systemic relapse, underscoring the aggressive clinical course that can occur. Further case accumulation with detailed clinicopathologic characterization is needed to refine diagnostic criteria and inform optimal multimodality management.

CONCLUSIONS

This case represents a rare occurrence of PML where, despite an extensive systemic workup and histopathologic confirmation supporting the diagnosis and curative-intent resection, rapid multi-organ recurrence developed shortly after surgery. The diagnosis of PML necessitates the rigorous exclusion of extrapulmonary primary sites through a comprehensive framework that integrates meticulous clinical evaluation, detailed histopathologic findings, and whole-body imaging. From a therapeutic perspective, further investigation into multidisciplinary strategies, particularly perioperative immunotherapy-based approaches, is warranted to improve outcomes for this aggressive malignancy.

DECLARATIONS

Funding

Not applicable.

Authors’ contributions

H.C. collected the clinical data, analyzed and interpreted the patient data, and drafted the manuscript.

T.T., H.S., M.K., T.I., and T.M. contributed to perioperative patient management and critically revised the manuscript.

K.H. performed the pathological evaluation and interpreted the histopathological findings.

S.O. contributed to dermatologic evaluation and postoperative systemic treatment and critically revised the manuscript.

J.O. supervised the study and critically revised the manuscript.

All authors read and approved the final manuscript.

Availability of data and materials

Not applicable.

Ethics approval and consent to participate

This work did not require ethical considerations or approval.

Consent for publication

Informed consent for publication of this case report was obtained from the patient.

Competing interests

The authors have no competing interests associated with this manuscript.