低分化肺腺癌同时表达 ALK 与 CD30：酷似 ALK 阳性间变性大细胞淋巴瘤的诊断陷阱Poorly differentiated lung adenocarcinomas with concurrent ALK and CD30 expression: a diagnostic pitfall mimicking ALK‐positive anaplastic large‐cell lymphoma

Scope: PMC open-access full text (CC BY-NC-ND 4.0) — abstract, introduction, materials and methods, results, discussion, author contributions, with Figures 1–3 and Tables 1–2. Conflict-of-interest statement, supporting information, data availability statement and references are not included (citation numbers are retained).

Abstract

Although ALK rearrangement is a well‐established alteration in lung adenocarcinoma (LUAD), CD30 expression in this setting remains poorly characterized. We analysed a series of poorly differentiated LUADs with concurrent ALK and CD30 expression that closely resemble ALK‐positive anaplastic large‐cell lymphoma (ALCL), focusing on their clinicopathological and molecular features as well as potential diagnostic and therapeutic implications.

We retrospectively searched multi‐institutional case databases and the published literature, ultimately identifying eight cases of poorly differentiated LUAD with concurrent ALK and CD30 expression. Clinical, histomorphological, immunohistochemical and molecular features, along with treatment responses and outcomes, were systematically recorded and analysed. Patients had a median age of 47.5 years (range, 26–67 years), with equal sex distribution and all presented with advanced‐stage disease (stage III–IV). The diagnostic challenge posed by these tumours was substantial: four cases were initially radiologically suspected to represent aggressive lymphoma, and one patient had been misdiagnosed and actively treated as ALCL at an outside institution before referral. Histologically, all tumours consisted of diffuse sheets of markedly pleomorphic large cells with prominent nucleoli and brisk mitotic activity, a morphology strikingly reminiscent of ALCL hallmark cells. Most cases showed diffuse or focal expression of epithelial markers, with variable TTF‐1 positivity. The Ki‐67 proliferation index ranged from 50% to 95%. All cases co‐expressed ALK and CD30, lacked lymphoma‐associated immunophenotypic markers and showed no evidence of clonal T‐cell receptor gene rearrangement (TCR). Molecular sequencing confirmed that all ALK fusion breakpoints corresponded to previously described LUAD‐associated variants, and all patients achieved marked responses to ALK‐directed therapy.

Poorly differentiated LUAD with dual ALK and CD30 expression is a rare but diagnostically challenging entity that can closely simulate ALK‐positive ALCL. Awareness of this mimic is critical to ensure accurate diagnosis and timely initiation of appropriate targeted therapy.

Introduction

Lung cancer is one of the two most common malignancies worldwide and a leading cause of cancer‐related death, with mortality concentrated predominantly among patients with advanced‐stage disease. Of all lung cancer cases, approximately 85% are classified as non‐small cell lung cancer (NSCLC) and lung adenocarcinoma (LUAD) accounts for roughly half of these. ¹ , ² , ³ , ⁴ , ⁵ LUAD exhibits six principal histological growth patterns: lepidic, papillary, acinar, cribriform, solid and micropapillary. ⁴ , ⁶ Tumours are graded as well, moderately or poorly differentiated based on the dominant pattern and its proportional extent within the neoplasm; those in which cribriform, solid or micropapillary architecture collectively exceeds 20% are designated poorly differentiated. ⁴ This grade carries a worse prognosis and a markedly higher risk of distant metastasis relative to well‐ or moderately differentiated tumours. ⁴ , ⁷ Immunohistochemically, LUAD characteristically expresses the epithelial markers pan‐keratin, EMA, CK7 and/or CK8/18, together with the lineage‐specific markers TTF‐1 and Napsin A, ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² although in poorly differentiated variants these markers may be substantially diminished or absent, a finding that can complicate pathological diagnosis. ⁹ , ¹³ At the molecular level, anaplastic lymphoma kinase (ALK) gene rearrangements represent one of the key oncogenic driver alterations in LUAD, occurring in 3%–5% of patients. ¹⁴ The prototypical rearrangement arises from an intrachromosomal inversion on chromosome 2p (bands p21–p23), fusing EML4 with ALK to generate the EML4::ALK oncoprotein. Additionally, less common fusion partners include KIF5B, KLC1, TPM3, STRN, ASXL2, ATP6V1B1, SPDYA and PRKAR1A. ¹⁵ , ¹⁶ Selective ALK tyrosine kinase inhibitors (TKIs) have since transformed the management of ALK‐positive LUAD and are now the established standard of care for this molecular subset. ¹⁷

CD30, a member of the tumour necrosis factor receptor (TNFR) superfamily, is an established diagnostic marker for several haematolymphoid neoplasms, most notably anaplastic large‐cell lymphoma (ALCL), classical Hodgkin lymphoma (cHL) and embryonal carcinoma ¹⁸ , ¹⁹ Among normal haematopoietic cells, activated B and T lymphocytes express CD30 strongly, while activated monocytes and eosinophils show only weak positivity. ¹⁹ , ²⁰ CD30 expression has additionally been reported in select epithelial solid tumours, including mesothelioma, nasopharyngeal carcinoma and poorly differentiated gastric carcinoma. ²¹ , ²² , ²³ In ALCL, CD30 is expressed in virtually all cases regardless of ALK status, and sustained CD30 signalling promotes tumour cell survival and proliferation while enabling immune evasion. ²⁰ ALK‐positive ALCL predominantly affects children and young adults, whereas ALK‐negative ALCL is more prevalent in middle‐aged and older patients, yet the two subtypes are cytomorphologically indistinguishable. ²⁰ , ²⁴ Despite encompassing morphological variants such as small cell, histiocyte‐rich and Hodgkin‐like patterns, ALK‐positive ALCL invariably contains hallmark anaplastic large cells. ²⁴ The t(2;5)(p23;q35) translocation, generating the NPM1‐ALK fusion oncoprotein, is the predominant molecular event; alternative, less common fusion partners include TPM3, ATIC, TFG, CLTC, MSN, TPM4, MYH9, RYH9 and RNF213. ²⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ The anti‐CD30 antibody–drug conjugate brentuximab vedotin (BV) is approved for ALCL and has shown considerable clinical efficacy. ¹⁸ , ³⁰ , ³¹ Whether CD30 is expressed in LUAD and whether it co‐occurs with ALK remains poorly understood.

Herein, we report eight cases of poorly differentiated LUAD composed predominantly of anaplastic cells, which closely mimic ALK‐positive ALCL in clinical presentation, histomorphology and dual CD30/ALK immunoreactivity. Through systematic analysis of the clinicopathological, morphological, immunohistochemical and molecular features of these cases, we seek to raise awareness of this diagnostically challenging LUAD variant, reduce the risk of misclassification and support timely, appropriate treatment.

Materials and methods

Case Identification and Literature Review

A total of 217 ALK‐positive LUADs diagnosed between January 2014 and October 2025 were retrospectively retrieved from the pathology archives of four tertiary medical centres in Guangdong, China, including 62 resection specimens and 155 biopsy specimens. CD30 immunohistochemistry was reviewed in all cases, and tumours showing moderate‐to‐strong CD30 expression were recorded. Seven poorly differentiated LUADs with concurrent ALK and CD30 expression were subsequently identified and included in the study. A complementary PubMed search using the terms ‘CD30’, ‘ALK’ and ‘lung’ in the title and abstract fields identified one additional eligible English‐language case report, and clinicopathological data were subsequently extracted. Cases with ALK and CD30 co‐expression but lacking expression of either epithelial markers or TTF‐1 were excluded from the study. The final study cohort comprised eight cases in total: seven from institutional archives and one from the published literature. Diagnostic material was obtained from the lung (n = 4), ileum (n = 1), cervical lymph node (n = 1), lumbar vertebra (n = 1) and thoracic vertebra with concurrent pulmonary involvement (n = 1); the ileal specimen was surgically resected, whereas all others were obtained by core needle biopsy. Clinicopathological characteristics, treatment regimens and survival outcomes were systematically collected and analysed. Written informed consent was obtained from all patients, and the study protocol was approved by the institutional ethics committee (approval no. SL‐G2023‐039‐01).

Histomorphological and Immunohistochemical Analysis

All specimens were formalin‐fixed, paraffin‐embedded, sectioned at 4 μm and stained with haematoxylin and eosin (H&E). Cases were included if all of the following diagnostic criteria were met: (1) presence of a primary pulmonary mass; (2) strong or focal positivity for epithelial markers, including pan‐cytokeratin (AE1/AE3), CK7 and/or CK8/18 (Case 8 showed low expression of the epithelial markers Claudin 4 and MOC31); (3) strong or focal TTF‐1 expression; (4) moderate‐to‐strong CD30 expression and strong ALK expression (clone D5F3); (5) absence of immunophenotypic markers associated with ALCL or other lymphoma subtypes; (6) no evidence of clonal T‐cell receptor gene rearrangement (TCR). All cases were independently reviewed by two pathologists; histological assessment encompassed architectural growth patterns, cytoplasmic features, nuclear morphology, mitotic activity, tumour necrosis and lymphoplasmacytic infiltration. Each immunohistochemical run was performed with concurrently processed positive and negative controls, and antibody clones, sources and working dilutions are detailed in Table S1.

Targeted Next‐Generation Sequencing (NGS)

Genomic DNA isolated from formalin‐fixed, paraffin‐embedded (FFPE) tumour specimens was processed for targeted NGS via the OncoScreen Plus panel (520 cancer‐associated genes; 1.64 Mb). Library construction, hybridization capture and amplification were carried out in accordance with the manufacturer's protocols. The resulting libraries were sequenced on an Illumina NextSeq 500 platform in paired‐end mode, at a mean sequencing depth of approximately 1000×.

Results

Clinicopathological Characteristics

Clinical data from all eight patients (including Case 8 from the literature ³² ) are summarized in Table 1. The cohort comprised four males and four females with a median age of 47.5 years (range, 26–67 years). Presenting symptoms varied widely, including chest pain, cough, cervical lymphadenopathy, low back pain, and, in patients with bowel metastasis, such as intestinal obstruction or gastrointestinal bleeding. Chest CT revealed a pulmonary mass in all cases, and lymphadenopathy involving regions commonly affected by lung cancer metastasis (mediastinal, hilar or supraclavicular nodal stations) was identified in seven of eight patients. Distant metastases were frequent: osseous involvement (iliac bone, sternum and/or lumbar vertebrae) was documented in four patients and gastrointestinal metastases affecting the small bowel, stomach or pancreas in three. All patients presented with stage III–IV disease. Seven had no prior oncological history; history was unavailable for the eighth. Lymphoma was the initial imaging impression in four cases. Case 2 had been misdiagnosed with ALCL at an outside institution and treated with four cycles of BV plus cyclophosphamide, doxorubicin, vincristine and prednisone (BV‐CHOP), after which the disease progressed. Following diagnostic revision to poorly differentiated LUAD, the patient was switched to alectinib and achieved marked symptomatic improvement.

AP, Adriamycin+Cisplatin; F, female; LN, lymph node; M, male; N, no; NA, not available; NED, no evidence of disease; PD, progressive disease; PR, partial response; SD, stable disease; Y, yes.

PMID: 38032068.

Histopathological Findings

All cases showed diffuse proliferation of anaplastic large cells; Case 8 alone also exhibited focal papillary architecture. The tumour cells uniformly displayed vesicular nuclei, conspicuous nucleoli and frequent binucleation (Figure 1A), with vesicular nuclei being especially prominent in Cases 3, 5 and 7 (Figure 1B). Case 4 was distinguished by marked nuclear pleomorphism spanning small, intermediate and large forms (Figure 1C). In the metastatic lymph nodes of Case 7, tumour cells adopted a predominantly sinus‐based, sheet‐like arrangement (Figure 1D). Focal cellular discohesion was noted in Cases 1 and 6 (Figure 1E), and characteristic hallmark cells with eccentric horseshoe‐shaped or reniform nuclei were identified in multiple cases (Figure 1F). Mitotic figures ranged from 3 to 12 per mm² across all seven evaluable cases (Figure 1G), and tumour necrosis was present in four of these seven (Cases 1, 2, 4 and 5) (Figure 1H). In Case 1, the sole surgical resection specimen, extensive lymphovascular invasion manifesting as intravascular tumour emboli was also observed (Figure 1I).

Immunophenotypic Profiles

The immunohistochemical findings are summarized in Table 2. All cases were positive for one or more epithelial markers (pan‐cytokeratin (AE1/AE3), CK7 and/or CK8/18), showing strong or focal staining (Figure 2A). TTF‐1 was strongly positive or focally positive in all cases (Figure 2B); Napsin A was expressed in Cases 2 and 7. ALK and CD30 were strongly co‐expressed in all cases (Figure 2C,D). All seven cases tested for CD45RO were negative (Figure 2E), and no case showed expression of T‐ or B‐cell–associated markers, including CD2, CD3, CD4, CD5, CD20, CD25, CD43, CD45, MUM‐1, TIA‐1, Perforin and granzyme B (Figure 2F,G,H; Table 2). The Ki‐67 proliferation index ranged from 50% to 95% in all seven evaluable cases (Figure 2I; data unavailable for Case 8). PD‐L1 expression was high (tumour proportion score >50%) in all three cases in which it was assessed (Cases 1, 5 and 7) (Table 2).

++, positive cells >80%; − indicates negative; ALK, ALK (D5F3).

Grb, granzyme B; NA, not available.

PMID: 38032068.

Molecular Findings

Targeted NGS was performed in Cases 1, 7 and 8; results are presented in Table S2. Case 1 carried somatic mutations in TP53, BRINP3, NOTCH3 and ERBB3 (VAFs: 42.39%, 18.29%, 11.64% and 9.66%, respectively), MYC and NFKBIA copy number amplification and a Class I PRKAR1A::ALK fusion. Case 7 carried mutations in ARID1B, CREBBP, ATM, AMER1, CSF1R, KMT2C, H3‐3B, NOTCH2, SIPA1, WRN, COL1A1 and TP53 (VAFs: 3.21%–9.41%), MYC amplification and a Class I EML4::ALK fusion. Case 8 carried mutations in NF1 and TP53 alongside a canonical EML4::ALK fusion. ³² Both Cases 1 and 7 exhibited low tumour mutation burden (TMB‐L) and microsatellite stability (MSS), with no germline alterations detected in either. ALK rearrangement was independently confirmed by fluorescence in situ hybridization (FISH) in Cases 3, 4 and 6. Cases 1–7 and the previously reported Case 8 all demonstrated a polyclonal TCR pattern. ³²

Treatment and Clinical Outcomes

Treatment and outcome data are summarized in Table 1. All patients presented with advanced‐stage disease (stage III–IV), and seven of eight did not undergo surgical resection. Case 1 required small bowel resection for intestinal obstruction, though the primary lung lesion was left in situ; this patient subsequently received alectinib and achieved stable disease. Case 2 had been misdiagnosed as ALCL at an outside institution and treated with four cycles of BV‐CHOP, after which the disease progressed. Following diagnostic revision to poorly differentiated LUAD, alectinib was initiated, and the patient achieved a partial response. Among the seven patients who received systemic therapy, six achieved a partial response. The predominant treatment modality was ALK‐directed therapy: five patients received alectinib, ensartinib or crizotinib, and one received lorlatinib combined with adriamycin and cisplatin (Table 1); no patient received immunotherapy. Follow‐up data  were available for seven patients (range, 5–45 months). Four patients (Cases 1, 2, 3 and 7) remained recurrence‐free throughout the follow‐up period. Case 4 developed recurrence at 25 months and was alive at the last follow‐up. Case 5 progressed at 12 months and died at 45 months; Case 6 progressed at 11 months and died at 36 months (Figure 3).

CD30 Expression Is a Rare Event in ALK‐Positive LUAD

CD30 positive expression was identified in 1 of 62 resected ALK‐positive LUADs (1.6%; Case 1) and in 6 of 155 ALK‐positive biopsy specimens (3.9%; Cases 2–7). Overall, CD30 expression was uncommon in ALK‐rearranged LUAD. All seven ALK+/CD30+ cases presented with advanced‐stage disease (stage III–IV) and exhibited anaplastic large‐cell morphology. Comparisons of clinicopathological characteristics between ALK+/CD30+ and ALK+/CD30− tumours are summarized in Table S3. Significant differences were observed between the two groups in disease stage (P = 0.034), histological subtype (P = 0.010), large‐cell morphology (P < 0.001), Ki‐67 proliferation index (P = 0.002) and PD‐L1 expression (P < 0.001).

Discussion

Although aberrant ALK expression and gene rearrangements are well documented in LUAD, concurrent CD30 and ALK co‐expression remains largely uncharacterized. To our knowledge, this study is among the first systematic analyses of this phenomenon. Through integration of clinical, radiological, histomorphological, immunohistochemical and molecular data, our findings expand the morphological and immunophenotypic spectrum of poorly differentiated LUAD and define a diagnostically challenging subset with direct therapeutic relevance.

CD30/ALK double‐positive LUAD is a rare but clinically consequential mimic of ALK‐positive ALCL. In our cohort, all patients presented with advanced‐stage disease accompanied by extensive lymphadenopathy, distant organ involvement and extranodal manifestations such as multifocal osteolytic lesions, findings that collectively and understandably raised an initial suspicion of lymphoma. This impression was further reinforced histologically by diffuse sheets of large anaplastic cells with conspicuous nucleoli and hallmark cells bearing eccentric horseshoe‐shaped or reniform nuclei and immunophenotypically by dual CD30 and ALK expression. ³³ Despite this overlap, several features consistently favour poorly differentiated LUAD. Demographically, ALK‐positive ALCL predominantly affects children and young adults and is often associated with systemic B symptoms (fever, night sweats and weight loss). In contrast, our cohort included young, middle‐aged and elderly individuals and lacked such symptoms. ³⁴ , ³⁵ Radiologically, a dominant pulmonary mass with hilar lymphadenopathy, present in all cases, is inconsistent with primary pulmonary ALCL, which is exceedingly rare in any form. ³⁶ Morphologically, all cases were predominantly composed of anaplastic large cells resembling those of anaplastic large‐cell lymphoma, highlighting the indispensable role of ancillary testing. ³⁵ Immunophenotypically, although most ALK‐positive ALCLs express epithelial membrane antigen (EMA), ³⁷ , ³⁸ and keratin expression has also been reported in a subset of cases, ³⁹ , ⁴⁰ the detection of epithelial markers should not be regarded as sufficient evidence for a diagnosis of poorly differentiated carcinoma. ⁴⁰ , ⁴¹ In our cases, diffuse or focal expression of epithelial markers (including pan‐keratin, CK7 and CK8/18) and the relatively LUAD‐specific marker TTF‐1, together with the absence of cytotoxic molecules and B‐ or T‐cell–associated antigens, supported a non‐haematolymphoid origin. Molecularly, whereas NPM1 is the predominant ALK fusion partner in ALK‐positive ALCL, EML4 and other LUAD‐associated genes predominate in ALK‐positive LUAD ¹⁵ , ¹⁶ , ²⁵ , ²⁹ ; in all three molecularly analysed cases (Cases 1, 7 and 8), the identified fusion partners fell exclusively within the LUAD‐associated spectrum. Moreover, TCR gene rearrangement analysis in all cases provided no evidence supporting a diagnosis of lymphoma. Given the potential for both false‐positive and false‐negative results in TCR gene rearrangement analysis, together with the frequent expression of epithelial membrane antigen (EMA) in ALK‐positive ALCL, ³⁷ , ³⁸ concurrent positivity for EMA, CD30 and ALK in small biopsy specimens from patients presenting with generalized lymphadenopathy should not be considered sufficient for a definitive diagnosis of ALK‐positive ALCL, particularly in the absence of convincing evidence of T‐cell lineage. The diagnostic challenge becomes even greater when admixed background lymphocytes express T‐cell markers. Therefore, accurate classification requires comprehensive integration of clinical, radiological, morphological, immunophenotypic and molecular findings.

CD30 is an established diagnostic marker for ALCL and classical Hodgkin lymphoma, yet its expression in epithelial malignancies, including lung cancer, has been documented only rarely. An exploratory analysis presented at the 2014 ASCO Annual Meeting identified no CD30 immunoreactivity in 17 ALK‐negative or 21 ALK‐positive NSCLC specimens. ⁴² Consistent with these findings, our data indicate that CD30 expression is uncommon in ALK‐rearranged LUAD. Notably, all CD30‐positive tumours in our cohort exhibited anaplastic large‐cell morphology. The rate of CD30 expression was even lower among ALK‐positive LUADs without anaplastic morphology. Although the morphological features of the previously reported cases were not described, ⁴² this observation may partly explain the absence of CD30 expression in that series. Furthermore, several lung cancer cell lines (H460, H358, A549, H1666 and H1299) have been experimentally shown to express CD30. ⁴³ This discrepancy likely reflects a combination of biological heterogeneity, the inherent limitations of in vitro models as surrogates for clinical tissue and the genuine rarity of CD30 expression in NSCLC. Despite these largely negative clinical data, sporadic reports have documented CD30 expression in lung cancer and other epithelial malignancies. In the setting of lung cancer, Siba El Hussein et al. described a poorly differentiated LUAD co‐expressing CD30 and ALK that closely simulated ALCL both morphologically and immunophenotypically, a scenario directly analogous to the cases presented here. ³² Beyond the lung, Grethlein et al. reported a CD30‐positive poorly differentiated gastric carcinoma with anaplastic cytological features, ²³ and Wakely and Chakhachiro documented CD30 expression in undifferentiated non‐keratinizing nasopharyngeal carcinoma, where tumour cells exhibited sinusoidal lymph node infiltration and overtly anaplastic morphology with marked nuclear enlargement and prominent nucleoli. ²² , ⁴⁴ In ALK‐positive ALCL, CD30 is characteristically diffuse and strong, involving the cell membrane and Golgi regions. ³⁰ , ⁴⁵ The cases in our cohort, by contrast, showed strong membranous staining. Although this difference may aid distinction in resection specimens with adequate sampling, it is easily overlooked in small biopsies, where intense CD30 positivity can render a poorly differentiated carcinoma virtually indistinguishable from ALK‐positive ALCL, a distinction with profound therapeutic consequences. Taken together, these findings suggest that CD30 expression is not lineage‐restricted but may instead be a feature associated with anaplastic cytomorphology across both epithelial and haematolymphoid malignancies. Accordingly, CD30 immunostaining should be interpreted as one component of a multiparametric diagnostic evaluation that incorporates morphological, immunophenotypic and molecular data, rather than as a standalone classifier.

In a published NGS analysis of 46 ALK‐positive ALCL cases, the most frequently mutated genes were TP53, KMT2 and CLMN (each 11%), followed by JUNB, GPS2, NTRK3 and PKD1 (each 7%) and MDM4, SETD2, TET1 and PREX2 (each 4%); ten additional genes (EP300, PRDM1, NOTCH1, ERBB4, ACSBG2, ARHGEF7, JAK3, SOCS6, DSP and FLG) were each mutated in a single case (2%). ⁴⁶ , ⁴⁷ In ALK‐positive NSCLC, Kron et al. reported a TP53 co‐mutation frequency of 23.8% in a cohort of 216 patients, with MET (5.4%), HER2 (2.9%), FGFR2 (1.8%), CTNNB1 (1.4%), PTEN (0.7%), DDR2 (0.7%) and BRAF (0.6%) as additional co‐mutated genes. ⁴⁸ In our cohort, all three sequenced CD30/ALK double‐positive cases harboured TP53 mutations, while no other mutated gene overlapped with the ALK‐positive ALCL mutation spectrum (Table S2). TP53 thus emerged as the sole recurrently mutated gene shared between the two tumour types, an observation that hints at a common molecular vulnerability despite otherwise divergent mutational landscapes, though confirmation in larger, comprehensively sequenced cohorts is required.

The tumours in our cohort were biologically aggressive, as evidenced by advanced stage at presentation, high metastatic burden and markedly elevated Ki‐67 indices. This aggressiveness, combined with the striking morphological and immunophenotypic overlap with ALK‐positive ALCL, makes accurate classification both challenging and critically important, since the two entities are treated fundamentally differently. ALK‐positive ALCL is managed primarily with combination chemotherapy, with or without the anti‐CD30 antibody–drug conjugate BV, whereas ALK‐TKIs are reserved for relapsed or refractory disease. ³⁰ , ³¹ , ⁴⁹ ALK‐positive LUAD, by contrast, is treated with ALK‐TKIs as the established first‐line standard of care ¹⁷ ; misclassification as ALCL would therefore deprive patients of a highly effective targeted therapy and expose them to inappropriate cytotoxic treatment. The single patient misdiagnosed as ALCL progressed on brentuximab vedotin plus CHOP, while all patients who received ALK‐TKI–based therapy achieved favourable responses, a contrast that directly demonstrates the therapeutic importance of correct diagnosis. Two additional therapeutic considerations arise from our cohort. First, whether BV might retain utility after ALK‐TKI resistance develops is a clinically relevant question that warrants prospective evaluation. Second, high PD‐L1 expression in Cases 1, 5 and 7, consistent with reports of elevated PD‐L1 in both ALK‐positive ALCL and ALK‐positive LUAD, ²⁴ , ⁵⁰ , ⁵¹ raises the possibility of immune checkpoint inhibition in selected patients, though its efficacy in this specific context remains to be established. Taken together, these findings highlight a diagnostically challenging but therapeutically actionable entity. Despite its morphological and immunophenotypic resemblance to ALK‐positive ALCL and its aggressive clinicopathological profile, poorly differentiated LUAD with dual CD30 and ALK expression responds favourably to ALK‐TKI therapy, provided the diagnosis is made correctly and promptly.

In summary, CD30/ALK double‐positive poorly differentiated LUAD is a rare but distinct diagnostic entity whose morphological and immunophenotypic resemblance to ALK‐positive ALCL poses a significant pitfall in thoracic pathology. Because the two tumours require fundamentally different treatments and carry different prognostic implications, accurate recognition is clinically indispensable. Correct classification relies on integrating clinical and radiological data with comprehensive immunophenotyping, encompassing both epithelial and lymphoid markers and molecular characterization of ALK fusion partners. Future studies should clarify the biological basis of CD30 expression in LUAD, its interplay with co‐occurring alterations such as TP53 mutations, and whether dual targeting of ALK and CD30 offers therapeutic benefit.

Author contributions

JN conceived and designed the study and provided overall supervision. CWJ, JJT and YQ analysed the data and drafted the manuscript. HCY and FYF collected and curated the clinical case data. YQ and CYS performed immunohistochemical staining and FISH. All authors reviewed the manuscript critically for important intellectual content and approved the final version for submission.