Anaplastic lymphoma kinase immunohistochemical positivity in high-grade pulmonary neuroendocrine carcinoma: an actionable signal or merely a shadow?

Abstract

Aims

Anaplastic lymphoma kinase (ALK) rearrangements are actionable drivers in non-small cell lung carcinoma (NSCLC), but the biological significance of ALK-immunohistochemistry (IHC) positivity in high-grade pulmonary neuroendocrine carcinoma (NEC) remains unclear. This study evaluated the diagnostic and therapeutic implications of discordant ALK IHC and genomic findings and the role of multimodal molecular testing in resolving them.

Methods

We retrospectively analysed eight South Asian patients (Indian and Nepali) with de novo high-grade pulmonary NEC and diffuse ALK immunoreactivity treated at a tertiary cancer centre in India. Comprehensive molecular profiling using DNA- and RNA-based next-generation sequencing (NGS) and ALK fluorescence in situ hybridisation (where tissue was adequate) was performed. Clinical outcomes and responses to ALK-targeted tyrosine kinase inhibitors (TKIs) were assessed.

Results

ALK IHC positivity was observed in 8 of 100 selectively tested cases among 319 high-grade pulmonary NECs diagnosed between 2019 and 2025. The cohort included seven SCLCs (one combined adenocarcinoma-SCLC) and one large-cell neuroendocrine carcinoma (LCNEC). Median age was 51 years; 75% were female and 87.5% never-smokers. Among five comprehensively profiled cases, true ALK rearrangements were confirmed in two (one LCNEC and one SCLC), both detectable only by RNA sequencing. Durable benefit from ALK-TKI therapy was seen only in the molecularly confirmed LCNEC case (>16 months), whereas ALK IHC-positive but NGS-negative cases progressed rapidly.

Conclusions

True ALK rearrangements in high-grade pulmonary NEC are rare but highly actionable. ALK IHC alone is an unreliable predictor of therapeutic benefit. RNA-based sequencing is essential for fusion detection. Comprehensive molecular confirmation, with RNA sequencing as the preferred modality, should precede any initiation of ALK-targeted therapy in high-grade pulmonary NEC.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Anaplastic lymphoma kinase (ALK) rearrangements are well-established, targetable oncogenic drivers in lung adenocarcinoma, where ALK immunohistochemistry (IHC) serves as a validated screening tool. In high-grade pulmonary neuroendocrine carcinomas (NECs); however, ALK IHC positivity has been reported without a confirmed underlying genomic rearrangement. The clinical significance of this finding and the optimal method for confirming it has remained unclear.

WHAT THIS STUDY ADDS

• In this cohort of eight ALK IHC-positive high-grade pulmonary NECs, true ALK rearrangements were confirmed in only two cases, both detectable exclusively by RNA-based sequencing and missed by DNA-based testing, one was also fluorescence in situ hybridisation (FISH)-negative despite a confirmed fusion. A durable response to ALK-directed therapy occurred only in the molecularly confirmed case, whereas empiric treatment based on IHC positivity alone produced rapid progression. We also report what is, to our knowledge, the first DCTN1-ALK fusion described in large-cell NEC.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• ALK IHC positivity alone should not guide ALK-directed therapy in high-grade pulmonary NEC and molecular confirmation is of paramount importance before starting treatment. RNA-based sequencing should be considered the preferred confirmatory modality in this setting, as both DNA sequencing and FISH failed to detect genuine rearrangements that RNA sequencing identified.

Introduction

Small-cell lung carcinoma (SCLC) is a highly aggressive, high-grade neuroendocrine carcinoma (NEC) that accounts for approximately 13%–15% of all lung cancer cases worldwide, and remains one of the most lethal thoracic malignancies due to its rapid growth, early dissemination and high relapse rates following initial therapy.¹ SCLC is strongly associated with tobacco exposure, and most patients present with extensive-stage disease at diagnosis owing to infrequent early symptoms and the tumour’s intrinsic rapid metastatic potential.² Although SCLC is initially sensitive to platinum-etoposide chemotherapy and radiation therapy, responses are typically short-lived. The addition of immune checkpoint inhibitors targeting PD-L1 to first-line platinum-etoposide regimens has modestly improved outcomes in extensive-stage SCLC, extending median overall survival to approximately 12–13 months, compared with historical outcomes of 8–10 months with chemotherapy alone.³ Despite these advances, 5 year survival rates remain poor, highlighting the urgency for biologically informed therapeutic strategies.⁴

Unlike non-small cell lung carcinoma (NSCLC), where routine comprehensive molecular testing has identified multiple actionable oncogenic drivers and facilitated effective targeted therapies (eg, EGFR, anaplastic lymphoma kinase (ALK), ROS1), SCLC and related high-grade NEC exhibit a distinct genomic landscape characterised by near-universal inactivation of the tumour suppressor genes TP53 and RB1. Large genomic profiling studies have demonstrated that biallelic inactivation of TP53 and RB1 is a defining feature of SCLC, whereas the recurrent actionable kinase fusions typical of NSCLC are exceedingly rare in de novo SCLC.^{5 6} Consequently, comprehensive molecular profiling has not been routinely incorporated into standard SCLC management outside research settings.

Nonetheless, isolated reports have described ALK positivity in treatment-naïve SCLC. A small number of case reports document true ALK gene rearrangements (eg, EML4-ALK fusions) detected by next-generation sequencing (NGS), occasionally associated with clinical responses to ALK tyrosine kinase inhibitors (TKIs).^{7 8} In contrast, several cohort studies have identified ALK protein expression by immunohistochemistry (IHC) in SCLC without confirmatory genomic rearrangement by fluorescence in situ hybridisation (FISH) or NGS, raising concerns regarding the specificity and clinical relevance of ALK IHC in this setting.^{9 10} This discordance between protein expression and bona fide genomic rearrangement creates a significant practical dilemma for clinicians and pathologists: whether ALK positivity in high-grade NEC represents a genuine, although rare, actionable driver alteration, or predominantly reflects non-driver protein expression with limited therapeutic relevance.¹⁰ To further evaluate this question, we report a retrospective case series of eight patients with de novo ALK IHC-positive high-grade pulmonary NEC, examining concordance between IHC and genomic findings and clinical outcomes following ALK-targeted therapy. We emphasise that we do not advocate routine ALK IHC testing in high-grade pulmonary NEC, which is not part of standard diagnostic practice. Rather, in our real-world setting, a subset of these patients present with features atypical for the histology, such as never-smoking status, younger age and female sex, prompting clinicians to search for potentially targetable alterations in the hope of improving otherwise poor outcomes (akin to a small percentage of oncogene addicted squamous cell carcinoma). We therefore performed molecular diagnostics selectively in such atypical cases to determine whether a genuine ALK driver was present in any of them, while recognising that IHC positivity alone is insufficient to guide therapy.

Materials and methods

Study design and patient selection

This retrospective, single-centre observational case series was conducted at Rajiv Gandhi Cancer Institute & Research Centre, New Delhi, India. Patients of South Asian (Indian and Nepali) origin diagnosed with de novo high-grade pulmonary NEC and IHC-confirmed ALK positivity between 2019 and 2025 were included. During this period, 319 high-grade pulmonary NEC cases were diagnosed at our institution, of which 100 underwent ALK IHC testing on the basis of atypical clinicopathological features and eight were ALK IHC-positive which constitute in this series. Clinical, pathological, treatment and outcome data were extracted from electronic medical records.

ALK immunohistochemistry and molecular profiling

All tumours were pathologically confirmed as SCLC or related high-grade NEC, including one combined adenocarcinoma-SCLC case and one large-cell neuroendocrine carcinoma (LCNEC). Diagnosis was established using morphology together with a neuroendocrine immunohistochemical panel (table 1) in most cases, comprising cytokeratin (CK), synaptophysin, INSM1, p40 and TTF-1. ALK protein expression was assessed using the Ventana D5F3 monoclonal antibody assay (Ventana Medical Systems, Tucson, Arizona, USA). ALK positivity was defined by the presence of diffuse, intense cytoplasmic staining, in accordance with established diagnostic criteria routinely applied in lung cancer practice. A representative photomicrograph demonstrating diffuse, strong cytoplasmic ALK staining is shown in (figure 1).

Table 1. Diagnostic immunohistochemistry and morphology.

Pt	Histology	Biopsy type/site	Synaptophysin	Chromogranin A	CD56	TTF-1	Napsin A	Ki-67 (%)	Mitotic rate	Other markers
1	SCLC	Endobronchial biopsy (Right hilar mass)	Positive (strong)	Positive	not done	not done	not done	not done	Brisk mitosis	not done
2	SCLC	USG guided core needle biopsy from left scapular mass	Positive	Negative	not done	Positive	not done	67%–80%	Intense	CK/EMA/P53 (strong) +ve and P40 −ve
3	SCLC	Liver biopsy	Positive	Not done	Not done	Positive	Not done	80%	n//a	CK (weak) +ve
4	SCLC	Right lung	Positive	not done	Not done	Negative	Not done	~40%	10–12/2 mm2	CK (interrupted pattern +ve), P16 +ve, P53 (mutational pattern +ve)/P40 and CD20 −ve
5	SCLC	Endobronchial and liver biopsy	Positive	Not done	not done	Positive	Not done	80%	n/a	CK +ve, INSM1 +ve, LCA and P40 −ve
6	Combined adenocarcinoma-SCLC	Left lung mass	Positive (in SCLC component)	Positive (in SCLC component)	not done	Positive (both)	not done	n/a	n/a	P53 mutant in both components
7	SCLC	Liver lesion	Positive	not done	not done	Positive	not done	~95%	n/a	CK7, INSM1 +ve/CK20 −ve and RB1 and TP53- concurrent mutation
8	LCNEC	Endobronchial biopsy from left upper lobe mass	Positive	not done	Positive	Positive	not done	~60%	significant	P40, PAX8, calcitonin, INSM1 −ve

LCNEC, large-cell neuroendocrine carcinoma; SCLC, small-cell lung carcinoma.

Figure 1. Anaplastic lymphoma kinase (ALK) immunohistochemistry in high-grade pulmonary neuroendocrine carcinoma (patient 8 -the DCTN1-ALK large-cell neuroendocrine carcinoma case): representative photomicrograph demonstrating diffuse, strong cytoplasmic staining for ALK in tumour cells using the Ventana D5F3 antibody clone.

Comprehensive molecular analysis was performed where tissue was adequate (five cases). DNA-based sequencing was performed using a 38-gene hotspot panel on the Ion Torrent S5 platform (Thermo Fisher Scientific, USA). RNA-based fusion testing was performed using the Oncomine Fusion RNA Assay (Thermo Fisher Scientific, Ion Torrent platform), which interrogates expressed fusion transcripts and is more sensitive than DNA hotspot panels for the detection of gene fusions. ALK FISH was performed using a break-apart probe in all cases with adequate residual tumour material (six cases), two cases had insufficient tissue for FISH and three for NGS.

Treatment and outcomes

ALK-directed TKI therapy was administered to selected patients, either in the first-line setting or following platinum-based chemotherapy. Treatment and outcome details are summarised in table 2. Radiologic response was assessed using RECIST V.1.1. Progression-free survival (PFS) was defined as the interval from initiation of ALK-TKI therapy to documented radiologic progression or death from any cause.

Table 2. Baseline clinicopathological characteristics, molecular findings and outcomes with TKI therapy.

Pt	Age/sex	Smoking	Histology	Stage	Brain mets (baseline)	Pleural effusion (baseline)	Liver mets (baseline)	Bone mets (baseline)	IHC (ALK D5F3)	NGS (DNA and RNA sequencing)	FISH	Treatment received	ALK-TKI	Best response (to systemic rx RECIST 1.1)	PFS on TKI (months)	Outcome/status
1	50s/F	Never	SCLC	Limited	No	Yes (cytology negative)	No	On progression	Positive	Not done (tissue inadequate)	Negative	ECx6+33# RT+Prophylactic Cranial Irradiation (PCI)	No	Significant response	n/a	Dead
2	70s/M	Yes	SCLC	Extensive	Yes	No	No	Yes	Positive	Not done (insufficient tissue)	Not done	No treatment	No	n/a	n/a	Dead
3	50s/F	Never	SCLC	Extensive	No	Yes	Yes	Yes	Positive	Not done (insufficient tissue)	Not done	No treatment	No	n/a	n/a	Dead
4	30s/F	Never	SCLC	Extensive	On progression	Yes	No	No	Positive	Negative	Negative	ECx6→ ceritinib → irinotecan → paclitaxel	Yes (ceritinib)	PD (SoD+63.1%; 8.4→13.7 cm)	~5	Dead
5	50s/F	Never	SCLC	Extensive	No	No	Yes	Yes	Positive	TP53 EML4(13)-ALK(20)	Negative	ECx3 → irinotecan-carboplatin x3 → weekly paclitaxel	No	PD	n/a	Dead
6	40s/F	Never	Combined adenocarcinoma-SCLC	Extensive	No	No	No	Yes	Positive	TP53+RB1	Negative	EC+PCI → lorlatinib maintenance → irinotecan	Yes (lorlatinib)	PD (SoD+22.2%; new liver mets)	~5	Dead
7	30s/F	Never	SCLC	Extensive	Yes	No	Yes	Yes	Positive	TP53+RB1	Negative	1st-line lorlatinib → planned WBRT+chemotherapy	Yes (lorlatinib, 1st-line)	PD (intracranial+extracranial)	~1	Dead
8	40s/M	Never	LCNEC	Extensive	No	No	No	Yes	Positive	DCTN1(26)-ALK(20)	Positive	1st-line lorlatinib	Yes (lorlatinib, 1st-line)	Good partial response (clinical)	16+	Alive, on treatment

ALK, anaplastic lymphoma kinase; FISH, fluorescence in situ hybridisation; IHC, immunohistochemistry; LCNEC, large-cell neuroendocrine carcinoma; NGS, next-generation sequencing; PFS, progression-free survival; SCLC, small-cell lung carcinoma; TKI, tyrosine kinase inhibitor.

Results

Baseline characteristics

Eight consecutive patients with de novo high-grade pulmonary NEC demonstrating diffuse ALK positivity by IHC were identified (median age 51 years, range 33–70 years). Seven patients had SCLC (including one combined adenocarcinoma-SCLC case), and one patient had LCNEC. The majority were women (75%), and most had a never-smoking history (87.5%). Most patients presented with extensive-stage disease and one patient had limited-stage disease (table 2).

Molecular findings and concordance between IHC and NGS

Comprehensive NGS (DNA and RNA) was available in five patients. Confirmatory ALK rearrangements were detected in two cases: a DCTN1(exon 26)::ALK(exon 20) fusion in the LCNEC patient (patient 8) and an EML4(exon 13)::ALK(exon 20) fusion in an SCLC patient (patient 5). Both fusions were identified on RNA-based sequencing. The DCTN1-ALK fusion was not detected on parallel DNA sequencing. The remaining three NGS-tested cases (patient 4, 6 and 7) did not demonstrate ALK rearrangement despite diffuse ALK expression on IHC, indicating low concordance between ALK IHC positivity and genomic fusion status. Among the ALK IHC-positive but rearrangement-negative cases, TP53 alterations were identified in two patients along with RB1 alteration (patient 6 and 7), consistent with canonical SCLC genomic patterns.¹¹ ALK FISH, performed in six cases, was positive only in the DCTN1-ALK LCNEC case (patient 8) and negative in all five remaining evaluable cases, including the EML4-ALK SCLC case (tables 1 and 2). To address the possibility of a combined small-cell carcinoma in the EML4-ALK case (patient 5), the block that revealed the fusion on RNA sequencing was re-evaluated; it showed small-cell morphology with positive synaptophysin and INSM1 expression and weak, dot-like CK immunoexpression, consistent with pure small-cell carcinoma, with no non-small-cell component identified on re-review.

Treatment and clinical outcomes

ALK-directed therapy was administered to selected ALK IHC-positive patients, either in the first-line setting or after platinum-etoposide-based therapy, and outcomes differed markedly according to molecular concordance (table 2).

The patient with concordant ALK positivity on IHC and RNA sequencing (DCTN1-ALK fusion, LCNEC-patient 8) received lorlatinib as first-line therapy and achieved a sustained partial response, remaining alive with ongoing benefit beyond 16 months. This represented the only case in the cohort demonstrating durable clinical activity of ALK inhibition. Notably, the second molecularly confirmed case (EML4-ALK SCLC-patient 5) did not receive ALK-targeted therapy due to logical and financial constraints. The concurrent FISH result was negative, and the patient was instead managed with chemotherapy.

In contrast, outcomes were poor among rearrangement-negative patients treated empirically with ALK-TKI on the basis of ALK IHC positivity alone, consistent with primary resistance. Patient 7 treated with first-line lorlatinib, despite a negative molecular work-up, progressed within approximately 1 month. Patient 4 received ceritinib following first-line etoposide-carboplatin and progressed at approximately four to 5 months. Patient 6 was treated with lorlatinib maintenance also progressed within months. Patients who did not receive ALK-targeted therapy were managed with standard chemotherapy-based regimens (patient 1 and 5). Overall, these findings indicate that ALK IHC positivity in high-grade pulmonary NEC frequently lacks a corresponding genomic ALK rearrangement and is not reliably predictive of clinical benefit from ALK-TKI therapy in the absence of molecular confirmation.

Discussion

ALK-IHC has transformed the management of advanced lung adenocarcinoma. However, there is lack of clarity and recommendation of its usefulness in high-grade pulmonary NEC remains incompletely defined. We retrospectively analysed eight ALK IHC-positive high-grade pulmonary NEC cases using a multimodal molecular diagnostic approach incorporating DNA, RNA-based NGS and FISH, to propose that IHC positivity in this histological context does not often reflect an underlying genomic rearrangement and is not independently predictive of ALK-targeted therapy response. Crucially, both confirmed ALK rearrangements identified in this series were detectable on RNA sequencing, providing direct in-series evidence that RNA-based sequencing is quintessential for the molecular characterisation of ALK IHC-positive high-grade pulmonary NEC.

ALK IHC positivity in high-grade pulmonary NEC: prevalence, demographics and testing context

ALK IHC positivity was identified in 8 of 100 tested high-grade pulmonary NEC cases (8%)(signifying pseudo-positivity) from a denominator of 319 cases diagnosed at our institution between 2019 and 2025 (figure 2). Testing was performed selectively on the basis of atypical clinical features which included young age, female sex and never- or light-smoker status, rather than as a reflex panel applied to all cases. Hence, these figures therefore cannot be extrapolated to reflect true population-level prevalence. The demographic profile of the IHC-positive cohort is notable: six of eight patients were female and seven were never-smokers. This demonstrates a phenotypic overlap with ALK-rearranged NSCLC^{9 12} as well as with squamous histology occurring in never- or light-smokers, a finding that is well-documented in the literature and guided our clinical decision to pursue ALK testing in these cases. Of all tested cases (100), true ALK rearrangements were confirmed in only 2 (2%). This is consistent with the rarity of ALK as a driver alteration in unselected NEC populations.¹¹

Figure 2. Consolidated Standards of Reporting Trials (CONSORT) flow diagram of ALK testing in high-grade pulmonary neuroendocrine carcinoma at a tertiary cancer centre, 2019 to 2025. ALK, anaplastic lymphoma kinase; FISH, fluorescence in situ hybridisation; IHC, immunohistochemistry; LCNEC, large-cell neuroendocrine carcinoma; NEC, neuroendocrine carcinoma; NGS, next-generation sequencing; SCLC, small-cell lung carcinoma; TKI, tyrosine kinase inhibitor.

This reinforces that, even in a clinically enriched and selectively tested cohort, the vast majority of ALK IHC-positive high-grade NEC cases do not harbour a genomic ALK rearrangement. Reliance on ALK IHC positivity alone may therefore lead to inappropriate treatment selection and delays in the initiation of effective therapy, with potentially detrimental consequences given the aggressive clinical course of these malignancies.

Physiological ALK expression in the neuroendocrine lineage: the biological basis of non-driver IHC positivity

An important unresolved question is the mechanism responsible for diffuse ALK immunoreactivity in tumours that are negative for ALK rearrangements by NGS. ALK is a receptor tyrosine kinase with established physiological roles in neural and neuroendocrine development, where it regulates neuronal differentiation, axonal growth and synaptogenesis.^{13 14} In the absence of an oncogenic rearrangement, residual ALK protein may persist in tumours of neuroendocrine lineage and is detectable by the highly sensitive D5F3 clone employed in this and most contemporary studies. The data from this series (table 2) support this interpretation directly: in all six FISH-evaluable cases (patient 1, 4, 5, 6, 7 and 8), it shows ALK IHC positivity and negative comprehensive NGS, FISH, excluding both chromosomal rearrangement and gene amplification as explanations for ALK protein expression. Physiological neuroendocrine lineage expression therefore represents the most plausible biological basis for IHC positivity in these cases. This is consistent with previously published observations in SCLC and pulmonary carcinoid tumours.^{9 10}

Novel DCTN1(exon 26)::ALK(exon 20) fusion in LCNEC

The index case of this series is a male in his late 40s with extensive-stage LCNEC (patient 8). This represents the central contribution and illustrates both the potential and the necessity of RNA-based molecular testing in ALK IHC-positive high-grade NEC. Molecular profiling using the Oncomine Fusion RNA Assay (ThermoFisher Scientific, Ion Torrent platform) identified a novel DCTN1(exon 26)::ALK(exon 20) fusion, which was not detected on parallel DNA-based sequencing.

DCTN1 (dynactin subunit 1) encodes a core component of the dynactin complex implicated in microtubule-based intracellular transport as a 5’ fusion partner and it contributes a dimerisation domain that drives constitutive ALK kinase activation. DCTN1-ALK is an exceptionally rare fusion, previously reported in a small number of cases across NSCLC adenocarcinoma, inflammatory myofibroblastic tumour, epithelioid fibrous histiocytoma, ALK-positive histiocytosis, gastric cancer and renal cell carcinoma.^15–18

To the best of our knowledge, this is the first reported case of DCTN1-ALK fusion in LCNEC and among the first in any thoracic malignancy. In the context of ALK-rearranged LCNEC, only 21 published cases have been reported in a systematic review of the literature, all characterised by EML4-ALK fusion.¹⁹ The DCTN1-ALK case reported here therefore extends the known spectrum of actionable ALK fusions in this rare histological entity beyond EML4. Treatment with first-line lorlatinib produced a good partial radiologic response with ongoing disease control exceeding 16 months in our patient (at the time of last follow-up). This clinical response supports the concept of ALK oncogene dependence and aligns with the demonstrated efficacy of third-generation ALK inhibitors in molecularly confirmed ALK-rearranged thoracic malignancies.^{20 21}

The failure of DNA-based sequencing to detect this fusion is mechanistically straightforward. Gene fusions typically arise at intronic breakpoints that fall outside the coverage of hotspot mutation panels, whereas RNA sequencing interrogates the expressed fusion transcript directly and is not constrained by breakpoint location.^{22 23}

This principle was corroborated by the second confirmed ALK-rearranged case in this series (patient 5), a female in her late 50s with extensive-stage SCLC in whom an EML4(exon 13)::ALK(exon 20) rearrangement was identified on RNA sequencing. In this patient, ALK FISH was negative despite a confirmed rearrangement, illustrating a recognised but clinically important limitation of break-apart FISH probes. Due to financial and logistical constraints, the patient did not receive ALK-targeted therapy and was instead commenced on chemotherapy (table 2). We do not draw any therapeutic conclusions from this case as the patient received sequential chemotherapy.

This is explained by the fact that EML4 and ALK are located only ~12 Mb apart on the short arm of chromosome 2, the paracentric inversion that generates the EML4-ALK fusion produces only a subtle separation of the 5′ and 3′ probe signals, which may fall below the threshold for a scored split signal and result in a false-negative interpretation.²³ A negative break-apart FISH result therefore cannot reliably exclude an ALK rearrangement, and RNA sequencing should be considered the more sensitive modality for confirming ALK fusion status in such settings.

Notably, our own institutional experience in ALK-rearranged NSCLC has similarly demonstrated that a substantial proportion of ALK IHC-positive cases are FISH-negative, with this discordant group showing inferior PFS, underscoring the limitations of FISH as a confirmatory modality.²⁴

Clinical outcomes of ALK-directed therapy in the absence of molecular confirmation

We had three patients (patient 4, 6 and 7) who received ALK-targeted therapy on the basis of IHC positivity alone, without confirmed rearrangement on comprehensive molecular testing. Among these, none derived clinical benefit. Ceritinib (patient 4) administered after platinum-etoposide resulted in RECIST-defined progression at 4 months, while lorlatinib given either as maintenance therapy (patient 6) or as first-line treatment (patient 7) led to disease progression within three and 1 months, respectively, including the development of new metastatic lesions and intracranial progression in one case (table 2). In all the above three cases, comprehensive molecular testing encompassing DNA-based sequencing, RNA-based fusion analysis, and FISH was negative for ALK rearrangement.

The absence of durable benefit across all three of our patients is consistent with the established principle that IHC positivity arising from physiological neuroendocrine lineage expression does not reflect oncogene dependence and therefore lacks predictive value for ALK inhibitor response.^{9 10} These findings support the conclusion that molecular confirmation of ALK rearrangement by RNA sequencing is the preferred modality and should be considered a prerequisite before initiating ALK-targeted therapy in high-grade pulmonary NEC.

Napsin A immunostaining was not performed because it is not included in our routine diagnostic panel and it is consistently negative in SCLC and is expressed only focally in a minority of LCNEC.²⁵ Hence, neuroendocrine differentiation in our cohort was established using the morphology and requisite IHC markers (table 1), including Synaptophysin and INSM1, which fulfil the diagnostic criteria for neuroendocrine differentiation.²⁶ Our comprehensive IHC panel comprised CK, Synaptophysin, INSM1, p40 and TTF-1, which was done for most of the NEC and LCNEC cases (table 1). We further note that Napsin A can be used to help distinguish LCNEC from adenocarcinoma with neuroendocrine differentiation. In our practice we rely on the limited, WHO-recommended panel²⁷ to preserve tissue for molecular testing, and in our experience Napsin A has not added diagnostic utility in this context. In the LCNEC case (patient 8), the neuroendocrine diagnosis was supported by tumour morphology together with synaptophysin and CD56 positivity, with the diagnosis resting on histomorphology supported by neuroendocrine IHC (figure 3).

Figure 3. (a) Sections show attenuated endobronchial lining (black arrow) with underlying stroma showing poorly differentiated carcinoma with crushing artefacts and (b) Tumour cells are large with moderate eosinophilic cytoplasm, vesicular chromatin and significant mitoses (red arrows).

Limitations

This study has several limitations inherent to its retrospective, single-centre design and small sample size. ALK IHC was performed selectively rather than systematically across all 319 cases, introducing ascertainment bias and precluding estimation of true population-level prevalence of ALK IHC positivity or genomic rearrangement in unselected high-grade NEC.

FISH was not feasible in two cases owing to insufficient residual tumour material. One patient did not receive ALK-TKI due to logical and financial constraints. No functional validation of the biological mechanism underlying non-driver ALK expression in neuroendocrine lineage cells was undertaken.

Notwithstanding these limitations, the multimodal molecular characterisation of this cohort encompassing DNA sequencing, RNA fusion analysis and FISH across eight cases provides a more comprehensive dataset than most published series in this space and supports the primacy of RNA-based testing in the molecular workup of ALK IHC-positive high-grade pulmonary NEC.

Conclusion

In conclusion, high-grade NEC and NSCLCs are biologically distinct entities, differing in their morphology, molecular landscape, pathogenesis, clinical behaviour, therapeutic vulnerabilities and patterns of oncogene dependence. Consequently, ALK positivity should not be viewed as a uniform biomarker across these histologies. While ALK expression in NSCLC often reflects an actionable oncogenic driver, diffuse ALK immunohistochemical positivity in high-grade NECs does not necessarily indicate the presence of a targetable ALK rearrangement or oncogene addiction. The clinical implications are clear: ALK IHC positivity alone should not be used to guide treatment decisions in high-grade NECs, and preferably RNA-based NGS sequencing remains essential before considering ALK-directed therapy.

Data availability statement

Data are available upon reasonable request.