Delta-like ligand 3 expression in isocitrate dehydrogenase-mutant and isocitrate dehydrogenase-wildtype glioma is largely retained at recurrence, supporting its potential as a therapeutic target

Rosa Luning; Sybren L N Maas; Pim J French; Youri Hoogstrate; Levi van Hijfte; Rania Head; Iris de Heer; Thierry P P van den Bosch; Adelheid Woehrer; Martin J van den Bent; Marjolein Geurts

doi:10.1093/noajnl/vdag027

. 2026 Mar 17;8(1):vdag027. doi: 10.1093/noajnl/vdag027

Delta-like ligand 3 expression in isocitrate dehydrogenase-mutant and isocitrate dehydrogenase-wildtype glioma is largely retained at recurrence, supporting its potential as a therapeutic target

Rosa Luning ^1,^✉, Sybren L N Maas ^2,³, Pim J French ⁴, Youri Hoogstrate ⁵, Levi van Hijfte ⁶, Rania Head ⁷, Iris de Heer ⁸, Thierry P P van den Bosch ⁹, Adelheid Woehrer ¹⁰, Martin J van den Bent ^11,^†,^✉, Marjolein Geurts ^12,^13,^†

¹ Department of Neurology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands

² Department of Pathology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, The Netherlands

³ Department of Pathology, Leiden University Medical Center, Leiden, the Netherlands

⁴ Department of Neurology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands

⁵ Department of Neurology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands

⁶ Department of Neurosurgery, University Clinic Erlangen, Erlangen, Germany

⁷ Department of Neurology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands

⁸ Department of Neurology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands

⁹ Department of Pathology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, The Netherlands

¹⁰ Institute of Neuropathology and Neuromolecular Pathology, Medical University of Innsbruck, Innsbruck, Austria

¹¹ Department of Neurology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands

¹² Department of Neurology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands

¹³ Department of Medical Oncology, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands

^†

These authors are shared senior authors.

^✉

Corresponding Authors: Martin J. van den Bent, MD, PhD, Department of Neurology, Erasmus Medical Center Cancer Center, Dr. Molewaterplein 40, 3015 GD Rotterdam, the Netherlands (mjvandenbent@neuro-oncology.nl); Rosa Luning, MD, Department of Neurology, Erasmus Medical Center Cancer Center, Dr. Molewaterplein 40, 3015 GD Rotterdam, the Netherlands (a.luning@erasmusmc.nl).

PMCID: PMC13014167 PMID: 41890363

Abstract

Background

Delta-like ligand 3 (DLL3), an inhibitory Notch ligand, is now explored as a therapeutic target in recurrent diffuse glioma. Since clinical trials may select patients with recurrent glioma based on DLL3 expression in the primary tumor, this study aims to determine whether initial DLL3 expression in glioma is retained at recurrence.

Methods

198 formalin fixed, paraffin-embedded (FFPE) samples of paired primary and recurrent gliomas were retrieved from the biobank in the Erasmus MC, Rotterdam, the Netherlands. The cohort included oligodendroglioma, isocitrate dehydrogenase mutant (IDHmt) and 1p/19q co-deleted (n = 36), astrocytoma IDHmt (n = 38) and glioblastoma IDH wildtype (IDHwt) (n = 124). DLL3 protein expression was assessed by immunohistochemistry, and samples were categorized as DLL3-negative (0%), DLL3-low (>0-10%), DLL3-medium (10%-50%) and DLL3-high (>50%) based on the estimated percentage of DLL3-positive cells within vital tumor tissue.

Results

DLL3 expression was detected in the majority of glioma samples (86%, 164/190). Expression levels were highest in oligodendroglioma (94% medium or high) and astrocytoma (78.4% medium or high), and lowest in glioblastoma (42% medium or high). 26/190 samples were DLL3-negative. In most patients, DLL3 expression stayed in the same category or increased at recurrence (oligodendroglioma 87.6%, astrocytoma 66.6%, glioblastoma 77.2%). Loss of initial DLL3-positivity at recurrence occurred in only 7/91 patients (7.7%).

Conclusion

DLL3 expression level is retained or increased in the majority of recurrent gliomas. Inclusion in trials with DLL3-targeting agents in recurrent glioma based on primary tumor material is justified, depending on the required level of DLL3 positivity for inclusion.

Keywords: DLL3, glioma, immunotherapy, Notch, recurrence

Key Points.

The inhibitory Notch ligand DLL3 is expressed in the majority of gliomas
DLL3 expression is retained at tumor recurrence in the majority of patients
We provide translational support for the evaluation of DLL3-targeted (immuno)therapies in clinical trials

Importance of the Study

There is a great need for new treatment strategies in patients with diffuse gliomas that progress beyond standard-of-care therapies. Several therapeutic constructs targeting the inhibitory Notch ligand DLL3 have been developed, and since DLL3 expression has previously been demonstrated in glioma, particularly in IDHmt gliomas, such treatments might also be considered for glioma patients. However, a major unanswered question before moving DLL3-targeted treatments forward into clinical trials with recurrent glioma patients is whether DLL3 expression is retained in recurrent tumors. Here, we apply a robust and ready-to-use immunohistochemistry method to show that DLL3 is expressed in the majority of patients with oligodendroglioma, IDHmt and 1p/19q co-deleted, astrocytoma IDHmt and glioblastoma IDHwt in a large and well-annotated clinical cohort, and that initial DLL3 expression is largely retained at recurrence. This helps fortify the rationale to conduct clinical trials using DLL3-targeting therapies in these patient populations.

Adult diffuse gliomas represent the majority of primary malignant brain tumors in adults, and remain invariably lethal.¹ The WHO 2021 classification of CNS tumors combines histological and molecular parameters to classify gliomas into subtype and grade.² The main subtypes are oligodendroglioma, isocitrate dehydrogenase mutant (IDHmt) and 1p/19q codeleted; astrocytoma, IDHmt; and glioblastoma, IDH wildtype (IDHwt); other types of primary glial brain tumors are rare in adults.¹ Standard of care for newly diagnosed gliomas usually consists of maximal safe resection followed by a combination of radiotherapy and/or chemotherapy, either with temozolomide³ or the PCV-regimen (Procarbazine, CCNU, and Vincristine).⁴ However, these treatments are not curative, and all patients eventually progress after first-line therapies. Treatment strategies at the time of progression are highly variable and of limited efficacy. The choice of treatment at progression usually depends on several factors, including tumor size, location, assumed chemosensitivity of the tumor, time between initial treatment and recurrence, clinical performance status of the patient and goals of care.¹ In many cases, treatment in clinical trials should be considered at this point.

DLL3 is an inhibitory ligand in the Notch signaling pathway, a highly conserved signaling pathway that plays a critical role in cell fate decisions during embryonic development. Under physiological conditions, DLL3 is only transiently expressed during embryonic development. DLL3 expression occurs mostly in the presomitic mesoderm, the precursor to adult vertebrae, and mutations in DLL3 can cause axial skeletal defects.⁵ Additionally, DLL3 expression plays an important role in neurogenesis, and has been documented in the fetal brain and spinal cord.⁶ Involvement of DLL3 and the Notch-pathway has been implicated in various biological processes related to malignancies, including cell proliferation, malignant transformation and the epithelial-to-mesenchymal transition.⁷

DLL3 is highly and homogeneously expressed on the cell surface in several malignancies, including small cell lung cancer (SCLC) and other neuro-endocrine tumors, whereas no DLL3 expression is observed in healthy adult tissues.⁸ This establishes DLL3 as a tumor-associated antigen and possible therapeutic target, and has led to the development of several targeted (immuno)therapies directed against DLL3, such as antibody-drug conjugates (ADCs), bispecific T cell engagers (TCEs), and chimeric antigen receptor (CAR)-T cells (as reviewed in⁸) as well as radio-immunotherapies, in which anti-DLL3 antibodies are conjugated to a radioisotope for imaging or therapeutic purposes.⁹ Of these therapies, the bispecific TCE tarlatamab has recently been approved by the FDA for treatment in chemotherapy-refractory, advanced-stage SCLC after a phase 2 clinical trial reported objective responses in 40% of patients, far exceeding the historical control benchmark of 15%.¹⁰^,¹¹ There is now considerable interest in expanding the application of DLL3-targeted therapies to other types of malignancies as well.¹²

While studying DLL3 in pulmonary neuro-endocrine tumors, Saunders and colleagues investigated RNA sequencing data from The Cancer Genome Atlas (TCGA) and found that DLL3 expression levels were elevated in diffuse glioma as well.¹³ Spino and colleagues report that IDHmt gliomas in the TCGA dataset harbor significantly higher DLL3 expression compared to IDHwt glioma, and using immunohistochemistry (IHC) for DLL3 in a cohort of 63 glioma patients with mostly primary tumor samples (58/63), they observe high, homogeneous DLL3 expression in most IDHmt gliomas, especially 1p/19q codeleted oligodendrogliomas, as opposed to patchy, weak DLL3 expression in IDHwt tumors.¹⁴ High protein expression of DLL3 in low grade gliomas was confirmed by Noor and colleagues in 2021, who reported that 72.9% (35/48) of low grade gliomas stained positive for DLL3 using IHC (67.7%, 21/31 in astrocytoma and 14/17, 82.3% in oligodendrogliomas).⁶ Thus, DLL3-targeted therapies have gained interest for the treatment of glioma as well, and a clinical trial with the bi-specific T cell engager (TCE) obrixtamig (BI 764532) is now ongoing in glioma patients with recurrence after first-line therapies (NCT05916313). This trial enrolls patients based on DLL3 protein expression using immunohistochemistry (IHC) on archival tissue from the primary tumor, which is a regular approach in neuro-oncology given the invasiveness and risk of biopsies at the time of recurrence; however, this requires the reasonable likelihood that the target is still present at the time of progression.¹⁵ Therefore, the aim of the present study was (1) to determine the extent of DLL3 expression in relation to IDH mutational status in a large and well-annotated cohort of paired primary and recurrent gliomas, and (2) to determine whether DLL3 expression status of the primary tumor is retained at recurrence.

Methods

Patient Inclusion

Patients were selected based on the availability of formalin-fixed and paraffin-embedded (FFPE) material of paired primary and recurrent gliomas, with an interval of at least six months between the first and second resection. Tumor classification was based on pathology reports and was done in accordance with the WHO 2021 classification.² FFPE samples were collected between 1990 and 2023, and were stored at the biobank in the Erasmus MC, Rotterdam, the Netherlands. All IDHmt glioma samples were derived from the GLASS-OD and GLASS-NL databases, as previously established through the Glioma Longitudinal Analysis Consortium (GLASS), with molecular characterization available including IDH mutational status.¹⁶ For all IDHmt tumors and most (45/62) IDHwt glioblastomas, a clinically targeted next-generation sequencing (NGS) panel was applied as previously described.¹⁷ The remaining 17 patients diagnosed with IDHwt glioblastoma were >55 years of age, and in accordance with WHO guidelines further molecular analysis beyond IDH-R132H IHC was not routinely performed. Clinical information was extracted from electronic patient records. The study was conducted according to institutional and national regulations and was approved by the institutional review board of the Erasmus MC METC (Rotterdam, The Netherlands, nr MEC-2022-0818-A-0002). All patients provided written informed consent.

Immunohistochemical Staining for DLL3 and IDH1-R132H

FFPE tissues were cut into 4 µm sections and placed on glass slides along with a tissue micro-array (TMA) of positive controls (SCLC) and negative controls (tonsil, kidney and liver tissue; Figure 2E). Immunohistochemistry (IHC) for DLL3 was performed using an automated, validated and accredited ready-to-use staining kit, which has been proven to give robust results and is already used for DLL3 diagnostics and for determining patient eligibility in clinical studies (Ventana Benchmark ULTRA, Ventana Medical Systems, Arizona, USA). In brief, following deparaffinization and heat-induced antigen retrieval, tissue samples were incubated with anti-DLL3 rabbit monoclonal antibody (clone SP347) for 32 minutes at a concentration of 7.40 µg/ml, followed by a haematoxylin II counterstain for 20 minutes and a blue coloring reagent for 8 minutes according to the manufacturer’s instructions (Ventana Medical Systems Inc, Arizona, USA). For detection the Optiview universal DAB detection kit was used (CC1, 80′). Consecutive 4 µm tissue sections were stained with haematoxylin & eosin (H&E), and in IDHmt tumors, consecutive slides were stained for IDH-R132H as a tumor marker using a mouse monoclonal antibody (clone H09, Dianova), following standard diagnostic practice at our institution. All stained slides were digitally imaged at 40X magnification using a NanoZoomer digital slide scanner (Hamamatsu Photonics K.K., Japan).

Figure 2. — Representative images of observed DLL3 staining patterns in different glioma subtypes. (A) DLL3 expression in a recurrent oligodendroglioma, largely following the same staining pattern as the IDH1-R132H staining. Scale bar: 2mm. (B) consecutive slide stained for IDH1-R132H of the same recurrent oligodendroglioma. Only the patch marked in the top right box does express IDH1-R132H but does not express DLL3. Scale bar: 2mm. (C) homogeneous DLL3 expression in an astrocytoma. Scale bar: 400µm. (D) more patchy and heterogeneous DLL3 expression in a glioblastoma. (E) TMA of positive and negative controls, from left to right and from top to bottom: tonsil, kidney, SCLC (3×), liver. (F-I) representative images of the different DLL3-scoring categories: negative, low, medium and high. Scale bar: 250 µm.

Determining and Categorizing DLL3 Positivity

DLL3 scoring was performed under supervision of a certified neuropathologist (S. M.) using standard light microscopy and was done blinded to the diagnosis, pathology report, and resection number. DLL3-positivity was defined as any cytoplasmic or membranous staining in tumor cells, with any intensities. We estimated the percentage of DLL3-positive cells within areas of vital tumor tissue based on visual inspection; areas containing necrosis or normal-appearing brain tissue were excluded based on the assessment of consecutive H&E-stained slides. Each sample was categorized into one of the following bins: DLL3-negative (0%), low (>0%-10%), medium (>10%-50%), or high (>50% positive cells). These cutoff points had previously been established in a consensus meeting with a multidisciplinary panel of experts, including neuro-oncologists, neuropathologists and senior clinical biomarker scientists. An overview of the workflow is shown in Figure 1, and representative images for each of the scoring categories are shown in Figure 2F-I. All statistical analyses were performed using R version 4.4.1¹⁸ and RStudio version 2024.9.0.375.¹⁹

Figure 1. — Schematic representation of the workflow. OD = oligodendroglioma IDHmt and 1p/19q co-deleted, A = astrocytoma IDHmt, GBM = glioblastoma IDHwt. This figure was created using BioRender and can be accessed through the following link: https://app.biorender.com/illustrations/67288664e7895662d16b856e? slideId=15dc6777-8591-4c3d-9ef4-61d19dc2648a (P. French, 2025).

Results

Patient Demographics

We collected 198 samples of paired primary and recurrent gliomas from n = 99 patients, including oligodendroglioma, IDHmt and 1p/19q codeleted (n = 36 samples from 18 patients), astrocytoma, IDHmt (n = 38 samples from 19 patients), and glioblastoma, IDHwt (n = 124 samples from 62 patients). 190 out of 198 samples were of sufficient quality for analysis, and 182 samples from n = 91 patients could be analyzed in matched pairs. An overview of DLL3 scores per patient for the first and second resection can be found in Supplement 1, and a summary of patient characteristics can be found in Table 1. The median age at diagnosis for patients with oligodendroglioma and astrocytoma was younger than those with glioblastoma (42.5, 29, and 52 years, respectively). Most patients with IDHmt gliomas had WHO grade 2 tumors at diagnosis (27/37 patients, 73%). The intervals between the first and second surgery were longer in oligodendroglioma and astrocytoma when compared to glioblastoma (62.5, 54, and 11 months, respectively), consistent with their slower growth. Treatment strategies between the first and second resection were heterogeneous, especially for patients with oligodendroglioma and astrocytoma, of whom 50% and 53% respectively did not receive any anti-tumor treatment between surgeries. Most patients with glioblastoma received radiotherapy with concomitant and adjuvant temozolomide according to the Stupp regimen³ (53/62, 85%).

Table 1.

Patient demographics

Diagnosis	Oligodendroglioma, IDHmt and 1p/19q codeleted	Astrocytoma, IDHmt	Glioblastoma, IDHwt
Number of patients (n, %)	18/99 (18%)	19/99 (19%)	62/99 (63%)
Gender
Male (n, %)	9/18 (50%)	6/19 (32%)	19/62 (31%)
Female (n, %)	9/18 (50%)	13/19 (68%)	43/62 (69%)
Age at diagnosis (Y)
Median (range)	42.5 (20-66)	29 (21-55)	52 (31-74)
WHO grade at diagnosis (n, %)
2	11/18 (61%)	16/19 (84%)	0
3	7/18 (39%)	2/19 (11%)	0
4	0	1/19 (5%)	62 (100%)
Months between resections
Median (range)	62.5 (13-198)	54 (19-110)	11 (6-46)
Treatment between R1 and R2
None	9/18 (50%)	10/19 (53%)	1/62 (2%)
RT only	7/18 (39%)	3/19 (16%)	3/62 (5%)
RT + chemo	2/18 (11%)	6/19 (32%)	53/62 (85%)
Other (eg RT + TKI or immunotherapy)	0	0	5/62 (8%)

Open in a new tab

Abbreviations: R1 = resection 1; R2 = resection 2; RT = radiotherapy; TKI = tyrosine kinase inhibitor.

DLL3 is Expressed in the Majority of Gliomas, and is Highest in IDHmt Tumors

In our cohort of glioma patients, we observed DLL3 expression in the majority of samples (86%). In IDHmt tumors, the staining patterns for IDH-R132H and DLL3 were largely concordant, further supporting that DLL3 is primarily expressed in malignant cells (Figure 2A and B). However, in IDHmt gliomas, not all IDH-positive cells were also DLL3-positive (as illustrated by the box in Figure 2A and B). DLL3 staining was more intense and more homogeneous in IDHmt gliomas, compared to more heterogeneous/patchy and less intense staining in IDHwt glioblastomas (representative images are shown in Figure 2C and D). In our cohort, DLL3 expression fell in any of the non-negative scoring categories in all IDHmt tumors, and in 86% of the total population (164/190 samples; Figure 3A and B). The 26/190 samples that were completely DLL3-negative were all IDHwt glioblastomas (Figure 3B). DLL3 expression was highest in oligodendroglioma (94% medium or high) and astrocytoma (78.4% medium or high), and lowest in glioblastoma (42% medium or high) (Figure 3B). High DLL3 expression (defined as >50% DLL3-positive cells in vital tumor tissue) was observed in 38% of oligodendrogliomas, 24% of astrocytomas and 5% of glioblastomas (Figure 4A, D, and G, respectively).

Figure 3. — DLL3 expression is significantly higher in IDHmt glioma compared to IDHwt glioma, and expression is largely concordant between primary and recurrent gliomas. (A) overview of DLL3 expression in the cohort (n = 190 samples from 99 patients with oligodendroglioma, astrocytoma or glioblastoma). In 86% of cases there was at least some DLL3 expression. (B) DLL3 expression stratified by glioma subtype. DLL3 expression is either medium or high in 94% of oligodendrogliomas (OD), 78.4% of astrocytomas (A), and 42% of glioblastomas (GBM). DLL3 expression in OD and A is significantly higher compared to GBM (P = 1.7 × 10⁻⁹ and P = 9.6 × 10⁻⁶, respectively), as determined using Dunn’s test with Benjamini-Hochberg correction. (C) alluvial plot of DLL3-score change between resection 1 and 2 in all gliomas. (D) DLL3-score category change in all gliomas; in 77% of gliomas, DLL3 expression stays in the same category or increases at recurrence.

Figure 4. — When stratified by glioma subtype, DLL3 is largely retained at recurrence in all subtypes. (A) DLL3 expression in oligodendroglioma. (B) alluvial plot of DLL3-score change between resection 1 and 2 in oligodendrogliomas. (C) DLL3-score category change in oligodendroglioma; in 87.6% of oligodendrogliomas, DLL3 expression stays in the same category or increases at recurrence. (D) DLL3 expression in astrocytoma. (E) alluvial plot of DLL3-score change between resection 1 and 2 in astrocytomas. (F) DLL3-score category change in astrocytoma; in 66.6% of astrocytomas, DLL3 expression stays in the same category or increases at recurrence. (G) DLL3 expression in glioblastoma. (H) alluvial plot of DLL3-score change between resection 1 and 2 in glioblastoma. (I) DLL3-score category change in glioblastoma; in 77.2% of glioblastomas, DLL3 expression stays in the same category or increases at recurrence.

To determine whether the observed difference between IDHmt and IDHwt tumors was statistically significant, we first applied the non-parametric Kruskal–Wallis test, appropriate for comparing more than two independent groups when the response variable is ordinal and may violate normality assumptions. This revealed a statistically significant difference in DLL3 expression levels across tumor types (P = 2.97 × 10⁻¹¹). Subsequent pairwise comparisons using Dunn’s test with Benjamini-Hochberg correction identified significant differences between glioblastoma and oligodendroglioma (P = 1.7 × 10⁻⁹) and glioblastoma and astrocytoma (P = 9.6 × 10⁻⁶), whereas no significant difference was observed between oligodendroglioma and astrocytoma (P = .135). These results indicate that DLL3 expression is significantly higher in oligodendroglioma, IDHmt and 1p/19q codeleted and astrocytoma, IDHmt when compared to IDHwt glioblastoma (Figure 3B).

DLL3 Expression is Largely Retained in Recurrent Tumors

As expected, primary tumor tissue sections proportionally contained more vital tumor tissue, whereas we observed more therapy-effects such as radionecrosis in tumors that recurred after treatment (data not shown). We found that in all tumor types, DLL3 expression within vital tumor tissue was largely retained between primary and recurrent tumors. In matched pairs, DLL3 expression either stayed in the same category or increased at recurrence in 77% of patients (70/91 patients; Figure 3C and D). When stratified by glioma subtype, DLL3 expression appeared to remain most constant in oligodendrogliomas IDHmt 1p/19q co-deleted; 14/16 patients (87.6%) stayed in the same category or increased at recurrence; Figure 4B and C. DLL3 expression was slightly more variable in astrocytomas IDHmt (12/18 patients, 66.6% stayed in the same category or increased at recurrence; Figure 4E and F) and glioblastomas IDHwt (44/57 patients, 77.2% stayed in the same category or increased at recurrence; Figure 4H and I). Furthermore, even in the patient group with decreased DLL3 expression at recurrence, most patients only decreased one category; thus, most recurrent tumors retained at least some DLL3 expression at recurrence. Only 7/91 tumors that previously expressed DLL3 became DLL3-negative at recurrence (7.7%; Figure 4h); all of these were glioblastoma, IDHwt. Conversely, 9/91 patients who were categorized as DLL3-negative at the primary resection became positive at recurrence; all of these were glioblastoma, IDHwt as well (9.9%; Figure 4H).

Discussion

Despite significant advancements in our understanding of glioma biology, including insights into the tumor microenvironment (TME), most of the recently investigated novel therapies have failed to improve clinical outcomes for glioma patients. One of the key challenges lies in the high temporal, intra- and intertumoral heterogeneity of gliomas; therapeutic targets that are widely and homogeneously expressed in glioma but absent in normal brain tissue are relatively rare. Since trials on targeted treatments in recurrent glioma typically depend on demonstration of target presence at initial diagnosis and rarely require confirmation at the time of progression, the persistence and stability of target expression in samples at the time of progression should be documented in series before the target is exploited in patients with recurrent disease without demonstrated expression at the time of recurrence.¹⁵^,²⁰

In this study, we assessed the expression of the tumor-associated antigen DLL3 in paired primary and recurrent gliomas in the largest patient cohort to date. Our results show that DLL3 is highly and consistently expressed across glioma subtypes, particularly in IDHmt gliomas, and that DLL3 expression is preserved after recurrence in the majority of patients. These data provide support for DLL3 as a potentially suitable and attractive treatment target and encourage the field to move forward with clinical trials for DLL3-targeted (immuno)therapies in recurrent glioma.

In accordance with previous studies, we find that DLL3 expression is significantly higher in IDHmt gliomas compared to IDHwt glioblastomas. However, our finding of universal DLL3-positivity in IDHmt astrocytomas and oligodendrogliomas differs from previously published IHC results by Noor et al, who reported DLL3-positivity in 21/31 astrocytomas (67.7%) and 14/17 (82.3%) oligodendrogliomas using a similar visual scoring method to our study.⁶ This discrepancy might partly reflect changes in the WHO classification of CNS tumors: notably, 6 of the 31 cases classified as astrocytomas in the study by Noor et al are IDHwt, and would now be reclassified as glioblastomas. However, even among cases with a confirmed IDH mutation in their dataset, 7/39 were reported as DLL3-negative. Immunohistochemical staining for DLL3 is known to be technically challenging, and differences in antibody sensitivity and staining protocols might contribute to this variation. In our work, we employed a validated and accredited IHC assay (Ventana, AZ, USA), which is widely used in clinical trials and has demonstrated high sensitivity and specificity for DLL3 detection; a comparative study by Brcic et al showed that the Ventana assay significantly outperforms most other commercially available antibodies.²¹^,²² Thus, it is possible that the use of a different antibody may account for the negative cases reported. Furthermore, tumor heterogeneity may introduce sampling bias and contribute to the differences between our studies, especially in glioblastoma, IDHwt samples.

Consistent with previous work, we find that DLL3 is most widely and homogeneously expressed in oligodendroglioma, IDHmt, and 1p/19q codeleted. Interestingly, the DLL3 gene is located on chromosome 19q13.13, which falls within the region of loss-of-heterozygosity (LOH) characteristic for this glioma type. A possible explanation might be that 1p/19q-LOH and DLL3 expression are functionally linked, for instance, through (compensatory) transcriptional or epigenetic upregulation of DLL3 protein expression as a result of 1p/19q LOH, although this remains speculative. Alternatively, given the important role of DLL3 in cell fate decisions during neurodevelopment and the transient nature of DLL3 expression during embryonic development, oligodendroglioma cells might more closely resemble developmental states where DLL3 is ‘naturally’ high, regardless of 1p/19q LOH. Tirosh et al. have previously proposed a cellular developmental hierarchy in oligodendrogliomas, where astrocytic (AC)-like, oligodendrocytic (OC)-like, and neural progenitor cell (NPC)-like cells are each characterized by different developmental programs and transcriptional signatures,²³ and the Notch-signaling pathway was shown to be involved in such lineage-differentiation decisions.²⁴ Inactivation of the Notch pathway has been associated with a more aggressive disease course in oligodendroglioma.²⁵ Interestingly, in glioblastoma, DLL3 expression has historically been considered a hallmark of the proneural variant, a subtype that is enriched for genes related to oligodendrocytic development.²⁶^,²⁷

Furthermore, previous work suggests that DLL3 expression correlates with changes in the tumor micro-environment (TME) in SCLC. Shirasawa et al find suppression of several immune-related pathways in SCLC tumors with high DLL3 expression (defined here as DLL3 expression above median based on transcriptome data), as opposed to DLL3-low tumors (DLL3 expression below median), including pathways related to antigen presentation and T cell trafficking. In their study, DLL3-high samples contained significantly smaller proportions of T cells, macrophages, and dendritic cells.²⁸ Noor et al also suggest that DLL3 protein expression and DLL3 methylation correlate with changes in the immune micro-environment in gliomas, although the effects they find differ per immune cell subset and differ for IDHmt and IDHwt gliomas.⁶ This might be explored in future studies.

There are some limitations in our study that need to be addressed, including tumor heterogeneity which might introduce sampling bias, and the use of a manual review method. While our manual review method may raise concerns about intra- and inter observer variability, the aim of this study was to resemble current practice as closely as possible, and IHC assessment and sample categorization by a single neuropathologist is still the gold standard for assessment of DLL3 expression in clinical routine, especially in the context of clinical trials. Furthermore, we think that our approach of categorization avoids over-interpretation of minor differences in expression that may not be biologically relevant. Another important aspect to consider is that the clinical relevance of our chosen DLL3 expression categories is currently unknown. Given the absence of established thresholds for DLL3 expression in glioma, we categorized samples into four expression groups (negative, low, medium, high), established based on multi-disciplinary expert consensus. However, a clear level of DLL3 expression required to derive therapeutic benefit from DLL3-targeted therapies has not been established yet and may differ per tumor type and therapeutic construct. For instance, the cutoff value for HER2 positivity in order to be eligible for treatment with the HER2-targeted monoclonal antibody trastuzumab in breast cancer is defined as 10% in clinical guidelines,²⁹ even though response is not linear to biomarker expression,³⁰ whereas for checkpoint inhibitors targeting PD-1 in non-small cell lung cancer, >1% expression is already sufficient to induce a clinically meaningful response.³¹ Given the high and homogeneous expression of DLL3 in SCLC, most clinical trials with DLL3-targeted therapies in SCLC to date were conducted regardless of DLL3 expression, and FDA approval for tarlatamab in SCLC is not contingent on DLL3 expression level.¹¹ Retrospective analyses of the DeLLphi-trials showed that in refractory SCLC, objective responses after treatment with tarlatamab were observed irrespective of DLL3 expression level.¹¹^,³² Thus, to what extent therapy response is dependent on DLL3 expression level in SCLC is currently unknown, and is likely to differ per therapeutic construct. Therefore, it is crucial that future studies determine the clinical relevance of DLL3 expression level in regard to therapy response and survival in glioma. Since data from trials with DLL3-targeting therapies in glioma are not available yet, this currently remains an open question. The intensity of expression level per cell might affect therapy response as well and should be considered in future studies. However, given the high potency of certain T cell engagers even at low antigen density,³³ perhaps even low or patchy DLL3 expression may suffice to induce a meaningful response to certain therapies, but this remains to be determined in future studies.

To conclude, we show that DLL3 is widely expressed across diffuse glioma subtypes in our cohort, especially in IDHmt glioma, and that expression is retained at recurrence in most patients. However, even though the stability of DLL3 expression is high overall, we do observe some variability across glioma subgroups. Particularly, the concordance rate of 66.6% in IDHmt astrocytoma means that a subset of these tumors (6/18 patients) do show a decrease in DLL3 expression at recurrence. Whether this is clinically relevant remains to be determined. Importantly, DLL3 expression was lost in only 7 patients in our cohort, all diagnosed with glioblastoma, IDHwt; and thus, DLL3 expression was never completely lost at recurrence in patients with IDHmt glioma.

Biomarker dynamics are important to consider for clinical trial design, as eligibility for recurrent-disease in neuro-oncological studies is often based on biomarker expression in archival tissue from the first resection. Even though our findings suggest that primary tumor material can serve as a reasonable surrogate for DLL3 status at recurrence for many patients, reliance on archival tissue should be aligned with the required DLL3-positivity threshold of the specific therapeutic approach that is investigated. Overall, the high prevalence and relative stability of DLL3 expression support continued development of DLL3-targeted therapies in both newly diagnosed and recurrent glioma.

Supplementary Material

vdag027_Supplementary_Data

vdag027_supplementary_data.xlsx^{(12KB, xlsx)}

Acknowledgements

The authors would like to thank all members of the GLASS consortium.

Contributor Information

Rosa Luning, Department of Neurology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands.

Sybren L N Maas, Department of Pathology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, The Netherlands; Department of Pathology, Leiden University Medical Center, Leiden, the Netherlands.

Pim J French, Department of Neurology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands.

Youri Hoogstrate, Department of Neurology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands.

Levi van Hijfte, Department of Neurosurgery, University Clinic Erlangen, Erlangen, Germany.

Rania Head, Department of Neurology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands.

Iris de Heer, Department of Neurology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands.

Thierry P P van den Bosch, Department of Pathology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, The Netherlands.

Adelheid Woehrer, Institute of Neuropathology and Neuromolecular Pathology, Medical University of Innsbruck, Innsbruck, Austria.

Martin J van den Bent, Department of Neurology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands.

Marjolein Geurts, Department of Neurology, Brain Tumor Center, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands; Department of Medical Oncology, Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, the Netherlands.

Supplementary Material

Supplementary material is available online at Neuro-Oncology Advances (https://academic.oup.com/noa).

Author Contributions

Conceptualization: R.L., S.M., P.F., A.W., M.G., M.B. Experimental work: T.B. and R.L. DLL3 scoring and analysis: S.M. and R.L. Writing (initial): R.L. Supervision and funding: M.G., M.B., P.F. Review & editing: all co-authors.

Conflict of Interest Statement

This study was supported by a research grant from Boehringer-Ingelheim (RL, MG, MvdB).

Funding

This study was funded by a research grant from Boehringer-Ingelheim (RL, MG, MvdB) and partially funded through The Netherlands Organisation for Health Research and Development (ZonMW; RL, MG).

Ethical Approval

All studies were conducted according to institutional and national regulations and were approved by the institutional review board of Erasmus MC (Rotterdam, The Netherlands, nr MEC-2022-0818-A-0002).

Data Availability

The data generated in this study are available upon reasonable request to the corresponding author.

References

1. van den Bent MJ, Geurts M, French PJ, et al. Primary brain tumours in adults. Lancet. 2023;402:1564-1579. 10.1016/S0140-6736(23)01054-1 [DOI] [PubMed] [Google Scholar]
2. Louis DN, Perry A, Wesseling P, et al. The 2021 WHO classification of tumors of the central nervous system: a summary. Neuro Oncol. 2021;23:1231-1251. 10.1093/neuonc/noab106 [DOI] [PMC free article] [PubMed] [Google Scholar]
3. Stupp R, Mason WP, van den Bent MJ, et al. , National Cancer Institute of Canada Clinical Trials Group Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med. 2005;352:987-996. 10.1056/NEJMoa043330 [DOI] [PubMed] [Google Scholar]
4. van den Bent MJ. Thirty years of progress in the management of low-grade gliomas. Rev Neurol (Paris). 2023;179:425-429. 10.1016/j.neurol.2023.03.001 [DOI] [PubMed] [Google Scholar]
5. Bulman MP, Kusumi K, Frayling TM, et al. Mutations in the human delta homologue, DLL3, cause axial skeletal defects in spondylocostal dysostosis. Nat Genet. 2000;24:438-441. 10.1038/74307 [DOI] [PubMed] [Google Scholar]
6. Noor H, Whittaker S, McDonald KL. DLL3 expression and methylation are associated with lower-grade glioma immune microenvironment and prognosis. Genomics. 2022;114:110289. 10.1016/j.ygeno.2022.110289 [DOI] [PubMed] [Google Scholar]
7. Peddio A, Pietroluongo E, Lamia MR, et al. DLL3 as a potential diagnostic and therapeutic target in neuroendocrine neoplasms: a narrative review. Crit Rev Oncol Hematol. 2024;204:104524. 10.1016/j.critrevonc.2024.104524 [DOI] [PubMed] [Google Scholar]
8. Owen DH, Giffin MJ, Bailis JM, Smit MD, Carbone DP, He K. DLL3: an emerging target in small cell lung cancer. J Hematol Oncol. 2019;12:61. 10.1186/s13045-019-0745-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
9. Tully KM, Tendler S, Carter LM, et al. Radioimmunotherapy targeting delta-like ligand 3 in small cell lung cancer exhibits antitumor efficacy with low toxicity. Clin Cancer Res. 2022;28:1391-1401. 10.1158/1078-0432.CCR-21-1533 [DOI] [PMC free article] [PubMed] [Google Scholar]
10. Dhillon S. Tarlatamab: first approval. Drugs. 2024;84:995-1003. 10.1007/s40265-024-02070-z [DOI] [PubMed] [Google Scholar]
11. Ahn MJ, Cho BC, Felip E, et al. , DeLLphi-301 Investigators Tarlatamab for patients with previously treated small-cell lung cancer. N Engl J Med. 2023;389:2063-2075. 10.1056/NEJMoa2307980 [DOI] [PubMed] [Google Scholar]
12. Chou J, Egusa EA, Wang S, et al. Immunotherapeutic targeting and PET imaging of DLL3 in small-cell neuroendocrine prostate cancer. Cancer Res. 2023;83:301-315. 10.1158/0008-5472.CAN-22-1433 [DOI] [PMC free article] [PubMed] [Google Scholar]
13. Saunders LR, Bankovich AJ, Anderson WC, et al. A DLL3-targeted antibody-drug conjugate eradicates high-grade pulmonary neuroendocrine tumor-initiating cells in vivo. Sci Transl Med. 2015;7:302ra136. 10.1126/scitranslmed.aac9459 [DOI] [PMC free article] [PubMed] [Google Scholar]
14. Spino M, Kurz SC, Chiriboga L, et al. Cell surface notch ligand DLL3 is a therapeutic target in isocitrate dehydrogenase-mutant glioma. Clin Cancer Res. 2019;25:1261-1271. 10.1158/1078-0432.CCR-18-2312 [DOI] [PMC free article] [PubMed] [Google Scholar]
15. Draaisma K, Chatzipli A, Taphoorn M, et al. Molecular evolution of IDH wild-type glioblastomas treated with standard of care affects survival and design of precision medicine trials: a report from the EORTC 1542 study. J Clin Oncol. 2020;38:81-99. 10.1200/JCO.19.00367 [DOI] [PubMed] [Google Scholar]
16. Consortium G. Glioma through the looking GLASS: molecular evolution of diffuse gliomas and the glioma longitudinal analysis consortium. Neuro Oncol. 2018;20:873-884. [DOI] [PMC free article] [PubMed] [Google Scholar]
17. Wijnenga MMJ, French PJ, Dubbink HJ, et al. The impact of surgery in molecularly defined low-grade glioma: an integrated clinical, radiological, and molecular analysis. Neuro Oncol. 2018;20:103-112. 10.1093/neuonc/nox176 [DOI] [PMC free article] [PubMed] [Google Scholar]
18. R: A Language and Environment for Statistical Computing. [Computer Program]. Vienna, Austria: R Foundation for Statistical Computing; 2024. [Google Scholar]
19. RStudio: Integrated Development Environment for R. [Computer Program]. Boston, MA: Posit Software, PBC; 2024. [Google Scholar]
20. van den Bent MJ, Gao Y, Kerkhof M, et al. Changes in the EGFR amplification and EGFRvIII expression between paired primary and recurrent glioblastomas. Neuro Oncol. 2015;17:935-941. 10.1093/neuonc/nov013 [DOI] [PMC free article] [PubMed] [Google Scholar]
21. Rudin CM, Pietanza MC, Bauer TM, et al. , SCRX16-001 investigators Rovalpituzumab tesirine, a DLL3-targeted antibody-drug conjugate, in recurrent small-cell lung cancer: a first-in-human, first-in-class, open-label, phase 1 study. Lancet Oncol. 2017;18:42-51. 10.1016/S1470-2045(16)30565-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
22. Brcic L, Kuchler C, Eidenhammer S, et al. Comparison of four DLL3 antibodies performance in high grade neuroendocrine lung tumor samples and cell cultures. Diagn Pathol. 2019;14:47. 10.1186/s13000-019-0827-z [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Tirosh I, Venteicher AS, Hebert C, et al. Single-cell RNA-seq supports a developmental hierarchy in human oligodendroglioma. Nature. 2016;539:309-313. 10.1038/nature20123 [DOI] [PMC free article] [PubMed] [Google Scholar]
24. Spitzer A, Gritsch S, Nomura M, et al. Mutant IDH inhibitors induce lineage differentiation in IDH-mutant oligodendroglioma. Cancer Cell. 2024;42:904-914 e909. 10.1016/j.ccell.2024.03.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
25. Halani SH, Yousefi S, Velazquez Vega J, et al. Multi-faceted computational assessment of risk and progression in oligodendroglioma implicates NOTCH and PI3K pathways. NPJ Precis Oncol. 2018;2:24. 10.1038/s41698-018-0067-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
26. Verhaak RG, Hoadley KA, Purdom E, et al. , Cancer Genome Atlas Research Network Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1. Cancer Cell. 2010;17:98-110. 10.1016/j.ccr.2009.12.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
27. Phillips HS, Kharbanda S, Chen R, et al. Molecular subclasses of high-grade glioma predict prognosis, delineate a pattern of disease progression, and resemble stages in neurogenesis. Cancer Cell. 2006;9:157-173. 10.1016/j.ccr.2006.02.019 [DOI] [PubMed] [Google Scholar]
28. Shirasawa M, Yoshida T, Shiraishi K, et al. Tumor microenvironment-mediated immune profiles and efficacy of anti-PD-L1 antibody plus chemotherapy stratified by DLL3 expression in small-cell lung cancer. Br J Cancer. 2023;129:2003-2013. 10.1038/s41416-023-02427-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Eisen A, Fletcher GG, Gandhi S, et al. , Members of the Early Breast Cancer Systemic Therapy Consensus Panel Optimal systemic therapy for early breast cancer in women: a clinical practice guideline. Curr Oncol. 2015;22:S67-81. 10.3747/co.22.2320 [DOI] [PMC free article] [PubMed] [Google Scholar]
30. Ronnlund C, Sifakis EG, Schagerholm C, et al. Prognostic impact of HER2 biomarker levels in trastuzumab-treated early HER2-positive breast cancer. Breast Cancer Res. 2024;26:24. 10.1186/s13058-024-01779-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
31. Shien K, Papadimitrakopoulou VA, Wistuba II. Predictive biomarkers of response to PD-1/PD-L1 immune checkpoint inhibitors in non-small cell lung cancer. Lung Cancer. 2016;99:79-87. 10.1016/j.lungcan.2016.06.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Mountzios G, Sun L, Cho BC, et al. , DeLLphi-304 Investigators Tarlatamab in Small-Cell lung cancer after Platinum-Based chemotherapy. N Engl J Med. 2025;393:349-361. 10.1056/NEJMoa2502099 [DOI] [PubMed] [Google Scholar]
33. Cao L, Leclercq-Cohen G, Klein C, Sorrentino A, Bacac M. Mechanistic insights into resistance mechanisms to T cell engagers. Front Immunol. 2025;16:1583044. 10.3389/fimmu.2025.1583044 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

vdag027_Supplementary_Data

vdag027_supplementary_data.xlsx^{(12KB, xlsx)}

Data Availability Statement

The data generated in this study are available upon reasonable request to the corresponding author.

[vdag027-B1] 1. van den Bent MJ, Geurts M, French PJ, et al. Primary brain tumours in adults. Lancet. 2023;402:1564-1579. 10.1016/S0140-6736(23)01054-1 [DOI] [PubMed] [Google Scholar]

[vdag027-B2] 2. Louis DN, Perry A, Wesseling P, et al. The 2021 WHO classification of tumors of the central nervous system: a summary. Neuro Oncol. 2021;23:1231-1251. 10.1093/neuonc/noab106 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B3] 3. Stupp R, Mason WP, van den Bent MJ, et al. , National Cancer Institute of Canada Clinical Trials Group Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med. 2005;352:987-996. 10.1056/NEJMoa043330 [DOI] [PubMed] [Google Scholar]

[vdag027-B4] 4. van den Bent MJ. Thirty years of progress in the management of low-grade gliomas. Rev Neurol (Paris). 2023;179:425-429. 10.1016/j.neurol.2023.03.001 [DOI] [PubMed] [Google Scholar]

[vdag027-B5] 5. Bulman MP, Kusumi K, Frayling TM, et al. Mutations in the human delta homologue, DLL3, cause axial skeletal defects in spondylocostal dysostosis. Nat Genet. 2000;24:438-441. 10.1038/74307 [DOI] [PubMed] [Google Scholar]

[vdag027-B6] 6. Noor H, Whittaker S, McDonald KL. DLL3 expression and methylation are associated with lower-grade glioma immune microenvironment and prognosis. Genomics. 2022;114:110289. 10.1016/j.ygeno.2022.110289 [DOI] [PubMed] [Google Scholar]

[vdag027-B7] 7. Peddio A, Pietroluongo E, Lamia MR, et al. DLL3 as a potential diagnostic and therapeutic target in neuroendocrine neoplasms: a narrative review. Crit Rev Oncol Hematol. 2024;204:104524. 10.1016/j.critrevonc.2024.104524 [DOI] [PubMed] [Google Scholar]

[vdag027-B8] 8. Owen DH, Giffin MJ, Bailis JM, Smit MD, Carbone DP, He K. DLL3: an emerging target in small cell lung cancer. J Hematol Oncol. 2019;12:61. 10.1186/s13045-019-0745-2 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B9] 9. Tully KM, Tendler S, Carter LM, et al. Radioimmunotherapy targeting delta-like ligand 3 in small cell lung cancer exhibits antitumor efficacy with low toxicity. Clin Cancer Res. 2022;28:1391-1401. 10.1158/1078-0432.CCR-21-1533 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B10] 10. Dhillon S. Tarlatamab: first approval. Drugs. 2024;84:995-1003. 10.1007/s40265-024-02070-z [DOI] [PubMed] [Google Scholar]

[vdag027-B11] 11. Ahn MJ, Cho BC, Felip E, et al. , DeLLphi-301 Investigators Tarlatamab for patients with previously treated small-cell lung cancer. N Engl J Med. 2023;389:2063-2075. 10.1056/NEJMoa2307980 [DOI] [PubMed] [Google Scholar]

[vdag027-B12] 12. Chou J, Egusa EA, Wang S, et al. Immunotherapeutic targeting and PET imaging of DLL3 in small-cell neuroendocrine prostate cancer. Cancer Res. 2023;83:301-315. 10.1158/0008-5472.CAN-22-1433 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B13] 13. Saunders LR, Bankovich AJ, Anderson WC, et al. A DLL3-targeted antibody-drug conjugate eradicates high-grade pulmonary neuroendocrine tumor-initiating cells in vivo. Sci Transl Med. 2015;7:302ra136. 10.1126/scitranslmed.aac9459 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B14] 14. Spino M, Kurz SC, Chiriboga L, et al. Cell surface notch ligand DLL3 is a therapeutic target in isocitrate dehydrogenase-mutant glioma. Clin Cancer Res. 2019;25:1261-1271. 10.1158/1078-0432.CCR-18-2312 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B15] 15. Draaisma K, Chatzipli A, Taphoorn M, et al. Molecular evolution of IDH wild-type glioblastomas treated with standard of care affects survival and design of precision medicine trials: a report from the EORTC 1542 study. J Clin Oncol. 2020;38:81-99. 10.1200/JCO.19.00367 [DOI] [PubMed] [Google Scholar]

[vdag027-B16] 16. Consortium G. Glioma through the looking GLASS: molecular evolution of diffuse gliomas and the glioma longitudinal analysis consortium. Neuro Oncol. 2018;20:873-884. [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B17] 17. Wijnenga MMJ, French PJ, Dubbink HJ, et al. The impact of surgery in molecularly defined low-grade glioma: an integrated clinical, radiological, and molecular analysis. Neuro Oncol. 2018;20:103-112. 10.1093/neuonc/nox176 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B18] 18. R: A Language and Environment for Statistical Computing. [Computer Program]. Vienna, Austria: R Foundation for Statistical Computing; 2024. [Google Scholar]

[vdag027-B19] 19. RStudio: Integrated Development Environment for R. [Computer Program]. Boston, MA: Posit Software, PBC; 2024. [Google Scholar]

[vdag027-B20] 20. van den Bent MJ, Gao Y, Kerkhof M, et al. Changes in the EGFR amplification and EGFRvIII expression between paired primary and recurrent glioblastomas. Neuro Oncol. 2015;17:935-941. 10.1093/neuonc/nov013 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B21] 21. Rudin CM, Pietanza MC, Bauer TM, et al. , SCRX16-001 investigators Rovalpituzumab tesirine, a DLL3-targeted antibody-drug conjugate, in recurrent small-cell lung cancer: a first-in-human, first-in-class, open-label, phase 1 study. Lancet Oncol. 2017;18:42-51. 10.1016/S1470-2045(16)30565-4 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B22] 22. Brcic L, Kuchler C, Eidenhammer S, et al. Comparison of four DLL3 antibodies performance in high grade neuroendocrine lung tumor samples and cell cultures. Diagn Pathol. 2019;14:47. 10.1186/s13000-019-0827-z [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B23] 23. Tirosh I, Venteicher AS, Hebert C, et al. Single-cell RNA-seq supports a developmental hierarchy in human oligodendroglioma. Nature. 2016;539:309-313. 10.1038/nature20123 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B24] 24. Spitzer A, Gritsch S, Nomura M, et al. Mutant IDH inhibitors induce lineage differentiation in IDH-mutant oligodendroglioma. Cancer Cell. 2024;42:904-914 e909. 10.1016/j.ccell.2024.03.008 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B25] 25. Halani SH, Yousefi S, Velazquez Vega J, et al. Multi-faceted computational assessment of risk and progression in oligodendroglioma implicates NOTCH and PI3K pathways. NPJ Precis Oncol. 2018;2:24. 10.1038/s41698-018-0067-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B26] 26. Verhaak RG, Hoadley KA, Purdom E, et al. , Cancer Genome Atlas Research Network Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1. Cancer Cell. 2010;17:98-110. 10.1016/j.ccr.2009.12.020 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B27] 27. Phillips HS, Kharbanda S, Chen R, et al. Molecular subclasses of high-grade glioma predict prognosis, delineate a pattern of disease progression, and resemble stages in neurogenesis. Cancer Cell. 2006;9:157-173. 10.1016/j.ccr.2006.02.019 [DOI] [PubMed] [Google Scholar]

[vdag027-B28] 28. Shirasawa M, Yoshida T, Shiraishi K, et al. Tumor microenvironment-mediated immune profiles and efficacy of anti-PD-L1 antibody plus chemotherapy stratified by DLL3 expression in small-cell lung cancer. Br J Cancer. 2023;129:2003-2013. 10.1038/s41416-023-02427-3 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B29] 29. Eisen A, Fletcher GG, Gandhi S, et al. , Members of the Early Breast Cancer Systemic Therapy Consensus Panel Optimal systemic therapy for early breast cancer in women: a clinical practice guideline. Curr Oncol. 2015;22:S67-81. 10.3747/co.22.2320 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B30] 30. Ronnlund C, Sifakis EG, Schagerholm C, et al. Prognostic impact of HER2 biomarker levels in trastuzumab-treated early HER2-positive breast cancer. Breast Cancer Res. 2024;26:24. 10.1186/s13058-024-01779-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B31] 31. Shien K, Papadimitrakopoulou VA, Wistuba II. Predictive biomarkers of response to PD-1/PD-L1 immune checkpoint inhibitors in non-small cell lung cancer. Lung Cancer. 2016;99:79-87. 10.1016/j.lungcan.2016.06.016 [DOI] [PMC free article] [PubMed] [Google Scholar]

[vdag027-B32] 32. Mountzios G, Sun L, Cho BC, et al. , DeLLphi-304 Investigators Tarlatamab in Small-Cell lung cancer after Platinum-Based chemotherapy. N Engl J Med. 2025;393:349-361. 10.1056/NEJMoa2502099 [DOI] [PubMed] [Google Scholar]

[vdag027-B33] 33. Cao L, Leclercq-Cohen G, Klein C, Sorrentino A, Bacac M. Mechanistic insights into resistance mechanisms to T cell engagers. Front Immunol. 2025;16:1583044. 10.3389/fimmu.2025.1583044 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Delta-like ligand 3 expression in isocitrate dehydrogenase-mutant and isocitrate dehydrogenase-wildtype glioma is largely retained at recurrence, supporting its potential as a therapeutic target

Rosa Luning

Sybren L N Maas

Pim J French

Youri Hoogstrate

Levi van Hijfte

Rania Head

Iris de Heer

Thierry P P van den Bosch

Adelheid Woehrer

Martin J van den Bent

Marjolein Geurts

Abstract

Background

Methods

Results

Conclusion

Key Points.

Methods

Patient Inclusion

Immunohistochemical Staining for DLL3 and IDH1-R132H

Figure 2.

Determining and Categorizing DLL3 Positivity

Figure 1.

Results

Patient Demographics

Table 1.

DLL3 is Expressed in the Majority of Gliomas, and is Highest in IDHmt Tumors

Figure 3.

Figure 4.

DLL3 Expression is Largely Retained in Recurrent Tumors

Discussion

Supplementary Material

Acknowledgements

Contributor Information

Supplementary Material

Author Contributions

Conflict of Interest Statement

Funding

Ethical Approval

Data Availability

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases