IGF2BP3 (IMP3) expression in angiosarcoma, epithelioid hemangioendothelioma, and benign vascular lesions

We selected all the examined cases which had been fixed in 4% neutral buffered formalin for 12–72 h at room temperature and embedded in paraffin for this study. All selected cases were diagnosed between 2010 and 2019. Under this condition, 30 cases of angiosarcoma and five cases of epithelioid hemangioendothelioma (EHE) were identified in the diagnostic database of Kyoto University Hospital. For comparison, benign vascular lesions (10 hemangiomas, 14 pyogenic granulomas, and 12 granulation tissues with proliferating capillaries) were selected from the most recent diagnostic sign-out cases under the same fixation condition. The clinical information is summarized in Table 1. All samples Patients signed the “Kyoto University Hospital Informed Consent Form for the Non-therapeutic Use of Histopathological Materials”, and signed forms were uploaded into each electronic health record.

Three-micrometer tissue sections were deparaffinized with xylene, rehydrated, and pretreated with 0.3% hydrogen peroxide for 5 min. For IGF2BP3 staining, after steam heating the sections for 20 min in pH 9.0 EDTA buffer, anti-IMP3 antibody (Ab; 1:75, mouse monoclonal clone 69.1, DAKO Cytomation, Glostrup, Denmark) was added and the sections were incubated for 15 min at room temperature following blocking of background staining using Protein Block (X0909; DAKO Cytomation). Staining was performed using a BOND III automated stainer (Leica Biosystems, Richmond, IL, USA) according to the manufacturer’s instructions. For PD-L1 staining, after steam heating the sections for 60 min in pH 8.5 citrate buffer, anti-PD-L1 Ab (1:200, rabbit monoclonal clone E1L3N, Cell Signaling Technology, Beverly, MA, USA) was added and the sections were incubated for 16 min at room temperature following blocking of background staining using Protein Block. Staining was performed using a Benchmark Ultra automated stainer (Ventana Medical Systems, AZ, USA) according to the manufacturer’s instructions. Stained sections were imaged under a BX45 microscope (Olympus, Tokyo, Japan) equipped with a DP26 digital camera (Olympus).

The degree of IGF2BP3 staining was scored, according to the proportion of the staining and regardless of the intensity of the staining as follows: negative (0% positive among endothelial/neoplastic cells), equivocal (1–25% positive), and positive (> 26%). PD-L1 staining was defined as positive if > 5% of membranous expression was observed at the tumor site, as reported previously [13].

Differences between groups were examined for statistical significance using Student’s t-test or Chi-squared test (Microsoft Excel, Redmond, WA, USA). A P value less than 0.05 indicated statistical significance.

IGF2BP3 staining was positive in eight of 30 angiosarcoma cases (26.6%; Table 1a and Fig. 1). Most cases (n = 17, 56.6%) were scored as equivocal (Table 1a and Fig. 1). Completely negative staining was seen in five cases (16.6%; Table 1a and Fig. 1). There was no difference in clinical parameters (age, gender, location, morphological classification, presence of metastatic foci, and local recurrence) between IGF2BP3-positive and –equivocal /-negative cases (Table 2).

Next, we assessed the association between IGF2BP3 expression and PD-L1 expression, which is a positive prognostic marker for angiosarcoma [13]. Among IGF2BP3-positive cases of angiosarcoma, two cases (25.0%) were PD-L1-negative and six (75.0%) were PD-L1-positive. Among negative or equivocal cases, one was a consultation from another hospital and no extra glass slide was available for PD-L1 analysis. Therefore, we assessed 21 angiosarcoma cases scored as negative or equivocal; four cases (19.0%) were PD-L1-negative and 17 (81.0%) were PD-L1-positive. Again, there was no statistical association between IGF2BP3 expression and PD-L1 expression among the angiosarcoma cases (Table 2).

Two of five (40%) cases of EHE were scored as positive for IGF2BP3 (Table 1b and Fig. 2). The remaining cases were equivocal (n = 2) or negative (n = 1) (Table 1b and Fig. 2).

No benign vascular lesion (n = 36) was scored as positive (Tables 1c – e and Fig. 3). Ten hemangioma cases and 14 granulation tissues showed completely negative staining (Tables 1c and e and Fig. 3). In contrast, six of 14 cases of pyogenic granuloma (42.9%) were scored as equivocal (Table 1d and Fig. 3).