Expresia imunohistochimică a SATB2 în tumorile osoase, cu accent pe tumora de celule gigantice şi malignităţi

Introduction

In bone and soft tissue pathology practice, diagnosing malignant tumors like osteosarcoma is a difficult task, especially in biopsy samples⁽¹⁾. But an accurate decision is crucial, as chemotherapy and surgery approaches are fundamentally distinct for different tumor types^(2-4).

Immunohistochemistry has been assessed as an additional tool to the morphological investigation, and according to diagnostic immunohistochemistry, a series of osteoblast-specific markers have been described in the literature, such as osteocalcin, osteonectin, bone morphogenetic protein (BMP), COL-I-C peptide, bone sialoprotein, bone glycoprotein, 75 decorin, osteopontin, proteoglycans I and II, type I collagen and RANK, but only osteocalcin and osteonectin have been used in diagnostic immune histologic studies, without a special resonance in clinical practice⁽⁵⁾.

But in the last decade a new series of antibodies has been being proposed as osteoblastic differentiation markers, such as RANK^(6,7), CD56⁽⁸⁾ and SATB2⁽⁹⁾. 

CD56 is a homophilic binding glycoprotein with a role in cell-cell adhesion. As an IHC marker, its main uses are in hematopoietic disorders, as a marker of NK cells, for detecting residual myeloma or differentiate plasma cells in myeloma⁽¹⁰⁾ from reactive plasmacytosis. Hughes, based on previous evidence of CD56 expression in osteoblast lineage, presented a study which concluded that CD56 may be a biologically important molecule in osteoblast function, and its expression is consistently strong in normal and neoplastic osteoblasts.

In 2012, Kallen detected p63 expression in osteoblastic tumors⁽¹¹⁾. Previously, p63 had been known as a member of p53 gene family⁽¹²⁾, but without being a tumor suppressor gene. It appeared to regulate the development of epithelial organs and may be a molecular switch for the initiation of “epithelial stratification program”, regulating human keratinocyte proliferation. Isoform deltaNp63 inhibits the transcription activation of p53 gene⁽¹³⁾.

Then, in 2013, Hornick suggested SATB2 as a potential marker for osteoblastic differentiation in bone and soft tissue tumors. It encodes at least six different proteins with different biologic functions⁽⁹⁾.

SATB2 – known as special AT-rich sequence-binding protein 2 – is an AT-rich DNA-binding protein that binds to nuclear matrix attachment regions involved in transcriptional regulation and chromatin remodeling⁽¹⁴⁾.

Initial studies showed that SATB2 protein is expressed in colon and rectum tissue and tumors⁽¹⁵⁾, brain, branchial arches and osteoblast-lineage cells^(16,17), kidney, bladder⁽¹⁸⁾, some lymphoid cells and ovarian tumors⁽¹⁹⁾. Despite been identified in multiple tissues, it has not been thoroughly investigated as a potential neoplastic marker in bone and soft tissue tumors.

Our purpose was to study the potential diagnostic applications of the three quoted antibodies as immunohistochemical markers in bone and soft tissue pathology, as well as evaluating positive expression scores and artefactual staining patterns.

Materials and method

We searched the Foişor Hospital database, for all the bone and soft tissue tumors from 2015 to 2018 and encountered 39 cases (Table 1 and Table 2). The cases were extracted according to the year of diagnosis, histological diagnosis, localization, and included data on gender, age of patients, site, morphological features, and the presence or absence of the primary tumor in patient’s history.
 

The tumors were subtyped according to diagnosis in 23 primary bone tumors (benign and malignant) as follows (Figure 1): two fibrosarcomas, two osteosarcomas, seven chondromas, ten giant cell tumors (GCT; two of them malignant), one chondromyxoid fibroma, one brown tumor; and 16 metastatic tumors: two lung carcinomas, five clear cell renal carcinoma, one prostate carcinoma, one colon adenocarcinoma, one hepatocarcinoma, two breast carcinomas, four unknown primary adenocarcinomas.
 

The hematoxylin-eosin slides for all the cases were reexamined, by three independent pathologists. We used Olympus CX41 microscopes equipped with digital Colorview III cameras for the optical examination, and for the morphological and comparative study, the Virtual Slide Microscope VS120 and the viewing software.
 

We constructed two tissue microarray paraffin blocks (TMA) using the technique^(20,21), as follows (Figure 2). All donor blocks for the study cases have been selected and the region of interest defined on the paraffin wax block. After the initial review and selection, the donor blocks were arranged and kept in the order that is represented in the TMA. The array construction was made manual and was accompanied by a computer file, containing the tissue block identification numbers. The TMA was divided into four rows designated by letters and six columns ordered by numbers. Two-mm diameter donor tissue core needles were sampled. Tissue cores were punched from the predefined region, and then transferred to a recipient paraffin wax block, into a readymade hole. The obtained paraffin blocks were incubated and processed as usual (Figure 3).
 

For the immunostaining, we followed the indirect protocol for paraffin embedded tissue sections. The tissue section slides were deparaffinized, rehydrated and incubated with a 3% hydrogen peroxide solution, and heat induced antigen retrieval (HIER). For the staining we used Avidin-Biotin Complex (ABC) method and counterstaining with Mayer hematoxylin.

For the IHC expression we used the immunoreactive score (IRS)⁽²²⁾. The immunoreactive score (IRS) uses a range of 0-12 as a product of multiplication between positive cells proportion score (0-4) and staining intensity score (0-3) (Table 3). The immunoreactive score was broadly used for the expression of a wide spectrum of IHC markers^(23-25).
 

Results

We evaluated the expression of p63, CD 56 and SATB2 on the 39 tissue samples.

The intensity of the reaction was variable, positive expression intensity ranging from weak signal (1) to strongly positive (3).

Using IHC, we found that SATB2 positive expression was highly restricted to nuclei. SATB2 was not detected in any adult normal non-osteoblastic tissue. Instead, we had experimented cases with nucleolar artefactual staining. In bone forming neoplastic tissue, SATB2 was detectable by IHC in 70% of 23 primary bone neoplasms examined.
 

The two osteosarcomas were negative for p63 and one presented moderate expression for CD56, while for SATB2 we noted moderate expression in ~55% (30-80%) of the tumor cells, in both cases. Remarkably, the fibrosarcoma tumors showed moderate expression in 40% of the tumor cells for SATB2 and negativity for p63 and CD56. Synovial giant cell tumor didn’t show any reactivity to the antibodies, while the bone counterpart both malignant and benign ones showed the highest percentage of stained cells, but the intensity of the signal was weak in the benign tumors and strong in the malignant ones. Chondroid tumors didn’t show any positive staining for SATB2 or p63 and CD56 (Table 4 and Table 5).
 

Of the 16 metastatic tumors (two lung carcinomas, five clear cell renal carcinoma, one prostate carcinoma, one colon adenocarcinoma, one hepatocarcinoma, two breast carcinomas, four unknown primary adenocarcinomas in our study), only the colorectal carcinoma showed significant labeling of the nuclear region, thus confirming the literature data⁽²⁶⁾.

Discussion and conclusions

The aim of the study was to compare the potency of markers to fulfill an objective classification of osteoblastic differentiation in mesenchymal tumors. Of these, SATB2 showed the highest numbers of cells marked and the highest intensity of the signal, while the CD56 showed the lowest numbers of marked cells. P63 lack of reactivity with high grade osteosarcoma makes it useful only for bone giant cell tumors and, when needed, as an indirect control for SATB2 positive tumors.

While SATB2 can be considered neither a sensitive marker nor a specific one, it can become a valuable tool in diagnosing mesenchymal tumors in association with p63 and CD56. As from Table 4 and Figure 5, using p63 is highly valuable in IHC staining diagnosis of giant cell tumors, while SATB2 and CD56 are more useful in the diagnosis of aggressive osteosarcomas and malignancy in giant cell tumors.
 

One bizarre SATB2 staining pattern was observed in several cases showing nucleolar/dot like fashion staining. The aspect was mentioned before in literature, but it was not investigated to our knowledge. The presence of a homogeneous nuclear staining pattern associated with a nucleolar staining pattern in the same tumor suggests that this should not be considered a mere artifact and deserves further investigation.  

Conflict of interests: The authors declare no conflict of interests.