Protein Expression of Giant Cell Lesions of the Maxillofacial Skeleton and Axial/Appendicular Skeleton: A Confirmative Study

Giant cell lesions (GCLs) affect both the maxillofacial (MF) and axial/appendicular (AA) skeleton. They are benign intraosseous tumors that may be locally destructive and frequently recur after treatment.  Controversy exists as to whether AA and MF GCLs are the same disease occurring in different anatomic areas or separate, unrelated entities. Both AA and MF lesions can be classified into aggressive (A) and non-aggressive (NA) lesions according to clinical and radiological features.¹

Currently, no histological or molecular markers exist to differentiate MF and AA lesions or A and NA lesions in either location. A recent gene expression analysis and protein expression study of aggressive and non-aggressive GCLs of the AA and MF skeleton at the Massachusetts General Hospital (MGH) identified potential biomarkers to characterize each of these lesions.²The aim of this study is to reproduce these earlier identified putative immunohistochemical markers of biological behaviour and localization in a new cohort of AA and MF GCLs from another geographical region.

This is a retrospective cohort study of patients with MF and AA GCLs treated at the University of Amsterdam. Subjects with histologically confirmed GCLs and complete records to classify behaviour were included. Subjects with a systemic disease or syndrome known to be associated with GCLs (e.g. Noonan syndrome and Cherubism), inadequate or unconfirmed histologic samples and inadequate follow up (<6 months) were excluded.

A tissue micro array (TMA) was constructed with 3 mm cores of representative tumor from the included subjects (AA: N=35, MF: N=39). Clinical behaviour (A vs. NA) was determined using existing classification systems, the Chuong and Kaban classification for MF lesions and the modified Enneking system for AA lesions.¹ Immunoreactivity to the putative biomarkers Matrix metalloproteinase 9 (MMP9), Cathepsin K (CTSK), and T-cell immune regulator 1 (TCIRG1) was assessed using immunohistochemistry. Aperio®digital pathology analysis was used to quantify optical density of staining. Comparisons were made between location (MF vs. AA) and behaviour (A vs. NA) using a one-way ANOVA and Bonferroni correction given multiple comparisons.

The cohort consisted of 28 aggressive and 11 non-aggressive MF GCLs and 25 aggressive and 10 non-aggressive AA GCLs.  In the MF lesions, mean staining intensity for MMP9, CTSK and TCIRG1, respectively was 175.200 ± 21.140, 186.102 ± 19.698, 205.886 ± 16.085 for multinucleated giant cells and 100.483 ± 0.801, 103.017 ± 3.116, 101.550 ± 1.414 for stromal cells. In the AA lesions, mean staining intensity for MMP9, CTSK and TCIRG1 respectively was 147.163 ± 28.446, 197.455 ± 19.149, 196.523 ± 21.217 for multinucleated giant cells and 100.711 ± 1.937, 105.218 ± 9.517, 102.316 ± 2.032 for stromal cells. The only significant and clinically relevant difference in staining intensity between the different groups was seen with MMP9. MMP9 staining showed a significant difference between MF A and MF NA (p=0.01).

The results of this study indicate that MMP9 might be a putative marker for aggressive behavior in maxillofacial giant cell tumors. The absence of differences in staining intensity between AA and MF GCLs supports the hypothesis that they are a similar disease in different localizations.  

1. Chuong, R., et al. Central giant cell lesions of the jaws: a clinicopathologic study. J Oral Maxillofac Surg, 1986. 44(9): p. 708-13.
2. Peacock ZS, et al. Genetic analysis of giant cell tumors of the axial/appendicular and maxillofacial skeleton. J Am Coll Surg, 2014. 219(3): S70