[1459] HLA-DR Immunohistochemistry Quantitation in Renal Allograft Biopsies: Objective Discrimination of Rejection and Other Pathologic Processes.

Alton B Farris, Jun Kong, Candace Chisolm, Howard M Gebel, Robert A Bray, Cynthia Cohen, Daniel J Brat, Joel H Saltz, Allan D Kirk, Randolph Hennigar. Emory University, Atlanta, GA; Emory University, Atlanta

Background: Histologic renal allograft biopsy (bx) assessment methods (e.g., the Banff criteria), although historically useful, largely rely on semi-quantitative, granular criteria, sometimes making the appreciation of subtle pathologic differences difficult. Human leukocyte antigen (HLA)-DR expression has been shown in renal allografts, particularly bxs with rejection. We investigated the utility of renal allograft bx assessment using HLA-DR immunohistochemistry (IHC) whole slide image (WSI) analysis.
Design: IHC was performed with a commercially-available HLA-DR antibody on bxs containing allograft rejection, polyomavirus nephritis (PVN), a borderline pattern, normal donor tissue, and stable allografts. HLA-DR WSIs obtained using an Olympus Nanozoomer scanner were quantitated with regard to total and cortical renal parenchymal staining using the Aperio Imagescope Positive Pixel Count algorithm, yielding a % of parenchyma with HLA-DR positivity.
Results: Highest HLA-DR positivity was seen in allograft rejection bxs [Table]. Elevated levels were also seen in PVN, followed by borderline cases. A baseline HLA-DR level was present in normal donor bxs, and the lowest levels were in stable allografts. Differences were not significant between rejection and polyomavirus nephritis; however HLA-DR levels in rejection and PVN were significantly higher than stable and normal donor bxs. Rejection but not PVN was significantly higher than borderline. Linear regression showed a direct correlation between total and cortical HLA-DR (R² = 0.98, P < 0.0001).

Type	HLA-DR total parenchymal positivity (%) [Mean ± S.D.]	Comparison (T-test)
Rejection (n = 15)	62.2 ± 16.2	***, *, •
Antibody-mediated rejection (n=3)	68.6 ± 9.3	***, *, •
Cellular rejection (n =12)	60.6 ± 17.4	***, *, •
PVN (n = 3)	60.6 ± 8.1	**, ^
Borderline (n = 2)	36.1 ± 10.4
Normal donor (n = 4)	33.6 ± 18.5
Stable allograft (n =3)	15.0 ± 4.4

S.D.: standard deviation, ***: P < 0.0005 vs. stable, **: P = 0.001 vs. stable, *: P < 0.01 vs. normal donor, ^: P = 0.03 vs. normal donor, •: P < 0.05 vs. borderline


Conclusions: Our data suggest that HLA-DR IHC quantitation has a potential for utility in objectively measuring the immunologic activation of allografts, particularly when combined with emerging WSI segmentation algorithms. Future studies using additional histologic, IHC, and molecular markers may prove useful.
Category: Kidney (does not include tumors)

Wednesday, March 2, 2011 1:00 PM

Poster Session VI # 269, Wednesday Afternoon