BANFF SCHEMA FOR GRADING PANCREAS ALLOGRAFT REJECTION 2007

BANFF SCHEMA FOR GRADING PANCREAS ALLOGRAFT REJECTION 2007 -2009 Meeting Summary

• Pancreas Tx are only performed in selected centers and pancreas biopsies are few in comparison to those of other organs. – Progress has been slow – Immunosuppression has improved – Indications have been re- and re-evaluated – Ongoing need for beta cell replacement

– Overall the number of whole pancreas tx has stabilized • New pancreas transplant programs continue to be established (explosive growth in some regions)

Banff 2007 • Pre-meeting discussions were concluded in the formal and “working lunch” sessions. • Consensus agreement – Definitions of the histological lesions – “cell mediated” grading schema – General histopathology guidelines including adequacy, recommendations for reporting, etc

• Completion of consensus manuscript through group e-mail discussions.

Banff 2009 Pancreas Session Chairs: Edward Kraus Brian Nankivell Additional working breakfast the following morning

Banff 2009 Pancreas Session 1. Gene expression based categorization of pancreas transplant biopsies Fu Luan - First attempt to correlate histological rejection with gene expression by using techniques and gene selection that has been used in the kidney - Demonstrated differential expression of a variety of genes that correlated with histological rejection grades and graft outcome.

Banff 2009 Pancreas Session AMR in pancreas allografts 3. Ingeborg Bajema 4. Jose Torrealba 5. Erika Bevilaqua Rangel

Banff 2009 Pancreas Session AMR in pancreas allografts Ingeborg Bajema: Retrospective experience on C 4 d IHC. A combination of DSA+ and C 4 d+ correlate with worse graft outcome. Jose Torrealba: Summarized their large experience with AMR in pancreas tx /SPK. Pattern of C 4 d staining. Erika Bevilaqua Rangel: AMR in a cohort of patients transplanted in San Paulo (Brazil). Largest experience with IF C 4 d. DSA available on few pts.

Banff 2009 Pancreas Session 6. Lois Arend: Controversial lesions in pancreas biopsy grading. Drug toxicity, acinar inflammation. Discussed experience with AMR cases and confirmed findings in previous presentations.

Proposal for Diagnosis of Acute Antibody Mediated Rejection • Consensus discussion – Importance of defining AMR in the pancreas. – Recommendation to evaluate C 4 d staining in all biopsies – Both IHC and IF work well: comparison between methods to be performed in the future. – Interacinar staining pattern strongly correlates with AMR (in contrast to other staining patterns).

Proposal for Diagnosis of Acute Antibody Mediated Rejection – Focal staining for C 4 d appears to be important: Agreement to recommend guidelines for reporting C 4 d as in the kidney schema.

2007 Proposed Banff Schema

2009 Proposal for Diagnosis of Acute Antibody Mediated Rejection Remove requirement for “graft dysfunction” for the pathology diagnosis of acute AMR.

Working Proposal for diagnosis of AMR Acute antibody mediated rejection 1. C 4 d+ 2. Morphologic evidence of chronic tissue injury (margination of inflammatory cells in interacinar capillaries (capillaritis), acinar cell damage, arteritis, etc). 3. DSA/HLA confirmatory testing Two of 3 criteria necessary for diagnosis

Rule out AMR - C 4 d+ with no tissue injury ? Recommendation to do CD 68 stain pending further discussions

• Chronic Active antibody mediated rejection category remains. Final wording will be discussed.

Acute AMR C 4 d+

2. Erica Bracamonte Preliminary results: Reproducibility of Banff schema for grading of acute pancreas rejection.

Reproducibility Study Materials • 12 Digital whole slides (Aperio system) • Only one case had C 4 d stain.

Participants • • • Lois Arend Ingeborg Bajema Jan Bruijn Helen Cathro Billie Fyfe-Kirschner Lillian Gaber Julio Goldberg Danielle Hollanda Ramesh Nair Jean Olson John Papadimitriou • • • Lorraine Racusen Karine Renaudin-Autain Finn P Reinholt Monica Patricia Revelo Phillip Ruiz Edward B Stelow Jose Torrealba Megan Troxell Matthew Turner Ludek Voska Matthew Yeh

Participants • 22, from 19 institutions • Level of experience (self reported) – Novice: <2 yrs experience (n=5) – Competent: 2 -5 yrs experience (n=12) – Expert: >5 yrs experience (n=4)

Preliminary Results Reproducibility study (E. Bracamonte) • Data still being collected • Rejection vs no rejection Kappa score 0. 55

Further studies of Banff Schema Reproducibility (E. Bracamonte) First Phase: Better Define Morphologic Features • Recognition of histological features should be tested. • Wording in schema should be simplified. Lesions grouped in categories for easier understanding.

Specific plan: • Images for each of the morphologic features to be posted at the University of Pittsburg website. • The selection of images will be done by “consensus” agreement with the goal of making more uniform the diagnostic interpretation of the cases.

• The images and web site will be also used for discussion of Banff scoring categories similar to kidney. • The material will be used for “training” of interested pathologists and clinicians (overlapping with the current didactic plan of the ASTS).

Second Phase – Retest Scoring of Cases • By consensus , another set of test cases should be passed around for restudy after the addition discussion of the morphologic features. • All cases will include a C 4 d Stain and more cases of humoral rejection will be included. • The biopsy worksheet will be revised and simplified (? online responses)

• “Gold standard” diagnoses to be defined. Probably by using a small panel of two or three pathologists who come to a consensus. • All agreement statistics will be re-run based on the second set of testing, and compared to the prior set already performed.

Main goal of the study • Assess reproducibility of the schema. • Ultimately, improve the Banff schema by decreasing complexity and increasing applicability for the average practicing pathologist.

• Many thanks to the Banff Conference Organizing committee that provided excellent support for the Pancreas sessions.