Grading Systems for CTA Rejection Proposals and Prospects

Grading Systems for CTA Rejection Proposals and Prospects David E. Kleiner, M. D. , Ph. D. National Cancer Institute

Setting the Stage • Grading system philosophy and choices of grading systems • Review of published grading system criteria for CTA acute rejection • Comparison of existing systems looking for common themes • Where do we go from here?

What Are the Goals in Creating a Grading System? • Stratification for treatment/protocol entry – Minimum hepatic fibrosis for HCV therapy • Prognostication – Cancer Staging and Grading • Structured pathology data collection – NASH-CRN Feature Scoring System

Some Definitions • Stage – Stratification of the level of progress of a disease to its final end-point (Clinical Tool) • Grade – Stratification of the severity of a disease or disease feature at a particular point in time (Clinical Tool) • Scoring – the assignment of quantitative or semiquantitative values to individual disease features (Research Tool) It is usually possible for therapeutic intervention to improve the Grade of a disease but it is usually difficult or impossible to improve the Stage of a disease

Organ Failure Cirrhosis Loss of Function (Stage) Rate (Grade) Onset of Disease Time The apparent rate may or may not be a good predictor of progression Death

Types of Grading Systems • Tiered Systems – Each grade is differentiated by the addition of a new lesion – Banff Renal Acute Cellular Rejection Grade I vs II • Progressive Severity Systems – Gradual worsening of one or more features with (arbitrary) thresholds – Banff Renal Acute Cellular Rejection Borderline vs Ia vs Ib • Composite Score Systems – Grade is a summation of scores of individual features – Hepatitis inflammation grading

Tiered Grading Systems Grade 1 Grade 2 Grade 3 Inflammation Epithelial Injury Vascular Injury + + (any) + • Advantages: Easy to use, Probably better reproducibility • Disadvantages: Doesn’t account well for variation in severity of features, especially when features seem inappropriately mild or negative

Progressive Severity System Grade 1 Grade 2 Grade 3 Inflammation Epithelial Injury Vascular Injury + ++ +++ • Advantages: Better system when features generally vary in parallel. Natural relationship to scoring individual features • Disadvantages: Need to define thresholds for each feature -> decreases reproducibility. Difficulties assigning grade if features are out of sync with one another.

Composite Score System Score Range Inflammation Epithelial Injury Vascular Injury 0 -3 0 -3 Sum the individual scores: 0: Grade 0; 1 -3: Grade 1; 4 -6: Grade 2; 7 -9: Grade 3 • Advantages: Most sophisticated system. Accounts well for individual variation between features. Relates well to scoring systems. Better for clinical trials • Disadvantages: Threshold problems. Implied weighting of features, therefore requires advanced knowledge of relative importance of features

Published Systems for Grading CTA Rejection • The Pathology of Full Thickness Cadaver Skin Transplant for Large Abdominal Defects – Bejarano et al. , Am. J. Surg. Pathol. 28: 670 -675; 2004 • Steroid- and ATG-Resistant Rejection After Double Forearm Transplantation Responds to Campath-1 H – Schneeberger et al. , Am. J. Transplant 4: 1372 -1374; 2004 • Pathological Score for the Evaluation of Allograft Rejection in Human Hand (Composite Tissue) Allotransplantation – Kanitakis et al. , Eur J. Dermatol. 15: 235 -8; 2005 • Composite Tissue Allotransplantation: Classification of Clinical Acute Skin Rejection – Cendales et al. , Transplantation 81: 418 -22; 2006

• Abdominal wall transplantation • 9 patients (5 adults, 4 children), 10 transplants • 22 specimens (17 punch biopsies, 3 graft excisions, 2 post-mortem) • Blind categorization (3 pathologists) of multiple histologic features related to inflammation, epidermal changes and stromal changes • Features were analyzed with respect to an overall clinico-pathologic determination of the presence of rejection

Histologic Associations with Rejection Non-Rejection (mean/mode) N=9 Rejection (mean/mode) N = 13 p 0. 1 (0. 2%) 12. 7 (35%) 0. 017 Intensity of Perivascular Infiltrates 0. 0 2. 0 0. 00007 Location of perivascular infiltrates None Upper dermis 0. 00011 Small Lymphocytes 0. 0 1. 7 0. 00007 Large Lymphocytes 0. 0 1. 0 0. 014 Eosinophils 0. 0 0. 6 0. 04 Endothelial Plumping 0. 4 1. 1 0. 011 Spongiosis 0. 0 1. 1 0. 03 No. infiltrated vessels

No Rejection Grade 0 No perivascular infitrates. Indeterminate for Rejection Grade 1 < 10% of vessels with perivascular infiltrates, no spongiosis, eosinophils, large lymphocytes, epidermal infiltrates or stromal inflammation. Mild Rejection Grade 2 11 -50% of vessels with perivascular infiltrates, +/mild spongiosis, eosinophils. No epidermal infiltrates, stromal infiltrates or large lymphocytes. Moderate Rejection Grade 3 >50% of vessels with perivascular infiltrates +/epidermal/stromal inflammation, eosinophils, endothelial plumping, large lymphocytes. No dyskeratosis. Severe Rejection Grade 4 >50% of vessels with perivascular infiltrates with dyskeratosis, heavier lymphocytic infiltrates in the epidermis, moderate to severe spongiosis, stomal inflammation extending to basal layer

• Case report of 41 M with double forearm transplantation • Grading system presented based on clinical experience with prior rejection episodes and on published literature of CTA rejection • Reported 2 Grade I rejections that were steroidresponsive and one Grade IVa rejection that was steroid-resistant and ATG-resistant but responded to Campath-1 H • Follow-up biopsies confirmed resolution of infiltrates

Minimal Rejection Grade I Perivascular lymphocytic and eosinophilic infiltrates Mild Rejection Grade II Additional interface reaction in epidermis and/or adnexal structures Moderate Rejection Grade III Diffuse lymphocytic infiltration of epidermis and dermis Marked Rejection Grade IVa Necrosis of single keratinocytes and focal dermal-epidermal separation Marked Rejection Grade IVb Necrosis and loss of the epidermis

• • Hand/Forearm Transplants 6 patients (all M), 33. 8 yrs (22 -48) 89 skin biopsies (punch or scalpel) Biopsies reviewed for a variety of epidermal, adnexal, inflammatory and vascular changes • Immunoperoxidase staining for lymphocyte phenotype, HLA, mast cells • Grading based on biopsy review, grouping similar biopsies together into 5 grades

No Rejection Grade 0 Normal or very low density infiltrates around a few vessels Mild Rejection Grade I Mild dermal lymphocytic infiltrate forming perivascular cuffs in upper and occasionally mid-dermis. The epidermis is unaffected. Moderate Rejection Grade II Moderately dense dermal infiltrate, perivascular and interstitially between collagen bundles, +/- mild degree of epidermal exocytosis, spongiosis. No necrotic keratinocytes Severe Rejection Grade III Dense lymphocytic infiltrate around blood vessels in upper, mid, lower dermis and eccrine sweat glands. Band-like infiltrate in papillary dermis with lichenoid epidermal changes, spongiosis, exocytosis, variable epidermal apoptosis Very Severe Rejection Grade IV Pandermal inflammation with large cuffs of lymphocytes, variable epidermal thickness with areas of necrosis, thinning, hyperplasia, basal cell vacuolization, epidermal apoptosis

• Forearm transplantation • 2 patients, 11 biopsies • Biopsies were ranked by overall severity of changes and grouped into categories to set definitions • Interobserver agreement tested on grading 18 additional biopsies from abdominal wall transplants

Nonspecific Changes Grade 0 No or only mild lymphocytic infiltration, no involvement of adnexa or epidermis Mild Rejection Grade 1 Superficial perivascular inflammation in upper dermis, no involvement of epidermis Moderate Rejection Features of Grade 1 plus inflammatory infiltration of adnexal structures Grade 2 Severe Rejection Grade 3 Bandlike superficial dermal infiltrate with more continuous involvement of the epidermis, mid and deep dermal perivascular infiltrate Necrotizing Rejection Grade 4 Grade 3 with necrosis of epidermis or other tissues

Common Themes • Perivascular lymphocytic infiltrates become progressively more intense and involve more vessels with increasing grades • Inflammation extends to involve dermal stroma, epidermis (including DEJ), adnexa at moderate to marked grades • Epidermal apoptosis/necrosis only at higher grades • All tiered grading systems with implied worsening inflammation with increasing grade

Grades from None/Non-specific to Mild/Moderate Nominal Grade Proposal Perivascular Dermal Inf Epidermal/ Adnexal Apoptosis/ Dysker. 0 1 - - 3 -/+ - - - 4 -/+ - - 1 + - - - 2 + - - 3 + - - 4 + - - 1 ++ - - - 2 + - +/- - - 3 + + +/- - - Necrosis 2 1 2 - - + + Proposal 1 st author: 1. Bejarano, 2. Schneeberger, 3. Kanitakis, 4. Cendales 4

Grades from Moderate to Very Severe Nominal Grade Proposal Perivascular Dermal Inf Epidermal/ Adnexal Apoptosis/ Dysker. 3 1 +++ +/- - 2 + + + - - 3 + +/- - 4 + + + 1 +++ + 2 + + - 3 + + + 4 + + + 2 + + + 4 5 Necrosis - + + + Proposal 1 st author: 1. Bejarano, 2. Schneeberger, 3. Kanitakis, 4. Cendales

Approximate Grade Equivalences First Author Bejarano Schneeberger Kanitakis Cendales 0 1 2 3 4 a 4 b 0 0 1 1 2 2 3 3 4 4

Conclusions and Challenges • There is already substantial agreement on basic grade stratification for acute rejection • Histologic features of rejection (especially at mild grades) are also seen in a large variety of non-rejection pathologies • Published experience using pathology grading in prospective studies is very limited • Future refinements may require prospective systematic evaluation of biopsy features, similar to other developed rejection classification systems • We should consider defining scoring thresholds for scaling individual features (inflammation, epidermal injury etc. )