**iii.** Focal versus diffuse C4d staining

It is generally accepted that the detection of C4d in renal allograft biopsies using immuno‐ fluorescence staining is more sensitive than immunohistochemical staining [42, 49]. The level of C4d staining appears to have prognostic significance and it is widely accepted that diffuse C4d staining involving >50% of PTC by either technique is considered positive and correlates much more strongly with adverse graft outcome compared to focal C4d staining involving <50% of PTC, but this remains controversial [50]. However, there are other studies suggesting that focal C4d staining is also associated with histological evidence of AMR including glomerulitis and/or peritubular dilatation [51].

### **iv.** Non-PTC C4d staining

Glomerular, arteriolar and/or erythrocyte C4d positivity often occurs in the absence of PTC C4d staining but the clinical significance of these patterns remains unclear. In a retrospective study of 539 indication renal allograft biopsies, *Kikic et al* demonstrated a poor correlation between arteriolar C4d staining and graft survival, whereas linear glomerular C4d staining was strongly associated with graft failure [52]. There has been considerable interest in the detection of erythrocyte C4d deposition (eC4d) by indirect immunofluorescence as a potential surrogate marker of disease activity in patients with systemic lupus erythematosus and may be useful for the monitoring of disease activity and/or response to treatment in these patients [53, 54]. In kidney transplantation, *Haidar et al* showed a greater amount of eC4d in PTC C4d positive samples compared to PTC C4d negative samples. The authors reported that the positive (PPV) and negative predictive value (NPV) of PTC C4d and eC4d for peritubular capillaritis were 28% and 46% for PPV and 93% and 94% for NPV respectively suggesting that monitoring of eC4d may be an useful non-invasive marker of AMR [55].
