Table 3.

Summary of the limitations and remaining questions of immunohistochemical studies on deposition of C5b-9 in human kidneys.

Inherent limitations of current studies in general

They cannot unravel whether deposits of C5b-9 are a cause of kidney disease or a consequence of kidney injury. They cannot distinguish between locally formed C5b-9 bound to cells, C5b-9 bound to or shedded as extracellular vesicles, and sC5b-9 originating in urine or blood. They cannot assess when deposits have arisen, so that, given their slow clearance, deposits may have chronically accumulated. They evaluate staining subjectively and semiquantitatively.

Specific limitations of current included studies

Included patients were generally ill-characterized. Staining techniques were often described very concisely. Different staining techniques and antibodies were seldomly compared. The method of evaluating staining was mostly undefined. The method of evaluating staining was variable. As examples, traces of staining were usually considered negligible but sometimes counted as positive (18) and scoring systems were used incidentally and incomparably (19, 31, 52, 83, 84, 87, 94, 116, 117, 135, 136, 139, 142, 209). Variability of staining among individual patients with the same kidney disease was rarely documented, while it might be large (57). Staining across different kidney diseases was directly compared in only few studies (44, 57, 67, 70, 72, 75, 79, 84, 96, 98, 126, 141–143). Colocalization with immunoglobulins and other complement factors, especially in tubules and vessels, was reported only briefly. Correlations between deposits and histological lesions or clinical characteristics were not studied systematically. Changes in staining were uncommonly tracked through time or treatment.

Remaining questions for future studies

Does staining of C5b-9 differ when directly comparing staining techniques and antibodies? Is staining more common in tissue obtained with autopsy than biopsy or nephrectomy? And can this be explained by a different selection of patients? How do deposits of C5b-9 differ between kidney diseases due to deposition of immune complexes, due to activation of the alternative pathway, and due to other mechanisms? How do deposits vary among patients with the same kidney disease? Are deposits dependent on the age and sex of patients? With which immunoglobulins and other complement factors do deposits colocalize in various localizations and in various kidney diseases? What structures are associated with deposits on immunoelectron microscopy? How fast are deposits cleared in various kidney diseases and across individual patients with the same kidney disease? How do deposits change through time and treatment? And how does the change relate to variable activation of complement pathways, for example in membranous nephropathy? Do deposits consistently predict prognosis and treatment effect? And how does this depend on their localizations and on the underlying kidney disease?