The pathology of lupus nephritis

Glomeruli

A normal glomerulus ( Figures 43.1 and 43.5 ) shows delicate patent capillary loops, and a mesangium that is no more than mildly conspicuous. LM evaluation begins with an assessment of glomerular cellularity. Increased glomerular cellularity (historically, but inaccurately called “proliferation”) is primarily due to an influx of exogenous inflammatory cells, although there may be a component of an increase in native cells as well. Hypercellularity may be noted in the mesangial, endocapillary or extracapillary (crescent) zones. The distribution of glomerular lesions is described within individual glomeruli (segmental vs. global) as well as their total population (focal vs. diffuse). Table 43.1 contains a glossary of basic pathological terminology.

The fundamental glomerular lesion of LN is mesangial IC deposition, with all other glomerular lesions superimposed on it. Mesangial IC deposition is identified by IF and EM ( Figures 43.6 and 43.7 ) and is usually associated with mesangial hypercellularity and matrical increase (4 or more cells in a region away from the vascular pole) best recognized on LM ( Figure 43.8 ). The correlation between the extent of mesangial hypercellularity and immunodeposition may be poor. Mesangial processes are typically expressed by hematuria although the mechanism for this remains unclear. Mesangial hypercellularity, by definition, does not involve the capillaries whose lumens remain patent and normocellular.

The now abandoned term “proliferative glomerulonephritis” in the context of LN refers to endocapillary hypercellularity which is recognized by the occlusion of the capillary lumens by increased cells ( Figure 43.9 ). Endocapillary hypercellularity can be variable and its distribution throughout the sample must be quantitated. Copious IC deposition, particularly involving the subendothelial space, is common in this setting ( Figure 43.10 ). Subendothelial deposits are the most accessible to the circulation and are therefore most able to activate complement, generate chemoattractants, and generate the influx of leukocytes. They may also lead to the development of tubulointerstitial tertiary lymphoid organs and intrarenal autoantibody production. These deposits are therefore associated with more damaging forms of the disease. When large, such deposits may be seen on LM as capillary wall thickening, and are referred to as “wire loops” ( Figure 43.11 ). Interestingly, glomeruli with such deposits often show less hypercellularity than ones without. Endolumenal IC plugs (so-called “fibrin thrombi,” although they are neither) are actually projections in continuity with subendothelial deposits ( Figure 43.12 ). True fibrin thrombi may be seen as a result of complement activation or when a thrombotic microangiopathy (TMA) is superimposed. Hematoxylin bodies are the tissue equivalent of the LE body and are comprised of basophilic nuclei coated with antinuclear antibodies. Although highly specific for SLE, they are also highly rare ( Figure 43.13 ).

Hypercellular endocapillary lesions are most associated with “active lesions” and particularly necrosis of the glomerular tuft ( Figures 43.14 and 43.15 ). Necrosis is usually segmental, and is identified by the presence of fibrin, neutrophilic exudation, apoptosis, karyorrhexis, and / or fragmentation of the basement membrane ( Figure 43.16 ). Severe inflammation and especially fibrinoid necrosis results in glomerular capillary rupture with hemorrhage into Bowman’s space and formation of cellular crescents (extracapillary hypercellularity) ( Figure 43.17 ). Cellular crescents, (a variable combination of proliferating epithelial cells and infiltrating mononuclear inflammatory cells), indicate severe glomerular injury and when persistent organize into fibrocellular and then fibrous crescents (often seen with glomerular obsolescence) ( Figure 43.18 ). More than one type of crescent can be seen in the same glomerulus. Despite their association, both segmental necrosis and cellular crescents may be seen alone. Extracapillary hypercellularity seen in association with collapsing lesions of the tuft is likely due to podocytes, and is not designated as a crescent.

Chronic subendothelial deposition may result in a membranoproliferative pattern, reflecting reactive basement membrane/“double contour” formation ( Figure 43.19 ). In addition to neo-membrane covering the subendothelial deposits, EM may reveal cells in that space, in continuity with the mesangium (so-called “mesangial interposition,” although they are likely to be leukocytes) ( Figure 43.20 ).

Scattered subepithelial IC deposition is common in LN and may be associated with reactive changes of the lamina densa. Such “membranous” changes are best identified on IF and EM, and when more extensive on Jones stain ( Figure 43.21 ). Deposits in this “protected” location, (walled off from the circulation), may activate complement, but are not associated with inflammatory cell influx (hypercellularity). When this pattern dominates, the appearance may suggest idiopathic membranous GN. Often however, subepithelial deposition is combined with endocapillary hypercellularity due to deposits at other sites.

The active changes noted above (immune deposition, hypercellularity, necrosis, etc.) may progress to chronic scarring. Serial biopsies suggest that segmental necrosis heals as segmental scars (sclerosis) ( Figure 43.22 ). More severe glomerular injury or crescent formation may result in global sclerosis. The percentage of sclerotic glomeruli (global and segmental) is an important metric of chronicity. Similar chronic changes may also result from unrelated processes such as hypertension and aging. Only glomerulosclerosis that can be attributed to post-inflammatory scarring is classified as LN involvement. Unfortunately, that distinction may be difficult to make on morphologic grounds.

EM usually shows frequent tubuloreticular structures (or inclusions), most commonly in glomerular endothelial cells ( Figure 43.23 ). These 24 nm interanastamosing structures located within the endoplasmic reticulum are associated with elevated levels of circulating interferon and can also be seen in viral infections. Their association with LN underscores the central role of interferon in this disease. Their presence in a case of otherwise primary membranous GN may portend the subsequent development of SLE.

Tubulointerstitium

Tubules and interstitium share a tight anatomical association and their lesions (either acute or chronic) are often considered together. Interstitial edema (appearing as clear space) and inflammation, predominantly comprised of mononuclear cells (lymphocytes, plasma cells, and macrophages) are acute changes seen to varying degrees in LN, and in association with tubulitis and other forms of acute tubular injury (degenerative and regenerative) including red cell and/or “pus” casts. While interstitial infiltrates are rarely seen alone, the extent of inflammation usually parallels the severity of glomerular disease. The degree of inflammation has been correlated with both reduction in GFR and serologic activity. Studies immunophenotyping the infiltrates have not yielded consistent results. Staining for CD45+ cells may improve assessment for intermediate grades of tubulointerstitial inflammation. Macrophage infiltration may correlate with both current and future renal function. Inflammation leads to fibrosis, and macrophages play a role in both. This complex subject is discussed more fully elsewhere in this text. A recent biopsy study using single-cell RNA sequencing reviews these issues and reveals the complexity of immune populations in LN kidneys. Work in recent years has led to the realization that tubulointerstitial inflammation in SLE may be an entirely independent process from the systemic autoimmunity that results in glomerular damage. Investigation in this important area has been hampered by the fact that two important features of human tubulointerstitial disease, the formation of tertiary lymphoid structures, and tubular basement membrane IC deposition, are not seen in most murine models.

IF and EM often reveal deposits (most commonly IgG) along tubular basement membranes (always suggestive of LN or another autoimmune condition) and even within the interstitium ( Figure 43.24 ). While tubulointerstitial deposits can rarely be seen in the absence of glomerular deposition, typically their extent parallels glomerular hypercellularity. Nevertheless, deposits along the TBM might arise from different pathogenic mechanisms than those resulting in glomerular deposition. This is underscored by data showing that TBM deposits are associated with a poor renal outcome especially in patients with non-proliferative glomerular lesions. Interstitial inflammation in lupus nephritis does not appear to be related to tubulointerstitial immune deposits but may be caused by autoantibodies to interstitial antigens, particularly vimentin. Finally, nuclei may show staining for IgG on IF (“tissue ANA,” perhaps an artifact of tissue sectioning) which when strong may obscure other deposits ( Figure 43.24 ).

Evaluation of interstitial fibrosis (and tubular atrophy) is critical to define the extent of chronic injury. It is best assessed on MT, where expansion of the normally inconspicuous interstitium appears as varying intensities of blue, reflecting the density of collagen deposition ( Figure 43.25 ). As in other renal diseases, the degree of tubulointerstitial scarring is the histologic parameter that best correlates with both current and future renal function. Morphometric evaluation has been shown to improve the assessment, but has not been adopted in clinical practice. As tubules progress from acute injury to chronic atrophy they undergo changes that result in diminished cytoplasm and diameter, thickening of basement membranes (best seen on PAS stain) and cast formation. Like interstitial fibrosis, these processes are thought to be irreversible.

Vessels

Nonspecific chronic lesions (secondary to hypertension, aging, etc.) are the most common changes seen. Lesions specific to SLE receive insufficient attention, in part because they (like tubulointerstitial changes) are absent from LN classifications and their terminology has not been standardized. Their clinical significance ranges from trivial to profound. A schema is presented in Table 43.2 . The most common lesion, uncomplicated vascular immune deposits, is recognized by irregular deposition of immunoreactants (most commonly IgG although other immunoglobulins and complement components may be present) in vessel walls (predominantly small arteries and arterioles) as seen by IF ( Figure 43.26 ). The presence of IgM and/or complement alone may reflect nonspecific injury and is not diagnostic of a lupus-related process. Vessel walls may display IC on LM but the lumen is not compromised by them. EM localizes the deposits to the intimal basement membrane and medial matrix. The degree of deposition roughly correlates with glomerular hypercellularity and tubulointerstitial deposition. These deposits are usually clinically silent and may not carry negative prognostic implications.

Lupus vasculopathy (noninflammatory necrotizing vasculopathy) in contrast, is far less common, and is associated with a bad prognosis. Seen most commonly in the setting of severe hypercellular LN, vessels (usually arterioles) show lumenal narrowing by IC deposition accompanied by endothelial and medial injury. Fibrin and necrosis is typically present but inflammation (by definition) is absent. ( Figure 43.27 )

Literal vasculitis, that is, fibrinoid necrosis of the vessel wall with associated inflammation is a very rare finding in LN. Its histologic appearance is indistinguishable from a systemic vasculitis of the ANCA type, and may represent the simultaneous occurrence of an unrelated systemic or “renal limited” vasculitis.

Thrombotic microangiopathy (TMA) in the lupus setting may be seen in association with any of its related clinical entities (such as malignant hypertension, systemic sclerosis, HUS/TTP), or with the antiphospholipid antibody nephropathy/lupus anticoagulant syndrome. However, TMA can be present without any systemic syndrome (“renal limited”). Small arteries and arterioles may show thrombosis, fibrinoid necrosis, and mucoid intimal hyperplasia (“onion skinning”) ( Figure 43.28 ). As opposed to lupus vasculopathy, lesions show no significant IgG immune deposition (IgM and C3 may be present, secondary to vascular injury). Glomeruli show typical TMA changes such as thrombosis, mesangiolysis, and double contour formation that may be superimposed on other glomerular changes of LN. Studies have shown a strong relationship between the presence of glomerular micro-thrombi and intensity of C4d staining; suggesting a role for the classical complement pathway in LN associated TMA. The presence of TMA in LN is associated with more severe clinical and histological activity, and significantly inferior long-term renal outcome.

Anti-phospholipid antibody syndrome may be primary or lupus (or “lupus-like”) associated, and is a clinically important cause of systemic and renal thrombosis, resulting in systemic and renal infarctions and subsequent organization and recanalization of the vessels. The acute form of the SLE associated variant is more likely to also involve smaller vessels and display more typical findings of TMA ( Figure 43.29 ). TMA lesions are significantly more common in lupus patients with anti-phospholipid antibodies. However, many patients may be antibody positive without demonstrating the syndrome. The chronic form, which may be clinically smoldering and unrecognized shows only chronic injury in the form of zonal cortical scarring and tubular thyroidization. This insensitive finding, when present, greatly increases the likelihood of an underlying phospholipid antibody syndrome. Recognition is important as anticoagulation may slow the progression of scarring.