The assessment of serum-mediated phagocytosis of necrotic material by polymorphonuclear leukocytes to diagnose and predict the clinical features of systemic lupus erythematosus: an observational longitudinal study

Systemic lupus erythematosus (SLE) is an autoimmune syndrome of unknown etiology, with a complex interplay between genetic and environmental factors [1, 2] and between the innate and the adaptive immune systems [3]. The disease commonly affects women of childbearing age and is characterized by chronic inflammation in several different organ systems with potentially life-threatening outcomes. Many clinical manifestations of SLE are caused by immune complex deposition in target organ and subsequent inflammation caused by complement activation and infiltration of immune cells. Among the different cells involved in the pathogenesis of SLE, the crucial role played by polymorphonuclear leukocytes (PMNs) has recently been emphasized [4‐12]. The immune complexes mostly contain nuclear material able to stimulate immune cells through Toll-like receptor 7 (TLR7) and TLR9, and are formed upon binding of autoantibodies to different antigens, often remnants of cells dying due to impaired clearance of apoptotic cells [8, 13].

Although SLE classification criteria have been formulated and periodically updated [14‐17], these criteria are constructed to classify patient materials for research purposes and are not diagnostic criteria meant to be used in clinical practice [18]. Furthermore, there is a great need for novel diagnostic and prognostic tools to complement the ones already in use, including assessment of antibodies against double-stranded DNA (dsDNA) and histone proteins, as well as consumption of components of the classical pathway of the complement system. The first laboratory test proposed as a diagnostic tool for SLE was the lupus erythematosus (LE) cell, described in 1948 as a specific finding in bone marrow leukocytes in patients affected by SLE [19]. Further studies have shown that LE cells consist of mature PMNs in which the nucleus has been dislocated to the periphery of the cell after engulfment of antibody- and complement-opsonized nuclear material [20, 21]. The presence of LE cells was included in the list of the classification criteria for SLE [14, 17] and related to more severe clinical manifestations [22, 23]. The LE cell was rarely observed in other rheumatic [24] and non-rheumatic disorders [25], but published data on the diagnostic accuracy of the LE cell test in unselected patients are still missing [22, 26, 27]. Previous studies showed that the phenomenon is dependent on anti-dsDNA antibodies [21] and anti-histone antibodies, in particular anti-histone H1 antibodies [28]. Because of its complexity, the assessment of LE cells by light microscopy was abandoned as a routine test in favor of other diagnostic tools, including anti-dsDNA antibody analysis.

More recently, a flow cytometry–based assay—the phagocytosis of necrotic cells (PNC) assay—was developed as an in vitro assessment and quantification of LE cells in patients with suspected SLE [29], but its use in clinical practice has never been validated. In a previous study [30], our group demonstrated that the outcome of the PNC assay is often positive in patients with SLE with increased disease activity. The phagocytosis seems to be associated with oxidative burst activity that is mediated by Fc γ-receptor IIA (FcγRIIA), FcγRIIIB, and complement receptor type 1 in combination. The phagocytosis is more efficient in the presence of high levels of different anti-histone antibodies and when the classical pathway of the complement system is functional and active [30].

In this study, we show that the PNC assay, by detecting serum-mediated phagocytosis of necrotic cell material (NCM), may be used as a tool in the process of diagnosing SLE. Furthermore, we found that the outcome of the PNC assay was related to clinical manifestations and, in combination with other established biomarkers, with disease activity.

Positivity in PNC assays is specific but not exclusively found in patients with SLE. To assess its diagnostic accuracy for SLE diagnosis, the PNC assay was performed in the 529 MRG patients. The outcomes of PNC assays in MRG patients are summarized in Table 1. The PNC assay was positive in 35 MRG patients (6.6 %) with recent onset of any rheumatic symptoms. Among these patients, the most common clinical diagnosis was SLE (13 cases), resulting in a sensitivity of the test at 0.20, specificity of the test at 0.95, and PPV of the test for SLE diagnosis at 0.37. Other, less common diagnoses in PNC-positive MRG patients were arthralgia (two cases, 5.9 % of MRG patients affected by arthralgia and 5.7 % of PNC-positive patients), osteoarthritis (six cases, 10.2 % of MRG patients affected by osteoarthritis and 17.1 % of PNC-positive patients), inflammatory joint disease (three cases, 2.7 % of MRG patients affected by inflammatory joint disease and 8.6 % of PNC-positive patients), non-rheumatic disease (six cases, 5 % of MRG patients affected by non-rheumatic disease and 17.1 % of PNC-positive patients), connective tissue disease (three cases, 5.6 % of MRG patients affected by connective tissue disease and 8.6 % of PNC-positive patients), systemic inflammatory disease (one case, 4.2 % of MRG patients affected by systemic inflammatory disease and 2.9 % of PNC-positive patients), and soft tissue rheumatism (one case, 4 % of MRG patients affected by soft tissue rheumatism and 2.9 % of PNC-positive patients). In the subgroup of 44 anti-dsDNA–positive MRG patients, 15 (34.1 %) had a positive PNC test result. Among them, 10 had SLE, 2 had a connective tissue disease, 2 had an inflammatory joint disease, and 1 had a non-rheumatic disease. The positivity of CLIFT alone resulted in a sensitivity at 0.32, specificity at 0.95, and PPV at 0.48. The combined positivity of CLIFT and the PNC assay increased the specificity to 0.99 and the PPV to 0.67 in patients with SLE (Table 3).

Overall, the positive outcomes of PNC assays were seen mostly in patients with SLE. The percentages of PNC-positive patients in the different subgroups and in healthy individuals are reported in Fig. 3.

Area under the ROC curve analysis plotting the diagnostic accuracy for SLE of PNC assay and its combination with CLIFT, in comparison with ANA and CLIFT alone, is shown in Fig. 4 and Table 4. ANA and CLIFT had the best AUC (0.7 and 0.64, respectively), suggesting a sharper and more immediate diagnostic role than PNC assay for SLE, in patients with recent onset of rheumatic symptoms.

Our data show that the PNC assay might partially improve the diagnostic accuracy of anti-dsDNA antibodies for SLE in patients with a recent onset of rheumatic symptoms. In patients with established SLE, its positive outcome was strongly correlated with ongoing hematological features such as leukopenia and thrombocytopenia. Furthermore, by itself and/or in combination with other laboratory analyses, the PNC assay was able to predict upcoming disease manifestations (within 2 months) in patients with SLE, including mucocutaneous features, alopecia, and glomerulonephritis. Altogether, the results of our investigation support the use of the PNC assay in clinical praxis to sharpen the diagnostics in undiagnosed patients and to improve the evaluation of prognosis in patients with SLE. Early identification of patients with negative prognostic factors, because of prediction of worsened outcome, should lead to earlier adjustment of their management and to overall decreased disease activity.

We started the investigation by assessing reference values of a PNC assay in apparently healthy individuals. The obtained reference values were evaluated in a Scandinavian group of patients with recent onset of any rheumatic symptoms. By comparing patients with SLE with patients affected by other disorders, we could demonstrate that SLE is by far the most common diagnosis related to a positive outcome of PNC assay, especially in anti-dsDNA antibody–positive patients (Table 1). One reason for a positive PNC assay result in patients with no SLE diagnosis and absence of detectable anti-dsDNA antibodies may be the presence in serum of autoantibodies against other parts of the NCM, including histones or nucleosomes.

In a realistic diagnostic setting, the physician is initially unaware of the correct diagnosis in patients with recent onset of rheumatic symptoms and the use of reliable laboratory tools plays an important diagnostic role. It is important to emphasize that the PPV for SLE after a cross-sectional analysis of the PNC assay is low. Other clinical diagnoses can be found in patients with a positive test, more frequently in anti-dsDNA–negative patients. We have previously reported that the PPV for SLE after a cross-sectional analysis of CLIFT alone in a similar population was 0.46 [32]. Nonetheless, the combined positivity of CLIFT and PNC assay in the present study resulted in a PPV of 0.67, suggesting an additive diagnostic role of the PNC assay in patients affected by SLE. Furthermore, although our study had a median follow-up of 4.7 years, it is possible that some of the individuals with positive PNC assay will still develop SLE in the future with or without any other overlapping rheumatic disease.

The second main aim of the present study was to investigate any potential predictive value of the PNC assay with regard to SLE-related clinical manifestations. In particular, we wanted to study whether the positive outcome of the test (alone or in combination with concomitant presence of anti-histone and/or anti-dsDNA antibodies and/or hypocomplementemia) coincides with or predicts the occurrence of any relevant clinical manifestations or changes in disease activity.

On one hand, the occurrence of glomerulonephritis was much more frequent in patients with SLE with persistent positive outcome of PNC assay, compared with those where the outcome was variable over time or always negative (Table 5). On the other hand, the occurrence of arthritis was much more frequent in patients with persistently negative PNC assay outcomes (Table 5). Although patients with PNC positivity had an increased SLEDAI-2K score, we could not show any significant association over time of the PNC assay with SLE disease activity. However, in combination with low complement levels, PNC positivity could predict upcoming flare, in particular active glomerulonephritis. The additive effect of low complement levels may be counterintuitive, considering the essential role of complements in the PNC assay [30]. However, low complement levels are often sufficient to activate the classical pathway of complements [33] and may also be a sign of ongoing immune complex–mediated disease, such as the ones detected in the PNC assay. Interestingly, positivity in the PNC assay also displayed an inverse association with development of arthritis, a clinical manifestation with much higher prevalence in patients with SLE with consistently negative PNC assay outcomes. It strongly suggests and supports the hypothesis that neutrophil abnormalities are related to a distinct lupus phenotype associated with severe manifestations, including hematological features and glomerulonephritis, similar to what has been described previously [30].

The PNC assay may help the physician to distinguish between different phenotypes in patients with SLE. The test displayed a highly significant correlation with concomitant hematological manifestations such as leukopenia and thrombocytopenia. It may suggest a potential role of the PNC assay as a marker for these relevant features. We do not have sufficient data at the moment to establish whether the hematological manifestations in our SLE cohort were related to active SLE or other causes, such as concomitant treatment with cytotoxic drugs, infections, or other. If our results are confirmed, the PNC assay could be a valid test that would provide the clinician with important information on whether the hematological manifestations are SLE-related. SLE-related hematological manifestations may require higher doses of immunosuppressive treatment, whereas a dose reduction or interruption of the treatment is often recommended whenever leukopenia and thrombocytopenia are supposed to be side effects of the treatment itself. To our knowledge, no other validated tests used in the management of patients with SLE are good enough to obtain the same information.

Limitations of the present investigation are mostly due to the cross-sectional analysis of the PNC assay in MRG patients and healthy individuals. Moreover, the laboratory analysis in MRG patients was limited to ANA and anti-dsDNA antibodies, besides the PNC assay. We have no data at the moment to discern anything about the variation over time of the outcome of the PNC test in healthy individuals or in patients without SLE with recent onset of rheumatic symptoms. This is an important aspect to analyze in future studies because the longitudinal analysis in the SLE cohort showed that the outcome of the PNC assay may influence the prevalence of crucial clinical features. Data that verify how anti-histone antibodies and the classical pathway of the complement system may influence the outcome of the PNC assay in these patients should also be included.

Recent discoveries, including the formation of neutrophil extracellular traps (NETs), a neutrophil cell death process in which several key autoantigens (including dsDNA and histones) are extruded from the neutrophil [34], as well as the identification of spontaneously NET-forming low-density granulocytes, an immature inflammatory neutrophil subset highly enriched in inflammatory conditions such as SLE [9, 10], have highlighted neutrophils as important immune cells in SLE pathogenesis [9, 11, 35]. Although NETs are cleared silently by macrophages in healthy individuals [36], patients with SLE have impaired clearance of NETs due to the presence of autoantibodies as well as decreased DNase I activity [6, 12, 37], enabling recognition of NETs by plasmacytoid dendritic cells and induction of type I interferons, key cytokines in SLE pathogenesis [38]. However, the clinical utility of assessing neutrophil function in the diagnosis and prognosis of patients with SLE has not been addressed. In the present study, we evaluated the diagnostic and prognostic value of the PNC assay, a functional test used to assess the ability of PMNs to engulf autoantibody- and complement-opsonized cell debris.

Assessment of serum-mediated phagocytosis of NCM by PMNs as determined by the PNC assay may be considered a useful supplementary laboratory tool in the management of patients affected by SLE. Either alone or in combination with other laboratory tests, the PNC assay may provide the clinician with additional information in the diagnostic process and in assessing current and, more importantly, risk for future disease activity, as well as occurrence of clinical manifestations and phenotypes, such as glomerulonephritis, hematological, and mucocutaneous manifestations. Long-term longitudinal assessment of PNC assay in larger cohorts of patients with SLE is needed to validate the test as a relevant tool in the management of the disease.