Background: Question emerged during COVID-19 pandemic to minimize the risk of nosocomial viral exposure. Pharmacokinetic modelling evidence supports extended-interval dosing of ICI in advanced NSCLC, showing equivalent pharmacological performance than standard schedule. However, there is a clear lack of clinical data. Methods: We performed an observational, retrospective study in a French university hospital. The extended-schedule of ICI administration began during the first pandemic period (from march to may 2020). We report here the clinical characteristics and early efficacy and safety signals, after a minimal follow-up of 6 months. Data (tumor response, adverse event) were collected based on medical records. Results: 25 patients received the extending-dose schedule (13 pembrolizumab 400 mg Q6W, 12 nivolumab 480 mg Q4W) during the inclusion period. Most of the malignancies were stage IV (21/25) adenocarcinoma (20/25) with 13/25 tumors showing a PD-L1 TPS>50%. Most of the patients were in 2nd or 3rd line of treatment (15/25). 3 patients started ICI with double dose-schedule, whereas 22 transitioned from a previous standard-dose regimen. Altogether, 13/25 (52%) patients presented or remained on partial response with extended-interval dosing schedule during follow-up, with 11/25 (44%) continuing this regimen on september 1st. The adverse events reported in the patient still on ICI were grade 1 diarrhea or arthralgia. The median duration of prior exposure to ICI for those patients was 278 days. 14 patients stopped the extended-interval dosing schedule including 7 for disease progression and 6 for immune-related adverse event. The main observed adverse events were asthenia (n = 4), diarrhea (n = 1) and arthralgia. The median duration of prior exposure to ICI for those patients was 178 days. 3 patients died during the follow-up period. No SARS-CoV2 infection was observed. Conclusions: This work based on real-life experience shows that extending the dose and interval of ICI treatment in advanced NSCLC is feasible. Early efficacy and safety signals appear encouraging. The adverse events reported were expected side-effects of immunotherapy and no grade 4–5 toxicity was observed. Legal entity responsible for the study: CHU Rennes. Funding: Has not received any funding. Disclosure: All authors have declared no conflicts of interest.

Background: The antiphospholipid syndrome (APS) is characterized by the persistent presence of anti-beta 2-glycoprotein I (beta 2-GPI) autoantibodies. beta 2-GPI can exist in two conformations. In plasma it is a circular protein, whereas it adopts a fish-hook conformation after binding to phospholipids. Only the latter conformation is recognized by patient antibodies. beta 2-GPI has been shown to interact with Streptococcus pyogenes. Objective: To evaluate the potential of S. pyogenes-derived proteins to induce anti-beta 2-GPI autoantibodies. Methods and results: Four S. pyogenes surface proteins (M1 protein, protein H, streptococcal collagen-like protein A [SclA], and streptococcal collagen-like protein B [SclB]) were found to interact with beta 2-GPI. Only binding to protein H induces a conformational change in beta 2-GPI, thereby exposing a cryptic epitope for APS-related autoantibodies. Mice were injected with the four proteins. Only mice injected with protein H developed antibodies against the patient antibody-related epitope in domain I of beta 2-GPI. Patients with pharyngotonsillitis caused by S. pyogenes who developed anti-protein H antibodies also generated anti-beta 2-GPI antibodies. Conclusions: Our study has demonstrated that a bacterial protein can induce a conformational change in beta 2-GPI, resulting in the formation of anti beta 2-GPI autoantibodies. This constitutes a novel mechanism for the formation of anti-beta 2-GPI autoantibodies.

SummaryThe antiphospholipid syndrome (APS) is a non-inflammatory autoimmune disease characterized by the presence of antiphospholipid antibodies (aPL) in the plasma of patients with vascular thrombosis, recurrent complications of pregnancy, or both (1, 2). The presence of aPL in plasma of patients can be detected with either a prolongation of phospholipid dependent coagulation tests (lupus anticoagulant, LAC), or with solid phase immune assays against the protein β2-glycoprotein I (β2-GPI) or the phospholipid cardiolipin (anti-β2-GPI antibody ELISA and anti-cardiolipin antibody ELISA, respectively) (3). For a long time there was a lot of confusion on who had the syndrome and who not. To solve this dispute, an international consensus meeting was organized in Sapporo in 1999 to formulate classification criteria for patients with the antiphospholipid syndrome (4). These criteria have been updated in 2004 at another international consensus meeting in Sydney (5). The classification criteria were defined for scientific purposes and were aimed to be used as inclusion criteria in patient related studies. They were specifically not defined for diagnostic purposes. However, current practice is that these criteria are used as a diagnostic tool. This is very unfortunate because the specificity of the different aPL assays to detect the clinical manifestations that characterize APS are disputable. One of the aims of defining the criteria was to initiate studies to determine the value of the different anti-phospholipid antibody assays to serve as biomarker for the risk of thrombosis and pregnancy morbidity. The recent progress made on this important topic will be discussed.

Summary The antiphospholipid syndrome (APS) is a non-inflammatory autoimmune disease characterized by the presence of antiphospholipid antibodies (aPL) in the plasma of patients with vascular thrombosis, recurrent complications of pregnancy, or both (1, 2). The presence of aPL in plasma of patients can be detected with either a prolongation of phospholipid dependent coagulation tests (lupus anticoagulant, LAC), or with solid phase immune assays against the protein β2-glycoprotein I (β2-GPI) or the phospholipid cardiolipin (anti-β2-GPI antibody ELISA and anti-cardiolipin antibody ELISA, respectively) (3). For a long time there was a lot of confusion on who had the syndrome and who not. To solve this dispute, an international consensus meeting was organized in Sapporo in 1999 to formulate classification criteria for patients with the antiphospholipid syndrome (4). These criteria have been updated in 2004 at another international consensus meeting in Sydney (5). The classification criteria were defined for scientific purposes and were aimed to be used as inclusion criteria in patient related studies. They were specifically not defined for diagnostic purposes. However, current practice is that these criteria are used as a diagnostic tool. This is very unfortunate because the specificity of the different aPL assays to detect the clinical manifestations that characterize APS are disputable. One of the aims of defining the criteria was to initiate studies to determine the value of the different anti-phospholipid antibody assays to serve as biomarker for the risk of thrombosis and pregnancy morbidity. The recent progress made on this important topic will be discussed.

The antiphospholipid syndrome (APS) is a non-inflammatory autoimmune disease characterized by the presence of antiphospholipid antibodies (aPL) in the plasma of patients with vascular thrombosis, recurrent complications of pregnancy, or both (1, 2). The presence of aPL in plasma of patients can be detected with either a prolongation of phospholipid dependent coagulation tests (lupus anticoagulant, LAC), or with solid phase immune assays against the protein beta2-glycoprotein I (beta2-GPI) or the phospholipid cardiolipin (anti-beta2-GPI antibody ELISA and anti-cardiolipin antibody ELISA, respectively) (3). For a long time there was a lot of confusion on who had the syndrome and who not. To solve this dispute, an international consensus meeting was organized in Sapporo in 1999 to formulate classification criteria for patients with the antiphospholipid syndrome (4). These criteria have been updated in 2004 at another international consensus meeting in Sydney (5). The classification criteria were defined for scientific purposes and were aimed to be used as inclusion criteria in patient related studies. They were specifically not defined for diagnostic purposes. However, current practice is that these criteria are used as a diagnostic tool. This is very unfortunate because the specificity of the different aPL assays to detect the clinical manifestations that characterize APS are disputable. One of the aims of defining the criteria was to initiate studies to determine the value of the different anti-phospholipid antibody assays to serve as biomarker for the risk of thrombosis and pregnancy morbidity. The recent progress made on this important topic will be discussed.

Summary The antiphospholipid syndrome (APS) is a non-inflammatory autoimmune disease characterized by the presence of antiphospholipid antibodies (aPL) in the plasma of patients with vascular thrombosis, recurrent complications of pregnancy, or both (1, 2). The presence of aPL in plasma of patients can be detected with either a prolongation of phospholipid dependent coagulation tests (lupus anticoagulant, LAC), or with solid phase immune assays against the protein β2-glycoprotein I (β2-GPI) or the phospholipid cardiolipin (anti-β2-GPI antibody ELISA and anti-cardiolipin antibody ELISA, respectively) (3). For a long time there was a lot of confusion on who had the syndrome and who not. To solve this dispute, an international consensus meeting was organized in Sapporo in 1999 to formulate classification criteria for patients with the antiphospholipid syndrome (4). These criteria have been updated in 2004 at another international consensus meeting in Sydney (5). The classification criteria were defined for scientific purposes and were aimed to be used as inclusion criteria in patient related studies. They were specifically not defined for diagnostic purposes. However, current practice is that these criteria are used as a diagnostic tool. This is very unfortunate because the specificity of the different aPL assays to detect the clinical manifestations that characterize APS are disputable. One of the aims of defining the criteria was to initiate studies to determine the value of the different anti-phospholipid antibody assays to serve as biomarker for the risk of thrombosis and pregnancy morbidity. The recent progress made on this important topic will be discussed.

The antiphospholipid syndrome (APS) is a non-inflammatory autoimmune disease characterized by the presence of antiphospholipid antibodies (aPL) in the plasma of patients with vascular thrombosis, recurrent complications of pregnancy, or both (1, 2). The presence of aPL in plasma of patients can be detected with either a prolongation of phospholipid dependent coagulation tests (lupus anticoagulant, LAC), or with solid phase immune assays against the protein beta(2)-glycoprotein I (beta(2)-GPI) or the phospholipid cardiolipin (anti-beta(2)-GPI antibody ELISA and anti-cardiolipin antibody ELISA, respectively) (3). For a long time there was a lot of confusion on who had the syndrome and who not. To solve this dispute, an international consensus meeting was organized in Sapporo in 1999 to formulate classification criteria for patients with the antiphospholipid syndrome (4). These criteria have been updated in 2004 at another international consensus meeting in Sydney (5) The classification criteria were defined for scientific purposes and were aimed to be used as inclusion criteria in patient related studies They were specifically not defined for diagnostic purposes However, current practice is that these criteria are used as a diagnostic tool. This is very unfortunate because the specificity of the different aPL assays to detect the clinical manifestations that characterize APS are disputable. One of the aims of defining the criteria was to initiate studies to determine the value of the different anti-phospholipid antibody assays to serve as biomarker for the risk of thrombosis and pregnancy morbidity. The recent progress made on this important topic will be discussed.

β2-Glycoprotein I (β2-GPI) is a highly abundant protein present in blood, but without a known physiological function. In 1990, β2-GPI became a protein of great interest as it was shown by different groups that the so-called antiphospholipid antibodies present in antiphospholipid syndrome (APS) are in fact directed against this plasma protein [1, 2]. It has been acknowledged that β2-GPI plays an important role in the thrombotic and pregnancy complications observed in APS. Thus, the correct biochemical characterization of β2-GPI is of pivotal importance [3, 4]. In 1979, Polz and Kostner [5] showed the distribution of β2-GPI over different human lipoproteins. Based on these observations, Lee et al. [6] designated β2-GPI as apolipoprotein H (apoH) [6]. Since then, the names β2-GPI and apoH have both been used for the same protein, and the official designation for the β2-GPI gene has become APOH. We were interested in whether the localization of β2-GPI on lipoproteins was influenced by the presence of antiphospholipid antibodies, and so we decided to reinvestigate the distribution of β2-GPI over the different lipoproteins and plasma fractions. We observed that after this original observation no other publications have confirmed the observed association of β2-GPI with lipoproteins. Blood was drawn from five healthy volunteers in a fasting state and 3 h after consuming a classic English breakfast (>1000 kcal), to repeat the original experiments by Polz and Kostner. Moreover, plasmas from two septic patients, two APS patients with antibodies against β2-GPI and pooled plasma from more than 200 healthy volunteers were also investigated. The Institutional Review Boards of the University Medical Centre Utrecht and Academic Medical Centre Amsterdam approved this study and informed consent was obtained from all patients or their caretakers. Citrated blood samples were centrifuged (15 min, 1200 ×g) and plasma was collected. Three millilitres of plasma were brought to a density (D) of 1.250 with KBr and layered with three KBr densities; D = 1.225, D = 1.100, and D = 1.006. A single-step ultracentrifugation (XL-90 Beckman, Beckman Coulter, Fullerton, CA, USA) was performed (96.000 ×g for 19 h at 10 °C), and fractions of 200 μL were collected with a fraction collector. Collected fractions were diluted at least 1000-fold in Tris-buffered saline (TBS; 50 mmol L−1 Tris, 150 mmol L−1 NaCl, 0.1% Tween-20, pH 7.4). Determination of β2-GPI was carried out by a homemade sandwich enzyme-linked immunosorbent assay (ELISA), using a mouse monoclonal antibody 3B7 as the capturing antibody and a rabbit polyclonal α-β2-GPI as a secondary antibody. Serial dilutions of normal pooled plasma (2.000× to 256.000× in TBS) were used as a standard curve. Very low-density lipoprotein (VLDL), low-density lipoprotein (LDL) and high-density lipoprotein (HDL) samples were determined by PAP 250 cholesterol enzymatic methods. Cholesterol reagent (Biomerieux, Le Fontanille, France) was added to 10 μL of the sample and measured on a spectrophotometer. As can be observed in Fig. 1A, no β2-GPI could be detected in the different lipoprotein fractions from a healthy volunteer. These results were confirmed with an additional four healthy volunteers, two septic patients and normal pooled plasma (Fig. 1B). The consumption of a classical English breakfast did not change the distribution of β2-GPI (data not shown). (A) Ultracentrifugation profile of subject 1. Cholesterol (open circles) and β2-glycoprotein I (β2-GPI; closed circles) are depicted. (B) β2-Glycoprotein I distribution over the different human plasma lipoproteins. Lipoproteins were separated after a one-step ultracentrifugation. Subjects 1–5 were normolipemic volunteers. Normal pooled plasma is from more than 200 volunteers. Blood from septic patients were drawn at the time they had sepsis. APS: patients with antibodies against β2-GPI. All fractions were also measured with surface plasmon resonance using a Biacore 2000 (Life Sciences, GE Healthcare, Uppsala, Sweden). To determine the binding of β2-GPI to the lipoproteins, anti-β2-GPI antibodies were coupled to a CM5-chip and the fractions were applied to the chip. In the fractions containing the different lipoproteins, no β2-GPI could be detected (data not shown). Antibodies directed against apoA1 and apoB were also coupled to a CM5-chip. The different lipoproteins were then directly captured from the ultracentrifugation fractions, followed by an injection of anti-β2-GPI antibodies to detect potentially formed complexes between β2-GPI and VLDL, LDL or HDL. No complexes were detected (data not shown). Subsequently, reconstituted HDL (CSL-111, Parkville, Victoria, Australia) was bound to an anti-apoA1 coupled chip and purified β2-GPI, from human plasma as described by Oosting et al. [7], was injected over the chip. No complex formation between purified β2-GPI and purified HDL could be observed (data not shown). To exclude the possibility that the separation technique for the lipoproteins could influence the outcome, lipoproteins from plasmas of three volunteers were separated using gel filtration on a Superose 6 HR 10/30 (Pharmacia Biotech, Uppsala, Sweden) column with inline fluorescence and ultraviolet detection. Fractions were diluted in BSA/TBS (20 mmol L−1 Tris, 150 mmol L−1 NaCl, 3% BSA) and 0.1% Tween-20 and the presence of β2-GPI was detected with an ELISA. Again, β2-GPI was only found in the plasma fractions, not in the fractions that contain the different lipoproteins (data not shown). Since Polz and Kostner published the presence of β2-GPI in human lipoproteins almost 30 years ago, no other studies have confirmed the distribution of β2-GPI over the different lipoprotein. Following this observation, Lee et al. [6] designated the name apoH for β2-GPI and from then on apoH and β2-GPI were used as synonyms for the same protein. Here we show with state-of-the-art techniques that β2-GPI is not present in appreciable quantities in the lipoprotein fractions, neither in fasting healthy persons nor postprandially. Also, when plasmas of two APS patients positive for anti-β2-GPI antibodies were subjected to lipoprotein separation, the presence of anti-β2-GPI antibodies did not result in redistribution of β2-GPI from the plasma fraction over the lipoprotein fractions (Fig. 1B). From these observations and the direct binding experiments using surface plasmon resonance, it is clear that there are no interactions between β2-GPI and LDL or HDL. It cannot be excluded that there is a possible weak interaction with VLDL. We have not studied the effects of oxidation of LDL [8] on the distribution of β2-GPI over the lipoproteins, because we do not know a patient cohort with proven oxidation of lipoproteins, and it is questionable if in vitro oxidation of LDL mimics a physiological condition. We conclude that apoH is not expected to be an integral part of lipoproteins and for this reason the name apolipoprotein H for β2-GPI is clearly a misnomer. We therefore suggest the use of the name β2-GPI only. The authors state that they have no conflict of interest.