Biomarkers of venous thromboembolism recurrence after discontinuation of low molecular weight heparin treatment for cancer-associated thrombosis (HISPALIS-study).

Citation: Otero, R.; Solier-López, A.; Sánchez-López, V.; Oto, J.; Arellano, E.; Marín, S.; Jara-Palomares, L.; Elías, T.; Asencio, M.I.; Blasco-Esquivias, I.; et al. Biomarkers of Venous Thromboembolism Recurrence after Discontinuation of Low Molecular Weight Heparin Treatment for Cancer-Associated Thrombosis (HISPALIS-Study). Cancers 2022,14, 2771. https://doi.org/10.3390/ cancers14112771 Academic Editor: David Wong Received: 10 May 2022 Accepted: 30 May 2022 Published: 2 June 2022 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). cancers Article Biomarkers of Venous Thromboembolism Recurrence after Discontinuation of Low Molecular Weight Heparin Treatment for Cancer-Associated Thrombosis (HISPALIS-Study) Remedios Otero 1,2,3,*, Aurora Solier-López 1, Verónica Sánchez-López 2,3, Julia Oto 4, Elena Arellano 2,3, Samira Marín1, Luis Jara-Palomares 1,2,3 , Teresa Elías 1,2 , María Isabel Asencio 1,2, Isabel Blasco-Esquivias 5, María Rodríguez de la Borbolla 6, JoséMaría Sánchez-Díaz 1, Macarena Real-Domínguez 7, Emilio García-Cabrera 8, Francisco Javier Rodríguez-Martorell 9and Pilar Medina 4 1Medical Surgical Unit of Respiratory Diseases, Hospital Universitario Virgen del Rocío, 41013 Seville, Spain; [email protected]g (A.S.-L.); [email protected] (S.M.); [email protected] (L.J.-P.); [email protected] (T.E.); [email protected] (M.I.A.); [email protected] (J.M.S.-D.) 2Instituto de Biomedicina de Sevilla (IBiS), Consejo Superior de Investigaciones Cientificas (CSIC), Universidad de Sevilla, 41013 Sevilla, Spain; [email protected] (V.S.-L.); [email protected] (E.A.) 3Centro de Investigación Biomédica en Red de Enfermedades Respiratorias (CIBERES), Instituto de Salud Carlos III, 28029 Madrid, Spain 4 Haemostasis, Thrombosis, Arteriosclerosis and Vascular Biology Research Group, Medical Research Institute Hospital La Fe (IIS La Fe), La Fe University and Polytechnic Hospital, 46026 Valencia, Spain; [email protected] (J.O.); [email protected] (P.M.) 5Internal Medicine, Emergency Service, Virgen Macarena Hospital, 41009 Seville, Spain; [email protected] 6Medical Oncology, Virgen de Valme Hospital, 41014 Seville, Spain; [email protected] 7Departamento de Medicina Preventiva y Salud Pública, Universidad de Málaga, 29071 Málaga, Spain; macarenar[email protected] 8Departamento de Medicina Preventiva y Salud Pública, Universidad de Sevilla, 41009 Seville, Spain; [email protected] 9Department of Hematology, Hospital Universitario Virgen del Rocio, 41013 Seville, Spain; fjavie.rodriguez.martor[email protected] *Correspondence: [email protected]; Tel.: +34-955013159 Simple Summary: The duration of anticoagulant treatment for venous thromboembolism in cancer patients is not well defined. The decision to suspend or re-introduce anticoagulation is based on an individual assessment between thrombotic and hemorrhagic risk. In this study, three biomarkers have shown the ability to help the clinician in decision making. Although these results should be evaluated in further studies, they open up a possibility of change in the management of cancer associated thrombosis patients. Abstract: The most appropriate duration of anticoagulant treatment for cancer-associated venous thromboembolism (CAT) remains unclear. We have conducted a prospective multicenter study in CAT patients with more than 6 months of anticoagulant treatment to predict the risk of venous thromboembolism (VTE) recurrence after anticoagulation discontinuation. Blood samples were obtained when patients stopped the anticoagulation, at 21 days and at 90 days. In each sample we assessed different coagulation-related biomarkers: D-dimer (DD), high-sensitivity C-reactive protein (hs-CRP), P-selectin (PS), phospholipids, soluble tissue factor, factor VIII and the thrombin generation test. It was evaluated 325 CAT patients and 166 patients were included in the study, mean age 64 ±17 years. VTE recurrence until 6 months after stopping anticoagulation treatment was 9.87% [95% confidence interval (CI): 6–15]. The biomarkers sub-distribution hazard ratios were 6.32 for ratio DD basal/DD 21 days > 2 (95% CI: 1.82–21.90), 6.36 for hs-CRP > 4.5 (95% CI: 1.73–23.40) and 5.58 for PS > 40 (95% CI: 1.46–21.30) after 21 days of stopping anticoagulation. This is the first study that has identified the DD ratio, hs-CRP and PS as potential biomarkers of VTE recurrence in cancer Cancers 2022,14, 2771. https://doi.org/10.3390/cancers14112771 https://www.mdpi.com/journal/cancers Cancers 2022,14, 2771 2 of 15 patients after the discontinuation of anticoagulation treatment. A risk-adapted strategy may allow the identification of the optimal time to withdraw the anticoagulation in each CAT patient. Keywords: biomarkers; venous thromboembolism; recurrence; cancer; cancer-associated thrombosis 1. Introduction Cancer occurrence in conjunction with its treatment and the intrinsic characteristics of the patients increases the risk of developing a thromboembolic disease. Indeed, the overall risk of venous thromboembolism (VTE) is increased 7-fold in oncologic patients compared with the general population [ 1 ]. Patients with cancer-associated thrombosis (CAT) have a 3-fold shorter one-year survival rate compared to patients with cancer without VTE [ 2 ]. The risk factors for VTE in cancer patients can be divided into individual patient risk factors (age, race, comorbidities, immobility, previous history of VTE), risk factors related to cancer (histology, stage, localization of neoplasm, time since diagnosis) and risk factors related to treatment (surgery and hospitalization, chemotherapy, angiogenesis inhibitors, central venous catheters) [ 3 ]. The most thrombogenic tumor types are pancreas, head/neck, central nervous system, upper gastrointestinal, endocrine, lung and colorectal [ 4 ], with the greatest VTE incidence occurring within the first year of follow-up in metastatic cancer [5]. The hazard of death is increased 3-fold among CAT patients with recurrent VTE [ 6 ], revealing that the prevention of VTE recurrence may be of paramount importance for long-term survival in these patients. Conversely, cancer patients also have a high risk of anticoagulant-associated major bleeding [ 6 ]. Consequently, the optimal dose and duration of anticoagulation in CAT patients remains unclear. Several clinical assessment scores have been proposed for thrombotic risk stratification in cancer patients [ 7 – 11 ]. These scores comprehend tumor-related factors, cancer subtype and baseline laboratory values. The most well-known score is the Khorana score and The Vienna Cancer and Thrombosis (CATS) Study proposes an expanded risk Khorana model incorporating two additional biomarkers, soluble P-selectin (sP-selectin) and D-dimer (DD), considerably improving the VTE risk prediction [ 12 ]. As for all clinical assessment scores, several limitations have been raised [ 13 – 15 ] and current tools for predicting and monitoring the risk of VTE are inadequate. Nonetheless, these risk assessment scores are ineffectual in cancer patients with a previous VTE event, where clinical decisions regarding anticoagulant therapy duration ought to be based on the estimation of the risk of VTE recurrences of each individual patient. Guidelines establish a duration of anticoagulant treatment in CAT patients up to 6 months. The anticoagulation beyond 6months often depends on a case-by-case basis. The incidence of VTE recurrence in CAT patients varies among studies, typically ranging from 10% to 20% [ 16 ]. The only available tool to specifically predict VTE recurrence in CAT patients is the Ottawa score [ 17 ]. This score, in its original or the modified version, classifies CAT patients as high or low risk to developing a VTE recurrence during the first 6 months of anticoagulation treatment, and comprehends the following variables: sex, tumor type, previous history of VTE and tumor stage. However, no tools are available to monitor the risk of VTE recurrences in CAT patients after anticoagulation withdrawal. Thus, the aim of our study was to identify a profile of clinical variables and coagulationrelated parameters capable of predicting the risk of VTE recurrences in CAT patients after anticoagulant withdrawal. 2. Materials and Methods 2.1. Design Longitudinal prospective multicenter study to detect VTE recurrences after withdrawal of anticoagulation in CAT patients. Approval was obtained from the Andalusian Bioethics Cancers 2022,14, 2771 3 of 15 Committee and all participants provided their written informed consent (clinicaltrials.gov identifier: NCT03134820). 2.2. Patients Patients were recruited in three university hospitals of Spanish Health Service at Seville (Hospital Virgen Macarena, Hospital de Valme, and Hospital Virgen del Rocío) between December 2013 and September 2016. All patients provided written informed consent before enrolment. Eligible patients were considered any active cancer patients (excluding nonmelanoma skin cancer) with objectively diagnosed pulmonary embolism (PE) and/or deep vein thrombosis (DVT) and treated using low-molecular-weight heparin (LMWH) for ≥ 6 months after the diagnosis. The exclusion criteria were: (1) a life expectancy of <6 months, (2) cancer in progression, (3) pregnancy, (4) residual vein thrombosis in DVT (defined as <0.3 cm of vein transversal diameter when a compression is exerted with the ultrasound transducer), (5) lupus anticoagulant or anticardiolipin antibodies (required two tests separated by 12 weeks), (6) no medical reason to continue treatment and (7) suspicion of chronic thromboembolic pulmonary hypertension (CTEPH), which was ruled out by the following questions: (a) Have you presented with dyspnea after PE event? (b) Have you presented with palpitations, chest tightness or fading without the known justification? (c) Have you had signs of congestive heart failure without the known etiology or justification (jugular engorgement, peripheral oedema, or ascites)? 2.3. Study Protocol The study plan is shown in Figure 1. Patients were treated with LMWH at a weightadjusted dosage. After at least 6 months of anticoagulation treatment, the clinician proposed the enrollment to those patients with no exclusion criteria. Once anticoagulation was stopped, blood samples were obtained from all patients to test biomarkers on the day of anticoagulation withdrawal (baseline), 21 and 90 days later. All blood samples were analyzed in the same laboratory (Instituto de Biomedicina de Sevilla) using identical procedures. Patients were informed to detect symptoms of VTE recurrence, and a telephone number was provided to contact the clinicians to consult any suggestive symptoms of VTE recurrence. All patients were also followed after 180 days. Episodes of VTE recurrence, primary outcome, bleeding and all causes of death were recorded for analysis. When a patient suffers a recurrence of VTE, it will mean the end of follow-up for the patient, anticoagulant treatment will be started again and subsequent biomarker determinations that the patient has scheduled will be suspended. 2.4. Blood Samples and Biomarker Testing Venous blood was collected with a 21-gauge needle (discarding the first 3 mL) using 3.5 mL of 9NC coagulation 3.2% sodium citrate Vacuette ® tubes (Greiner Bio-One, NC, USA). Citrated blood samples were centrifuged at 1500 × gfor 30 min at 4 ◦ C within one hour of venipuncture. Plasma aliquots were stored at − 80 ◦ C until further use. Biomarkers were performed according to manufacturer’s instructions: high-sensitivity C-reactive protein (hs-CRP) (N High Sensitivity CRP, Dade Behring ® ; normal range: <5 mg/L) on a Behring Nephelometer II System; DD (Innovance ® D-Dimer Siemens Medical Solutions Diagnostics, Deerfield, IL, USA, lower limit of detection 0.099 mg/L fibrinogen equivalent units (FEU)); sP-Selectin (Human sP-selectin/CD62P Immunoassay, sensitivity: 0.121 ng/mL, specificity: recombinant and human P-selectin, R&D Systems, Minneapolis, MN, USA); and tissue factor (TF) (Imubind ® human Tissue Factor ELISA, sensitivity: 10 pg/mL approximately, specificity: TF-apo, TF and TF-VII complex and is designed such as there is no interference from other coagulation factors or inhibitors of procoagulant activity, American Diagnostica, Stanford, CT, USA). Factor (F) VIII activity was measured by coagulometric method on a Sysmex CA 7000 analyzer using FVIII-deficient plasma and APTT reagents (all from Siemens Healthcare Diagnostics, Newark, DE, EEUU). All assays Cancers 2022,14, 2771 4 of 15 were performed according to the manufacturer’s specifications and standard laboratory methods. The phospholipid procoagulant activity (STA-Procoag-PPL, Diagnostica Stago, Asnieres, France) assay was used to evaluate the PPL activity. Results are expressed as coagulation time (s), the shorter the coagulation time, the higher the PPL activity. Figure 1. Plan of the HISPALIS Study. The thrombin generation test (TGT) was performed using the Calibrated Automated Thrombogram at 37 ◦ C in a Fluoroskan Ascent reader (Thermo Labsystems) with 390/460 filters. The TG curves and the different parameters were obtained using the software Thrombinoscope (Thrombinoscope BV). A typical TGT reaction consisted of 80 µ L platelet-poor plasma, 20 µ L PPP reagent (Thrombinoscope BV) that includes 5 pM TF and 4 µ M phospholipids and 20 µ L fluorogenic substrate FluCa-Kit (Thrombinoscope BV) in the presence of CaCl 2 . Each reaction was performed in duplicate. In parallel, an additional reaction containing 20 µL thrombin calibrator (Thrombinoscope BV) was also performed for normalization. 2.5. Study Variables Data were collected at the time of the enrolment concerning the patients’ demographic characteristics, VTE risk factors, VTE diagnosis (DVT and/or PE), anticoagulation treatment and oncological treatment before inclusion. The biomarker results (at baseline, 21 or 90 days) were collected. The main outcome variable was VTE recurrences up to 6 months after stopping the anticoagulant treatment. All-cause mortality and VTE-related mortality were determined using objective methods or via clinical consensus among the oncologists and pulmonologist or internist. Recurrent PE was defined with at least one of the following findings: (1) a new intraluminal filling defect in the sub-segmental or more proximal branches that were detected during spiral computed tomography; (2) a new intraluminal filling defect, the extension of an existing defect or a new sudden occlusion of vessels with a diameter of >2.5mm that was detected during pulmonary angiography; (3) a new perfusion defect ( ≥ 75% of a segment) with a local normal ventilation result (high probability) on the ventilation/perfusion lung scan; or (4) inconclusive findings from computed tomography, pulmonary angiography or ventilation/perfusion lung scan but with compression ultrasonography or venography revealing new or extended DVT in the lower extremities. Recurrent DVT was defined as (1) abnormal findings from compression ultrasonography with previous normal compression, (2) a substantial increase in diameter ( ≥ 4 mm) of the thrombus during full compression for previously non-compressible areas or (3) extension of an intraluminal filling defect, a new intraluminal filling defect, or an extension of venous non-visualization in the presence of a sudden cutoff during venography. Cancers 2022,14, 2771 5 of 15 2.6. Statistical Analysis and Sample Size Continuous variables were reported as mean ± standard deviation (SD) or median and interquartile range (IQR) according to normal or non-normal distribution, respectively. Categorical variables were reported as number and frequency. We evaluated the optimal levels of DD, hs-CRP and sP-selectin for predicting VTE recurrence using a receiver operating characteristic (ROC) curve, the area under the curve (AUC) was >0.5 based on the standard error obtained using DeLong’s method. A multivariate logistic regression model evaluated the predictive values of these biomarkers. Competing risk regression analysis of time to VTE recurrence was performed for significant biomarkers (Fisher’s exact test, p< 0.05). Wald’s test was used to assess the effect of a variable within the competing risk regression model. All analyses were performed using the IBM SPSS software (version 20), EPIDAT software (version 4.1) and R software (version 3.0.1) with the “survival” and “cmprsk” packages. We assumed a recurrence rate of 10% during the first 6 months after discontinuation of anticoagulation treatment. Based on a safety level of 95% (1 −α ), statistical power of 90% and a 15% mortality rate, the minimum required sample size was 113 patients. 3. Results A total of 377 CAT patients, with LMWH beyond 6 months, were evaluated; finally, 166 patients were enrolled in the study. Figure 2shows the flowchart of patients enrolled in the HISPALIS study. The clinical characteristics of the patients studied are summarized in Table 1, while the clinical characteristics of the excluded patients can be seen in Table S1. The mean age of the enrolled participants was 64 ± 17 years and 53.6% (N = 89) were men. There was a higher rate of recurrences in men (11/89; 12.4%) than in women (5/77; 6.5%) but this difference was not significant. Tumors were located at the following sites: gastrointestinal (N = 35, 21.1%), breast and ovarian (N = 30, 18.1%), hematologic (N = 28, 16.9%), urinary (N = 26, 15.6%), lung (N = 19, 11.4%), cerebral (N = 3, 1.8%) and other sites (N = 25, 15.1%). Nearly 40% of the patients had metastases. VTE presented as follows: 53.6% DVT, 28.3% PE, 11.4% PE plus DVT and 6.1% atypical VTE (N = 4 in inferior vena cava and iliac territory, N = 2 upper limb DVT and N = 1 splanchnic vein thrombosis). Figure 2. Flowchart of HISPALIS Study. Cancers 2022,14, 2771 6 of 15 The incidence of VTE recurrence within 3 months after anticoagulation withdrawal was 4.8%, six patients suffered DVT and two patients PE plus DVT. The VTE recurrence within 6 months was 9.6%, eleven patients suffered DVT and five patients PE plus DVT. Table 1. Clinical characteristics of the patients studied. Total N = 166 VTE Recurrence at 3 Months N=8 VTE Recurrence at 6 Months N = 16 Sex male, N (%) 89 (53.6) 6 (75.0) 11 (68.8) Age (y), median (IQR) 64 (17.0) 64.5 (23.5) 63.5 (16.0) Weight (Kg), median (IQR) 71 (20.0) 82.0 (19.0) 79.5 (14.0) Cancer location, N (%) Gastrointestinal 35 (21.1) 0 (0.0) 1 (6.2) Breast and Ovarian 30 (18.1) 1 (12.5) 3 (18.8) Hematologic 28 (16.9) 1 (12.5) 2 (12.5) Urinary 26 (15.6) 1 (12.5) 3 (18.8) Lung 19 (11.4) 3 (37.5) 5 (31.2) Cerebral 3 (1.8) 1 (12.5) 1 (6.2) Other 25 (15.1) 1 (12.5) 1 (6.2) VTE, N (%) DVT 89 (53.6) 6 (75.0) 11 (68.8) PE 47 (28.3) 0 (0.0) 0 (0.0) DVT + PE 18 (10.8) 2 (25.0) 5 (31.2) * Atypical VTE 12 (7.2) 0 (0.0) 0 (0.0) * PT g.20210G>A, N (%) 6 (5.0) 0 (0.0) 1 (10.0) FV Leiden, N (%) 10 (6.0) 2 (25.0) 2 (20.0) Anti-phospholipid antibody, N (%) 2 (1.7) 0 (0.0) 0 (0.0) Situation at anticoagulant therapy discontinuation, N (%) ECOG status ≥2 16 (9.6) 1 (12.5) 3 (18.8) Active metastasis 65 (39.2) 5 (62.5) 9 (56.2) Active cancer treatment Hormone therapy Alkylating agents Anti-metabolites Immune checkpoint inhibitors Monoclonal antibodies 56 (33.7) 20 (12.0) 3 (1.8) 15 (9.0) 5 (3.0) 8 (4.8) 4 (50.0) 2 (25.0) 1 (12.5) 0 (0.0) 2 (25.0) 0 (0.0) 9 (56.2) * 3 (18.8) 1 (6.3) 3 (18.8) 2 (12.5) 0 (0.0) CV Catheter 12 (7.2) 1 (12.5) 1 (6.2) Months of anticoagulant treatment, median (IQR) 10 (11.0) 18 (15.0) 12 (13.0) DVT resolution, N (%) 72/107 (67.2) 5/8 (62.5) 9/16 (56.2) PE resolution, N (%) 53/65 (81.5) 2/2 (100.0) 3/5 (60.0) *p< 0.05; PT g.20210G>A, FV Leiden and anti-phospholipid antibody analyses were performed in 120 patients. VTE: venous thromboembolism; IQR: interquartile range; DVT: deep vein thrombosis; PE: pulmonary embolism; PT: prothrombin; CV: central venous. A patient could receive several drugs of different types of cancer treatment. Of the 166 patients studied, six (3.61%) had bleedings before stopping the anticoagulant treatment, two of whom had major bleeding. The median ± IQR for the duration of anticoagulant treatment in all patients was 10 ± 11 months, with no differences between those with or without VTE recurrences (12 ± 14 vs. 10 ± 11 months; p= 0.102). No association was found between VTE recurrences and anticoagulation treatment for <12 months (6/61) vs. ≥12 months (4/53) (p= 0.75). Cancers 2022,14, 2771 7 of 15 3.1. Incidence of VTE Recurrences after Anticoagulation Withdrawal In order to evaluate the existence of specific periods of higher VTE recurrence after anticoagulant withdrawal, we performed a Kaplan–Meier analysis. Figure 3depicts the cumulative incidence of VTE recurrences for the follow-up time in our cohort of CAT patients. Following 180 days of anticoagulant withdrawal, 8 VTE recurrences were registered and at the end of follow-up, 16 recurrences were registered. The cumulative rate of all follow-up period was 9.87% (95% CI: 5.1–14.7). All recurrences were objectively diagnosed but anticoagulant treatment was restarted without objective diagnosis in nine patients (5.42%). The reason was an extremely high DD value according to the doctor’s interpretation that made a reintroduction of the anticoagulant treatment. Figure 3. Kaplan–Meier curve of VTE recurrences in cancer patients after anticoagulation withdrawal. 3.2. Evolution of Biomarkers after Anticoagulation Withdrawal In each blood sample obtained from CAT patients (0 days, 21 days and 90 days after anticoagulation withdrawal) different parameters were measured and their levels were compared between patients with and without VTE recurrence. As seen in Figure 4, we found significantly increased DD levels at recruitment, at 21 days and 90 days after anticoagulation withdrawal. We also found a significant increase in sP-selectin at recruitment and at 21 days but not at 90 days after anticoagulation withdrawal. Additionally, we found higher levels of hs-CRP in those patients who suffered VTE recurrence 21 days and 90 days after anticoagulation withdrawal but not at recruitment. Finally, we found an increase in FVIII levels 90 days after anticoagulation withdrawal. No significant differences were evidenced in TF levels or PPL activity during follow-up. We also performed the TGT in each blood sample obtained and we compared the results of each TGT parameter between patients with and without VTE recurrence. We found a significant increase in the lagtime 21 days and 90 days after anticoagulant withdrawal. We also observed an increase in the ttPeak and start tail 90 days after anticoagulant withdrawal. No significant differences occurred in ETP, peak or VelIndex during follow-up (Figure 5). Cancers 2022,14, 2771 8 of 15 Figure 4. Longitudinal profile of coagulation-related parameters measured in CAT patients. D-dimer (DD), soluble P-selectin (sP-selectin), high-sensitivity C reactive protein (hs-CRP), factor VIII (FVIII), tissue factor (TF) and phospholipid procoagulant activity (PPL). Blue boxes refer to CAT patients with VTE recurrence and red boxes refer to patients without VTE recurrence. Parameters were measured at recruitment (the day of anticoagulation discontinuation), 21 days and 90 days after discontinuation of anticoagulation. Boxplots show median and interquartile range. Cancers 2022,14, 2771 9 of 15 Figure 5. Longitudinal profile of the different parameters obtained from the thrombin generation test in CAT patients. Endogen thrombin potential (ETP), velocity index (Vel Index) and time to peak (ttPeak). Blue boxes refer to CAT patients with VTE recurrence and red boxes refer to patients without VTE recurrence. TGT parameters were measured at recruitment (the day of anticoagulation discontinuation), 21 days and 90 days after discontinuation of anticoagulation. Boxplots show median and interquartile range.