Evaluation of the Khorana Score for Venous Thromboembolism Prediction in Multiple Myeloma (2022) - PDF

Received: 21 July 2021 | Revised: 7 October 2021 | Accepted: 19 October 2021 DOI: 10.1002/rth2.12634 ORIGINAL ARTICLE Evaluation of the Khorana score for prediction of venous thromboembolism in patients with multiple myeloma Kristen M. Sanfilippo MD, MPHS1,2 | Kenneth R. Carson...

Received: 21 July 2021 | Revised: 7 October 2021 | Accepted: 19 October 2021 DOI: 10.1002/rth2.12634 ORIGINAL ARTICLE Evaluation of the Khorana score for prediction of venous thromboembolism in patients with multiple myeloma Kristen M. Sanfilippo MD, MPHS1,2 | Kenneth R. Carson MD, PhD3 | Tzu-­Fei Wang MD, MPH4 | Suhong Luo MS1,2 | Natasha Edwin MD5 | Nicole Kuderer MD6 | Jesse M. Keller MPD, MPHS1,2 | Brian F. Gage MD, MSc1 1 Washington University School of Medicine in St. Louis, St. Louis, Missouri, Abstract USA Background: Guidelines recommend thromboprophylaxis for patients with multiple 2 St. Louis Veterans Affairs Medical Center, St. Louis, Missouri, USA myeloma (MM) at high risk for venous thromboembolism (VTE). However, the optimal 3 Rush University Medical Center, Chicago, risk prediction model for VTE in MM remains unclear. Khorana et al developed a VTE Illinois, USA risk score (Khorana score) in ambulatory cancer patients receiving chemotherapy. We 4 University of Ottawa at The Ottawa aimed to evaluate the predictive ability of the Khorana score in patients with MM. Hospital and Ottawa Hospital Research Institute, Ottawa, Ontario, Canada Methods: We identified patients with MM within the Veterans Affairs health care 5 ThedaCare Regional Cancer Center, system between 2006 and 2013. The Khorana score was calculated before treatment Appleton, Wisconsin, USA 6 initiation. Using logistic regression, the relationship between risk group and VTE was Advanced Cancer Research Group, Kirkland, Washington, USA assessed at 3 and 6 months. We tested model discrimination using the concordance statistic. Correspondence Kristen M. Sanfilippo, Hematology Results: In the cohort of 2870 patients with MM, there were 1328 at low risk (0 Division, Campus Box 8125, Washington points), 1521 at intermediate risk (1-­2 points), and 21 at high risk (≥3 points) for VTE University School of Medicine, 660 South Euclid Avenue, St. Louis, MO 63110, USA. by the Khorana score. The 6-­month cumulative incidence of VTE was 5.1% (95% con- Email: [email protected] fidence interval [CI], 4.0%-­6.4%) in low risk, 3.9% (95% CI, 3.0%-­5.0%) in intermediate Funding information risk, 4.8% (95% CI, 0.3%-­20.2%) in high risk. The Khorana score did not strongly dis- This work was supported by the National criminate between patients who did and did not develop VTEs at 3 or 6 months (con- Heart, Lung, and Blood Institute at the National Institutes of Health cordance statistic, 0.58; 95% CI, 0.54-­0.63; and 0.53, 95% CI, 0.50-­0.57, respectively. (1K01HL136893-­01, 5K12 HL087107-­95, Conclusions: In conclusion, in this cohort of 2870 patients with MM, the Khorana and the NIH Loan Repayment Program to KMS). score did not predict VTE. Our study supports the need to use myeloma-­specific risk models to predict VTE risk in patients with MM. Handling Editor: Dr Lana Castellucci KEYWORDS cancer-­associated thrombosis, Khorana score, multiple myeloma, risk prediction, venous thromboembolism This is an open access article under the terms of the Creat​ive Commo​ns Attri​butio​n-­NonCo​mmerc​ial-­NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-­commercial and no modifications or adaptations are made. © 2022 The Authors. Research and Practice in Thrombosis and Haemostasis published by Wiley Periodicals LLC on behalf of International Society on Thrombosis and Haemostasis (ISTH). Res Pract Thromb Haemost. 2022;6:e12634.  wileyonlinelibrary.com/journal/rth2 | 1 of 7 https://doi.org/10.1002/rth2.12634 2 of 7 | SANFILIPPO et al. Essentials The optimal prediction model to quantify risk of thrombosis in multiple myeloma (MM) is unknown. We evaluated the performance of one potential model, the Khorana Score, in patients with MM. In our cohort of 2870 patients, the Khorana Score did predict thrombosis (concordance statistic, 0.58). MM-­specific prediction models should be used to predict thrombosis in patients with MM. 1 | I NTRO D U C TI O N and Washington University School of Medicine institutional review boards approved this study. With >100 000 Americans dying from venous thromboembolism We obtained data using the VA Informatics and Computing (VTE) annually, VTE is one of the most preventable causes of death.1,2 Infrastructure platform. Using height and weight assessed within Patients with cancer have an increased risk of VTE.3 This risk varies 1 month of diagnosis, body mass index (BMI) was calculated. with type of malignancy, and patients with multiple myeloma (MM) Hemoglobin (HGB), white blood cell (WBC) count, and platelet (PLT) have a 9-­fold increased risk of VTE compared to patients without count were obtained before treatment initiation but within 2 months MM.4,5 Thromboprophylaxis in patients with MM at high risk of VTE of MM diagnosis. When height, weight, HGB, WBC count, or PLT may reduce the incidence of VTE and improve patient survival.6-­8 count were not available electronically, the missing variables were However, identification of high-­risk patients in MM has remained abstracted manually from unstructured medical records, and if not challenging. The Myeloma XI trial incorporated International Myeloma available the patient was excluded. The primary outcome was the first Working Group (IMWG) guidance to identify patients with MM at high episode of VTE that occurred within 6-­months of MM-­treatment ini- risk of VTE and guide thromboprophylaxis.9 Despite this risk stratifi- tiation. Using a previously validated algorithm that combined ICD-­9 cation, incidence of VTE remained high, exceeding 10% at 6 months diagnostic codes with prescription for anticoagulation or placement after treatment initiation. Better risk prediction models are needed. of an inferior vena cava filter, we identified VTE.14 All VTEs were The Khorana score, a validated risk prediction model developed manually validated through chart abstraction. Patients who devel- primarily in patients with solid tumors and lymphoma, identifies am- oped VTE between MM diagnosis and start of chemotherapy were 10 bulatory patients with cancer at high risk of VTE. Recently, two excluded, as the Khorana score was developed for patients starting randomized controlled trials, AVERT and CASSINI, showed that pro- chemotherapy. Using the Pharmacy Benefits Management database, phylactic doses of direct oral anticoagulants (DOACs) reduced the data were obtained on medication utilization, including antineoplas- risk of VTE in ambulatory patients with cancer at intermediate to tic therapy, aspirin, warfarin, and low-­molecular-­weight heparin high risk for VTE as determined by a Khorana score of ≥2, without (LMWH). Given that dexamethasone is sometimes started before a significant increase in the risk of major bleeding.11-­13 However, pa- chemotherapy, patients were considered dexamethasone users if tients with MM were underrepresented in the derivation cohort for they received a prescription between MM diagnosis and chemother- the Khorana score. Although CASSINI excluded patients with MM, apy start date. Since the VA often prescribes more than a month’s the AVERT trial did include patients with MM as a “high-­risk” cancer supply per prescription of aspirin, patients were considered aspirin assigned 1 point.12,13 Therefore, there is a need to quantify the VTE users if they received a prescription within 90 days before MM di- risk discrimination of the Khorana score in patients with MM. Here, agnosis up to chemotherapy start date. For anticoagulant therapy, we aimed to assess the discrimination of the Khorana score in pa- patients were considered users if they received a prescription within tients with MM using a large cohort of US veterans. 30 days before MM diagnosis up to the start of chemotherapy. We calculated the Khorana score as developed by assigning 1 point for the following variables: PLT ≥ 350 000/µL, HGB < 10 g/ 2 | M E TH O D S dL and/or use of erythropoiesis-­stimulating agents (ESAs), WBC > 11 000/µL, and BMI ≥ 35 kg/m2.10 Patients with 0 points We identified patients with newly diagnosed MM treated in the were classified as low risk, 1 to 2 points as intermediate-­risk, and Veterans Affairs (VA) health care system between June 29, 2006, ≥3 points as high risk for VTE.10 Logistic regression was used to and December 31, 2013, using the International Classification quantify the odds ratio between the Khorana risk group and the in- of Diseases (ICD) O3 code 9732/3 within the VA Central Cancer cidence of VTE at 3 and 6 months following MM diagnosis while Registry, and followed the cohort through December 31, 2014. To adjusting for the use of aspirin and anticoagulant therapy (warfarin exclude patients with monoclonal gammopathy of undetermined or LMWH). We quantified model discrimination using the area under significance, solitary plasmacytoma, and/or smoldering myeloma, the c-­statistic with a range of 0.5 (no discriminative ability) to 1.0 we excluded patients who did not receive treatment within 6 months (perfect discriminative ability). Using 200 bootstrapped samples, of MM diagnosis. Based on a sample size of 2870 patients, the power 95% confidence intervals (CIs) for each c-­statistic were generated. to reject the null hypothesis (ie, a concordance [c]-­statistic of 0.50) We assessed the association between risk score and development of was 96%. Prior to cohort assembly, the St. Louis VA Medical Center VTE within 6 months after start of chemotherapy using a competing SANFILIPPO et al. | 3 of 7 risk model to adjust for the competing risk of non-­V TE death.15 A 0.3%-­20.2%) in the high-­risk group. Within 6 months of MM diagno- sensitivity analysis adjusting for putative thrombotic risk factors sis, 128 patients developed VTE with cumulative incidence of 5.1% present at the time of initiation of chemotherapy (lenalidomide; tha- (95% CI, 4.0%-­6.4%) in the low-­risk group, 3.9% (95% CI, 3.0%-­5.0%) lidomide; history of VTE prior to MM diagnosis; and prescription of in the intermediate-­risk group, and 4.8% (95% CI, 0.3%-­20.2%) in the aspirin, warfarin, or LMWH) was performed. All medications were high-­risk group. The 6-­month cumulative incidence of VTE stratified analyzed as time-­varying variables. by Khorana score is listed in Table S1. Of the VTE events, 31% of pa- We performed three additional sensitivity analyses. First, pa- tients had a pulmonary embolism, with the majority of events lower-­ tients were categorized as high risk if their Khorana score was ≥2, extremity deep vein thromboses (Table 2). similar to the AVERT and CASSINI trials.12,13 In the second sensi- There was no significant difference between the risk of VTE in tivity analysis, we added 1 point to all patients for their diagnosis the high-­ or intermediate-­risk groups versus the low-­risk group at 3 of MM and categorized patients as high risk with a score of ≥2 (and or 6 months (Table 3). The c-­statistics were 0.58 (95% CI, 0.54-­0.63) low risk otherwise), as in the AVERT trial.12 Third, all patients re- for 3 months and 0.53 (95% CI, 0.50-­0.57) for 6 months. After ex- ceiving anticoagulation at the start of chemotherapy, regardless of cluding patients taking anticoagulants at the start of chemotherapy, dose, were excluded, and the association of Khorana score and VTE results remained unchanged (Table 3), with c-­statistics of 0.57 at 3 at 3 and 6 months was assessed using logistic regression. Analyses months and 0.56 at 6 months. were performed using SAS version 9.2 software (SAS Institute, Cary, After adjusting for putative VTE risk factors, including chemo- NC, USA) and R statistical software (R Foundation for Statistical therapy, VTE history, aspirin, or anticoagulants, a competing risk Computing, Vienna, Austria). analysis found no increased risk for VTE with increasing Khorana score (adjusted hazard ratio [aHR], 0.82 per 1-­point increase; 95% CI, 0.63-­1.08, P =.17). In the competing risk model, use of thalidomide 3 | R E S U LT S or lenalidomide (P =.02) and dexamethasone (high dose, P =