Venus thromboembolism and risk stratification and hematological malignancies

In recent years, the incidence of venous thromboembolism (VTE) has been rising in patients with cancer. A Danish cohort indicated that the risk of VTE increased threefold over the last 20 years; this rate is higher than what is seen in the general population [1].

Looking specifically at patients with hematologic malignancies, the risk varies based on the type of the malignancy: myelodysplastic syndrome and chronic myeloid leukemia have a lower risk of VTE, while acute lymphoid leukemia, Hodgkin lymphoma, and multiple myeloma have the highest risk [2].

In addition, the risk is highest within this first 6-month period; some of the higher-risk hematologic malignancies have an increased risk of about 10−20-fold in the first 6 months, with a subsequent decrease over the lifespan of cancer [2].

It is often believed that patients with central venous catheters tend to have a higher risk of VTE, but this does not seem true. A recent study showed that in patients with hematological malignancy, the initial presenting event was pulmonary embolism in 51% of cases, followed by deep vein thrombosis (DVT) in 47% of cases. Furthermore, most of the DVT events (32%) were lower extremity deep vein thrombosis and not catheter-associated thrombosis, which only represented a small portion (12%) [3].

Additionally, anticoagulant therapy in patients with hematological malignancy carries a higher risk of recurrent VTE (7.67 [5.82−9.92] events per 100 patients-years) and bleeding (5.69 [4.12−7.66] events per 100 patients-years) than in the general population [3].

What can be done to prevent VTE events, their consequences, and anticoagulant therapy risks in hematological cancer patients?

A meta-analysis performed on the results of the AVERT and CASSINI trials showed that, in high-risk-VTE ambulatory cancer patients, thromboprophylaxis significantly decreases the risk of VTE by almost 50% with no significant increase in the risk of bleeding compared to placebo [4].

High-risk ambulatory cancer patients were defined by a Khorana score equal to or higher than two. In the AVERT trial, patients with multiple myeloma were assigned a point for the known risk associated with the disease [4].

The Khorana score is the most validated clinical risk prediction model for cancer-associated thrombosis (CAT). However, only a small percentage of the patients enrolled in clinical trials have hematological malignancy. For example, in the CASSINI and AVERT trial, 12% of the derivation cohort comprised patients with lymphoma. Patients with acute leukemia were excluded from the derivation cohorts, along with other hematologic malignancies, such as myeloma and acute leukemia [4].

A few studies investigated the Khorana in the context of hematologic malignancies [5-11].

Considering the c-statistic score as the reference to grade the score efficacy, the Khorana score showed mixed results in predicting VTE risk across lymphoma cancer patients. In some studies, the Khorana score was moderate (c-statistic of 0.6); in others, it was more mediocre (c-statistics of 0.5). Khorana score was found to perform in the moderate range in cancer patients with multiple myeloma (c-statistic of 0.58) [5-11].

C-statistic is a measure to discriminate which risk model can perfectly discriminate which patients will develop thrombosis. For example, a c-statistic score of 1 indicates that the model is 100% predicting the risk of developing VTE, while a c-statistic value of 0.5 is compatible with flipping a coin.

Generally, a score of six would be considered moderate, well-performing, whereas a score of 0.5 is considered mediocre.

One of the reasons why the Khorana score may not perform well in the hematologic malignancy patient population is that they have unique risk factors for VTE (different from the solid tumor); some of those could be treatment-related, such as asparaginase for acute lymphoblastic leukemia (ALL), or tumor type-related, with some hematological cancer (e.g., Burkitt’s lymphoma) carrying a higher risk than other more indolent lymphomas (e.g., follicular lymphoma) as previously mentioned.

A few attempts were to develop risk scores specific for hematological malignancies, such as the ThroLy score, TiC-LYMPHO, and Dharmarvaram score [12-14].

These scores incorporate common variables, such as immobility, history of VTE or arterial-venous thromboembolism (ATE), mediastinal involvement, and extranidal disease. The TiC-LYMPHO score also incorporates genetic risk, ThroLy considers neutrophil and hemoglobin counts with BMI, and Dharmarvaram considers white blood cell count and albumin.

In developing the ThroLy score, the study considered ATE, VTE, and superficial as composite outcomes. At the same time, VTE was the only considered outcome for developing the other two scores. During internal validation, all the three scores performed very well (c-statistics between 0.78 and 0.86), indicating a good predicting ability [13-14].

Only the ThroLy score was externally validated with two studies that focused on VTE only as an outcome. In the two external validation cohorts, the ThroLy score lost some of its predictive ability showing a reduction in performance from a c-statistics of 0.86 to a c-statistics of 0.55 or 0.57 [15, 16]. It is evident that the scores naturally perform better within the cohort for which they were developed.

Specifically for leukemia, one model was developed in adults with acute leukemia, the Al-Ani score, a relatively simple score with three variables (ALL, history of VTE/ATE, and platelet count) that includes all three types of acute leukemia. At internal validation, the Al-Ani score performed moderately well with a c-statistics of 0.66 [17].

In addition, a pediatric ALL risk score was developed that considers three variables: treatment, presence of central venous line, and thrombophilia [18].

Unfortunately, external validation has not yet been performed for these two models in patients with acute leukemia.

In 2016, a study assessed the predictive ability of D-dimer alone in acute leukemia patients founding that D-dimer of over 4mg/L is particularly well predictive of the risk of VTE with a c-statistic of 0.72 [19].

Myeloma-specific models

The first most widely used as an expert consensus developed by the International Myeloma Working Group regarding myeloma-specific models. The model suggested that high-VTE risk patients should receive low molecular weight heparin prophylaxis (or more intense therapy), and patients with lower risk were recommended to receive aspirin therapy [20].

In 2019, two new scores for multiple myeloma were published, IMPEDE VTE and SAVED, with a series of overlapping risk factors like Asian race being a protective factor and the dose of dexamethasone a risk factor [21,22]. In addition, the IMPEDE-VTE score also incorporated anticoagulant therapy.

Looking at the external validation of three models using the SEER-Medicare data and US Veterans cohorts, the International Myeloma Working Group performed mediocre, with a c-statistics of 0.52 and 0.55, respectively. The SAVED score was internally validated using the US-veterans cohort and maintained its c-statistics in the external validation using the SEER-Medicare cohort (c-statistics of 0.61 vs 0.60, respectively). IMPEDE-VTE score performed moderately well in its internal validation with a c-statistics of 0.66. It maintained its performance in external validations (SEER-Medicare cohort c-statistics of 0.65, MELISSE cohort c-statistics of 0.65, and cohort of patients treated at Cleveland Clinic c-statistics of 0.68) [21-24].

Is thromboprophylaxis effective in hematological malignancies?

As mentioned, most patients enrolled in thromboprophylaxis trials have solid tumors. A subgroup analysis of 156 patients that participated in the AVERT trial did not find a significant reduction of VTE in patients treated with apixaban thromboprophylaxis versus placebo [25].

Some smaller, prospective cohort studies or pilot studies in multiple myeloma also investigated prophylaxis in patients with hematological cancer. In one of these, patients considered to be at higher risk received apixaban 2.5 mg twice daily, and those at lower risk received low-dose aspirin. The rate of VTE was very low in both groups, but all the patients involved in the study were at their third-plus-line therapy [26]. As discussed, the risk of VTE is higher in the first 6 months and decreases slowly over time, yet stays higher than in the normal population. Risk models might lose some ability to stratify patients in more advanced stages of the disease.

Similar assessments have been done in patients with acute leukemia receiving asparaginase. The THROMBOTECT trial randomized patients among three groups, unfractionated heparin, low molecular weight heparin, and antithrombin, for thromboprophylaxis during induction therapy of ALL in children and adolescents. Results indicate that unfractionated heparin performed the best [27].

The PREVAPIX-ALL study will hopefully provide additional data on this. In this study, the patients are randomized for prophylaxis with apixaban vs no prophylaxis. Preliminary results (after 512 patients enrolled) show a possible trend toward a decreased VTE incidence in patients receiving apixaban [28].

Guidelines recommend prophylactic anticoagulant therapy for ambulatory cancer patients using the Khorana score as a risk model. In addition, most guidelines have separated recommendations for multiple myeloma patients, suggesting considering the type of treatment [29-31].

NCCN guidelines were updated after the publication of IMPEDE VTE or SAVED and suggested these two as risk factor models for people with multiple myeloma, recommending to decide prophylaxis based on those [30].

Conclusion

The Khorana score is the only one that we have clinical data trial on (AVERT and CASSINI) for hematological malignancy. No reliable validation for any score in haematological malignancy exists at the moment.

References

1. Mulder FI, Horváth-Puhó E, van Es N, et al. Venous thromboembolism in cancer patients: a population-based cohort study. Blood. 2021;137(14):1959-1969. doi:10.1182/blood.2020007338
2. Adelborg K, Corraini P, Darvalics B, et al. Risk of thromboembolic and bleeding outcomes following hematological cancers: A Danish population-based cohort study. J Thromb Haemost. 2019;17(8):1305-1318. doi:10.1111/jth.14475
3. Lecumberri R, Ruiz-Artacho P, Tzoran I, et al. Outcome of Cancer-Associated Venous Thromboembolism Is More Favorable among Patients with Hematologic Malignancies than in Those with Solid Tumors [published online ahead of print, 2022 Jun 13]. Thromb Haemost. 2022;10.1055/a-1777-4006. doi:10.1055/a-1777-4006
4. Wang TF, Zwicker JI, Ay C, et al. The use of direct oral anticoagulants for primary thromboprophylaxis in ambulatory cancer patients: Guidance from the SSC of the ISTH. J Thromb Haemost. 2019;17(10):1772-1778. doi:10.1111/jth.14564
5. Mirza AS, Yun S, Ali NA, Shin H, O’Neil JL, Elharake M, Schwartz D, Robinson K, Nowell E, Engle G, Badat I, Brimer T, Kuc A, Sequeira A, Mirza S, Sikaria D, Vera JD, Hackney N, Abusrur S, Jesurajan J, Kuang J, Patel S, Khalil S, Bhaskar S, Beard A, Abuelenen T, Ratnasamy K, Visweshwar N, Komrokji R, Jaglal M. Validation of the Khorana score in acute myeloid leukemia patients: a single-institution experience. Thromb J. 2019 Jul 2;17:13. doi: 10.1186/s12959-019-0202-z. PMID: 31303864; PMCID: PMC6604148.
6. Santi RM, Ceccarelli M, Bernocco E, Monagheddu C, Evangelista A, Valeri F, Monaco F, Vitolo U, Cortelazzo S, Cabras MG, Spina M, Baldini L, Boccomini C, Chiappella A, Bari A, Luminari S, Visco C, Calabrese M, Limberti G, Levis A, Contino L, Ciccone G, Ladetto M. Khorana score and histotype predicts incidence of early venous thromboembolism in non-Hodgkin lymphomas. A pooled-data analysis of 12 clinical trials of Fondazione Italiana Linfomi (FIL). Thromb Haemost. 2017 Apr 27. doi: 10.1160/TH16-11-0895. Epub ahead of print. PMID: 28447710.
7. Dharmavaram G, Cao S, Sundaram S, Ayyappan S, Boughan K, Gallogly M, Malek E, Metheny L, Tomlinson B, Otegbeye F, Lazarus HM, Cooper B, Fu P, de Lima M, Caimi PF. Aggressive lymphoma subtype is a risk factor for venous thrombosis. Development of lymphoma – specific venous thrombosis prediction models. Am J Hematol. 2020
8. Aug;95(8):918-926. doi: 10.1002/ajh.25837. Epub 2020 May 8. PMID: 32311162.
9. Rupa-Matysek J, Gil L, Kaźmierczak M, Barańska M, Komarnicki M. Prediction of venous thromboembolism in newly diagnosed patients treated for lymphoid malignancies: validation of the Khorana Risk Score. Med Oncol. 2017;35(1):5. Published 2017 Dec 4. doi:10.1007/s12032-017-1065-4
10. Bastos-Oreiro M, Ortiz J, Pradillo V, Salas E, Marínez-Laperche C, Muñoz A, Buño I, Diéz-Martin JL, Soria JM, Pascual Izquierdo C. Incorporating genetic and clinical data into the prediction of thromboembolism risk in patients with lymphoma. Cancer Med. 2021 Nov;10(21):7585-7592. doi: 10.1002/cam4.4280. Epub 2021 Oct 1. PMID: 34598309; PMCID: PMC8559493.
11. Sanfilippo KM, Carson KR, Wang TF, et al. Evaluation of the Khorana score for prediction of venous thromboembolism in patients with multiple myeloma. Res Pract Thromb Haemost. 2022;6(1):e12634. Published 2022 Jan 9. doi:10.1002/rth2.12634
12. Antic D, Milic N, Nikolovski S, et al. Development and validation of multivariable predictive model for thromboembolic events in lymphoma patients. Am J Hematol. 2016;91(10):1014-1019. doi:10.1002/ajh.24466
13. Dharmavaram G, Cao S, Sundaram S, et al. Aggressive lymphoma subtype is a risk factor for venous thrombosis. Development of lymphoma – specific venous thrombosis prediction models. Am J Hematol. 2020;95(8):918-926. doi:10.1002/ajh.25837
14. Bastos-Oreiro M, Ortiz J, Pradillo V, Salas E, Marínez-Laperche C, Muñoz A, Buño I, Diéz-Martin JL, Soria JM, Pascual Izquierdo C. Incorporating genetic and clinical data into the prediction of thromboembolism risk in patients with lymphoma. Cancer Med. 2021 Nov;10(21):7585-7592. doi: 10.1002/cam4.4280. Epub 2021 Oct 1. PMID: 34598309; PMCID: PMC8559493.
15. Rupa-Matysek J, Brzeźniakiewicz-Janus K, Gil L, Krasiński Z, Komarnicki M. Evaluation of the ThroLy score for the prediction of venous thromboembolism in newly diagnosed patients treated for lymphoid malignancies in clinical practice. Cancer Med. 2018 May 15;7(7):2868–75. doi: 10.1002/cam4.1540. Epub ahead of print. PMID: 29761831; PMCID: PMC6051175.
16. Bastos-Oreiro M, Ortiz J, Pradillo V, Salas E, Marínez-Laperche C, Muñoz A, Buño I, Diéz-Martin JL, Soria JM, Pascual Izquierdo C. Incorporating genetic and clinical data into the prediction of thromboembolism risk in patients with lymphoma. Cancer Med. 2021 Nov;10(21):7585-7592. doi: 10.1002/cam4.4280. Epub 2021 Oct 1. PMID: 34598309; PMCID: PMC8559493.
17. Al-Ani F, Wang YP, Lazo-Langner A. Development of a Clinical Prediction Rule for Venous Thromboembolism in Patients with Acute Leukemia. Thromb Haemost. 2020 Feb;120(2):322-328. doi: 10.1055/s-0039-3400303. Epub 2020 Jan 1. PMID: 31893562.
18. Mitchell L, Lambers M, Flege S, et al. Validation of a predictive model for identifying an increased risk for thromboembolism in children with acute lymphoblastic leukemia: results of a multicenter cohort study. Blood. 2010;115(24):4999-5004. doi:10.1182/blood-2010-01-263012
19. Libourel EJ, Klerk CPW, van Norden Y, et al. Disseminated intravascular coagulation at diagnosis is a strong predictor for thrombosis in acute myeloid leukemia. Blood. 2016;128(14):1854-1861. doi:10.1182/blood-2016-02-701094
20. Palumbo A, Rajkumar SV, Dimopoulos MA, et al. Prevention of thalidomide- and lenalidomide-associated thrombosis in myeloma. Leukemia. 2008;22(2):414-423. doi:10.1038/sj.leu.2405062
21. Li A, Wu Q, Luo S, et al. Derivation and Validation of a Risk Assessment Model for Immunomodulatory Drug-Associated Thrombosis Among Patients With Multiple Myeloma. J Natl Compr Canc Netw. 2019;17(7):840-847. doi:10.6004/jnccn.2018.7273
22. Sanfilippo KM, Luo S, Wang TF, et al. Predicting venous thromboembolism in multiple myeloma: development and validation of the IMPEDE VTE score. Am J Hematol. 2019;94(11):1176-1184. doi:10.1002/ajh.25603
23. Chalayer E, Teste A, Guyotat D, Elalamy I, Leleu X, Tardy B. Predicting the risk of venous thromboembolism in newly diagnosed myeloma with immunomodulatory drugs: External validation of the IMPEDE VTE score. Am J Hematol. 2020;95(1):E18-E20. doi:10.1002/ajh.25658
24. Covut F, Ahmed R, Chawla S, et al. Validation of the IMPEDE VTE score for prediction of venous thromboembolism in multiple myeloma: a retrospective cohort study. Br J Haematol. 2021;193(6):1213-1219. doi:10.1111/bjh.17505
25. Nayak AL, Zahrai A, Mallick R, et al. Efficacy of primary prevention of venous thromboembolism among subgroups of cancer patients undergoing chemotherapy: A post- hoc analysis of the AVERT trial. Thromb Res. 2021;208:79-82. doi:10.1016/j.thromres.2021.10.018
26. Sayar Z, Gates C, Bristogiannis S, et al. Safety and efficacy of apixaban as thromboprophylaxis in myeloma patients receiving chemotherapy: A prospective cohort study. Thromb Res. 2022;213:27-29. doi:10.1016/j.thromres.2022.03.001
27. Greiner J, Schrappe M, Claviez A, et al. THROMBOTECT – a randomized study comparing low molecular weight heparin, antithrombin and unfractionated heparin for thromboprophylaxis during induction therapy of acute lymphoblastic leukemia in children and adolescents. Haematologica. 2019;104(4):756-765. doi:10.3324/haematol.2018.194175
28. O’Brien SH, Li D, Mitchell LG, et al. PREVAPIX-ALL: Apixaban Compared to Standard of Care for Prevention of Venous Thrombosis in Paediatric Acute Lymphoblastic Leukaemia (ALL)-Rationale and Design. Thromb Haemost. 2019;119(5):844-853. doi:10.1055/s-0039-1679938
29. Key NS, Khorana AA, Kuderer NM, et al. Venous Thromboembolism Prophylaxis and Treatment in Patients With Cancer: ASCO Clinical Practice Guideline Update. J Clin Oncol. 2020;38(5):496-520. doi:10.1200/JCO.19.01461
30. Kumar SK, Callander NS, Adekola K, et al. Multiple Myeloma, Version 3.2021, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw. 2020;18(12):1685-1717. Published 2020 Dec 2. doi:10.6004/jnccn.2020.0057
31. Lyman GH, Carrier M, Ay C, et al. American Society of Hematology 2021 guidelines for management of venous thromboembolism: prevention and treatment in patients with cancer [published correction appears in Blood Adv. 2021 Apr 13;5(7):1953]. Blood Adv. 2021;5(4):927-974. doi:10.1182/bloodadvances.2020003442