AUTHORS
Mariano Arribas, Kevin Guanwen Shim, MD PhD2, Anand Pabmanabhan, MBBS, PhD3, Surbhi Shah, MBBS, MD4, Leslie Padrnos, MD5, Oluseyi Abidoye, MBBS6, Cody Eslinger, MD, MS7, Kenneth Barker, MD8, Haley Meyer, MD9, J Erin Wiedmeier-Nutor, MD,MPH10, Udit Yadav, MBBS11, Saurabh Chhabra, MD, MBBS12, P. Leif Bergsagel, MD13, Rafael Fonseca, MD14
AFFILIATIONS
1 Mayo Clinic Arizona, Phoenix, Arizona, United States
2 Mayo Clinic Arizona, Phoenix, Arizona, United States
3 Mayo Clinic Arizona, Phoenix, Arizona, United States
4 Mayo Clinic Arizona, Phoenix, Arizona, United States
5 Mayo Clinic Arizona, Phoenix, Arizona, United States
6 Mayo Clinic Arizona, Phoenix, Arizona, United States
7 Mayo Clinic Arizona, Phoenix, Arizona, United States
8 Mayo Clinic Arizona, Phoenix, Arizona, United States
9 Mayo Clinic Arizona, Phoenix, Arizona, United States
10 Mayo Clinic Arizona, Phoenix, Arizona, United States
11 Mayo Clinic Arizona, Phoenix, Arizona, United States
12 Mayo Clinic Arizona, Phoenix, Arizona, United States
13 Mayo Clinic Arizona, Phoenix, Arizona, United States
14 Mayo Clinic Arizona, Phoenix, Arizona, United States
ABSTRACT
Background
While venous thromboembolism (VTE) is a common complication of multiple myeloma (MM), it mostly arises as a consequence of treatment with immunomodulatory drugs (Rajkumar, Mayo Clin Proc 2005) or with immobility in patients with significant pain. A paraprotein-driven hypercoagulable state (“monoclonal gammopathy of thrombotic significance”) described in MGUS (Gkalea et al, Cancers 2023) raises the question of whether a similar mechanism contributes to MM even before treatment begins. Elevated levels of factor VIII, von Willebrand factor, and reduced protein S have been reported in untreated MM (Auwerda et al, Haematologica 2007; van Marion et al, Leuk Res 2008), though their ability to predict individual VTE risk remains unclear. Here we aimed to evaluate the frequency of VTE preceding and following MM diagnosis using real-world evidence (RWE).
Methods
We retrospectively reviewed 2774 patients diagnosed with MM seen at Mayo Clinic in Arizona, using the Mayo Data Explorer (MDE, a RWE repository). Patients with MGUS and smoldering MM were excluded on this analysis. Diagnosis dates were individually and manually verified rather than abstracted only from coded electronic health record data (EHR). VTE were identified using ICD-10 I82 (venous embolism and thrombosis) and I26 (pulmonary embolism) codes and classified as occurring before or after MM diagnosis. To keep follow-up comparable with prior population-based work in MM (Kristinsson et al, Blood 2010), post-diagnosis events occurring more than 10 years after diagnosis were excluded.
For comparison, we also examined rates of thrombosis using the same institutional real-world database (MDE) among solid tumor cohorts with recognized thrombotic risk: kidney (n=94, localized/non-metastatic disease), colorectal (n=315), and biliary tract cancer (n=286, including a mixture of locally advanced and metastatic). Diagnosis dates were individually verified by the corresponding clinical departments. Proportions were compared using two-sided Fisher’s exact tests, with Bonferroni correction for multiple comparisons (α=0.0083).
Results
Among 2774 MM patients, 313 (11.3%) had a documented VTE, but most occurred after MM diagnosis (297/313, 94.9%). Sixteen events (5.1% of all VTE; 0.6% of all MM patients) took place prior to diagnosis. Of these, 13 patients (81.2%) had provoked VTE, often occurring months to years prior to MM diagnosis, while 3 patients (18.8%) had unprovoked VTE prior to MM diagnosis. In the pre-diagnosis group, the median interval from VTE to MM diagnosis was 1.8 months (IQR 0.2–39.5). In the post-diagnosis group, the median interval from MM diagnosis to VTE was 12.1 months (IQR 5.7–50.2).
In contrast, the overall proportion of patients with thrombosis was much higher in kidney cancer (30/94, 31.9%), colorectal cancer (91/315, 28.9%), and biliary tract cancer (113/286, 39.5%). Kidney cancer showed the opposite temporal distribution, with most thrombotic events occurring before cancer diagnosis (20/30, 66.7%). This differed significantly from MM (p=2.3×10?¹5) and is consistent with the high prevalence of venous tumor thrombus in renal cell carcinoma at diagnosis described by Kaptein et al (JACC CardioOncol 2022).
Conclusions
In this real-world cohort, VTE in MM were seen mainly after diagnosis and presumed to be mostly a consequence of therapy with immunomodulatory drugs. Unlike most large RWE databases, diagnosis dates here were manually confirmed for each patient, reducing the risk of temporal misclassification. Our data show that solid tumors had a substantially higher incidence of thrombosis than MM. Lower tissue factor expression in hematologic malignancies compared with solid tumors (Fonseca et al, J Clin Oncol 2017) may partially explain this difference, alongside other disease-specific mechanisms. While cancer is often cited as a risk for thrombosis, these findings suggest that the relationship between malignancy and thrombosis may differ between solid tumors and hematologic cancers.
DISCLOSURES
Mariano Arribas: No relevant conflicts of interest to declare.