Risk Factors Predictive of Occult Cancer Detection in Patients with Unprovoked Venous Thromboembolism

Background: Venous thromboembolism (VTE) may be the earliest sign of cancer. Risk factors associated with the presence of an occult cancer in patients with a first acute unprovoked VTE are unknown. We sought to assess the risk factors predictive of occult cancer detection in patients with a first unprovoked symptomatic VTE.

Methods: Post-hoc, pre-defined analyses of the multicenter open-label randomized controlled trial – Screening for Occult Malignancy in Patients with Idiopathic Venous Thromboembolism (SOME) trial (Carrier M et al. N Engl J Med 2015). The trial compared comprehensive computed tomography (cCT) of the abdomen and pelvis in addition to limited occult-cancer screening (complete history and examination, basic laboratory testing, chest radiography, and breast, cervical and prostate cancer screening) with limited occult-cancer screening alone in patients with a first unprovoked episode of VTE. Cox proportional hazard models were used to analyze the effect of specific risk factors on the outcome of occult cancer within 12 months of a diagnosis of unprovoked VTE. Multivariable analysis was performed using Cox proportional hazard models that included all variables that achieved a p value of < 0.20 in univariate analyses.

Results: A total of 854 patients were randomized to limited occult cancer screening only, or limited occult cancer screening in combination with a cCT. The mean age was 54 years and 67.4% were males. A total of 33 (3.9%; 95% C.I. 2.8-5.4) patients received a new diagnosis of cancer at 12 months follow-up. Age ≥ 60 years, compared to age < 60 years, was a predictor of cancer with a corresponding hazard ratio (HR) of 2.90 (95% C.I. 1.44-5.83, p=0.003). A previous provoked VTE in patients was also associated with a higher risk of developing cancer (HR=3.57, 95% C.I. 1.38-9.25, p=0.009). Patients with an unprovoked deep vein thrombosis (DVT), compared to either those with a pulmonary embolism (PE) only or both DVT and PE, seemed more likely to have a diagnosis of cancer. However, this trend was not statistically significant. (Table 1) These results were confirmed on multivariable analysis. Patients exhibiting one of these characteristics had a three-fold higher risk of occult cancer compared with patients without these characteristics. (Table 1)

Conclusion: Age at unprovoked VTE diagnosis (≥ 60 years) and prior provoked VTE are predictors of occult cancer, and could potentially be used to identify a group of patients with unprovoked VTE at high risk of underlying cancer.

Table 1.  Risk factors of occult malignancy among patients with a first unprovoked symptomatic VTE.
	Patients without cancer (%) (n = 821)	Patients with cancer (%) (n = 33)	Univariate analysis Hazard Ratio (95% C.I.)	P value	Multivariable analysis  Hazard Ratio (95% C.I.)	P value
Age at diagnosis ≥ 60 years	288 (35.1)	20 (60.6)	2.90 (1.44-5.83)	0.003	3.0 (1.47-5.99)	0.002
Male sex	555 (67.6)	21 (63.6)	0.72 (0.35-1.46)	0.358	-	-
Prior provoked VTE	42 (5.1)	5 (15.2)	3.57 (1.38-9.25)	0.009	3.8 (1.46-10.03)	0.006
Type of current VTE
DVT only	444 (54.3)	24 (72.7)	1.91 (0.89-4.12)	0.097	2.1 (0.97-4.51)	0.061
PE only	271 (33.1)	7 (21.2)	0.60 (0.26-1.38)	0.229	-	-
DVT + PE	103 (12.6)	2 (6.1)	0.54 (0.13-2.24)	0.392	-	-
Baseline medications
Oral contraceptive pill	48 (5.8)	0 (0.0)	-	-	-	-
Exogenous estrogen	18 (2.2)	1 (3.0)	1.51 (0.21-11.07)	0.685	-	-
Antiplatelet agent	39 (4.8)	1 (3.0)	0.62 (0.09-4.56)	0.641	-	-
Oral anticoagulant	688 (83.8)	26 (78.8)	0.66 (0.29-1.53)	0.337	-	-
LMWH	391 (47.7)	15 (45.5)	0.68 (0.34-1.36)	0.275	-	-

VTE, venous thromboembolism; DVT, deep vein thrombosis; PE, pulmonary embolism; LMWH, low molecular weight heparin