Predictive Efficacy of the Khorana Score in Black Patients: Retrospective Cohort Study

Background

The Khorana score(KS) is a predictive score used to determine the risk of venous thromboembolism (VTE) in cancer(CA) patients(pts). According to the original study, a score of 1-2 was consistent with an intermediate risk(IR) for VTE (VTE risk(VTER)) of 1.8-2.0% at 2.5 months). A score of >3 was associated with a high risk(HR) of VTE (6.7-7.1% VTER at 2.5 months). Several studies have noted an increased risk of VTE in Black CA pts. The original KS study did not stratify pts based on race. We sought to determine if the Khorana score was appropriately predictive of VTE in Black CA pts at our institution, using pre-chemotherapy lab values to calculate the score.

Methods

Retrospective study. Inclusion: any CA between 01/01/2015-12/31/2021, aged 18-100 years. Exclusion: history of anticoagulation 6 months prior to CA diagnosis, history of chronic VTE, history of arterial thromboembolism. Follow-up period: 1 year post cancer diagnosis or until death if sooner.

Discussion

The population was composed of 233 patients. Black pts(AA) composed 39.9%; White pts(WP) composed 60.1%. Mean age of the population was 61.3 years (standard deviation (SD) 12.4). Male patients composed 49.8%. At the time of the study 39.9% were deceased; the mean duration from CA to death was 14.5 months (SD 15.0). Most common CA stage was stage IV (21.03%); breast cancer was most prevalent (15.9%). The median follow-up time was 12.0 months (IQR 7.9). VTE occurred in 39.9%; mean cancer-to-VTE duration was 4.4 months (SD 3.2). The mean cancer-to-VTE duration in AA was 4.5 months (SD 3.4); 4.3 months (SD 3.2) in WPs.

The average Charlson comorbidity index(CCI) for the population was 3.3 (SD 2.9); there were no differences between groups when CCI was evaluated (Table 1).

The population mean Khorana score was 1.3 (SD 1.2). The Khorana score for: AA was 1.4 (SD 1.1) and 1.3 (SD 1.2) in WPs. The average Khorana scores were: AA without VTE 1.4 (SD 1.1), AA with VTE 1.5 (SD 1.1), WP without VTE 1.1 (SD 1.1), WP with VTE 1.5 (SD 1.2). There were no differences between any of these average Khorana scores (Table 1). The average population Khorana score for VTE regardless of race was 1.5 (SD 1.2) and 1.3 (1.2) for patients without VTE; this was not a significant finding (p=0.130).

The positive predictive value (PPV) of the KS was determined for each KS group by race (i.e., HR score >3, IR score 1-2, low risk(LR) score 0). AA: HR PPV 13%, IR PPV 31%, LR PPV 17%. WP: HR PPV 65%, IM PPV 54%, LR PPV 34%. These values were significantly different between AA and WP for all risk groups (p<0.001) except LR (p=0.11).

Distant metastasis occurred in 22/233 (9.4%) of the total study population. More patients with VTE had distant metastasis (16/93, 17.2%) than patients without VTE (6/140, 4.3%); (OR: 4.641, 95% CI 3.7 -5.6; p=0.001).

Conclusion

Most patients developed VTE around 4 months following CA diagnosis. Our study shows that the KS may not be a good predictor of VTE in AA. While there was a stepwise increase in the PPV of the KS for VTE in WPs, such a pattern was not seen in AA. Pts with distant metastasis were 4.6 times more likely to have VTE.