Does Khorana Score Accurately Predict VTE in Patients With Uterine and Ovarian Cancer?

At the 2021 ASCO Annual Meeting, researchers shared results from two studies examining the Khorana score for predicting venous thromboembolism (VTE) in patients with gynecologic cancers.

The Khorana score assigns risk according to cancer type (with gynecologic cancers defined as high-risk cancers), hemoglobin level, body mass index, and pre-chemotherapy platelet and leukocyte counts. Patients with higher scores are more likely to experience VTE. Clinical practice guidelines recommend that patients with Khorana scores ≥2 who are on systemic chemotherapy receive thromboprophylaxis.

However, both studies found that the tool, which has been validated to identify patients with cancer who are at high risk for VTE and would benefit from thromboprophylaxis, did not accurately predict VTE in this patient population.

In the first study, Ellen Marcus, BS, a medical student from Albert Einstein College of Medicine and Montefiore Medical Center in New York, and colleagues, examined the predictive ability of the Khorana score in 427 patients who were diagnosed with ovarian cancer at her institution between 2000 and 2020. Scores were retrospectively calculated for all patients who received chemotherapy based on their pre-chemotherapy lab values.

Patients in the cohort had a median follow-up of 33.7 months (range = 13.3-61.1). During that time, 142 patients (31%) underwent diagnostic testing to rule out VTE, and 62 (15%) were diagnosed with VTE. The median time to VTE presentation was 8.7 months (range = 2.7-30.3). Factors significantly associated with VTE included stage III-IV disease, epithelial histology, open surgery, radical tumor debulking, and presence of residual disease after surgery.

The researchers found that, compared to patients with a score of 1, patients with Khorana scores of 2 or 3 were not significantly more likely to develop VTE during chemotherapy (odds ratios [ORs] = 1.73 and 0.89). However, patients with scores of 4 were nearly seven times more likely to develop VTE while on chemotherapy.

In addition, in a multivariable model, “a Khorana score of 2 or higher could explain only 0.86% of the variability in predicting VTE,” the authors wrote. Only stage III-IV disease and hyperlipidemia were significantly associated with VTE after adjustment.

Overall, the results from this analysis suggest that, although patients with ovarian cancer are at high risk of developing VTE many months after diagnosis and initiation of chemotherapy, a higher Khorana score did not predict the patients most likely to develop VTE. “Future models should be developed and validated in large population-based cohorts to determine if there is a more accurate strategy to identify women with ovarian cancer who are at risk for VTE,” the authors concluded.

In the second study, Vincent Wagner, MD, from The Ohio State University Wexner Medical Center and James Cancer Hospital in Columbus, and colleagues, reviewed the predictive value of the Khorana score in another retrospective cohort of 265 patients with uterine cancer who were receiving either neoadjuvant or adjuvant chemotherapy. Researchers evaluated the incidence of VTE in the 12 months following the first cycle of chemotherapy.

Most patients in the cohort were obese (60.4%) and ≥60 years old (70.9%). A total of 24 patients (9.1%) developed VTE, which was higher, but not statistically different, among patients receiving neoadjuvant compared with adjuvant chemotherapy (13.5% vs. 8.3%; p=0.35). The authors found no statistically significant correlations between the development of VTE and pathologic data, demographics, or treatment factors such as hysterectomy or radiation treatment.

The proportion of patients with high Khorana scores (defined as either ≥2 or ≥3) was similar between groups. In the whole cohort, high Khorana score did not significantly predict VTE; however, the model using ≥3 was more predictive than the ≥2 model typically used in clinical practice guidelines (OR=1.15; p=0.7326). For patients who received neoadjuvant chemotherapy, neither model was predictive of VTE, the researchers added, while a Khorana score ≥3 was a better prediction model in those who received adjuvant chemotherapy (OR=1.56; p=0.4213).

“Although validated in other cancer types, the Khorana score was found to be a poor predictor of VTE in this population,” the authors concluded. “At this time, use of the Khorana score to guide routine thromboprophylaxis in patients undergoing chemotherapy for uterine cancer should be used with caution.”

Study authors report no relevant conflicts of interest.


  1. Marcus E, Kuo DYS, Nevadunsky NS, et al. Association of the Khorana Score with development of venous thromboembolism in ovarian cancer. Abstract #5555. Presented at the 2021 ASCO Annual Meeting, June 4-8, 2021.  
  2. Wagner V, Levine M, Piver R, et al. Use of Khorana score to predict VTE in patients undergoing chemotherapy for uterine cancer. Abstract #5589. Presented at the 2021 ASCO Annual Meeting, June 4-8, 2021.