Khorana Risk Score for VTE in Cancer Patients

Learn the Khorana score (0–7): tumor site, platelets, hemoglobin or ESAs, WBC, and BMI for ambulatory cancer patients starting chemotherapy. Risk strata, clinical use, and limitations.

Khorana VTE risk score

Enter pre-chemotherapy (or current outpatient) labs and BMI as used in the original model. Select the category that best matches the primary cancer site. ESAs include erythropoietin and similar agents.

Primary cancer site *

Platelet count (×10⁹/L) *

1 point if ≥350 ×10⁹/L

Hemoglobin (g/dL) *

1 point if <10 g/dL or ESA use (single point)

Leukocyte count (×10⁹/L) *

1 point if ≥11 ×10⁹/L

BMI (kg/m²) *

1 point if ≥35 kg/m²

Erythropoiesis-stimulating agent (ESA) in use? *

NoYes (counts as 1 point for the hemoglobin variable)

Disclaimer: Educational support only. The Khorana model estimates relative VTE risk in defined oncology populations; it is not a diagnosis and does not mandate treatment. Apply current guidelines, bleeding risk, and shared decision-making for thromboprophylaxis.

Khorana score for cancer-associated VTE

Overview

The Khorana score is a clinical prediction rule for symptomatic venous thromboembolism (VTE) in ambulatory adults with solid tumors or lymphoma who are starting chemotherapy. It combines cancer site with five laboratory and anthropometric variables available before treatment. The total ranges from 0 to 7.

Purpose: Stratify baseline VTE risk to inform discussion of thromboprophylaxis alongside bleeding risk and guideline recommendations—not as a stand-alone treatment mandate.

Variables and points

Variable	Criterion	Points
Site of primary tumor	Gastric or pancreatic Lung, lymphoma, myeloma, gynecologic, GU (renal, bladder, testicular; not prostate), other hematologic Breast, colorectal, head and neck, prostate, or other solid as appropriate	2 1 0
Platelet count	≥350 ×10⁹/L	1
Hemoglobin or ESAs	Hemoglobin <10 g/dL and/or use of erythropoiesis-stimulating agents	1 (max)
Leukocyte count	≥11 ×10⁹/L	1
BMI	≥35 kg/m²	1

Wording follows Khorana et al., J Clin Oncol 2008. Local laboratories may report platelets as ×10³/μL (350 ×10⁹/L ≈ 350 ×10³/μL).

Risk strata (original publication)

0 — Low risk
1–2 — Intermediate risk
≥3 — High risk

Observed VTE rates rise across these groups in the derivation/validation ambulatory chemotherapy cohorts; exact percentages depend on population and follow-up.

When the score is most appropriate

Oncology outpatients initiating systemic therapy where the original studies enrolled similar patients
Risk discussions combined with ASCO, NCCN, ISTH, or regional guidance on thromboprophylaxis
Research and quality metrics requiring a standardized VTE risk estimate

Limitations

Reduced accuracy in purely surgical cohorts, long-term hormonally treated breast cancer, or some novel immunotherapies.
Does not incorporate prior VTE, inherited thrombophilia, central lines, or immobility—factors that may independently raise risk.
Site assignment can be ambiguous for unknown primary or multiple primaries; document assumptions.
Intermediate scores (1–2) have variable management across guidelines; avoid over-interpreting small score differences.

Background: venous thromboembolism in people with cancer

Venous thromboembolism (VTE), including deep vein thrombosis and pulmonary embolism, occurs more often in patients with malignancy than in the general population. Active cancer, systemic therapy, surgery, hospitalization, immobility, and central venous access can all contribute to thrombotic risk. Because pharmacologic thromboprophylaxis carries bleeding and operational complexity, clinicians benefit from tools that stratify risk in defined populations so prophylaxis discussions can be targeted, proportional, and aligned with current guidelines.

The Khorana risk score is one of the most widely used clinical prediction rules for symptomatic VTE specifically in ambulatory adults with cancer who are beginning chemotherapy. It is not a universal “cancer VTE score” for every oncology scenario; its performance characteristics are tied to the types of patients enrolled in its derivation and validation workstreams.

What the Khorana score estimates

The model produces an integer from 0 to 7 by assigning points for the anatomic category of the primary tumor and for five routinely available baseline measures: platelet count, hemoglobin (or use of erythropoiesis-stimulating agents), leukocyte count, and body mass index (BMI). The score stratifies patients into low, intermediate, and high risk groups that, in the original ambulatory chemotherapy cohorts, corresponded to meaningfully different observed rates of symptomatic VTE over subsequent months of follow-up.

The score is intended to support shared decision-making about thromboprophylaxis in appropriate settings. It does not diagnose VTE, does not replace assessment of bleeding risk, renal function, drug interactions, or patient values, and should not be interpreted as a mandate to treat in isolation.

Score components (how each variable is counted)

1. Site of the primary tumor

Tumor site is weighted because certain primaries associate with higher baseline thrombotic risk in patients initiating cytotoxic or targeted therapy in outpatient regimens. In the original formulation, gastric and pancreatic cancers receive the highest site-related weight (2 points). A middle tier (1 point) includes categories such as lung cancer, lymphoma, multiple myeloma, gynecologic primaries, genitourinary sites such as renal, bladder, or testicular malignancy (with prostate cancer typically excluded from that intermediate tier in the classic mapping), and other hematologic malignancies (for example leukemia) when they fit the published grouping. Tumor types such as breast, colorectal, head and neck, and prostate cancer, along with other solid tumors that fall outside the higher-weight categories, are generally assigned 0 points for the site variable.

Practical tip: when the primary site is uncertain, metastatic disease of unknown primary, or multiple synchronous primaries, teams should document the mapping assumption used, because small changes in classification can change the score.

2. Platelet count

Thrombocytosis can reflect inflammatory or pro-thrombotic biology in advanced malignancy. The model awards 1 point when the pre-chemotherapy platelet count is at or above 350 × 10⁹/L (often reported in labs as 350 K/μL or similar, depending on local units). Values below that threshold contribute 0 points for this item.

3. Hemoglobin and erythropoiesis-stimulating agents (ESAs)

This item captures anemia burden and exposure to agents used to treat chemotherapy-associated anemia. The score assigns 1 point if hemoglobin is below 10 g/dL or if the patient is receiving an erythropoiesis-stimulating agent (for example erythropoietin or related therapies), reflecting the original variable definition. Importantly, only one point is assigned for this domain even if both low hemoglobin and ESA use are present; the intent is a single composite signal, not double counting.

4. Leukocyte count

Leukocytosis may track with systemic inflammatory state and has been incorporated as a simple laboratory risk signal. 1 point is assigned when the white blood cell count is at or above 11 × 10⁹/L on the pre-chemotherapy sample used for scoring.

5. Body mass index (BMI)

Obesity is associated with alterations in coagulation and inflammation and was included as an easily obtained anthropometric factor. 1 point is added when BMI is at or above 35 kg/m².

Risk strata and how they are used clinically

After summing points, patients are commonly grouped as follows in teaching and guideline discussions:

Low risk: total score 0
Intermediate risk: total score 1–2
High risk: total score ≥ 3

In the original ambulatory chemotherapy populations, VTE event rates increased across these strata. Exact event rates depend on the underlying cancer mix, chemotherapy exposures, intensity of ascertainment, and length of follow-up, so numeric risk estimates from one publication should not be over-interpreted as patient-specific probabilities without local validation context.

Many contemporary oncology pathways discuss outpatient thromboprophylaxis when the Khorana score is high, particularly in patients without unacceptably high bleeding risk and when anticoagulant choice and monitoring can be managed safely. Intermediate scores often prompt individualized review: some patients will have additional risk amplifiers (prior VTE, prolonged immobility, certain molecular therapies, central venous devices) that are not captured in the score itself. A low score reduces relative risk compared with higher strata but does not reduce risk to zero; surgery, hospitalization, infection, and drug interactions remain relevant.

Where the model fits—and where it does not

The Khorana score is best viewed as a baseline, pre-chemotherapy risk estimate for symptomatic VTE in patients similar to those studied in its development: primarily ambulatory adults initiating systemic therapy. Performance may differ in hospitalized patients, perioperative-only cohorts, patients receiving long-term endocrine therapy without cytotoxics, or regimens that were uncommon at the time of early validation.

The score also does not explicitly encode several factors clinicians routinely weigh, including a personal history of VTE, known thrombophilia, extensive residual disease burden, performance status, concurrent anticoagulation or antiplatelet therapy, renal or hepatic dysfunction affecting drug clearance, and social determinants that influence adherence or monitoring. For these reasons, the numeric result should be one input among many in a structured risk–benefit conversation.

Operational considerations for accurate scoring

Use a clear rule for which laboratory draw counts as “pre-chemotherapy” when therapy has already started or when labs are repeated frequently.
Confirm unit consistency for platelets and leukocytes (×10⁹/L versus ×10³/μL) before entering values into any calculator.
Document ESA exposure at the time of assessment; intermittent or planned future use should be handled per local protocol.
For BMI, use measured height and weight when possible; self-reported values can bias the score in either direction.
When tumor site mapping is ambiguous, record the rationale so subsequent providers can reproduce the score.

Disclaimer

This article supports education and documentation. It is not individualized medical advice. Thromboprophylaxis decisions must follow current specialty society guidance, institutional policy, bleeding-risk assessment, and the patient’s preferences and comorbidities.