Abstract
Introduction
Several risk scores have been proposed to predict venous thromboembolism (VTE) in hospitalized patients. However, their predictive performances in lung cancer patients receiving immune checkpoint inhibitors (ICIs) is unclear. We aimed to validate and compare their performances of the Caprini, Padua and Khorana risk scores in lung cancer patients receiving ICIs.
Methods
This was a retrospective cohort study of patients with lung cancer treated with ICIs at West China Hospital between January 2018 and March 2022. The primary outcome was VTE during 12 months of follow-up from the first day of treatment with ICIs. The predictive performances of risk scores was determined using receiver operating characteristic (ROC) curve analysis.
Results
Among the 1115 eligible patients with lung cancer who received ICIs, 105 patients (9.4%) experienced VTE during the 12-month follow-up period. There was a statistically significant difference in the cumulative incidence of VTE between the different risk levels as determined by Caprini and Padua scores (all P < 0.001). However, no significant difference was observed for the Khorana score (P = 0.488). The Caprini and Padua scores demonstrated good discriminative performances (AUC 0.743, 95% CI 0.688-0.799 for Caprini score; AUC 0.745, 95% CI 0.687‐0.803 for Padua score), which were significantly better than that of the Khorana score (AUC 0.553, 95% CI, 0.493‐0.613) (P < 0.05).
Conclusion
In our study, the Caprini and Padua risk scores had better discriminative ability than the Khorana score to identify lung cancer patients treated with ICIs who were at high risk of VTE.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12890-024-03323-z.
Keywords: Lung cancer, Immune checkpoint inhibitors, Venous thromboembolism, Risk assessment scores
Introduction
Lung cancer is one of the most prevalent malignancies and is the primary cause of cancer death worldwide [1]. In recent years, immune checkpoint inhibitors (ICIs), a new milestone for cancer treatment, mainly target the protein programmed cell death 1 (PD-1) receptor and its ligand (PD-L1), which have made important and impactful contributions to cancer medicine [2–5]. A subset of lung cancer patients receiving ICIs, either used alone or in combination with chemotherapy, have a realistic chance of achieving better overall survival and response rates [5]. Venous thromboembolism (VTE) is a frequent complication of lung cancer, with reported 12-month incidence rates as high as 6.5-13.1% [6–8]. In addition, the reported incidence of VTE varies from 4.5 to 17.8% in hospitalized patients with lung cancer receiving ICIs [9–11]. The development of VTE strongly correlates with a poor prognosis, leading to more severe health conditions and worse overall survival [12].
VTE, consisting of deep venous thrombosis (DVT) and pulmonary embolism (PE), is preventable, with common prophylaxes including mechanical methods and pharmacological agents [13]. The American College of Chest Physicians (ACCP) and the National Comprehensive Cancer Network (NCCN) guidelines both recommend thromboprophylaxis for lung cancer patients who are at high risk of VTE [14, 15]. However, the VTE thromboprophylaxis for lung cancer patients may come with some side effects, such as major bleeding and heparin-induced thrombocytopenia. Therefore,
the evaluation of the risk of VTE is essential to select an appropriate prophylaxis strategy in lung cancer patients receiving ICIs.
Several risk assessment scores for VTE have been derived and validated in hospitalized patients. The Caprini, Padua and Khorana scores are the most widely studied, but there has been no consensus on which one is the best in patients with lung cancer [16–18]. The Caprini risk assessment score was originally developed and validated for predicting postoperative VTE in surgical patients. It contained 36 items and has been adopted to cancer patients [19–21]. The Padua score could help select patients at particularly high risk of VTE and has been recommended to evaluate the VTE risk among medical inpatients by the ACCP guideline [14]. However, the small number of VTE events in the derivation cohort and suboptimal validation led to limited clinical application of this risk score [22]. The Khorana score identified 5 different variables that were found to be independent predictors of chemotherapy-associated VTE in cancer patients [18]. Guidelines suggested that cancer patients with a Khorana score of 2 or higher might need primary pharmacological prophylaxis for VTE [23]. However, the predictive performances of these risk scores were not clear in patients with lung cancer receiving ICIs.
In this study, we aimed to validate and compare the performances of the Caprini, Padua and Khorana risk scores in patients with lung cancer receiving ICIs.
Methods
Study design and patients
This was a retrospective cohort study of lung cancer hospitalized patients at West China Hospital between January 2018 and March 2022. The inclusion criteria were as follows: histologically confirmed lung carcinoma, age ≥ 18 years, and had received at least one dose of an approved ICIs. The exclusion criteria were as follows: diagnosis of VTE within 48 hours following admission, receiving thromboprophylaxis or treatment with antiplatelets when treated with ICIs, and incomplete follow-up data. This study was approved by the Ethics Committee on Biomedical Research, West China Hospital of Sichuan University in accordance with the Declaration of Helsinki. As all analyses were conducted using anonymized, de-identified patient data, the requirement for written informed consent was waived.
Data collection
The following parameters were collected: demographics (age, gender, body mass index (BMI)), cancer-related variables (past or active smoking status, histological subtype, TNM stage, Eastern Cooperative Oncology Group Performance Status (ECOG-PS), PD-L1 status, number of distant metastatic sites, molecular alterations, cancer treatments), medical history (hypertension, diabetes, hyperlipemia, abnormal pulmonary function, severe lung disease, respiratory failure, atrial fibrillation, rheumatological disorder, arterial vascular events), variables associated with VTE risk (varicose veins, swollen legs, immobilization, recent (≤ 1 month) surgery, history of VTE), and laboratory findings (hemoglobin, white blood cell (WBC), platelet count). Patients were followed from the first day of treatment with ICIs until 12 months or death by rehospitalization, outpatient visits, or telephone.
Risk assessment scores
For the present study, the Caprini, Padua and Khorana risk scores were calculated using available data by trained physicians. The risk factors included in the three risk assessment scores, the points assigned, and the risk stratification were depicted in Supplemental Table S1. Based on commonly accepted definitions, patients with a Caprini score of 0 were at very low risk, 1–2 were low risk, 3–4 were moderate risk and ≥ 5 were high risk. The Padua score classified patients into low-risk (0–3) and high-risk (≥ 4) groups based on 11 risk factors. The Khorana risk score is composed of five objective parameters: type of cancer (1 score for lung, lymphoma, gynecologic, bladder, testicular, or 2 scores for stomach, pancreas), pre-treatment platelet counts ≥ 350 × 109/L (1 score), hemoglobin < 10 g/dL (1 score), pre-treatment WBC count > 11 × 109/L (1 score), and BMI ≥ 35 kg/m2 (1 score). Cancer patients were categorized into low-risk (score 0), intermediate-risk (score 1–2), and high-risk (score ≥ 3) groups based on the Khorana risk score.
Study outcomes
The main study outcome was VTE during 12 months of follow-up from the first day of treatment with ICIs. VTE was defined as symptomatic or asymptomatic, PE or DVT based on standardized radiographic findings. The clinical records and radiology reports of included patients were reviewed during follow-up.
Statistical analysis
Categorical variables were presented as frequencies and percentages and the chi-squared test was used to analyze group differences. Continuous variables were expressed as mean values with standard deviations or median plus quartile interval and were compared by the independent Student’s t test (normal distribution) and the Mann–Whitney U test (skewed distribution). The risk factors for VTE in the cohort were explored by multivariate logistic regression and the odds ratio (OR) values and 95% confidence interval (CI) were reported. Kaplan–Meier curves were used to describe the time-courses for the occurrence of VTE after receiving ICIs in lung cancer patients with different risk levels by the Caprini, Padua and Khorana risk scores, and the log-rank test was used to make group comparisons. We estimated the discriminative power of each score in predicting VTE by calculating the area under the receiver operating characteristic curve (AUC). All collected data were analyzed with SPSS (Version 25.0), and a P value < 0.05 indicated a statistically significant difference.
Results
Patient characteristics and VTE incidence
Overall, 1253 patients with lung cancer receiving ICIs were initially identified for inclusion in this study, of which 138 patients were excluded for the following reasons: (1) diagnosis of VTE within 48 hours of admission (n = 39), (2) received thromboprophylaxis or treated with antiplatelets when treated with ICIs (n = 10), and (3) incomplete follow-up data for 12 months (n = 89). Ultimately, 1115 patients were enrolled (Fig. 1). Overall, 105 patients (9.4%) experienced VTE during the 12-month follow-up period, of which 10 (9.5%) had definite isolated PE, 91 (86.7%) had isolated distal DVT and 4 (3.8%) had DVT complicated with PE. The median time from initial ICI treatment until VTE occurrence was 3 months.
Fig. 1.
Flow chart of the study. (Abbreviations: ICIs = immune checkpoint inhibitors; VTE = venous thromboembolism; PE = pulmonary embolism; DVT = deep venous thrombosis)
Table 1 showed the baseline characteristics of patients with VTE and non-VTE cases. The mean age at diagnosis was 61.70 ± 9.36 years, and 942 (84.5%) patients were men. Patients with adenocarcinoma tended to have a higher prevalence of VTE within one year, whereas the opposite tendency was observed in squamous cell carcinoma. Patients who had PS ECOG ≥ 2 at start had an higher probability of developing VTE at 12 months. There was no significant difference in age, sex, BMI, smoking status, TNM stage, or PD-L1 expression between the two groups. Patients with higher number of metastatic sites before ICIs start and higher number of previous lines had a higher probability of developing VTE within 12 months. Among molecular subtypes, the KRAS-mut subset showed a significantly higher prevalence of VTE within one year . Additionally, among patients previously treated with platinum-based chemotherapy, and with bevacizumab, the prevalence of VTE was higher (both P < 0.05). VTE within one year was more prevalent in patients with abnormal pulmonary function, severe lung disease, and respiratory failure (all P < 0.05). In terms of variables associated with VTE risk, known risk factors for VTE as swollen legs, immobilization and personal history of VTE were associated with higher risk of VTE at one year. With respect to laboratory tests, VTE tended to occur more frequently in patients with anemia (defined as hemoglobin < 10 g/dL) and hyperleukocytosis (defined as WBC > 11 × 10⁹/L). The median risk scores were significantly higher in the VTE group for all the risk assessment scores (all P < 0.05).
Table 1.
Patient characteristics at baseline
| Characteristics | All, N = 1115 | Venous thromboembolism | P-Value | |
|---|---|---|---|---|
| No, N = 1010 | Yes, N = 105 | |||
| Age, y | 61.70 ± 9.36 | 61.63 ± 9.41 | 62.37 ± 8.93 | 0.439 |
| Age ≥ 70 y | 215(19.3) | 194(19.2) | 21(20.0) | 0.845 |
| Male | 942(84.5) | 855(84.7) | 87(82.9) | 0.628 |
| BMI | 23.05 ± 3.19 | 23.06 ± 3.21 | 22.99 ± 3.06 | 0.827 |
| BMI ≥ 30 kg/m2 | 23(2.1) | 22(2.2) | 1(1.0) | 0.716 |
| Smoking (current or past) | 719(65.0) | 646(64.5) | 73(69.5) | 0.308 |
| Histological subtype | ||||
| Squamous cell carcinoma | 416(37.3) | 387(38.3) | 29(27.6) | 0.031 |
| Adenocarcinoma | 427(38.3) | 376(37.2) | 51(48.6) | 0.023 |
| Small cell Carcinoma | 197(17.7) | 182(18.0) | 15(14.3) | 0.340 |
| Other | 75(6.7) | 65(6.4) | 10(9.5) | 0.229 |
| TNM stage | 0.110 | |||
| II-III | 442(39.6) | 408(40.4) | 34(32.4) | |
| IV | 673(60.4) | 602(59.6) | 71(67.6) | |
| ECOG PS | <0.001 | |||
| 0–1 | 1037(93.0) | 960(95.0) | 77(73.3) | |
| ≥ 2 | 78(7.0) | 50(5.0) | 28(26.7) | |
| PD-L1 (TPS) | 0.063 | |||
| < 50% | 286(84.4) | 255(60.9) | 31(75.6) | |
| ≥ 50% | 174(15.6) | 164(39.1) | 10(24.4) | |
| Number of distant metastatic sites | 0.014 | |||
| 0–1 | 766(68.7) | 705(69.8) | 61(58.1) | |
| ≥ 2 | 349(31.3) | 305(30.2) | 44(41.9) | |
| Molecular driver | ||||
| KRAS+ | 94(20.0) | 77(18.5) | 17(32.1) | 0.020 |
| ALK+ | 4(0.9) | 3(0.7) | 1(1.9) | 0.939 |
| EGFR+ | 76(16.2) | 66(15.9) | 10(18.9) | 0.576 |
| ROS1+ | 2(0.4) | 2(0.5) | 0(0.0) | 1.000 |
| ICI regimen given in line | 0.019 | |||
| 1 | 731(65.6) | 673(66.6) | 58(55.2) | |
| ≥ 2 | 384(34.4) | 337(33.4) | 47(44.8) | |
| Previous treatment before ICIs | ||||
| Platinum-based chemotherapy | 298(26.7) | 257(25.4) | 41(39.0) | 0.003 |
| Chest radiotherapy | 86(7.7) | 75(7.4) | 11(10.5) | 0.265 |
| Surgical resection | 143(12.8) | 128(12.7) | 15(14.3) | 0.638 |
| Targeted therapy | 85(7.6) | 75(7.4) | 10(9.5) | 0.441 |
| Bevacizumab | 27(2.4) | 20(2.0) | 7(6.7) | 0.003 |
| Comorbidities | ||||
| Hypertension | 269(24.1) | 241(23.9) | 28(26.7) | 0.523 |
| Diabetes | 144(12.9) | 132(13.1) | 12(11.4) | 0.633 |
| Hyperlipemia | 153(13.8) | 137(13.6) | 16(15.2) | 0.647 |
| Abnormal lung function | 19(1.7) | 13(1.3) | 6(5.7) | 0.001 |
| Severe lung disease | 31(2.8) | 16(1.6) | 15(14.3) | <0.001 |
| Respiratory failure | 21(1.9) | 9(0.9) | 12(11.4) | <0.001 |
| Atrial fibrillation | 19(1.7) | 16(1.6) | 3(2.9) | 0.413 |
| Rheumatological disorder | 4(0.4) | 4(0.4) | 0(0.0) | 1.000 |
| Arterial vascular events | 71(6.4) | 64(6.3) | 7(6.7) | 0.895 |
| Risk factors for VTE | ||||
| Varicose veins | 27(2.4) | 24(2.4) | 3(2.9) | 0.735 |
| Swollen legs | 14(1.3) | 7(0.7) | 7(7.6) | <0.001 |
| Immobilization | 12(1.1) | 4(0.4) | 8(7.6) | <0.001 |
| Recent (≤ 1 month) surgery | 11(1.0) | 8(0.8) | 3(2.9) | 0.077 |
| History of VTE | 34(3.0) | 13(1.3) | 21(20.0) | <0.001 |
| Central venous access | 561(50.3) | 503(49.8) | 58(55.2) | 0.289 |
| Laboratory tests | ||||
| Hemoglobin (g/L) | 125.69 ± 18.43 | 126.31 ± 18.12 | 119.70 ± 20.38 | <0.001 |
| Hemoglobin < 10 g/dL | 95(8.5) | 79(7.8) | 16(15.2) | 0.010 |
| WBC(×10⁹/L) | 7.08 ± 2.93 | 7.07 ± 2.88 | 7.13 ± 3.35 | 0.862 |
| WBC > 11 × 10⁹/L | 88(7.9) | 73(7.2) | 15(14.3) | 0.011 |
| Platelet count(×10⁹/L) | 229.27 ± 95.05 | 229.79 ± 93.25 | 224.31 ± 111.26 | 0.575 |
| Platelet count ≥ 350 × 10⁹/L | 95(8.5) | 86(8.5) | 9(8.6) | 0.987 |
| Caprini score | 5(4–6) | 5(4–6) | 6(5–7) | <0.001 |
| Padua score | 3(3–4) | 3(3–4) | 4(3–6) | <0.001 |
| Khorana score | 1(1–1) | 1(1–1) | 1(1–2) | 0.034 |
Note Categorical variables were presented as number of patients (%); Continous variables were presented as mean (standard deviation) or median (interquartile range). Severe lung disease included symptomatic asthma, bronchiectasis, idiopathic pulmonary fibrosis, pneumoconiosis, active or destructive pulmonary tuberculosis, and pulmonary arterial hypertension
Abbreviations BMI = body mass index; ECOG PS = Eastern Cooperative Oncology Group Performance Status; PD-L1 = programmed cell death ligand 1; ICIs = immune checkpoint inhibitors; VTE = venous thromboembolism; WBC = white blood cell
Risk factors for VTE after initiation of ICIs
The clinical risk factors for VTE in lung cancer patients with ICIs therapy was explored. We performed a multivariate logistic regression analysis with histological subtype, ECOG PS, number of distant metastatic sites, previous treatment before ICIs, abnormal pulmonary function, severe lung disease, respiratory failure, swollen legs, immobilization, history of VTE, hemoglobin < 10 g/dL, and WBC > 11 × 10⁹/L as candidate predictive factors based on the results of the univariate analysis (all P < 0.05). We found that ECOG PS ≥ 2 (OR = 4.465; 95% CI: 2.393–8.331, P<0.001), previous platinum-based chemotherapy (OR = 1.759; 95% CI: 1.058–2.923, P = 0.029), severe lung disease (OR = 6.906; 95% CI: 2.774–17.195, P<0.001), respiratory failure (OR = 3.835; 95% CI: 1.262–11.651, P = 0.018), swollen legs (OR = 4.980; 95% CI: 1.354–18.313, P = 0.016), immobilization (OR = 12.122; 95% CI: 2.490-59.016, P = 0.002) and history of VTE (OR = 19.380; 95% CI: 8.723–43.058, P<0.001) were independent predictors of VTE (Table 2).
Table 2.
Multivariate analysis for predictors of VTE in lung cancer patients receiving ICIs
| Risk factors | OR | 95% CI | P-Value |
|---|---|---|---|
| Adenocarcinoma | 1.375 | 0.850–2.225 | 0.195 |
| ECOG PS ≥ 2 | 4.465 | 2.393–8.331 | <0.001 |
| Number of distant metastatic sites ≥ 2 | 1.351 | 0.824–2.215 | 0.233 |
| Previous chemotherapy | 1.759 | 1.058–2.923 | 0.029 |
| Abnormal lung function | 2.862 | 0.766–10.700 | 0.118 |
| Severe lung disease | 6.906 | 2.774–17.195 | <0.001 |
| Respiratory failure | 3.835 | 1.262 − 11.651 | 0.018 |
| Swollen legs | 4.980 | 1.354–18.313 | 0.016 |
| Immobilization | 12.122 | 2.490–59.016 | 0.002 |
| History of VTE | 19.380 | 8.723–43.058 | <0.001 |
| Hemoglobin < 10 g/dL | 1.209 | 0.582–2.514 | 0.611 |
| WBC > 11 × 10⁹/L | 1.827 | 0.866 − 3.855 | 0.114 |
Abbreviations VTE = venous thromboembolism; ICIs = immune checkpoint inhibitors; ECOG PS = Eastern Cooperative Oncology Group Performance Status; OR = odd ratio; CI = confidence interval
Incidence of VTE by risk levels and the three risk assessment scores
Table 3 showed the incidence of VTE based on risk levels of Caprini, Padua and Khorana scores in the cohort. For the Caprini risk assessment score, 0 (0%) patients were in the lowest-risk group, 5 (0.4%) patients were in the low-risk group, 370 (33.2%) patients were in the moderate-risk group, and 740 (66.4%) patients were in the high-risk group. The respective VTE incidence in the different risk levels were 0%, 3.5%, and 12.4%. There was a significant difference in the increased VTE incidence (P < 0.001). For the Padua risk score, 54.3% of the study population were classified as low risk with 0–3 score, and 45.7% as high risk with more than 4 cumulative risk score. A higher risk level indicated a greater incidence of acquired VTE within 12 months: 4.5% of low-risk patients acquired a VTE; among the high-risk patients, 15.5% acquired a VTE (P < 0.001). However, based on the Khorana score, 95.6% of the study patients with lung cancer were categorized as moderate risk, and 9.3% acquired VTE. There were no statistically significant differences in VTE incidence between moderate-risk and high-risk levels for the Khorana score (P = 0.488) (Fig. 2).
Table 3.
Incidence of 12-month VTE based on risk levels of Caprini, Padua and Khorana scores
| Score | Risk level | n, % | 12-month VTE | P-Value |
|---|---|---|---|---|
| Caprini | ||||
| Lowest risk (0) | 0 (0%) | - | < 0.001 | |
| Low risk (1–2) | 5(0.4%) | 0/5(0.0%) | ||
| Moderate risk (3–4) | 370(33.2%) | 13/370(3.5%) | ||
| High risk (≥ 5) | 740(66.4%) | 92/740(12.4%) | ||
| Pauda | ||||
| Low risk (0–3) | 605(54.3%) | 26/605(4.5%) | < 0.001 | |
| High risk (≥ 4) | 510(45.7%) | 79/510(15.5%) | ||
| Khorana | ||||
| Low risk (0) | 0 (0%) | - | 0.488 | |
| Moderate risk (1–2) | 1066(95.6%) | 99/1066(9.3%) | ||
| High risk (≥ 3) | 49(4.4%) | 6/49(12.2%) |
Abbreviations VTE = venous thromboembolism
Fig. 2.
Incidence of 12-month VTE by risk levels of Caprini, Padua and Khorana scores. (a) Caprini score (b) Padua score (c) Khorana score. (Abbreviations: VTE = venous thromboembolism)
Performances of risk assessment scores in the overall cohort
The 12-month cumulative incidences of VTE based on the risk levels of Caprini, Padua and Khorana risk scores were shown in Fig. 3. In the time-to-event analysis, there was a significant difference in the cumulative incidence of VTE between patients with different risk levels by Caprini and Padua scores (all P < 0.001). However, for the Khorana score, the cumulative incidence of VTE did not show a significant difference between the patients with moderate-risk scores and those with high-risk scores (P > 0.05).
Fig. 3.
Cumulative incidence of VTE in patients with different risk levels of Caprini, Padua and Khorana scores (P-value by the log-rank test for comparison of the outcome among groups at 12 months) (a) Caprini score (b) Padua score (c) Khorana score. (Abbreviations: VTE = venous thromboembolism)
Figure 4 presented the discriminative performances of the Caprini, Padua and Khorana scores. The AUC decreased for almost all scores compared with the original cohorts, with AUCs ranging from 0.553 to 0.745. The Caprini and Padua scores yielded good discriminative performances (AUC 0.743, 95% CI 0.688 - 0.799 for Caprini score; AUC 0.745, 95% CI 0.687 ‐ 0.803 for Padua score). The AUC of the Khorana score was 0.553 (95% CI, 0.493 ‐ 0.613), which was significantly lower than that of the Caprini and Padua scores (P < 0.05).
Fig. 4.
Receiver operating characteristic (ROC) curve of the Caprini, Padua and Khorana scores
Performances of risk assessment scores in the subgroup of ICIs monotherapy
A total of 152 patients received ICIs monotherapy in the study; 11 (7.2%) experienced VTE during the 12-month follow-up period, of which 1 (9.1%) had isolated PE, 9 (81.8%) had isolated distal DVT and 1 (9.1%) had DVT complicated with PE. The median time until VTE occurrence was 5 months. Based on the ROC curve analysis, the AUC was 0.762 (95% CI, 0.583–0.942) for the Caprini score, which exhibited good performance in predicting 12-month VTE in lung cancer patients receiving ICIs monotherapy. However, the discrimination performances of the Padua and Khorana scores for predicting VTE were poor (AUC 0.656, 95% CI 0.458-0.854 for Padua score; AUC 0.549, 95% CI 0.367‐0.731 for Khorana score) (Supplementary Figure S1).
Discussion
In this study, we validated and compared the performances of the Caprini, Padua and Khorana risk scores for predicting VTE in patients with lung cancer receiving ICIs. We found that the Caprini and Padua scores performed better than the Khorana score in identifying patients who were at high risk of VTE. Additionally, the Caprini score showed the best accuracy for predicting 12-month VTE in lung cancer patients receiving ICIs monotherapy. These findings indicated that the Khorana score may be not suitable for evaluating VTE in lung cancer patients receiving ICIs. On the other hand, the Caprini score may be useful in predicting VTE in lung cancer patients receiving ICIs monotherapy.
The incidence of VTE was reported to be relatively high in cancer patients treated with ICIs, most likely because ICIs enhanced the prothrombotic state by activation of coagulation and platelets and impairment of fibrinolysis [24–26]. An Israeli observational cohort study observed that the 6-month cumulative incidence of VTE in a cohort of 176 lung cancer patients receiving single-agent ICIs was 4.5% [9]. In a large investigation using HealthCore Integrated Research Environment - Oncology data, the VTE incidence rate (95% CI) per 100 person-years was 17.8 (95% CI 16.0 to 19.5) with a median follow-up of 9.1 months [11]. In our study, the 12-month VTE incidence was 9.4% in lung cancer patients who received immunotherapy. This was consistent with the findings of a recent monocentric observational study conducted in France, which reported that the incidence of VTE was 9.8% in lung cancer patients treated with ICIs [27]. VTE development during ICIs treatment had significant impact on overall survival. Therefore, more attention should be focused on VTE during ICIs therapy in order to optimize both the prevention and management of VTE.
Most guidelines recommend risk stratification for VTE using clinical risk scores in hospitalized cancer patients, followed by appropriate prophylaxis measures [14, 15]. Several risk assessment scores have been proposed and validated for risk stratification in cancer patients. The most widely used clinical risk scores were Caprini, Padua and Khorana scores, but they were not equivalent. The three risk scores for VTE were targeted to different populations, the Caprini score concerned patients undergoing surgery, Padua score was realized for inpatients, whereas Khorana score for ambulatory patients with cancer [17, 18, 28]. The Caprini score was originally developed for surgical patients and could be used to risk stratify patients effectively [29]. Zhou et al.. attempted to preliminarily assess the predictive ability of the Caprini risk assessment score among unselected hospitalized patients with VTE and found that the score provided an accurate, practical, and effective method for selection of high-risk patients who may benefit from thromboprophylaxis [30]. However, Paul et al. validated the Caprini score in 63,548 medical patients across 48 Michigan hospitals, and found that the Caprini risk assessment score had limited ability to identify medical patients who were at high risk for VTE [31]. Barbar et al. developed the Padua risk assessment score based on a single cohort study conducted in Italy [17]. Some flaws in this score have been mentioned by some researchers, for example, there were limited VTE events, the model performance characteristics were not presented, and external validation is suboptimal [32, 33]. The Khorana score, an established tool, helps identify patients receiving chemotherapy who were at risk for VTE [18]. The score is heavily weighted on the type of cancer; thus, all patients with lung cancer were classified as (at least) moderate risk for VTE [34]. The three risk scores are widely used in clinical practice, but it is not yet clear which one is the best among lung cancer patients receiving ICIs. In our study, we compared the performance of these risk scores and found that the Caprini and Padua scores performed better than the Khorana score in identifying lung cancer patients receiving ICIs who were at high risk of VTE. In this cohort of lung cancer patients receiving ICIs, the incidence of VTE increased with higher risk levels according to the Caprini and Padua risk scores. These results revealed that both the Caprini and Padua risk scores could be effectively used to stratify VTE risk among lung cancer patients receiving ICIs. Additionally, for the subgroup of lung cancer patients receiving ICIs monotherapy, the Caprini score showed the best accuracy for predicting VTE compared with the other scores. There were mainly three explanations for this phenomenon. Firstly, the Caprini score covered a broad range of risk factors and provided a personalized and quantifiable strategy for assessing VTE risk. The independent risk factors for VTE in patients with lung cancer receiving ICIs, such as swollen legs, severe lung disease, immobilization, and history of VTE were included in the Caprini score. Secondly, approximately 28.4–39.8% of lung cancer patients were complicated with chronic obstructive pulmonary disease (COPD) [35]. The factor of COPD was unique to the Caprini score, which was also found to be an independent risk factor for VTE in patients with lung cancer [36]. Third, the Caprini score performed a more detailed stratification of VTE risk, which might obtain a better effect on thromboprophylaxis. These findings suggested that the Caprini score may be applied as a useful tool in lung cancer patients receiving ICIs monotherapy.
The Khorana score has been used for VTE risk assessment in cancer patients, but its predictive validity to identify high-risk patients among lung cancer patients treated with ICIs remains unclear [34, 37, 38]. Recent studies have acknowledged that the Khorana score is associated with VTE in patients treated with ICIs. For example, Gong et al. found that patients with VTE had a higher Khorana risk score, and a Khorana score ≥ 2 could indicate cancer patients at high risk of VTE for thromboprophylaxis [39]. In a retrospective cohort study of cancer patients receiving ICIs therapy with a median follow-up of 27.3 months, a Khorana score ≥ 1 was reported to be significantly associated with VTE [40]. In contrast, the Khorana score showed poor performance with poor accuracy in our study population. It may be due to the difference in the study population, our study only enrolled hospitalized cancer patients but did not include outpatients. Alma et al. reached the same conclusion with our study [27]. No significant difference was observed in the rate of VTE between patients with a Khorana score < 2 and ≥ 2, the Khorana score did not effectively predict VTE in lung cancer patients receiving ICIs [27]. Thus, it is crucial to improve the Khorana score for VTE risk stratification and to guide primary prevention strategies.
Strengths and limitations
To our knowledge, this study was the first attempt to validate and compare the Caprini, Padua and Khorana risk scores among patients with lung cancer receiving ICIs. The data quality in our study was relatively high because of large number of patients and integral 12-month follow-up data. Nevertheless, our study was associated with several limitations. First, this was a retrospective cohort study of patients with lung cancer, and some variables, including PD-L1 expression levels and laboratory tests, were not available for all patients. Fortunately, the proportion of patients was small, and the impact on our results can be neglected. Second, our study was performed in a single center, and the majority of patients we included had previously received chemotherapy for locally advanced or metastatic cancer; thus, its generalizability might be limited. Therefore, multicenter prospective studies should be conducted to further validate the performance of existing risk stratification tools in lung cancer patients receiving ICIs.
Conclusion
In summary, the Caprini and Padua scores showed better performances in predicting the risk of VTE in lung cancer patients receiving ICIs than the Khorana score. Further studies were needed to develop and validate an ICI-specific risk assessment score in patients with lung cancer.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Acknowledgements
Thanks to all authors for their contributions to this manuscript.
Abbreviations
- VTE
venous thromboembolism
- ICIs
immune checkpoint inhibitors
- PD-1
programmed cell death 1
- PD-L1
programmed cell death ligand 1
- DVT
deep venous thrombosis
- PE
pulmonary embolism
- ACCP
American College of Chest Physicians
- NCCN
National Comprehensive Cancer Network
- BMI
body mass index
- ECOG-PS
Eastern Cooperative Oncology Group Performance Status
- WBC
white blood cell
- OR
odds ratio
- 95% CI
95% confidence interval
- AUC
area under the receiver operating characteristic curve
Author contributions
JZ and YX contributed to the curation and analysis of the data and drafted the manuscript. DL contributed to the design of the study and revised or critically reviewed the article. LY, MY, and RX all contributed to the acquisition and interpretation of the data and gave final approval of the version to be published.
Funding
This study was supported by the National Natural Science Foundation of China (355 82173182) and the Sichuan Science and Technology Program (2023NSFSC1939).
Data availability
The original data of this manuscript are available on reasonable request by contacting the correspondence author.
Declarations
Ethics approval and consent to participate
This study was approved by the institutional review board of the West China Hospital of Sichuan University in accordance with the Declaration of Helsinki. The informed consent was waived by the Ethics Review Committee of West China Hospital of Sichuan University.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Jiarui Zhang and Yufang Xie contributed equally to this work.
References
- 1.Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global Cancer statistics 2020: GLOBOCAN estimates of incidence and Mortality Worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71(3):209–49. [DOI] [PubMed] [Google Scholar]
- 2.Rittmeyer A, Barlesi F, Waterkamp D, Park K, Ciardiello F, von Pawel J, Gadgeel SM, Hida T, Kowalski DM, Dols MC, et al. Atezolizumab versus Docetaxel in patients with previously treated non-small-cell lung cancer (OAK): a phase 3, open-label, multicentre randomised controlled trial. Lancet. 2017;389(10066):255–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Borghaei H, Paz-Ares L, Horn L, Spigel DR, Steins M, Ready NE, Chow LQ, Vokes EE, Felip E, Holgado E, et al. Nivolumab versus Docetaxel in Advanced Nonsquamous Non-small-cell Lung Cancer. N Engl J Med. 2015;373(17):1627–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Herbst RS, Baas P, Kim DW, Felip E, Pérez-Gracia JL, Han JY, Molina J, Kim JH, Arvis CD, Ahn MJ, et al. Pembrolizumab versus Docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial. Lancet. 2016;387(10027):1540–50. [DOI] [PubMed] [Google Scholar]
- 5.Brahmer J, Reckamp KL, Baas P, Crinò L, Eberhardt WE, Poddubskaya E, Antonia S, Pluzanski A, Vokes EE, Holgado E, et al. Nivolumab versus Docetaxel in Advanced squamous-cell non-small-cell Lung Cancer. N Engl J Med. 2015;373(2):123–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Lee AY, Levine MN. Venous thromboembolism and cancer: risks and outcomes. Circulation. 2003;107(23 Suppl 1):I17–21. [DOI] [PubMed] [Google Scholar]
- 7.Khorana AA, Francis CW, Culakova E, Lyman GH. Risk factors for chemotherapy-associated venous thromboembolism in a prospective observational study. Cancer. 2005;104(12):2822–9. [DOI] [PubMed] [Google Scholar]
- 8.Hill H, Robinson M, Lu L, Slaughter D, Amin A, Mileham K, Patel JN. Venous thromboembolism incidence and risk factors in non-small cell lung cancer patients receiving first-line systemic therapy. Thromb Res. 2021;208:71–8. [DOI] [PubMed] [Google Scholar]
- 9.Icht O, Darzi N, Shimony S, Jacobi O, Reinhorn D, Landman Y, Mutai R, Averbuch I, Shochat T, Spectre G, et al. Venous thromboembolism incidence and risk assessment in lung cancer patients treated with immune checkpoint inhibitors. J Thromb Haemost. 2021;19(5):1250–8. [DOI] [PubMed] [Google Scholar]
- 10.Moik F, Chan WE, Wiedemann S, Hoeller C, Tuchmann F, Aretin MB, Fuereder T, Zöchbauer-Müller S, Preusser M, Pabinger I, et al. Incidence, risk factors, and outcomes of venous and arterial thromboembolism in immune checkpoint inhibitor therapy. Blood. 2021;137(12):1669–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Khorana AA, Palaia J, Rosenblatt L, Pisupati R, Huang N, Nguyen C, Barron J, Gallagher K, Bond TC. Venous thromboembolism incidence and risk factors associated with immune checkpoint inhibitors among patients with advanced non-small cell lung cancer. J Immunother Cancer 2023, 11(1). [DOI] [PMC free article] [PubMed]
- 12.Bjørnhart B, Kristiansen C, Asmussen J, Hansen KH, Wedervang K, Jørgensen TL, Herrstedt J, Schytte T. Clinical impact of venous thromboembolism in non-small cell lung cancer patients receiving immunotherapy. Thromb Res. 2023;221:164–72. [DOI] [PubMed] [Google Scholar]
- 13.Agnelli G, Sonaglia F. Prevention of venous thromboembolism. Thromb Res. 2000;97(1):V49–62. [DOI] [PubMed] [Google Scholar]
- 14.Kahn SR, Lim W, Dunn AS, Cushman M, Dentali F, Akl EA, Cook DJ, Balekian AA, Klein RC, Le H, et al. Prevention of VTE in nonsurgical patients: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2012;141(2 Suppl):eS195–226. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Elshoury A, Schaefer JK, Lim MY, Skalla DP, Streiff MB. Update on guidelines for the Prevention of Cancer-Associated thrombosis. J Natl Compr Canc Netw 2022, 20(13). [DOI] [PubMed]
- 16.Caprini JA. Thrombosis risk assessment as a guide to quality patient care. Dis Mon. 2005;51(2–3):70–8. [DOI] [PubMed] [Google Scholar]
- 17.Barbar S, Noventa F, Rossetto V, Ferrari A, Brandolin B, Perlati M, De Bon E, Tormene D, Pagnan A, Prandoni P. A risk assessment model for the identification of hospitalized medical patients at risk for venous thromboembolism: the Padua Prediction score. J Thromb Haemost. 2010;8(11):2450–7. [DOI] [PubMed] [Google Scholar]
- 18.Khorana AA, Kuderer NM, Culakova E, Lyman GH, Francis CW. Development and validation of a predictive model for chemotherapy-associated thrombosis. Blood. 2008;111(10):4902–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Hachey KJ, Hewes PD, Porter LP, Ridyard DG, Rosenkranz P, McAneny D, Fernando HC, Litle VR. Caprini venous thromboembolism risk assessment permits selection for postdischarge prophylactic anticoagulation in patients with resectable lung cancer. J Thorac Cardiovasc Surg. 2016;151(1):37–e4431. [DOI] [PubMed] [Google Scholar]
- 20.Sterbling HM, Rosen AK, Hachey KJ, Vellanki NS, Hewes PD, Rao SR, Pinjic E, Fernando HC, Litle VR. Caprini Risk Model decreases venous thromboembolism rates in thoracic surgery Cancer patients. Ann Thorac Surg. 2018;105(3):879–85. [DOI] [PubMed] [Google Scholar]
- 21.Lin Y, Zeng Z, Lin R, Zheng J, Liu S, Gao X. The Caprini thrombosis risk model predicts the risk of peripherally inserted central catheter-related upper extremity venous thrombosis in patients with cancer. J Vasc Surg Venous Lymphat Disord. 2021;9(5):1151–8. [DOI] [PubMed] [Google Scholar]
- 22.Moumneh T, Riou J, Douillet D, Henni S, Mottier D, Tritschler T, Le Gal G, Roy PM. Validation of risk assessment models predicting venous thromboembolism in acutely ill medical inpatients: a cohort study. J Thromb Haemost. 2020;18(6):1398–407. [DOI] [PubMed] [Google Scholar]
- 23.Key NS, Khorana AA, Kuderer NM, Bohlke K, Lee AYY, Arcelus JI, Wong SL, Balaban EP, Flowers CR, Francis CW, et al. Venous thromboembolism prophylaxis and treatment in patients with Cancer: ASCO Clinical Practice Guideline Update. J Clin Oncol. 2020;38(5):496–520. [DOI] [PubMed] [Google Scholar]
- 24.Esmon CT. Inflammation and thrombosis. J Thromb Haemost. 2003;1(7):1343–8. [DOI] [PubMed] [Google Scholar]
- 25.Engelmann B, Massberg S. Thrombosis as an intravascular effector of innate immunity. Nat Rev Immunol. 2013;13(1):34–45. [DOI] [PubMed] [Google Scholar]
- 26.Franco AT, Corken A, Ware J. Platelets at the interface of thrombosis, inflammation, and cancer. Blood. 2015;126(5):582–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Alma S, Eloi D, Léa V, Julie C, Valérie M, Pierre G, Hilgers W, Philippe G, Christine Z, Philippe D. Incidence of venous thromboembolism and discriminating capacity of Khorana score in lung cancer patients treated with immune checkpoint inhibitors. J Thromb Thrombolysis. 2022;54(2):287–94. [DOI] [PubMed] [Google Scholar]
- 28.Caprini JA, Arcelus JI, Hasty JH, Tamhane AC, Fabrega F. Clinical assessment of venous thromboembolic risk in surgical patients. Semin Thromb Hemost. 1991;17(Suppl 3):304–12. [PubMed] [Google Scholar]
- 29.Obi AT, Pannucci CJ, Nackashi A, Abdullah N, Alvarez R, Bahl V, Wakefield TW, Henke PK. Validation of the Caprini venous thromboembolism risk Assessment Model in critically Ill Surgical patients. JAMA Surg. 2015;150(10):941–8. [DOI] [PubMed] [Google Scholar]
- 30.Zhou HX, Peng LQ, Yan Y, Yi Q, Tang YJ, Shen YC, Feng YL, Wen FQ. Validation of the Caprini risk assessment model in Chinese hospitalized patients with venous thromboembolism. Thromb Res. 2012;130(5):735–40. [DOI] [PubMed] [Google Scholar]
- 31.Grant PJ, Greene MT, Chopra V, Bernstein SJ, Hofer TP, Flanders SA. Assessing the Caprini score for Risk Assessment of venous thromboembolism in Hospitalized Medical patients. Am J Med. 2016;129(5):528–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Zhou H, Hu Y, Li X, Wang L, Wang M, Xiao J, Yi Q. Assessment of the risk of venous thromboembolism in Medical inpatients using the Padua Prediction score and Caprini Risk Assessment Model. J Atheroscler Thromb. 2018;25(11):1091–104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Greene MT, Spyropoulos AC, Chopra V, Grant PJ, Kaatz S, Bernstein SJ, Flanders SA. Validation of Risk Assessment models of venous thromboembolism in Hospitalized Medical patients. Am J Med. 2016;129(9):e10011009–18. [DOI] [PubMed]
- 34.Mansfield AS, Tafur AJ, Wang CE, Kourelis TV, Wysokinska EM, Yang P. Predictors of active cancer thromboembolic outcomes: validation of the Khorana score among patients with lung cancer. J Thromb Haemost. 2016;14(9):1773–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Mouronte-Roibás C, Leiro-Fernández V, Fernández-Villar A, Botana-Rial M, Ramos-Hernández C, Ruano-Ravina A. COPD, emphysema and the onset of lung cancer. A systematic review. Cancer Lett. 2016;382(2):240–4. [DOI] [PubMed] [Google Scholar]
- 36.Lei H, Tao D, Zhang N, Sun M, Sun L, Yang D, Jiang Y, Zhou W, Xie Y, Wang Y. Nomogram prediction for the risk of venous thromboembolism in patients with lung cancer. Cancer Cell Int. 2023;23(1):40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.van Es N, Ventresca M, Di Nisio M, Zhou Q, Noble S, Crowther M, Briel M, Garcia D, Lyman GH, Macbeth F, et al. The Khorana score for prediction of venous thromboembolism in cancer patients: an individual patient data meta-analysis. J Thromb Haemost. 2020;18(8):1940–51. [DOI] [PubMed] [Google Scholar]
- 38.Tsubata Y, Kawakado K, Hamai K, Furuya N, Yokoyama T, Saito R, Nakamura A, Masuda T, Hamaguchi M, Kuyama S, et al. Identification of risk factors for venous thromboembolism and validation of the Khorana score in patients with advanced lung cancer: based on the multicenter, prospective Rising-VTE/NEJ037 study data. Int J Clin Oncol. 2023;28(1):69–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Gong J, Drobni ZD, Alvi RM, Murphy SP, Sullivan RJ, Hartmann SE, Gilman HK, Lee H, Zubiri L, Raghu VK, et al. Immune checkpoint inhibitors for cancer and venous thromboembolic events. Eur J Cancer. 2021;158:99–110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Sussman TA, Li H, Hobbs B, Funchain P, McCrae KR, Khorana AA. Incidence of thromboembolism in patients with melanoma on immune checkpoint inhibitor therapy and its adverse association with survival. J Immunother Cancer 2021, 9(1). [DOI] [PMC free article] [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The original data of this manuscript are available on reasonable request by contacting the correspondence author.