Abstract
Background
Anaplastic lymphoma kinase-positive (ALK+) and ROS proto-oncogene 1 (ROS1)-positive (ROS1+) lung cancers have been reported to be associated with an elevated risk of thromboembolic events. This study aimed to assess the long-term risk of developing thromboembolism (TE) in ROS1+ lung cancer and to compare it with other oncogenic drivers in the Asian population.
Materials and methods
We retrospectively enrolled a cohort of ROS1+ lung adenocarcinoma in a medical center in Taiwan and a comparison cohort of ALK+ and epidermal growth factor receptor-positive (EGFR+) lung cancers. Venous and arterial TEs were identified throughout the cancer course, and the incidence rate was calculated.
Results
We enrolled 44 ROS1+, 98 ALK+, and 168 EGFR+ non-small-cell lung cancer (NSCLC) patients. A total of 11 (25%), 36 (36.7%), and 38 (22.6%) patients in the ROS1, ALK, and EGFR cohorts, respectively, were diagnosed with thromboembolic events throughout the follow-up course of the disease (P = 0.042). The incidence rates were 99.0, 91.9, and 82.5 events per 1000 person-years for the ROS1, ALK, and EGFR cohorts, respectively. The majority of thrombosis events in the ROS1 (91.6%) and ALK (85.4%) cohorts were venous. On the contrary, 43.2% of thromboembolic events were arterial in the EGFR cohort. A higher proportion of thromboembolic events were noted during cancer diagnosis in the ROS1 cohort (36.3%) than in the ALK (16.7%) and EGFR (10.5%) cohorts. The stage was the only clinical variable associated with thromboembolic risk. There was a significant difference in survival between patients with and without TE in the EGFR cohort, but not in the ALK and ROS1 cohorts.
Conclusions
Although ROS1+ and ALK+ NSCLCs had a higher cumulative incidence of TE than EGFR+ NSCLC, the person-year incidence rates were similar among the three groups. EGFR-mutated NSCLC had more arterial events. Nevertheless, ALK+ lung cancer had higher venous events than EGFR-mutated lung cancer.
Key words: ROS1 rearrangement, thromboembolism, non-small-cell lung cancer, ALK, EGFR
Highlights
-
•
ROS1+ and ALK+ NSCLCs have a higher cumulative incidence of TE.
-
•
The incidence rates calculated by events per 1000 person-years for ROS1+, ALK+, and EGFR-mutated NSCLCs are similar.
-
•
The peri-diagnostic thromboembolic risk is especially high in the ROS1 population.
-
•
EGFR-mutated NSCLC has more arterial events and ALK+ lung cancer has higher venous events than EGFR-mutated lung cancer.
Introduction
Cancer-associated thrombosis is a common complication in patients with cancers. Generally, the risk of venous thromboembolism (TE) appears to increase four- to sevenfold in cancer patients and varies widely by cancer type.1, 2, 3 The absolute incidence rates of venous TE in cancer patients vary from 0.5% to 20% and are influenced by cancer type, stage of disease, age, proximity to diagnosis, and treatment modality.4 Lung cancer is a malignancy associated with a substantial risk of thrombotic events. Primary lung cancer has been reported to cause venous TE in 7%-15% of patients.5,6 The incidence of thrombosis in adenocarcinoma is three- to fourfold higher than that in squamous cell carcinoma and small-cell lung cancer.7,8
Recently, some studies reported that anaplastic lymphoma kinase-positive (ALK+) and ROS proto-oncogene 1 (ROS1)-positive (ROS1+) lung cancers were associated with an elevated risk of thromboembolic events.9, 10, 11 The reported incidence rate of thromboembolic events was 34.7% in a ROS1 cohort in the USA and China during the peri-diagnostic period.10 However, the incidence of TE is well known to have ethnic differences. Studies conducted in Asia have consistently reported lower rates of venous TE in Asians than in Caucasians.12, 13, 14, 15 The incidence of TE in ROS1+ lung adenocarcinoma has not yet been clearly reported. In addition, the prospective randomized controlled clinical trials of ROS1 tyrosine kinase inhibitor did not demonstrate a higher incidence of thromboembolic events.16, 17, 18, 19, 20, 21, 22 However, the population of clinical trials was highly selective. Thrombotic events occur predominantly during the peri-diagnostic or cancer progression periods, which are outside the observation period of clinical trials. Thus, observation in clinical trials may not reflect the true incidence of thromboembolic events in clinical practice in the real world.
We conducted a retrospective analysis of patients with ROS1+ lung adenocarcinoma and compared them with cohorts of epidermal growth factor receptor-positive (EGFR+) and ALK+ lung adenocarcinomas in a tertiary hospital in Taiwan. This study aimed to clarify the association between TE and ROS1+ lung cancer and to compare it to lung cancer with other oncogenic drivers in the Asian population. Furthermore, we focused on the long-term risk of developing TE around and after the diagnosis period.
Materials and methods
Study population
Patients diagnosed with ROS1+ lung adenocarcinoma between December 2001 and April 2020 at the National Taiwan University Hospital were retrospectively enrolled as the primary cohort. ROS1 rearrangement was screened by immunohistochemistry and confirmed by FISH, or reverse transcription-quantitative PCR. We then collected two separate cohorts of EGFR+ and ALK+ lung adenocarcinomas diagnosed during a similar period to compare the risk of thrombosis among lung cancer patients with different driver oncogenes. We included as many cases with ROS1 or ALK mutations that could be identified through medical records as possible because of their relatively low prevalence in non-small-cell lung cancer (NSCLC). No patients were excluded by particular reasons. Since EGFR-mutated NSCLC patients were much more than NSCLC patients with ROS1 or ALK mutations, we did not enroll whole population of EGFR-mutated patients to reduce the computational burden. EGFR cohorts were formed by a 1 : 4 covariate case-control matching. Case patients included those with NSCLC harboring ROS1 mutations. The control patients were selected from the cohort of NSCLC patients with common EGFR mutations which was analyzed and reported in the previously studies.23,24 After enrollment of one case, eligible control patients with the same sex, smoking history, and stage were recruited sequentially until four control patients were individually matched to each case patient (Supplementary Figure S1, available at https://doi.org/10.1016/j.esmoop.2022.100742).
Data on clinical characteristics, treatment response, and thromboembolic events were obtained from electronic medical records. Thromboembolic events that occurred within 1 year before cancer diagnosis to death or end of the study were identified by reviewing computed tomography and sonography reports, as well as outpatient and admission medical records.
The collection of human tissue samples and clinical data was approved by the Institutional Review Board of National Taiwan University Hospital (202012161RINA).
Statistics
Survival times are different in NSCLC patients with different driver mutations. Thromboembolic events could accumulate more if the survival times are longer. Thus, the person-time rate used in this study was more revealing than simple cumulative incidence with regard to the comparison of TE risk. Person-time at risk commenced 1 year before cancer diagnosis and ended at the time of thromboembolic events, death, or the last follow-up date (30 May 2021). The absolute rates of thromboembolic events (per 1000 person-years) were calculated by dividing the number of people with TE by the person-time at risk. Some patients experienced multiple thromboembolic events. However, only the first thromboembolic event in a specific patient was included in the calculation of the absolute incidence rate, analysis of risk factors, and survival analysis.
Each participant’s last follow-up date and living status were captured for survival analysis. Overall survival (OS) was calculated as the interval between the date of lung cancer diagnosis and death or the last follow-up date. Survival was estimated using the Kaplan–Meier method, and the log-rank test was used to compare the survival between groups. Univariate and multivariate Cox proportional hazards models were used to adjust for potentially interacting covariates and determine the factors associated with the longitudinal risks of thromboembolic events. Statistical significance was set at P < 0.05.
Statistical analyses were carried out using IBM SPSS Statistics, version 22.00 (SPSS Inc., Chicago, IL).
Results
Study population
There were 44, 98, and 168 patients in the ROS1, ALK, and EGFR cohorts, respectively. The baseline demographic characteristics are presented in Table 1. The mean age at diagnosis was significantly lower in the ROS1 (53.06 ± 15.41 years) and ALK (52.77 ± 11.30 years) cohorts than in the EGFR cohort (65.84 ± 12.67 years; P ˂ 0.0001). The majority of patients were diagnosed at an advanced stage and were nonsmokers.
Table 1.
Demographic characteristics of patients with ROS1+, ALK+, and EGFR+ lung adenocarcinoma
| N | ROS1 |
ALK |
EGFR |
|---|---|---|---|
| 44 | 98 | 168 | |
| Sex (%) | |||
| Female | 30 (68.2) | 49 (50) | 118 (70.2) |
| Male | 14 (31.8) | 49 (50) | 50 (29.8) |
| Age, mean ± SD | 53.06 ± 15.41 (19-90) | 52.77 ± 11.30 (26-80) | 65.84 ± 12.67 (36-89) |
| Smoking (%) | |||
| Nonsmoker | 35 (79.5) | 70 (71.4) | 141 (83.9) |
| Former/current smoker (all tobacco) | 9 (20.5) | 28 (28.6) | 27 (16.1) |
| Staging (%) | |||
| I | 1 (2.3) | 3 (3.0) | 19 (11.3) |
| II | 2 (4.5) | 3 (3.0) | 8 (4.8) |
| III | 8 (18.2) | 28 (28.6) | 17 (10.1) |
| IV | 33 (75) | 64 (65.3) | 124 (73.8) |
| T stage (%) | |||
| T1 | 7 (15.9) | 19 (19.4) | 25 (14.9) |
| T2 | 11 (25) | 24 (24.5) | 47 (28) |
| T3 | 4 (9.1) | 12 (12.2) | 15 (8.9) |
| T4 | 21 (47.7) | 40 (40.8) | 81 (48.2) |
| NA | 1 (2.3) | 3 (3.1) | 0 (0) |
| N stage | |||
| N0 | 3 (6.8) | 15 (15.3) | 48 (28.6) |
| N1 | 6 (13.6) | 8 (8.2) | 18 (10.7) |
| N2 | 14 (31.8) | 37 (37.8) | 49 (29.2) |
| N3 | 20 (45.5) | 35 (35.7) | 53 (31.5) |
| NA | 1 (2.3) | 3 (3.1) | 0 (0) |
| M stage (%) | |||
| M0 | 11 (25) | 34 (34.7) | 68 (40.5) |
| M1 | 33 (75) | 64 (65.3) | 100 (59.5) |
| Median overall survival (months) | 60.9 | 63.1 | 24.5 |
ALK, anaplastic lymphoma kinase; EGFR, epidermal growth factor receptor; SD, standard deviation.
The median follow-up times for the ROS1, ALK, and EGFR cohorts were 60.9 [interquartile range (IQR): 16.0-136.3] months, 63.1 (IQR: 28.1-113.0) months, and 24.53 (IQR: 16.3-46.2) months, respectively. The difference in survival time between the three molecular driver cohorts led to disparities in the follow-up time.
Characteristics and incidence of thromboembolic events
The total number of events was 12, 41, and 44 for the ROS1, ALK, and EGFR cohorts, respectively. Some patients had more than one thromboembolic event during the disease course. Only the first thromboembolic event in a specific patient was included in the following calculation of the absolute incidence rate. Venous thromboembolic events consisted of pulmonary embolism, deep vein thrombosis of the upper or lower extremities, renal vein thrombosis, hepatic vein thrombosis, and peripherally inserted central catheter-related thrombosis. Arterial thromboembolic events were mainly cerebral and coronary thromboses. Two patients in the ROS1 cohort were diagnosed with Trousseau's syndrome and presented with multiple, repeated, and medication-resistant thrombosis. The characteristics of thromboembolic events are shown in Table 2.
Table 2.
Characteristics of thromboembolic events
| Mutation cohort | ROS1 | ALK | EGFR | P valuea |
|---|---|---|---|---|
| Number of patients | 44 | 98 | 168 | |
| Number of patients with thromboembolism | 11 (25%)b | 36 (36.7%)c | 38 (22.6%)d | 0.042 |
| Sex | ||||
| Male | 2 | 18 | 13 | |
| Female | 9 | 18 | 25 | |
| Type | ||||
| Arterial events | 1 | 6 | 19 | 0.10 |
| Location | ||||
| Coronary | 0 | 3 | 2 | |
| Cerebral | 1 | 3 | 12 | |
| Both coronary and cerebral | 0 | 0 | 2 | |
| Other | 0 | 0 | 3 | |
| Venous events | 11 | 35 | 25 | <0.001 |
| Location | ||||
| PE | 3 | 10 | 4 | |
| DVT | 3 | 20 | 13 | |
| Both PE and DVT | 2 | 2 | 5 | |
| Trousseau syndrome | 2 | 0 | 0 | |
| Other | 1 | 3 | 3 |
ALK, anaplastic lymphoma kinase; DVT, deep vein thrombosis; EGFR, epidermal growth factor receptor; PE, pulmonary embolism.
By chi-square test.
One patient had both arterial and venous thrombosis.
Two patients had both arterial and venous thrombosis, one of whom had both coronary and cerebral embolism. Two patients had recurrent venous thrombosis.
Four patients had more than one arterial thromboembolic event. Two patients had both arterial and venous thromboses. One patient had recurrent venous thrombosis.
Throughout the follow-up course of the disease, 25% (n = 11), 36.7% (n = 36), and 22.6% (n = 38) of patients in the ROS1, ALK, and EGFR cohorts, respectively, had arterial or venous TEs (X2 = 6.35; df = 2; P = 0.042). To calculate the incidence of thromboembolic events, a total of 11, 36, and 38 patients in the ROS1, ALK, and EGFR cohorts had a subsequent diagnosis of thromboembolic events in 111.1, 391.8, and 460.6 person-years, respectively (Table 3). The incidence rates were 99.0 [95% confidence interval (CI) 56.5-173.4)], 91.9 (95% CI: 67.3-125.4), and 82.5 (95% CI: 60.8-111.9) events per 1000 person-years for the ROS1, ALK, and EGFR cohorts, respectively. The thromboembolism events in person-years incidence rate were not significantly different when comparing the ALK or EGFR cohorts to the ROS1 cohort in multivariate analysis after adjusting for age, sex, smoking, and stage. [hazard ratio (HR) 1.19, 95% CI 0.60-2.36, P = 0.613; HR 0.96, 95% CI 0.47-1.94, P = 0.90, respectively] (Supplementary Table S1, available at https://doi.org/10.1016/j.esmoop.2022.100742).
Table 3.
Absolute and relative rates of thromboembolic events in different molecular driver cohorts
| N | Patients with TE | Totala events | Person-time (years) | Incidence rate—arterial and venous (per 1000 person-years) | 95% CI |
Relative incidence rateb | 95% CI |
Incidence rate—venous (per 1000 person-years) | Incidence rate—arterial (per 1000 person-years)c | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Lower | Upper | Lower | Upper | |||||||||
| ROS1 | 44 | 11 (25%) | 12 | 111.1 | 99.0 | 56.5 | 173.4 | 1 | — | — | 99.0 | 9.0 |
| ALK | 98 | 36 (36.7%) | 41 | 391.8 | 91.9 | 67.3 | 125.4 | 1.192 (P = 0.61) | 0.60 | 2.36 | 79.1 | 12.8 |
| EGFR | 168 | 38 (22.6%) | 44 | 460.6 | 82.5 | 60.8 | 111.9 | 0.956 (P = 0.90) | 0.47 | 1.94 | 49.9 | 34.7 |
ALK, anaplastic lymphoma kinase; CI, confidence interval; EGFR, epidermal growth factor receptor; TE, thromboembolism.
The total events number was higher than the number of patients with events because some patients encountered more than one thromboembolic event.
Hazard ratios were estimated using Cox proportional hazards regression adjusted for sex, age, smoking, and stage.
Only the first thromboembolic event in one specific patient was included in the analysis of absolute rate. One patient each from the ROS1 and EGFR group, respectively, representing arterial TE and venous TE at the same time as their first event.
Difference in the incidence of arterial and venous thrombosis
The majority of thromboembolic events in the ROS1 (n = 11, 91.6%) and ALK (n = 35, 85.4%) cohorts were venous. Arterial thromboembolic events were more common in the EGFR cohort (n = 19, 43.2%). The cumulated arterial thromboembolic events throughout the follow-up course among the three cohorts were not significantly different (X2 = 4.66; df = 2; P = 0.10). However, there was significant difference in the cumulated venous thromboembolic events throughout the follow-up course among the three cohorts (X2 = 15.34; df = 2; P < 0.001) (Table 2).
The incidence rates by 1000 person-years for venous and arterial events separately are shown in Table 3. The arterial TE in person-years incidence rate was not significantly different between the ROS1 cohort and ALK or EGFR cohorts (Table 4) (HR 1.75, 95% CI 0.209-14.99, P = 0.61; HR 4.21, 95% CI 0.56-31.78, P = 0.16, respectively, in univariate analysis). The venous TE was not significant when comparing the ROS1 to the EGFR group in univariate analysis (HR 0.51, 95% CI 0.24-1.05, P = 0.07). When the ALK cohort was compared to the EGFR cohort in additional analysis, the arterial TE was not significantly different (HR 0.42, 95% CI 0.15-1.14, P = 0.09). Nevertheless, the venous TE was significantly higher in the ALK cohort than in the EGFR cohort (HR 1.96, 95% CI 1.12-3.43, P = 0.02).
Table 4.
Univariate and multivariate analysis of risk factors predicting venous and arterial thrombotic events among NSCLC patients with different molecular drivers
| Variable | Arterial thrombosis |
Venous thrombosis |
||||||
|---|---|---|---|---|---|---|---|---|
| Univariate analysis |
Multivariate analysis |
Univariate analysis |
Multivariate analysis |
|||||
| HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | |
| ROS1 versus ALK | 1.749 (0.20-14.99) | 0.61 | 1.669 (0.19-14.46) | 0.64 | 0.991 (0.50-1.98) | 0.98 | 1.036 (0.52-2.07) | 0.92 |
| ROS1 versus EGFR | 4.205 (0.56-31.78) | 0.16 | 3.745 (0.47-30.04) | 0.21 | 0.505 (0.24-1.05) | 0.07 | 0.548 (0.25-1.18) | 0.13 |
| EGFR versus ALK | 0.42 (0.15-1.14) | 0.09 | — | — | 1.96 (1.12-3.43) | 0.02 | — | — |
| Age (years, ≦60 versus >60) | 0.294 (0.12-0.75) | 0.01a | 0.376 (0.14-1.02) | 0.05a | 0.817 (0.49-1.36) | 0.44 | 1.022 (0.59-1.77) | 0.94 |
| Sex (male versus female) | 1.388 (0.60-3.22) | 0.44 | 1.827 (0.61-4.69) | 0.21 | 0.823 (0.49-1.39) | 0.47 | 0.600 (0.32-1.13) | 0.11 |
| Smoking (smoker versus nonsmoker) | 0.804 (0.27-2.38) | 0.69 | 0.679 (0.20-2.27) | 0.53 | 1.111 (0.62-1.99) | 0.72 | 1.343 (0.67-2.68) | 0.40 |
| Stage (per one stage increase) | 2.908 (1.07-7.90) | 0.04a | 2.825 (1.10-7.26) | 0.03a | 1.410 (1.02-1.94) | 0.04a | 1.373 (0.98-1.92) | 0.06 |
ALK, anaplastic lymphoma kinase; CI, confidence interval; EGFR, epidermal growth factor receptor; HR, hazard ratio; NSCLC, non-small-cell lung cancer.
P < 0.05.
Time of occurrence of thromboembolic events
A total of 45.5% (n = 5), 63.9% (n = 23), and 71.1% (n = 27) of patients in the ROS1, ALK, and EGFR cohorts had thromboembolic events during disease progression (Table 5). A few thromboembolic events occurred even when the disease was under control. A numerically higher proportion of thromboembolic events was noted during cancer diagnosis in the ROS1 cohort (n = 4, 36.3%) than in the ALK (n = 6, 16.7%) and EGFR (n = 4, 10.5%) cohorts (X2 = 4.14; df = 2; P = 0.13). The detailed occurrence times of the thromboembolic events are shown in Table 5. Recurrent thrombosis was noted in 37.5%, 13.9%, and 13.2% of patients in the ROS1, ALK, and EGFR cohorts, respectively (X2 = 1.41; df = 2; P = 0.49).
Table 5.
Disease status and occurrence time of the first thrombosis event
| Mutation cohort | ROS1 | ALK | EGFR |
|---|---|---|---|
| Clinical status of thrombosis events | |||
| At diagnosis | 4 | 6 | 4 |
| Partial response | 1 | 0 | 1 |
| Stable disease | 1 | 7 | 6 |
| Disease progression | 5 | 23 | 27 |
| Treatment received during thrombosis events | |||
| Untreated | 3 | 8 | 4 |
| Operation | 1 | 0 | 0 |
| Chemotherapy | 3 | 11 | 15 |
| TKI | 4 | 17 | 16 |
| TKI + chemotherapy | 0 | 0 | 3 |
| Recurrent thrombosis | |||
| Yes | 3 | 5 | 5 |
| No | 8 | 31 | 33 |
ALK, anaplastic lymphoma kinase; EGFR, epidermal growth factor receptor; TKI, tyrosine kinase inhibitor.
Factors associated with thromboembolic events
The HR of thromboembolic events in the ROS1 cohort was not significantly different from that in the ALK (HR 1.192, 95% CI 0.69-2.36, P = 0.613) and EGFR (HR 0.956, 95% CI 0.47-1.94, P = 0.901) cohorts. The stage of disease was the only clinical variable associated with thromboembolic risk in both the univariate and multivariate analyses. More advanced stages resulted in significantly higher HRs for thromboembolic events (P = 0.003). On an average, for each unit increase in the stage, the HR of thromboembolic events was 1.62 times higher [odds ratio (OR) 1.62, 95% CI 1.18-2.22] (Supplementary Table S1, available at https://doi.org/10.1016/j.esmoop.2022.100742).
To examine the specific factors associated with arterial or venous thromboembolic events, we further separated these two types of TEs in the analysis. In venous thrombosis, stage is the only risk factor associated with thromboembolic events. Regarding arterial events, age > 60 and advanced stage of disease had an increased risk of thromboembolic events (Table 4).
Thromboembolic events and survival
In the ROS1 cohort, patients with and without thromboembolic events observed during the disease course did not show a significant difference in OS [P = 0.53, HR 1.3 (range 0.54-3.28), log-rank test; Supplementary Figure S2A, available at https://doi.org/10.1016/j.esmoop.2022.100742]. The results were similar after adjusting for risk factors, such as age, sex, and stage [P = 0.49, HR 1.4 (range 0.52-3.84)]. The median survival time for patients who had thromboembolic events was 49.3 months (95% CI 10.6-88.0 months). The median survival time for patients without thromboembolic events was 77.9 months (95% CI 0-158.2 months). The survival difference was also insignificant in the ALK cohort [P = 0.12, HR 1.48 (range 0.90-2.42); Supplementary Figure S2B, available at https://doi.org/10.1016/j.esmoop.2022.100742]. NSCLC patients harboring EGFR mutations with TE had significantly poorer survival than those without TE [P = 0.04, HR 1.52 (range 1.02-2.27)], after adjusting for age, sex, and stage (Supplementary Figure S2C, available at https://doi.org/10.1016/j.esmoop.2022.100742).
Discussion
Data describing the longitudinal thrombogenic risk profile of patients with NSCLC and oncogenic drivers are scarce, especially in the Asian population. Our study described the long-term risk of thromboembolic events peri and post the cancer diagnosis, focusing on the ROS1, ALK, and EGFR-mutated lung cancers. Regarding the cumulative rate of thromboembolic events, patients with ROS1+ and ALK+ lung cancers had elevated risks of thromboembolic events during the cancer course compared to those with EGFR-mutated lung cancer. However, EGFR-mutated lung cancer patients had a higher rate of arterial TE. The person-years incidence rates of thromboembolic events were similar in these three molecular driver lung cancers because the survival time was longer in ROS1+ and ALK+ lung cancers than in EGFR+ lung cancer. However, ROS1+ and ALK+ lung cancers still had twice the person-years incidence rate of venous TEs compared to EGFR+ lung cancer, which contained half the arterial TEs. Venous TE was statistically more frequent in ALK+ lung cancer, but not ROS1+, compared to EGFR+ lung cancer.
Approximately 7%-15% of advanced NSCLC patients are estimated to experience venous TE during the course of their disease.5,6,25 Previous reports have focused mainly on venous thrombosis. The event rate could be higher when arterial events are considered. In addition, NSCLC patients with ALK or ROS1 mutations have recently been reported to have remarkably elevated rates of TE.10,11,26,27 Higher venous TE risk in NSCLC patients with EGFR mutations than in those with wild-type EGFR has also been reported. A recent study from a Chinese cohort reported that the rate of venous TE in lung adenocarcinoma with EGFR mutations after surgical pneumonectomy was 17%.28 The cumulative rates of TE in ALK+ and EGFR+ NSCLC in a US cohort were 45.3% and 21.2%, respectively, with a median follow-up time of 33.1 months.27 Two other ALK+ cohorts in Canada and Israel reported a 42% and 28% cumulative rate of venous TE within 22 and 13 months of follow-up, respectively.26 A recent report by Al-Samkari et al. showed that the overall rate of venous TE in the ALK group (42.7%) was significantly higher than that in the EGFR group (26.2%). However, the rates were similar for arterial thrombosis in the ALK (5%) and non-ALK populations (4.4%).29 Ng et al. reported that the peri-diagnostic incidence rates of TEs were 34.7%, 22.3%, and 13.7% for the ROS1, ALK, and EGFR cohorts, respectively.10 The incidence of TEs throughout the course of cancer in a ROS1 cohort in Spain was 46.6% with a median follow-up of 19 months.30 Generally, ROS1+ and ALK+ lung cancers had a comparable elevated risk of TE than in EGFR lung cancer. Our study results are compatible with the findings of the aforementioned studies. However, the cumulative rates of TEs were still lower in our study than in studies with Caucasian participants. The cumulative rates of TEs in our study were 25%, 36.7%, and 22.6% for the ROS1, ALK, and EGFR cohorts, respectively, with a median follow-up time of 60.9, 62.1, and 24.5 months. The follow-up time was much longer than that in the aforementioned studies. This observation is also consistent with a previous report, which found that Asian ethnicity was associated with an ∼60% lower risk of developing venous TE development in patients with primary lung cancer.8
The cumulative rates were highly dependent on the follow-up time. The longer the follow-up time, which was also influenced by the survival time, the higher the rates of TE. Survival was better in the ROS1 and ALK populations than in the EGFR population.21,22 The survival time partially contributed to the higher rates of TE in the ALK and ROS1 populations. The incidence rates in our study cohort were 99.0, 91.9, and 82.5 per 1000 person-years for the ROS1, ALK, and EGFR cohorts, respectively, which were not significantly different. Nevertheless, the incidence rates of venous TE were numerically different. A study in Taiwan showed that the incidence of venous TE in lung cancer was 39.5 per 1000 person-years.31 Compared to the venous TE incidence rate of 49.9 per 1000 person-years reported in our analysis, EGFR+ lung cancer had a slightly higher incidence than overall lung cancer. Furthermore, our study showed that ALK+ and ROS1+ lung cancers had remarkably higher venous TE incidence rates of 79.1 and 99.0 per 1000 person-years, respectively.
The mechanism of hypercoagulable state related to ALK and ROS1 rearrangement has been postulated to be abundant mucin production, a known prothrombotic factor.32,33 A higher proportion of TE events in the ROS1 cohort (36.4%) than the ALK (16.7%) and EGFR (10.5%) cohorts occurred at diagnosis in our study. Moreover, recurrent thrombosis despite anticoagulation was observed more frequently in the ROS1 cohort (27%) than in the ALK (14%) and EGFR (13%) cohorts (Table 5). The different timings of the occurrence of TE and different recurrent rates of TE may reflect the distinct nature of TE in lung cancer with different driver mutations. The cancer-specific mechanisms and molecular alterations may be associated with the underlying mechanism and play an intrinsic role in thrombus formation. The exact mechanism driving the TE process in lung cancer with different driver mutations still requires further investigation.
There are some limitations to our study. This was a retrospective cohort study. The number of patients in the ROS1 cohort was small because of the low prevalence of the driver oncogene in NSCLC. The survival difference between patients with and without TEs was demonstrated in the EGFR cohort, but not in the ROS1 and ALK cohorts. Due to the potentially insufficient sample size, we were unable to conclude the survival findings. One of the strengths of our study is the long follow-up period and focus on the longitudinal thromboembolic risk during the course of NSCLC with molecular driver mutations. We observed and collected the data on TEs occurring from peri-diagnosis until death and demonstrated the life-term TE risk since lung cancer diagnosis. In addition, we included and recorded both arterial and venous thromboembolic events, which had some similarities and have sometimes been viewed as separate pathophysiological entities. Our study revealed higher venous events in ALK+ than in EGFR-mutated lung cancer and elevated arterial events in patients with EGFR-mutated lung cancer. The patients in the EGFR cohort were significantly older than those in the ROS1 and ALK cohorts, both in our study and in the real world. This partially explains the observation of higher arterial TEs in EGFR-mutated lung cancers. We demonstrated the different TE risks in these two entities of thrombosis in NSCLC with different molecular drivers.
Conclusion
The cumulative incidence of venous and arterial TEs throughout the cancer course was higher in patients with ROS1+ and ALK+ lung cancers. The incidence rates calculated by events per 1000 person-years for the three groups were not significantly different. The peri-diagnostic thromboembolic risk was especially high in the ROS1 population. When arterial and venous events were calculated separately, ROS1+ and ALK+ lung cancers had a higher incidence rate of venous TEs than EGFR+ lung cancer. EGFR-mutated NSCLCs had more arterial TEs, which was probably related to the older age of this population.
Acknowledgments
Funding
This work was supported by the National Taiwan University Hospital Yunlin branch [grant number NTUHYL110.S015] (H-Y. Wang). The funding source had no role in the study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.
Disclosure
J-YS has served as an advisory board member from AstraZeneca, Roche, Boehringer Ingelheim, Eli Lilly, Pfizer, Novartis, Merck Sharp & Dohme, Chugai Pharma, Ono Pharmaceutical, Takeda, CStone Pharmaceuticals, Janssen, and Bristol-Myers Squibb and received speaking honoraria from AstraZeneca, Roche, Boehringer Ingelheim, Eli Lilly, Pfizer, Novartis, Merck Sharp & Dohme, Chugai Pharma, Ono Pharmaceutical, and Bristol-Myers Squibb, as well as grant from Roche. S-GW has received speaking honoraria from Roche, AstraZeneca, and Pfizer. Y-TL has received speaking honoraria from ACT Genomics, AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Chugai Pharmaceutical, Eli Lilly, Janssen, Manudipharma, Merck Sharp; Dohme, Novartis, Pfizer, Roche, Takeda and TTY Biopharm; and travel expense from Pfizer. All other authors have declared no conflicts of interest.
Supplementary data
References
- 1.Noble S., Pasi J. Epidemiology and pathophysiology of cancer-associated thrombosis. Br J Can. 2010;102(suppl 1):S2–S9. doi: 10.1038/sj.bjc.6605599. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Walker A.J., Card T.R., West J., Crooks C., Grainge M.J. Incidence of venous thromboembolism in patients with cancer—a cohort study using linked United Kingdom databases. Eur J Cancer (Oxford, England: 1990) 2013;49:1404–1413. doi: 10.1016/j.ejca.2012.10.021. [DOI] [PubMed] [Google Scholar]
- 3.Blom J.W., Doggen C.J., Osanto S., Rosendaal F.R. Malignancies, prothrombotic mutations, and the risk of venous thrombosis. JAMA. 2005;293:715–722. doi: 10.1001/jama.293.6.715. [DOI] [PubMed] [Google Scholar]
- 4.Horsted F., West J., Grainge M.J. Risk of venous thromboembolism in patients with cancer: a systematic review and meta-analysis. PLoS Med. 2012;9 doi: 10.1371/journal.pmed.1001275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Tagalakis V., Levi D., Agulnik J.S., Cohen V., Kasymjanova G., Small D. High risk of deep vein thrombosis in patients with non-small cell lung cancer: a cohort study of 493 patients. J Thorac Oncol. 2007;2:729–734. doi: 10.1097/JTO.0b013e31811ea275. [DOI] [PubMed] [Google Scholar]
- 6.Hicks L.K., Cheung M.C., Ding K., et al. Venous thromboembolism and nonsmall cell lung cancer: a pooled analysis of National Cancer Institute of Canada Clinical Trials Group trials. Cancer. 2009;115:5516–5525. doi: 10.1002/cncr.24596. [DOI] [PubMed] [Google Scholar]
- 7.Blom J.W., Osanto S., Rosendaal F.R. The risk of a venous thrombotic event in lung cancer patients: higher risk for adenocarcinoma than squamous cell carcinoma. J Thromb Haemost. 2004;2:1760–1765. doi: 10.1111/j.1538-7836.2004.00928.x. [DOI] [PubMed] [Google Scholar]
- 8.Chew H.K., Davies A.M., Wun T., Harvey D., Zhou H., White R.H. The incidence of venous thromboembolism among patients with primary lung cancer. J Thromb Haemost. 2008;6:601–608. doi: 10.1111/j.1538-7836.2008.02908.x. [DOI] [PubMed] [Google Scholar]
- 9.Nichetti F., Russo G.L., Prelaj A., et al. ALK/ROS1 rearrangements: a real hallmark for thromboembolic events in cancer patients? Thromb Res. 2020;194:176–177. doi: 10.1016/j.thromres.2020.06.041. [DOI] [PubMed] [Google Scholar]
- 10.Ng T.L., Smith D.E., Mushtaq R., et al. ROS1 gene rearrangements are associated with an elevated risk of peridiagnosis thromboembolic events. J Thor Oncol. 2019;14:596–605. doi: 10.1016/j.jtho.2018.12.001. [DOI] [PubMed] [Google Scholar]
- 11.Alexander M., Pavlakis N., John T., et al. A multicenter study of thromboembolic events among patients diagnosed with ROS1-rearranged non-small cell lung cancer. Lung Cancer. 2020;142:34–40. doi: 10.1016/j.lungcan.2020.01.017. [DOI] [PubMed] [Google Scholar]
- 12.Sakuma M., Nakamura M., Yamada N., et al. Venous thromboembolism: deep vein thrombosis with pulmonary embolism, deep vein thrombosis alone, and pulmonary embolism alone. Circ J. 2009;73:305–309. doi: 10.1253/circj.cj-08-0372. [DOI] [PubMed] [Google Scholar]
- 13.Jang M.J., Bang S.M., Oh D. Incidence of venous thromboembolism in Korea: from the Health Insurance Review and Assessment Service database. J Thromb Haemost. 2011;9:85–91. doi: 10.1111/j.1538-7836.2010.04108.x. [DOI] [PubMed] [Google Scholar]
- 14.Silverstein M.D., Heit J.A., Mohr D.N., Petterson T.M., O’Fallon W.M., Melton L.J., 3rd Trends in the incidence of deep vein thrombosis and pulmonary embolism: a 25-year population-based study. Arch Intern Med. 1998;158:585–593. doi: 10.1001/archinte.158.6.585. [DOI] [PubMed] [Google Scholar]
- 15.Lee C.H., Lin L.J., Cheng C.L., et al. Incidence and cumulative recurrence rates of venous thromboembolism in the Taiwanese population. J Thromb Haemost. 2010;8:1515–1523. doi: 10.1111/j.1538-7836.2010.03873.x. [DOI] [PubMed] [Google Scholar]
- 16.Solomon B.J., Mok T., Kim D.-W., et al. First-line crizotinib versus chemotherapy in ALK-positive lung cancer. N Eng J Med. 2014;371:2167–2177. doi: 10.1056/NEJMoa1408440. [DOI] [PubMed] [Google Scholar]
- 17.Shaw A.T., Kim D.-W., Mehra R., et al. Ceritinib in ALK-rearranged non–small-cell lung cancer. N Eng J Med. 2014;370:1189–1197. doi: 10.1056/NEJMoa1311107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Gettinger S.N., Bazhenova L.A., Langer C.J., et al. Activity and safety of brigatinib in ALK-rearranged non-small-cell lung cancer and other malignancies: a single-arm, open-label, phase 1/2 trial. Lancet Oncol. 2016;17:1683–1696. doi: 10.1016/S1470-2045(16)30392-8. [DOI] [PubMed] [Google Scholar]
- 19.Drilon A., Siena S., Dziadziuszko R., et al. Entrectinib in ROS1 fusion-positive non-small-cell lung cancer: integrated analysis of three phase 1-2 trials. Lancet Oncol. 2020;21:261–270. doi: 10.1016/S1470-2045(19)30690-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Shaw A.T., Ou S.-H.I., Bang Y.-J., et al. Crizotinib in ROS1-rearranged non–small-cell lung cancer. N Eng J Med. 2014;371:1963–1971. doi: 10.1056/NEJMoa1406766. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Michels S., Massutí B., Schildhaus H.U., et al. Safety and efficacy of crizotinib in patients with advanced or metastatic ROS1-rearranged lung cancer (EUCROSS): a European phase II clinical trial. J Thorac Oncol. 2019;14:1266–1276. doi: 10.1016/j.jtho.2019.03.020. [DOI] [PubMed] [Google Scholar]
- 22.Wu Y.L., Yang J.C., Kim D.W., et al. Phase II study of crizotinib in East Asian patients with ROS1-positive advanced non-small-cell lung cancer. J Clin Oncol. 2018;36:1405–1411. doi: 10.1200/JCO.2017.75.5587. [DOI] [PubMed] [Google Scholar]
- 23.Wu S.G., Yu C.J., Tsai M.F., et al. Survival of lung adenocarcinoma patients with malignant pleural effusion. Eur Respir J. 2013;41:1409–1418. doi: 10.1183/09031936.00069812. [DOI] [PubMed] [Google Scholar]
- 24.Wu S.G., Chang Y.L., Yu C.J., Yang P.C., Shih J.Y. The role of PIK3CA mutations among lung adenocarcinoma patients with primary and acquired resistance to EGFR tyrosine kinase inhibition. Sci Rep. 2016;6 doi: 10.1038/srep35249. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Connolly G.C., Menapace L., Safadjou S., Francis C.W., Khorana A.A. Prevalence and clinical significance of incidental and clinically suspected venous thromboembolism in lung cancer patients. Clin Lung Cancer. 2013;14:713–718. doi: 10.1016/j.cllc.2013.06.003. [DOI] [PubMed] [Google Scholar]
- 26.Zer A., Moskovitz M., Hwang D.M., et al. ALK-rearranged non-small-cell lung cancer is associated with a high rate of venous thromboembolism. Clin Lung Cancer. 2017;18:156–161. doi: 10.1016/j.cllc.2016.10.007. [DOI] [PubMed] [Google Scholar]
- 27.Roopkumar J., Poudel S.K., Gervaso L., et al. Risk of thromboembolism in patients with ALK- and EGFR-mutant lung cancer: a cohort study. J Thromb Haemost. 2021;19:822–829. doi: 10.1111/jth.15215. [DOI] [PubMed] [Google Scholar]
- 28.Wang J., Hu B., Li T., et al. The EGFR-rearranged adenocarcinoma is associated with a high rate of venous thromboembolism. Ann Transl Med. 2019;7:724. doi: 10.21037/atm.2019.12.24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Al-Samkari H., Leiva O., Dagogo-Jack I., et al. Impact of ALK rearrangement on venous and arterial thrombotic risk in NSCLC. J Thorac Oncol. 2020;15:1497–1506. doi: 10.1016/j.jtho.2020.04.033. [DOI] [PubMed] [Google Scholar]
- 30.Muñoz-Unceta N., Zugazagoitia J., Manzano A., et al. High risk of thrombosis in patients with advanced lung cancer harboring rearrangements in ROS1. Eur J Cancer. 2020;141:193–198. doi: 10.1016/j.ejca.2020.10.002. [DOI] [PubMed] [Google Scholar]
- 31.Chou S.-C., Pai C.-H., Lin S.-W., Tien H.-F. Incidence and risk factors for venous thromboembolism in a cohort of Taiwanese patients with lung, gastric, pancreatic cancers or lymphoma. J Formos Med Assoc. 2022;121(1 Pt 2):360–366. doi: 10.1016/j.jfma.2021.04.025. [DOI] [PubMed] [Google Scholar]
- 32.Wahrenbrock M., Borsig L., Le D., Varki N., Varki A. Selectin-mucin interactions as a probable molecular explanation for the association of Trousseau syndrome with mucinous adenocarcinomas. J Clin Invest. 2003;112:853–862. doi: 10.1172/JCI18882. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Kim H., Jang S.J., Chung D.H., et al. A comprehensive comparative analysis of the histomorphological features of ALK-rearranged lung adenocarcinoma based on driver oncogene mutations: frequent expression of epithelial-mesenchymal transition markers than other genotype. PLoS One. 2013;8 doi: 10.1371/journal.pone.0076999. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.