Feasibility and impact of screening for venous thromboembolism in treatment-naive lung cancer patients–Results of a prospective cohort study
Valliappan Muthu1, Ramesh L Narasimhan1, Kuruswamy T Prasad1, Jasmina Ahluwalia2, Mandeep Garg3, Digambar Behera1, Navneet Singh1
1 Department of Pulmonary Medicine, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India
2 Department of Hematology, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India
3 Department of Radiodiagosis, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India
Correspondence Address:
Navneet Singh
Department of Pulmonary Medicine, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh
India
Source of Support: None, Conflict of Interest: None
CheckDOI: 10.4103/ijc.IJC_678_19
Background: Venous thromboembolism (VTE) in cancer remains underdiagnosed. This prospective study aimed to evaluate the feasibility of screening for VTE in lung cancer (LC) patients. We assess the incidence of VTE, its risk factors, and effects on overall survival (OS).
Methods: Consecutive treatment-naive LC patients were screened for deep venous thrombosis (DVT) with compression ultrasonography and pulmonary thromboembolism (PTE) with computed tomography pulmonary angiography (CTPA) at diagnosis and after 3 months of treatment. The incidence rate of VTE (DVT and/or PTE) was calculated. Risk factors associated with VTE were assessed using logistic regression analysis. All participants were followed-up to 1 year after enrollment. OS was compared in LC subjects with and without VTE, using the Cox proportional hazard analysis.
Results: Around 301 subjects with LC (stages IIIB-IV accounted for 83.1%) were enrolled, of which 16 had VTE (5.3%). The incidence rate of VTE was 90 per 1000 person-years (PY). PTE was asymptomatic in 27.3% of cases while all DVT episodes were symptomatic. The incidence rate of asymptomatic PTE identified during the screening was 17 per 1000 PY. The median duration from LC diagnosis to the VTE event was 96.5 days. Median OS was significantly less in VTE patients [161 versus 311 days; P = 0.007] and death was attributable to VTE in 50%. After adjusting for covariates, VTE (hazard ratio [HR] = 2.1), smoking (HR = 1.7), and Eastern cooperative oncology group performance status ≥2 (HR = 1.6) were independently associated with poor OS in LC.
Conclusions: VTE occurs in approximately 1 in 20 newly-diagnosed patients with LC and is associated with decreased OS. Screening for PTE may be considered even in resource-limited settings.
Keywords: Chemotherapy, embolism, lung cancer, overall survival, venous thrombosis
Key Message:
Venous thromboembolism remains underdiagnosed in lung cancer patients and is associated with decreased survival. Screening for pulmonary thromboembolism using computerized tomography pulmonary angiography at diagnosis and after 3 months of treatment is feasible and may be useful.
Valliappan Muthu and Ramesh L Narasimhan are contributed equally to the study.
IntroductionThe association between cancer and venous thromboembolism (VTE) is well-known. Cancers produce a hypercoagulable state and VTE may antedate the malignancy.[1] Lung cancer (LC) also has an increased risk of VTE.[2] LCs associated with VTE have an advanced stage at presentation and have a decreased overall survival.[2],[3] The prognosis of LC patients with VTE is worse than those without VTE.[4],[5] This underscores the need to identify asymptomatic VTE. However, most of the data on LC-related VTE are from retrospective studies, which have inherent limitations,[6],[7] and not all LC patients were screened in these studies.[2],[8] Evaluation of VTE was undertaken only in symptomatic individuals and the data was acquired from cancer registries.[2] Therefore, the reported incidence may be underestimated. Moreover, only one study had screened all subjects with LC for VTE. However, this study from China included only hospitalized subjects.[9] Further, data from Asia and, in particular, from India is sparse.[9],[10],[11],[12],[13] Thus, it remains uncertain whether screening all treatment-naive LC patients for VTE is feasible and necessary in the real-world scenario. In this study, we screened all newly-diagnosed LC patients for deep vein thrombosis (DVT) and pulmonary thromboembolism (PTE) at baseline and during follow-up.
MethodsStudy design
A prospective observational study conducted for 18 months.
Study population
Consecutive newly diagnosed patients with histologically/cytologically proven LC aged ≥18 years attending the outpatient clinic or hospitalized at our institute were enrolled (initial 6 months of the study) and followed-up for 1 year.
Subjects with advanced renal failure (estimated creatinine clearance of <30 mL/min), pregnant woman, and those who failed to provide informed consent were excluded. The study protocol was approved by the institutional ethics committee.
Estimated sample size
Approximately 500 new patients with LC register at our clinic annually, and we estimated that at least 250 subjects would be recruited in the initial 6 months (enrollment phase). In addition, we also planned to recruit hospitalized subjects with newly-diagnosed LC.
Objectives
The primary objective was to screen all included participants and ascertain the incidence rates of DVT, PTE (symptomatic and asymptomatic), and VTE (DVT and/or PTE). The secondary objectives were: (a) to identify the risk factors associated with VTE, and (b) to compare the clinical profile and survival of LC subjects with and without VTE.
Study protocol
We recorded the demographic variables, smoking status, a detailed history (including asymmetric limb swelling, chest pain, syncope, and sudden-onset breathlessness), and the presence of comorbid illnesses. Smoking status was quantified using the smoking index (the number of beedis, and/or cigarettes smoked multiplied by the number of years of smoking).[14],[15] We enquired about the history of VTE in the past or in the family. Risk factors such as central venous catheterization and hospitalizations were documented. A histological subtype of the tumor, tumor (T), nodes (N), and metastases (M) (TNM) stage (seventh edition), and performance status (Eastern cooperative oncology group [ECOG] and Karnofsky performance scale [KPS]) were recorded.[16],[17],[18],[19]
All subjects underwent complete blood count, renal function, and liver function tests at baseline and before each chemotherapy cycle. We also performed baseline fasting blood glucose, lipid profile, and coagulation profile (prothrombin time [PT], activated partial thromboplastin time [aPTT], and qualitative d-dimer).
Chemotherapy was administered at our hospital's daycare center by trained oncology nurses. Standard histology-guided platinum doublet protocols were used for chemotherapy. Patients with EGFR-mutation or ALK-rearrangements were treated appropriately with tyrosine kinase inhibitors (TKI).
Screening for DVT and PTE
The study participants underwent gray-scale ultrasonography with Doppler study of lower limbs at enrollment, to screen for DVT. Upper limb veins were evaluated if there was a clinical suspicion of DVT or there was a history of central venous catheterization. All participants were subjected to computed tomography with pulmonary angiography (CTPA) at baseline to screen for PTE (using either Siemens Somatom Definition Flash 128-slice CT scanner or Philips Brilliance iCT 256-slice CT scanner). The Doppler ultrasonography and CTPA were repeated after 3 months of treatment (at initial response assessment) or earlier if there was a clinical suspicion of VTE.
Response assessment and follow-up
We used the Response Evaluation Criteria in Solid Tumors (RECIST) criteria and the Common Toxicity Criteria for Adverse Events (version 3) to assess the radiological response and toxicity, respectively.[20] The CTPA performed at 3 months served to assess treatment response as well as to detect PTE. Accordingly, tumor response was graded as complete response (CR), partial response (PR), stable disease, or progressive disease (PD). The hematological and gastrointestinal toxicity were assessed at each visit, whereas the other toxicities were recorded after the completion of three chemotherapy cycles.[21],[22]
Calculation of incidence rate
The incidence rate was calculated as the number of new VTE (DVT and/or PTE) cases during the study period divided by the time each person was observed, totaled for all persons. The incidence rate was expressed as a number of cases per 1000 person-years (PY).
Statistical analysis
Statistical analysis was performed using the statistical package software (SPSS for Windows, version 22.0; IBM SPSS Inc; Armonk, NY: IBM Corp). We compared all parameters of interest between subjects with and without VTE. Categorical variables were compared using Chi-square or Fisher's exact test; while the paired t-test (normally distributed data) or Wilcoxon signed-rank test (for data not normally distributed) was used for continuous variables. Multivariate analysis was carried out with stepwise logistic regression analysis and Cox proportional hazard ratio models to determine the independent predictors of VTE and mortality (overall survival [OS]-calculated from date of initiation of chemotherapy till death or last follow-up), respectively. Kaplan-Meier survival curves were constructed to assess the effect of VTE on survival and the two groups were compared using the log-rank test. A P value of <0.05 was considered statistically significant.
ResultsPrimary objective
VTE in LC
VTE (DVT and/or PTE) was detected in 16 of the total 301 subjects (prevalence of 5.3%). Five subjects had only DVT, two had only PTE while nine had both DVT and PTE. At baseline, DVT and PTE were present in four and three subjects, respectively while the others developed VTE during follow-up. Three subjects (27.3%) with PTE were asymptomatic while all subjects with DVT were symptomatic.
The incidence rate of VTE was 90 per 1000 PY [Table 1]. The median duration from LC diagnosis to the VTE event was 96.5 days. The incidence rates for symptomatic and asymptomatic PTE were 45 and 17 per 1000 PY, respectively. All 16 subjects with VTE received therapeutic anticoagulation while thrombolysis was performed in four subjects with PTE [Table 2]. On a multivariate analysis [Table 3], chronic obstructive pulmonary disease [COPD] (Hazard ratio [HR] = 5.20, 95% confidence interval [CI] = 1.60–16.91, P = 0.006) an increased number of extrathoracic metastases (HR = 1.93, 95% CI 1.05–3.52, P = 0.033) were independently associated with the occurrence of VTE.
Table 3: Logistic regression analysis to determine the predictors of VTE in our cohort of subjects with LCSecondary objectives
Comparison of LC with and without VTE
The study participants (n = 301) included primarily outpatients (n = 286, 95%) and a small proportion were inpatients (n = 15, 5%). Subjects with VTE had a higher prevalence of COPD and a higher proportion of subjects with poor ECOG-PS [Table 4]; while the remaining baseline parameters were similar in subjects with (n = 16) and without VTE (n = 285). The mean (standard deviation) age of the study cohort was 59.6 (10.2) years (range 27-82 years), years, and the majority were men (n = 255, 84.7%). Most participants were current or former smokers (n = 234, 77%) with a median smoking index of 480 (interquartile range 292–745.2). Advanced disease at presentation was seen in the majority (stage IIIB and IV in 31.9% and 51.2%, respectively). The most common histology was adenocarcinoma (43.9%) followed by squamous cell carcinoma (34.6%). The treatment offered and the responses to the therapy were comparable in subjects with or without VTE [Table 5]. Chemotherapy (with or without radiotherapy) was the most common treatment modality; two patients underwent surgery and none received bevacizumab.
Table 4: Baseline characteristics and outcome of LC subjects with and without VTETable 5: The treatment offered and the responses in LC patients with or without VTESurvival
LC subjects with VTE had a significantly lower median survival [Figure 1] as compared to those without VTE (161 days, [95% CI = 79–243] versus 311 days, [95% CI = 270–352], P = 0.007). In the former group, the death was attributable to VTE in 50%. The median OS in subjects with symptomatic (n = 8; 170 days [95% CI = 84.5–256]) and asymptomatic PTE (n = 3; 230 days [95% CI = 61.4–444.6]) was similar. After adjusting for other covariates, the independent predictors of mortality in the study cohort were the presence of VTE, lower serum albumin levels, poor ECOG PS, and history of smoking [Table 6].
Figure 1: Kaplan-Meier survival curve demonstrating the difference in the survival of subjects with and without venous thromboembolismTable 6: Cox proportional hazards analysis to determine the factors influencing survival in subjects with LC DiscussionIn this study, we found a high incidence rate (90 per 1000 PY), and prevalence (5.3%) of VTE. VTE occurred in nearly one out of 20 newly-diagnosed LC patients. Screening could identify 27% more cases of PTE (asymptomatic). It is feasible to detect asymptomatic PTE with CT scans performed at diagnosis and during follow-up, without the need for additional scans. This is particularly important, as VTE in LC was found to be associated with a poor OS, irrespective of whether the VTE was symptomatic or not.
The reported incidence rate(40 to 110 per 1000 PY),[2],[3],[23],[24],[25],[26],[27] and prevalence of VTE (1.5 –21.5%) in LC varies widely across studies.[6],[28] This wide variation may be attributed to the methodological differences such as the study design, setting (ambulant versus hospitalized), inclusion criteria (non-small cell lung cancer [NSCLC] or SCLC), duration of follow-up, and the method used to diagnose VTE. The only prospective study to have screened all included subjects for VTE showed a higher prevalence (13.2%); however, this study included only hospitalized subjects.[9] In contrast, the prevalence of VTE in ambulant LC patients was 5.6% in Behrendt et al., 6.1% in Kuderer et al., and 4.8% in Joshi et al.[13],[23],[29]; which is similar to our data (5.3%).
The risk factors for VTE in LC include patient-related, tumor-related, and treatment-related factors (e.g. bevacizumab). The patient-related factors include the presence of multiple comorbid illnesses, obesity, thrombophilic states, and smoking status.[8] An advanced tumor stage and metastatic disease have also been consistently associated with an increased risk of VTE.[2],[3] NSCLC (adenocarcinoma higher risk than squamous) have a higher chance of developing VTE than SCLC.[2],[24] In our cohort, histology did not predict VTE. Rather, the presence of COPD and metastatic disease was associated with a higher risk of VTE after adjusting for stage, performance status, and smoking.
VTE has been an independent predictor of mortality in several studies, including ours.[2],[4],[5],[8],[10],[26],[30] In our study, >25% of the PTE were asymptomatic, while in another study 89% of PTE remained unsuspected.[9] PTE affects mortality regardless of whether it is, (a) symptomatic or not, and (b) central or peripheral.[31] Though some authors have observed a smaller effect size with an asymptomatic PTE than symptomatic PTE, both are associated with decreased survival.[5] Identifying PTE in LC thus becomes essential not only for its prognostic role but also because untreated PTE can increase the risk of death by four-fold.[10] Furthermore, cancer patients with incidentally detected PE are at increased risk of recurrence.[32] Thus, the importance of screening for VTE cannot be overemphasized. Scoring systems have been developed and validated to predict VTE in several cancers (including LC).[29],[33] In other cancers, predicting the occurrence of PTE is essential to identify high-risk patients and subsequently confirm the diagnosis. LC offers a unique opportunity where the imaging used for diagnosis and response assessment can be used to diagnose PTE, without any additional cost. Previous studies have demonstrated that incidental PTE could be identified by reviewing the multidetector CT or positron emission tomography-CT (PET-CT) images acquired during staging and restaging.[30],[34] Contrast-enhanced CT scan of the thorax would be required at baseline as well as during response assessment, and utilizing the same for prospectively detecting an unsuspected PTE would be a viable option, as demonstrated in the current study. On the contrary, DVT was always symptomatic in our cohort; hence, the additional benefit of screening asymptomatic subjects for DVT is questionable.
The strengths of our study include the prospective design and screening of all consecutive newly-diagnosed LC patients for VTE. Our study is limited by its small sample size obtained from a single center. We did not employ pulmonary embolism rule-out criteria (PERC) or any other screening questionnaire for VTE in our study.[35] We have not explored the role of tumor biology and genetic factors, which could be contributing to VTE.[25],[34],[36],[37] Majority of our subjects were treated with chemotherapy; whether the incidence of VTE would vary with different modalities of treatment (e.g., immunotherapy or TKI) remains to be seen.[25] Though VTE is known to be associated with poor OS, it is not certain whether interventions aimed at early diagnosis or treatment would improve survival.
In conclusion, the incidence rate of VTE is high, and VTE is associated with decreased OS in LC patients. Approximately, half of the deaths in LC patients with VTE were attributable to the thromboembolic event. Screening for PTE using a CT scan (the same scan used for diagnosis and treatment response assessment) can be considered and may aid in detecting additional cases of asymptomatic PTE.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
References
留言 (0)