|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 24, No 30 (October 20), 2006: pp. 4928-4932 © 2006 American Society of Clinical Oncology. DOI: 10.1200/JCO.2006.06.5870 Unsuspected Pulmonary Emboli in Cancer Patients: Clinical Correlates and Relevance
From the Division of Hematology, Department of Medicine; Department of Radiology; and the Department of Preventive Medicine, University of Southern California-Keck School of Medicine, Los Angeles, CA Address reprint requests to Howard A. Liebman, MD, Division of Hematology, Kenneth Norris Jr Cancer Center, Rm 3466, 1441 Eastlake Ave, Los Angeles, CA 90033; e-mail: liebman{at}hsc.usc.edu
PURPOSE: Advances in computed tomography (CT) scanning have led to the detection of unsuspected pulmonary emboli (PE) on routine cancer staging scans. We hypothesized that these patients had signs or symptoms suggestive of PE that may have been overlooked by their health care providers. PATIENTS AND METHODS: A retrospective chart review was performed on 59 patients found on routine cancer staging CT scans to have unsuspected PE. Information on patient demographics, malignancy characteristics, risk factors for venous thromboembolism (VTE), and symptoms was recorded. A retrospective case-control analysis was then performed using two age- and stage-matched control patients for each patient who had similar staging CT scans performed during the same period. RESULTS: Fifty-two patients with unsuspected PE were identified. Forty-four percent had signs or symptoms commonly associated with PE; when fatigue was included, 75% were symptomatic. Ninety-two control patients were identified for 46 of the case patients. Patients with unsuspected PE were significantly more likely to have had a prior history of VTE (20% v 3%; P = .007). The patients with PE were significantly more likely than control patients to complain of fatigue (54% v 20%; P = .0002) and shortness of breath (22% v 8%; P = .02). There was no difference between the groups in administration of chemotherapy within 30 days, central venous catheter use, or erythropoietin therapy. CONCLUSION: Seventy-five percent of patients found to have unsuspected PE on cancer staging CT scans were symptomatic. Fatigue and shortness of breath were significantly more common in patients with unsuspected PE than in control patients.
The association between cancer and the development of symptomatic venous thromboembolism (VTE) is well documented.1-3 In addition to hemostatic activation induced by the malignancy, cancer patients are exposed to a variety of exogenous factors that place them at increased risk for VTE.2-7 The true incidence of VTE in cancer patients remains uncertain, despite the plethora of data linking VTE to morbidity and mortality in cancer patients. Autopsy studies and oncologist survey data suggest that it remains an underdiagnosed and undertreated phenomenon.8-10 Furthermore, recent advances in the sensitivity of computed tomography (CT) scans used for routine scanning of the chest have revealed an increase in the detection of nonfatal pulmonary emboli (PE) particularly among cancer patients that were not suspected clinically by the ordering physician.11-13 It is unknown whether the PE in these patients are clinically relevant. Since late 2002, our institution has been using a 16-slice multidetector CT (MDCT) scan for routine cancer staging. During a 20-month period between May 2003 and January 2005, we identified patients prospectively in whom routine staging scans revealed PE unsuspected by their referring physician. We then performed a retrospective chart review of these patients and a set of age- and stage-matched controls who underwent similar staging studies, attempting to characterize the patients' risk factors for VTE, malignancy characteristics, and evidence of signs or symptoms suggestive of PE.
Patients All patients with PE reported in this study were diagnosed by the department of Radiology at the Norris Comprehensive Cancer Center (Los Angeles, CA) from all adult cancer patients undergoing cancer staging CT scans between May 2003 and January 2005. The diagnosis of PE in these patients was confirmed by independent but nonblinded review by one of the authors (W.B.). A second independent blinded review of all patients with subsegmental PE and their matching controls was performed by a second radiologist (V.D.). The study was approved by the Institutional Review Board of the University of Southern California, Health Sciences Campus. Patients were included if their staging CT scan revealed a new PE and if their medical chart was available for review. Patients with a known VTE within 1 year of the CT study, patients undergoing anticoagulation therapy for known VTE, or whose studies were for suspected PE, were excluded from this analysis. Patient information was codified to remain anonymous after the initial data gathering.
Clinical Data VTE risk factors were obtained from consultation histories. A prior history of VTE included any deep vein thrombosis or PE occurring more than 1 year before the diagnostic CT scan. The patient was considered to have had recent surgery if major surgery was performed within 2 months of the date of the CT scan. Immobilization was considered present if the patient had been confined to a bed more than 75% of the time for at least 2 weeks before the CT scan. An indwelling central line was considered present if a line was reported on the CT scan. The use of chemotherapy was identified and recorded based on chemotherapy administration records. Chemotherapy was recorded as administered within 30 days, between 30 and 90 days, or more than 90 days before the diagnostic CT scan. Chemotherapeutic agents, as well as hormonal therapy and thalidomide, were included in the data analysis. Use of short- or long-acting erythropoietin was recorded as present if such medications were administered within 1 month before the staging CT scan. An attempt was made to identify two age- and stage-matched controls with the same primary malignancy for each case patient. These controls were drawn from radiology records spanning the same time period as the case patients. Controls were excluded if they were receiving anticoagulation therapy for VTE or if they were diagnosed with any VTE within 1 year of the diagnostic CT scan. Stage was defined as the TNM staging that befit the extent of the patient's tumor at the time of the CT scan. If it was not possible to identify a stage-matched patient whose age was within 5 years of the case patient, a patient of the same stage but older age was selected as the control. In a few instances, case patients had early-stage disease. If similarly staged control patients could not be found, age-matched patients with more advanced-stage tumors were selected as controls.
Radiologic Data
Statistical Methods
Patient and Control Selection Fifty-nine cancer patients with unsuspected PE were identified by the attending radiologist (W.B.) between May 2003 and January 2005 at the Norris Cancer Center. Seven patients were excluded because they were being treated actively with anticoagulation for a previously diagnosed VTE. We were able to identify two matched controls (n = 92) who had staging CT scan performed during the same period for each of 46 case patients. Six case patients were excluded from the case-control analysis because age- and stage-matched controls with the same primary malignancy could not be identified. We were also unable to find controls within 5 years of age for seven case patients, and therefore selected 10 control participants who were not within 5 years of age of the matching case, but were matched on tumor type and stage. One case patient with early-stage disease could not be matched for stage and another could only be matched with one control; therefore, three controls with the same primary but more advanced stage were selected.
Clinical Characteristics of Patients and Controls
The reported risk factors for VTE in the case and control patients are presented in Table 2. Patients with unsuspected PE were significantly more likely to have had a prior history of VTE (20% v 3%; P = .007). The case patients were also more likely to have had major surgery within 2 months of their diagnostic CT scan (11% v 2%; P = .06), although this did not reach statistical significance. There were no significant differences between case and control patients in their history of immobilization, presence of a central venous line, use of estrogen or thalidomide therapy, or use of erythropoietin. There was also no difference between case patients and controls in the number of patients who had received chemotherapy within 30 days.
The reported signs and symptoms for the case patients and controls are listed in Table 3. The patients with unsuspected PE were significantly more likely than control patients to complain of fatigue (54% v 20%; P = .0002) and shortness of breath (22% v 8%; P = .02). The case patients, however, did not have a higher incidence of tachycardia, limb pain or swelling, or chest pain.
The only significant difference in hematologic characteristics between the case and control patients was in mean platelet count, which was higher in the case patients (240 x 109/L v 220 x 109/L; P = .04). The median hemoglobin levels were not significantly different between case and control patients (11.9 v 12.2 g/dL; P = .11). The prevalence of anemia with a hemoglobin less than 10 g/dL was slightly greater among case patients than controls (13% v 8%; P = .36), but this difference was not statistically significant. The median hemoglobin level of patients with shortness of breath was not significantly different from that of patients without shortness of breath (11.9 v 12.1 g/dL; P = .55). To assess whether fatigue was actually related to anemia rather than to PE, we performed a multiple logistic regression model correcting fatigue for age and anemia. In this model, fatigue was still associated independently with PE, conferring an odds ratio of 4.82 (P = .0002; data not shown). Two patients in this study with unsuspected PE did not receive antithrombotic therapy at the discretion of their treating oncologists. Repeat CT studies performed 7 and 8 months later found new multiple thrombi in both patients, with symptomatic right and left main PE in one patient.
Radiologic Data Thirty-one patients (60%) had PE in more than one segment of the pulmonary arterial tree. The most proximal locations of PE were as follows: eight (15%) main pulmonary artery thrombi, 19 (37%) lobar thrombi, 11 (21%) segmental thrombi, and 14 (27%) subsegmental thrombi.
The use of helical MDCT has improved visualization of the pulmonary vasculature in the middle and peripheral lung zones.14 Using MDCT, Storto et al11 found a more than two-fold higher rate of detection of unsuspected PE compared with historical reports obtained using single-detector CT scanners (3.4% v 1.5%). A striking number of those patients (70%) had malignancies, but a thorough analysis of their risk factors, malignancy characteristics, and symptoms was not provided. The question thus remains whether these patients were truly asymptomatic or whether the unsuspected finding of PE represented a lack of clinical suspicion on the part of the ordering physician. In fact, autopsy data suggest that clinicians may fail to correctly diagnose fatal PE in as many as 70% of hospitalized patients.8 In one series, although the percentage of patients dying from PE who had an underlying cancer diagnosis increased during a 24-year period, the clinical detection of PE as the immediate cause of death declined.9 Furthermore, VTE is the most common cause of postoperative death in cancer patients undergoing surgery,15 and cancer patients with a documented VTE are more likely to have a recurrent clot16; nonetheless, fewer than 10% of British oncologists surveyed in 2002 reported initiating prophylactic therapy in their surgical patients, and fewer than half initiated prophylaxis in cancer patients with a prior history of VTE.10 The use of a 16-slice MDCT scan for routine cancer staging at our institution has also increased the finding of unsuspected PE significantly. Between May 2003 and January 2005, 59 patients were identified (predominantly outpatients). In contrast, the total number of hospitalizations at our institution for symptomatic VTE during the 2 year period, preceding the acquisition of the MDCT, was 53 patients, of which only 19 patients had symptomatic PE. Patients with PE were significantly more likely to have had prior VTE and recent surgery than were control patients. Other risk factors such as central venous catheters, hormonal therapy, thalidomide therapy, and erythropoietin therapy were not more prevalent among patients with unsuspected PE. Cancer is a well-known risk factor for postoperative VTE, increasing the risk of fatal PE more than three-fold compared with that in noncancer patients undergoing similar surgeries.17 For this reason, the American College of Chest Physicians recommends routine prophylaxis for cancer patients undergoing surgery.18 Recent surgery was associated with unsuspected PE in our study. We hypothesized that the cases of unsuspected PE were actually symptomatic; we therefore sought to evaluate the malignancy characteristics and risk factors in this population. Our retrospective review of patient medical records revealed a high incidence of PE-related symptoms. Of the initial 52 case patients, 44% were actually symptomatic with chest pain, shortness of breath, tachycardia, and/or limb pain or swelling. The percentage increased to 75% when fatigue was included in the symptom screening. To better ascertain whether the incidence of such symptoms was as high in cancer patients without PE, 92 age- and stage-matched control patients without evidence of PE and not undergoing therapy for VTE were identified for 46 of the case patients. Cancer patients with unsuspected PE were significantly more likely to have complained to their nurse or physician of fatigue and shortness of breath than were cancer patients without PE. Since cancer treatment in our institution is specific for tumor type and organ system, both the case patients and control patients report their symptoms to the same treating physicians and nurses. Neither metastases to the lungs nor anemia accounted for the higher incidence of shortness of breath in cancer patients with PE. Forty-three percent of the case patients and 46% of the controls had lung metastases (P = .72). The median hemoglobin of case patients with shortness of breath was 11.9 g/dL, and that of controls without shortness of breath was 12.1 g/dL. Fatigue was more common among cancer patients with unsuspected PE than among the control patients. Patients with fatigue had lower median hemoglobin levels than patients without fatigue (11.5 v 12.5 g/dL; P = .0006); however, fatigue remained a significant predictor of PE even after further adjustment for age and anemia (P = .0002). Fourteen of the 52 patients found to have unsuspected PE in our study had isolated subsegmental clots. However, 15% of the patients with PE had main pulmonary artery clots and 60% had multiple emboli. The clinical significance of small pulmonary emboli and of isolated subsegmental emboli remains uncertain.19 Our results must be viewed with the usual caveats inherent to retrospective data collection. We may have underestimated the treating physicians' suspicion for PE, given that we extracted information from the medical record. In addition, data extraction was performed by investigators who were not blinded to the CT scan results, and although criteria for data extraction had been developed before medical record review, we cannot exclude a potential bias in the assessment of patient symptoms. Our study suggests that the high sensitivity of MDCT scanning may reveal an inadequate index of suspicion for clinically significant PE among providers caring for cancer patients. In addition, the presence of these unsuspected PE results in an increase in patient symptoms and a reduction in quality of life for the cancer patient. The delineation of fatigue as a manifestation of PE could represent an inability of patients to differentiate fatigue from other symptoms such as dyspnea with exertion. However, these symptoms should increase the level of suspicion in medical oncologists for PE in their cancer patients.
The authors indicated no potential conflicts of interest.
Supported by a grant from the Vanoff Family Foundation and Max Gondon. Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Nordstrom M, Lindblad B, Anderson H, et al: Deep venous thrombosis and occult malignancy: An epidemiological study. BMJ 308:891-894, 1994 2. Otten HMB, Mathijssen J, Cate HT, et al: Symptomatic venous thromboembolism in cancer patients treated with chemotherapy: An underestimated phenomenon. Arch Intern Med 164:190-194, 2004 3. Levine M, Gent M, Hirsh J, et al: The thrombogenic effect of anticancer drug therapy in women with stage II breast cancer. N Engl J Med 318:404-407, 1988[Abstract] 4. von Tempelhoff G, Dietrich M, Hommel G, et al: Blood coagulation during adjuvant epirubicin/cyclophosphamide chemotherapy in patients with primary operable breast cancer. J Clin Oncol 14:2560-2568, 1996[Abstract] 5. Rickles FR, Hancock WW, Edwards RI, et al: Antimetastatic agents, the role of cellular procoagulants in the pathogenesis of fibrin deposition in cancer and in the use of anticoagulants and/or antiplatelet drugs in cancer treatment. Semin Thromb Hemost 14:88-94, 1988[Medline] 6. Sutherland DE, Weitz I, Liebman H: Thromboembolic complications of cancer: Epidemiology, pathogenesis, diagnosis and treatment. Am J Hematol 72:43-52, 2003[CrossRef][Medline] 7. Khorana AA, Francis CW, Culakova E, et al: Risk factors for chemotherapy-associated venous thromboembolism in a prospective observational study. Cancer 104:2822-2829, 2005[CrossRef][Medline] 8. Stein PD, Henry JW: Prevalence of acute pulmonary embolism among patients in a general hospital and at autopsy. Chest 108:978-981, 1995 9. Karwinski B, Svendsen E: Comparison of clinical and postmortem diagnosis of pulmonary embolism. J Clin Pathol 42:135-139, 1989 10. Kirwan CC, Nath E, Byrne GJ, et al: Prophylaxis for venous thromboembolism during treatment for cancer: Questionnaire survey. BMJ 327:597-598, 2003 11. Storto ML, DiCredico A, Guido F, et al: Incidental detection of pulmonary emboli on routine MDCT of the chest. AJR Am J Roentgenol 184:264-267, 2005 12. Gosselin MV, Rubin GD, Leung AN, et al: Unsuspected pulmonary embolism: Prospective detection on routine helical CT scans. Radiology 208:209-215, 1998 13. Sebastian AJ, Paddon AJ: Clinically unsuspected pulmonary embolism- an important secondary finding in oncology CT. Clin Radiol 61:81-85, 2006[CrossRef][Medline] 14. Raptopoulos V, Boiselle PM: Multi-detector row spiral CT pulmonary angiography: Comparison with single-detector row spiral CT. AJR Am J Roentgenol 221:606-613, 2001 15. Agnelli G, Bolis G, Capussotti L, et al: A clinical outcome-based prospective study on venous thromboembolism after cancer surgery: The @RISTOS project. Ann Surg 243:89-95, 2006[CrossRef][Medline] 16. Prandoni P, Lensing AWA, Cogo A, et al: The long-term clinical course of acute deep venous thrombosis. Ann Intern Med 125:1-7, 1996 17. Kakkar AK, Haas S, Wolf H, et al: Evaluation of perioperative fatal pulmonary embolism and death in cancer surgical patients: The MC-4 cancer substudy. Thromb Haemost 94:867-871, 2005[Medline] 18. Geerts WH, Pineo GF, Heit JA, et al: Prevention of venous thromboembolism: The seventh ACCP conference on antithrombotic and thrombolytic therapy. Chest 126:338-400, 2004[CrossRef] 19. Goodman L: Small pulmonary emboli: What do we know? Radiology 234:654-658, 2005 Submitted March 13, 2006; accepted August 15, 2006.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2006 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 P values. All data were analyzed using SAS, Version 9.0 (SAS Institute, Cary, NC). 
