|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 23, No 22 (August 1), 2005: pp. 5211-5216 © 2005 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.00.745 Abnormal Cytology Predicts Poor Prognosis in Cancer Patients With Pericardial EffusionFrom the Division of Cardiovascular Medicine, Brigham and Women's Hospital, Boston, MA Address reprint requests to Heather L. Gornik, MD, Cardiovascular Medicine Division, Brigham and Women's Hospital, 75 Francis St, Boston, MA 02115; e-mail: hgornik{at}partners.org
PURPOSE: Pericardial tamponade is a life-threatening disorder caused by varying medical conditions. Malignancy and complications of its treatment are a common cause of pericardial effusion. The natural history of pericardial effusion remains largely unknown. We investigated the association of malignancy with adverse outcomes after pericardiocentesis. PATIENTS AND METHODS: Consecutive patients undergoing pericardiocentesis at a single institution between January 1, 1999, and January 31, 2003, were included. Death was confirmed with the Social Security Death Index. Survival estimates were obtained by the Kaplan-Meier method. Cox regression was performed to determine the clinical characteristics associated with death. RESULTS: Two hundred nineteen patients underwent pericardiocentesis during the study period. The effusion was cancer-related in 43.8% of cases. Median survival was 59.6 weeks (95% CI, 24.3 to 94.8 weeks). During the follow-up period, 47.9% of patients died. Cancer-related pericardial effusion was associated with decreased survival (median, 15.1 weeks). Abnormal fluid cytology was further associated with poor prognosis among patients with malignancy (median survival, 7.3 v 29.7 weeks; P = .022). Patients with cancer-related pericardial effusion were more likely to require repeat pericardiocentesis (OR = 6.0; P = .001) and pericardial surgery (odds ratio [OR] OR = 5.7; P < .001). Cancer-related effusion and abnormal cytology were independent predictors of death in a multivariate model. CONCLUSION: Malignancy is the most common cause of pericardial effusion in a tertiary care center. Cancer-related pericardial effusion is associated with adverse outcomes, and abnormal cytology further worsens prognosis. The poor survival among cancer patients with pericardial effusion and abnormal fluid cytology may have important implications for management.
Pericardial tamponade is a life-threatening disorder caused by excessive fluid accumulation in the pericardial space leading to extracardiac compression and hemodynamic compromise.1 Pericardial effusion is associated with many underlying medical disorders, among which malignancy, infection, and complication of a recent cardiovascular procedure are the most common.2-7 The most common etiology of a large pericardial effusion varies among case series, primarily determined by the patient population of the reporting institution. In a tertiary-care academic center, malignancy is the most common cause of pericardial effusion and subsequent tamponade.4-6 Indeed, pericardial disease is one of the more common indications for cardiovascular consultation among hospitalized cancer patients. Case series, many including patients from decades past, have established that malignant pericardial effusion is associated with a high rate of recurrence and poor prognosis.7,8 In the modern era with current oncologic therapeutics, the natural history of malignant pericardial effusion is largely undefined. The additional prognostic value of pericardial fluid cytology in cancer patients with pericardial effusion remains controversial, with conflicting results in the medical literature.9-13 Symptomatic pericardial effusions are typically managed with percutaneous drainage.8,14 No randomized, controlled studies have established the optimal initial approach to the management of symptomatic pericardial effusion. In most cases, pericardiocentesis is both therapeutic and diagnostic.6,8,15 Surgery is usually reserved for cases of inadequate percutaneous drainage or for rapid fluid reaccumulation.8,14 Less common approaches to the management of recurrent pericardial effusion include serial pericardiocentesis procedures or prolonged drainage with an indwelling catheter, percutaneous balloon pericardiotomy, and intrapericardial sclerosis.8,14 We sought to investigate the natural history of pericardial effusion due to malignancy ("cancer-related pericardial effusion") in contrast to effusion of other etiology ("cancer-unrelated pericardial effusion"). We hypothesized that cancer-related pericardial effusion portends poor outcome, and that the presence of abnormal cytology in the pericardial fluid is further predictive of adverse prognosis.
We identified all consecutive patients who underwent fluoroscopy-guided pericardiocentesis at Brigham and Women's Hospital between January 1, 1999, and January 31, 2003. Pericardiocentesis is performed almost exclusively under fluoroscopic guidance at our institution. Patients were excluded from the study if a pericardiocentesis was attempted but was technically unsuccessful or if the procedure was performed for an indication other than the drainage of pericardial fluid (eg, to allow for access to the pericardial space for a cardiac electrophysiology study). The protocol was approved by the institutional review board of Partners Healthcare Systems Inc. The electronic medical record of each patient was reviewed to obtain baseline demographic and clinical data. History of malignancy, either active or in remission, was ascertained. For patients with a history of malignancy, the type of malignancy, as well as any exposure to radiation therapy (XRT) was ascertained. Relevant XRT exposure included radiation to the chest, neck, breast, mantle XRT for lymphoma, and total-body irradiation for bone marrow transplantation. In addition, clinical presentation and symptoms at the time of periocardiocentesis were obtained. Pericardial fluid cytolopathology and microbiology analyses were recorded when available. All medical records were reviewed by one investigator for consistency in the ascertainment of data. After review of all available medical record data, the most likely etiology of the effusion was determined. The effusion was classified as "definitely malignant" if pericardial fluid cytology demonstrated malignant, suspicious, or atypical cells in a patient with active malignancy, and if no other cause of the effusion was identified. Pericardial effusions were classified as "probably malignant" if the effusion occurred in a patient with a known active malignancy, if no other cause was identified, and if cytology did not reveal malignant cells. Effusions were classified as due to radiation therapy ("post-XRT") if the patient had received radiation therapy to the chest or neck in the past, if the patient's underlying malignancy was in remission or cured, and if pericardial fluid cytology was normal. Effusions were classified as due to bone marrow transplantation (BMT; "post-BMT") if the patient had undergone bone marrow transplantation for hematologic malignancy at any time, if the cytology was without evidence of recurrent hematologic malignancy, and if no other etiology of the effusion could be identified. Pericardial effusion following BMT may be due to the toxicities of conditioning chemotherapy or total-body irradiation, graft-versus-host disease, or cytomegalovirus infection.16 Effusions determined to be "definitely malignant," "probably malignant," "post-XRT," or "post-BMT," were classified as cancer-related pericardial effusions. Cancer-related pericardial effusions were further classified by underlying malignancy type into one of four categories: lung cancer; breast cancer; hematologic malignancy; and other solid tumors, including cancer of unknown primary. All other pericardial effusions were categorized by etiology, including post–cardiothoracic surgery (within 6 months), iatrogenic (ie, complication of cardiac catheterization or electrophysiology procedure), infectious, post–myocardial infarction (post-MI), traumatic, or rheumatologic. If the etiology of the effusion could not be determined or if there were a number of possible etiologies, the effusion was classified as being of "indeterminate" etiology. This category included a number of effusions that would have been classified as idiopathic or of viral etiology in other case series. Clinical follow-up through January 31, 2004, was ascertained for all patients using the electronic medical record. Vital status as of January 31, 2004, was confirmed by query of the Social Security Death Index, an online version of which has been validated as a tool for clinical research.17,18 Follow-up was defined as the time from pericardiocentesis until death or last documented follow-up visit for patients who were alive as of January 31, 2004. In addition to mortality, patients were followed-up for the need for repeat pericardiocentesis and referral for pericardial surgery. Pericardial surgical procedures included subxiphoid pericardiostomy ("pericardial window"), video-assisted thoracosocopic pericardiectomy, and pericardiectomy ("pericardial stripping").
Continuous variables were compared using the independent t test. Categorical variables were compared using the
Demographics Two hundred twenty-nine patients were identified. Ten patients were excluded. A total of 219 patients were included in the analysis. Of these 219 patients, 186 (84.9%) had documented symptoms (eg, chest discomfort, dyspnea, syncope) or signs (tachycardia, paradoxical pulse, hypotension) attributable to the pericardial effusion. The mean age of patients (± standard deviation) was 57.5 ± 1.0 years (range, 20.8 to 93.5 years), and 53.0% of patients were male. Of the 219 patients, 57.5% (126 of 219) were currently undergoing treatment for cancer or had a history of malignancy, and 29.2% of all patients had received XRT. Of the 10 patents excluded from the analysis, pericardiocentesis was unsuccessful in seven cases and did not drain the pericardial effusion. Two of these failed procedures were undertaken on an emergent basis for patients who developed hemodynamic compromise during a cardiac procedure (electrophysiology study and coronary angioplasty). In the three remaining cases, the pericardial space was entered to allow for epicardial access for electrophysiology study, without the presence of a pericardial effusion
Pericardial Effusion Etiology
Pericardial Fluid Cytologic analysis was available for 182 pericardial effusions (83.1% of cases). In 37 cases (16.9%), pericardial fluid was not sent to the cytology laboratory for analysis. The majority of pericardial effusions with missing cytology (34 of 37) were cancer-unrelated, most commonly iatrogenic or due to cardiothoracic surgery. Pericardial fluid cytology was positive for malignant cells in 28.6% of (52 of 182) samples, and contained atypical or suspicious cells in an additional 8.8% (16 of 182). The pericardial fluid cytology was normal in the remaining 62.6% of (114 of 182) samples. Among patients with active malignancy or a history of cancer, pericardial fluid cytology was abnormal in 54.1% of cases.
Survival
Estimated Survival for Cancer-Related Versus Cancer-Unrelated Effusions There was no difference in age or sex among patients with cancer-related and cancer-unrelated pericardial effusion. Median survival among patients with cancer-related pericardial effusion was 15.1 ± 4.8 weeks. Median survival among patients with cancer-unrelated effusion could not be calculated using the Kaplan-Meier method given inadequate number of events during the follow-up period, but was greater than 200 weeks (Fig 2). Cancer-related pericardial effusion was associated with a 72.5% reduction in estimated mean survival compared with cancer-unrelated effusion (45.6 v 166.0 weeks, log-rank P < .0001). Cancer type was further associated with survival. Patients with lung cancer and other solid tumors had reduced median survival (11.1 and 5.1 weeks, respectively) compared with patients with breast cancer and hematologic malignancies (43.0 and 45.6 weeks, respectively; log-rank P = .0012).
Among patients with cancer-related pericardial effusion, the presence of abnormal cytology further reduced survival (Fig 3). Median survival among patients with cancer-related pericardial effusion and abnormal cytology was 7.3 v 29.7 weeks among patients with cancer-related effusion and normal cytology (P = .0221). Among patients with cancer-related pericardial effusion due to lung cancer, abnormal cytology was also associated with reduced median survival (40.4 weeks normal cytology v 6.1 weeks abnormal cytology; P = .0139).
In a multivariate model of survival following pericardiocentesis, etiology of the effusion (cancer-related v cancer-unrelated) and abnormal cytology were independent predictors of survival (Table 3).
Repeat Procedures Patients with cancer-related effusion had a fivefold greater likelihood of requiring repeat percutaneous procedure and referral for pericardial surgery when compared with patients with cancer-unrelated effusions (Fig 4).
We found malignancy and complications of its treatment to be the most common cause of pericardial effusion requiring pericardiocentesis drainage in a tertiary care referral center. The majority of cases were cytology-positive malignant pericardial effusions or pericardial effusions that were probably malignant but did not meet formal cytopathologic criteria. In the appropriate clinical setting, normal cytology does not exclude the possibility of cancer-related effusion as a single pericardial cytology specimen has only moderate sensitivity for the detection of malignant cells (67% to 92% in case series).15,22,23 Radiation therapy and bone marrow transplantation accounted for a small percentage of cancer-related pericardial effusions, with too few cases to analyze independently for survival outcomes. Postprocedural complication was the second most common cause of effusion, either from postpericardiotomy syndrome of cardiothoracic surgery or as an iatrogenic complication of an invasive cardiac procedure. Infection accounted for only a small percentage of cases of pericardial effusion in our series. In nearly one quarter of cases in our series, the effusion was due to multiple potential etiologies or was deemed idiopathic. Our finding of malignancy as the most common cause of pericardial effusion is consistent with published case series from tertiary-care centers, including a case series of patients treated with ultrasound-guided pericardiocentesis at the Mayo Clinic.4-6 Other case series in the literature have found a greater prevalence of pericardial effusion of infectious and idiopathic etiology, though the diagnosis of infectious pericardial effusion was typically made by extensive microbiologic testing not routinely performed in clinical practice.2,3,24 The most likely etiology of a pericardial effusion at a given center is highly dependent on the patient population under study. As our institution is closely affiliated with a cancer referral center, the finding of malignancy as the most likely etiology is not surprising. Cancer-related pericardial effusion was associated with a dramatic and significant decrease in estimated survival compared with patients with noncancer-related effusions. The median survival among patients with cancer-related pericardial effusion was 15.1 weeks. The median survival among patients with cancer-unrelated pericardial effusion was greater than 200 weeks, but could not be calculated due to the relatively few deaths during the follow-up period. A similarly poor survival among 275 patients with malignant pericardial effusion has been reported by the Mayo clinic (median survival, 19.3 weeks), though the cohort included cancer patients treated before the modern era of chemotherapy (1979 to 1998).10 Similar findings have been reported in patients with malignant pericardial effusion treated with a primary surgical approach.25-30 Our study advances these findings and suggests that advances in oncology have had little impact on the prognosis of malignant pericardial effusion due to chemotherapy-insensitive tumors (eg, lung cancer, unknown primary) in contrast to chemosensitive tumors such as breast cancer and lymphoma (see Discussion). We identified the presence of malignant or atypical cells within a cancer-related pericardial effusion as a further harbinger of poor prognosis. The estimated median survival among such patients was extremely poor (7.3 weeks). Pericardial cytology remained an independent predictor of death in a multivariate model. Percardial effusions due to lung cancer or solid tumors other than breast cancer were associated with limited survival compared with effusions due to breast cancer or hematologic malignancies. Our findings are consistent with most other case series, but opposite of a recent study that found no difference in survival among patients with non–small-cell lung cancer and pericardial tamponade with and without malignant cytology.6,9 Lung cancer accounted for one third of cases of cancer-related pericardial effusion in our study, and abnormal cytology was further predictive of adverse prognosis in these patients (median survival, 6.1 weeks). In addition to an impact on survival, patients with cancer-related pericardial effusion were much more likely to undergo repeat pericardiocentesis or to be referred for pericardial surgery, such as subxiphoid pericardial window, video-assisted thoracoscopic pericardiectomy, or pericardial stripping procedure. Given the limited survival of patients with cytology-positive pericardial effusion, nonsurgical measures for palliation of recurrence should be considered, such as echocardiographic surveillance and serial pericardiocenteses or percutaneous balloon pericardiotomy. Most importantly, there should be an ongoing discourse between the oncology team and cardiovascular specialists with regard to the optimal management of these patients. Lack of adequate information regarding prognosis has been identified as a barrier to utilization of palliative care and hospice services by oncologists.31,32 We found malignancy to be the most common cause of pericardial effusion in a tertiary care center. Cancer-related pericardial effusion markedly reduced survival, and this trend is aggravated by the presence of abnormal cytology. The grim prognosis among cancer patients with pericardial effusion and abnormal fluid cytology may have important implications for management and may lead to consideration of a palliative approach to care.
The authors indicated no potential conflicts of interest.
Supported by the Clinical Investigator Training Program (H.L.G.): Harvard/MIT Health Sciences and Technology—Beth Israel Deaconess Medical Center, in collaboration with Pfizer Inc. Authors' disclosures of potential conflicts of interest are found at the end of this article.
1. Spodick DH: Acute cardiac tamponade. N Engl J Med 349: 684-690, 2003 2. Sagrista-Sauleda J, Merce J, Permanyer-Miralda G, et al: Clinical clues to the causes of large pericardial effusions. Am J Med 109: 95-101, 2000[CrossRef][Medline] 3. Levy PY, Corey R, Berger P, et al: Etiologic diagnosis of 204 pericardial effusions. Medicine 82: 385-391, 2003[Medline] 4. Corey GR, Campbell PT, Van Trigt P, et al: Etiology of large pericardial effusions. Am J Med 95: 209-213, 1993[CrossRef][Medline] 5. Colombo A, Olson HG, Egan J, et al: Etiology and prognostic implications of a large pericardial effusion in men. Clin Cardiol 11: 389-394, 1988[Medline]
6. Tsang TS, Enriquez-Sarano M, Freeman WK, et al: Consecutive 1127 therapeutic echocardiographically guided pericardiocenteses: Clinical profile, practice patterns, and outcomes spanning 21 years. Mayo Clin Proc 77: 429-436, 2002 7. Lindenberger M, Kjellberg M, Karlsson E, et al: Pericardiocentesis guided by 2-D echocardiography: The method of choice for treatment of pericardial effusion. J Intern Med 253: 411-417, 2003[CrossRef][Medline]
8. Vaitkus PT, Herrmann HC, LeWinter MM: Treatment of malignant pericardial effusion. JAMA 272: 59-64, 1994
9. Wang PC, Yang KY, Chao JY, et al: Prognostic role of pericardial fluid cytology in cardiac tamponade associated with non-small cell lung cancer. Chest 118: 744-749, 2000 10. Tsang TS, Seward JB, Barnes ME, et al: Outcomes of primary and secondary treatment of pericardial effusion in patients with malignancy. Mayo Clin Proc 75: 248-253, 2000[Abstract] 11. Edoute Y, Kuten A, Ben-Haim SA, et al: Symptomatic pericardial effusion in breast cancer patients: The role of fluid cytology. J Surg Oncol 45: 265-269, 1990[Medline] 12. Edoute Y, Malberger E, Kuten A, et al: Symptomatic pericardial effusion in lung cancer patients: The role of fluid cytology. J Surg Oncol 45: 121-123, 1990[Medline]
13. Cullinane CA, Paz IB, Smith D, et al: Prognostic factors in the surgical management of pericardial effusion in the patient with concurrent malignancy. Chest 125: 1328-1334, 2004
14. Soler-Soler J, Sagrista-Sauleda J, Permanyer-Miralda G: Management of pericardial effusion. Heart 86: 235-240, 2001
15. Meyers DG, Meyers RE, Prendergast TW: The usefulness of diagnostic tests on pericardial fluid. Chest 111: 1213-1221, 1997 16. Seber A, Khan SP, Kersey JH: Unexplained effusions: Association with allogeneic bone marrow transplantation and acute or chronic graft-versus-host disease. Bone Marrow Transplant 17: 207-211, 1996[Medline] 17. Schisterman EF, Whitcomb BW: Use of the Social Security Administration Death Master File for ascertainment of mortality status. Popul Health Metr 2: 2, 2004[CrossRef][Medline]
18. Sesso HD, Paffenbarger RS, Lee IM: Comparison of National Death Index and World Wide Web death searches. Am J Epidemiol 152: 107-111, 2000 19. Kaplan E, Meier P: Non-parametric estimation from incomplete observations. J Am Stat Assoc 53: 457-481, 1958[CrossRef] 20. Mantel N: Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Treat Rep 50: 163-170, 1966 21. Cox D: Regression models and life-tables. J R Stat Soc B 34: 187-220, 1972 22. Wiener HG, Kristensen IB, Haubek A, et al: The diagnostic value of pericardial cytology: An analysis of 95 cases. Acta Cytol 35: 149-153, 1991[Medline] 23. Edoute Y, Malberger E, Kuten A, et al: Cytologic analysis of pericardial effusion complicating extracardiac malignancy. Am J Cardiol 69: 568-571, 1992[CrossRef][Medline] 24. Merce J, Sagrista-Sauleda J, Permanyer-Miralda G, et al: Should pericardial drainage be performed routinely in patients who have a large pericardial effusion without tamponade? Am J Med 105: 106-109, 1998[CrossRef][Medline]
25. Moores DW, Allen KB, Faber LP, et al: Subxiphoid pericardial drainage for pericardial tamponade. J Thorac Cardiovasc Surg 109: 546-552, 1995 26. Wilkes JD, Fidias P, Vaickus L, et al: Malignancy-related pericardial effusion: 127 cases from the Roswell Park Cancer Institute. Cancer 76: 1377-1387, 1995[CrossRef][Medline]
27. Campbell PT, Van Trigt P, Wall TC, et al: Subxiphoid pericardiotomy in the diagnosis and management of large pericardial effusions associated with malignancy. Chest 101: 938-943, 1992 28. Olsen PS, Sorensen C, Andersen HO: Surgical treatment of large pericardial effusions: Etiology and long-term survival. Eur J Cardiothorac Surg 5: 430-432, 1991[Abstract]
29. Dosios T, Theakos N, Angouras D, et al: Risk factors affecting the survival of patients with pericardial effusion submitted to subxiphoid pericardiostomy. Chest 124: 242-246, 2003 30. Van Trigt P, Douglas J, Smith PK, et al: A prospective trial of subxiphoid pericardiotomy in the diagnosis and treatment of large pericardial effusion: A follow-up report. Ann Surg 218: 777-782, 1993[Medline] 31. Daugherty CK: Examining ethical dilemmas as obstacles to hospice and palliative care for advanced cancer patients. Cancer Invest 22: 123-131, 2004[CrossRef][Medline]
32. Lamont EB, Christakis NA: Complexities in prognostication in advanced cancer: "To help them live their lives the way they want to". JAMA 290: 98-104, 2003 Submitted September 1, 2004; accepted December 1, 2004.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2005 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 test. Survival estimates were obtained by the Kaplan-Meier method, and survival curves were compared using the log-rank test.
