Abstract
BACKGROUND:
Patients who have pleural effusions typically undergo thoracentesis with examination of pleural fluid in their initial assessment. However, limited data are available on the diagnostic yield of pleural fluid bacterial cultures, fungal and acid fast bacilli (AFB) smear and cultures in cancer patients.
METHODS:
We performed a retrospective cohort study of consecutive patients who had new onset pleural effusions and underwent an initial thoracentesis. The primary outcome was diagnostic yield of pleural fluid bacterial cultures, fungal and acid fast bacilli smear and cultures.
RESULTS:
Of 1637 patients, 1547 (94%) had evidence of active malignancy and 1359 (83%) had evidence of metastatic disease. Of the 1637 patients, 542 (33%) had high clinical suspicion of pneumonia within 14 days prior to thoracentesis. Only 14 patients (1.1%) had positive pleural fluid bacterial cultures, and only 6 of these positive cultures met the criteria for true pleural space infection.
CONCLUSIONS:
In cancer patients, pleural fluid bacterial cultures, fungal and acid fast bacilli smear and cultures should be ordered only selectively—when infection is suspected—because the incidence of positive results in cancer populations is very low.
Keywords: pleural effusion, thoracentesis
Introduction
Thoracentesis with pleural fluid analysis, including microbiological studies, is the most frequently used test for diagnosing pleural effusion that cannot be explained by clinical history.1 In patients with malignancy, pleural effusion may indicate spread of the disease; however, other causes such as congestive heart failure, infections, and paramalignant effusions are considered in the differential diagnosis.2 It is commonly recommended that microscopic examination of Gram-stained pleural fluid sediment be performed on all pleural fluid samples obtained by thoracentesis.3 For microbiological study of pleural fluid, it is recommended to inoculate all samples, including those that are not purulent, into both a sterile tube and an anaerobic blood culture vial, as this will increase the diagnostic yield.4
Even though the diagnostic yield for pleural fluid cultures has been studied in the general population in the past, there is very little information on the role of microbiological studies of pleural fluid in the cancer patient population.5–8
The aim of our study was to determine the frequency of positive results from pleural fluid bacterial cultures, fungal and acid fast bacilli (AFB) smear and cultures in patients with an existing or suspicious diagnosis of cancer.
Materials and Methods
This was a retrospective cohort study of patients with pleural effusions who underwent an initial thoracentesis at our institution from January 1, 2010, to December 31, 2013. Our Institutional Review Board (Committee 4 PA19–0159) approved this study.
We included all patients 18 years or older who had their first thoracentesis at our institution. We excluded patients with history of pleural procedures such as chest tube placement, pleurodesis, and surgery on the same side and patients who had prior thoracentesis on the same side.
We classified our patients as those with biopsy proven cancer, active within past 12 months, patients with proven malignancy whose cancer had been in remission for more than 12 months to 5 years; patients with a history of cancer whose disease had been in remission for more than 5 years; patients with no prior or current cancer (referred to us for suspicion of malignancy). Strong clinical evidence of metastatic disease was defined in one of our prior publications, and we used the same definition for this paper.9. Briefly, we defined strong clinical evidence of metastatic disease as imaging showing multiple metastases in a typical clinical pattern. In addition high clinical suspicion of infection was defined a patient with clinical suspicion of pneumonia within 14 days prior to thoracentesis on the same side as the pleural infection.
All the procedures were performed by trained pulmonologist using care fusion T-centesis catheter drainage 6 French trays and wall suction.
The primary outcome was the diagnostic yield of pleural cultures. The final determination of pleural effusion was established by additional diagnostic studies, in some cases biopsies or imaging studies, or by the clinical information available to the treating physician and pulmonary research team. Descriptive statistics were used to summarize the data. We used means and standard deviations for normally distributed continuous data and medians and interquartile ranges for non-normally distributed data. Frequencies were used for categorical data.
Results
A total of 2133 patients met inclusion criteria. We excluded 496 patients (402 had prior thoracentesis and 94 had prior chest tube placement). Of the remaining 1637 patients, 1547 (94%) had evidence of active malignancy and 1359 (83%) had strong clinical evidence of metastatic disease (Table 1). Of the 1637 patients, 542 (33%) had high clinical suspicion of pneumonia within 14 days and pleural infection before thoracentesis. All patients but 1 had pleural fluid sent for bacterial cultures, fungal and acid fast bacilli smear and cultures.
Table 1.
Clinical and Radiological Characteristics of Patients (N = 1637)
| Characteristic | n (%) |
|---|---|
| Age, years | |
| Mean ± SD | 59.92 ± 13.80 |
| Median (range) | 55 (18–92) |
| Sex | |
| Female | 867 (53.0) |
| Male | 770 (47.0) |
| Cancer diagnosis at the time of the thoracentesis | |
| Biopsy-proven cancer, active within past 12 months | 1547 (94.0) |
| Biopsy-proven cancer, remission for >12 months to 5 years | 63 (3.9) |
| History of cancer, in remission > 5 years | 9 (0.5) |
| No history of prior/current cancer | 25 (1.5) |
| Strong clinical evidence of metastatic disease | |
| No | 278 (17.0) |
| Yes | 1359 (83.0) |
| Cancer type | |
| Leukemia | 222 (13.5) |
| Lymphoma | 160 (9.7) |
| Solid non-lung | 932 (57) |
| Lung cancer | 379a (24.3) |
| High clinical suspicion of pneumonia 14 days prior to thoracentesis | |
| No | 1095 (66.9) |
| Yes | 542 (33.1) |
| Chemotherapy within 30 days prior to thoracentesis | |
| No | 726 (44.5) |
| Yes | 911 (55.5) |
| Thoracic radiation within 30 days prior to thoracentesis | |
| No | 1463 (89.3) |
| Yes | 174 (10.7) |
| Size of pleural effusion by chest x-ray | |
| 0 - no discernable effusion present | 22 (1.3) |
| 1 - blunting of costophrenic angle, but at least part of diaphragm is still visible | 119 (7.2) |
| 2 - effusion greater than 1, and up to the inferior border of the vascular pedicle | 473 (2.9) |
| 3 - effusion greater than 2, and up to the top of the cardiac silhouette | 454 (2.7) |
| 4 - effusion greater than 3, and up to the top of the aortic arch | 182 (1.1) |
| 5 - effusion greater than aortic arch | 82 (0.5) |
| X - loculated/unable to measure | 305b (18.6) |
| Pleural fluid bacterial cultures sent | |
| No | 1 (< 0.1) |
| Yes | 1636 (99.9) |
| Fungal cultures sent | |
| No | 1 (< 0.1) |
| Yes | 1636 (99.9) |
| AFB cultures sent | |
| No | 1 (< 0.1) |
| Yes | 1636 (99.9) |
| Positive pleural cultures | |
| No | 1623 (99.1) |
| Yes | 14 (0.9) |
Some patient are included in more than one category
Nine patients had a loculated effusion, and 296 patients had an effusion that was unable to be measured.
Only 14 patients (1.1%) had positive pleural fluid bacterial cultures (Table 2), and only 6 of these positive cultures were found to be true pleural space infection. All 6 of these patients had a clinical suspicion of pneumonia and were treated with antibiotics prior to thoracentesis. Interestingly 174 patients (10.6 %) were also deemed as having a parapneumonic effusions, but none of these patients had positive findings on pleural microbiological studies. These patients as well were on antibiotics for at least 24 hours prior to thoracenteses. Patients’ final diagnoses are shown in Table 3.
Table 2.
Characteristics of Patients with Positive Pleural Fluid Culture (N = 14)
| Infection | Suspicion of Infection Prior to Thoracentesis | Culture Treated | Final Diagnosis |
|---|---|---|---|
| Clostridium tertium | No | No | Contaminant |
| Clostridium tertium | No | No | Contaminant |
| Cryptococcus neoformans | Yes | Yes | Disseminated Cryptococcus |
| Fusobacterium species | Yes | Yes | Empyema |
| Klebsiella pneumoniae | Yes | Yes | Empyema |
| Propionibacterium acnes | No | No | Contaminant |
| Propionibacterium acnes | No | No | Contaminant |
| Propionibacterium acnes | No | No | Contaminant |
| Klebsiella pneumoniae | Yes | Yes | Empyema |
| Staphylococcus aureus | Yes | Yes | Empyema |
| Staphylococcus aureus | No | Yes | Empyema |
| Staphylococcus epidermidis | No | No | Contaminant |
| Staphylococcus epidermidis | No | No | Contaminant |
| Staphylococcus saccharolyticus | No | No | Contaminant |
Table 3.
Final Diagnoses of Patients with Pleural Effusions (N = 1637)
| Diagnosis | n |
|---|---|
| Malignant pleural effusion | 977 |
| Parapneumonic effusion | 174 |
| Empyema | 6 |
| Volume overload | 92 |
| Paramalignant | 62 |
| Drug-induced | 36 |
| Heart failure | 49 |
| Hemothorax | 12 |
| Hepatic hydrothorax | 18 |
| Postobstructive | 21 |
| Nephrotic syndrome | 3 |
| Pancreatitis | 1 |
| Pulmonary embolism | 6 |
| Sepsis | 3 |
| Trapped lung | 12 |
| Radiation induced | 5 |
| Unknown | 160 |
Discussion
We found that the diagnostic yield of pleural fluid bacterial cultures, fungal and acid fast bacilli smear and cultures at initial thoracentesis in patients with active cancer is very low. Only 14 (1.1%) of our patients had positive pleural fluid bacterial cultures, and only 6 of these were deemed as true pleural space infections and treated as such. Regardless of clinical suspicion, all patients but 1 had pleural fluid sent for bacterial cultures, fungal cultures, and acid fast bacilli smear and cultures.
Our findings are corroborated by a retrospective study by Walshe and colleagues, who reviewed 150 cases of pleural effusion in general population in which thoracentesis was used and found that most of the sampled pleural fluid was sent for microbiological investigation (91%).10 Of the patients whose fluid was sampled and sent for microbiological studies, only 5 (3.5%) had positive findings. In 4 cases, a bacterial pathogen was found, and there was 1 case of tuberculosis.10 Similarly, in a prospective study by Ferrer et al, only 15.5% of pleural fluid samples were positive for a bacterial pathogen.4 However, in patients with high likelihood of empyema, the pleural fluid cultures were positive in up to 69.4%.4 At our institution, routine submission of pleural fluid samples for microbiology have been removed as a result of this study, and they are only sent in those for which we suspect infection.
One major issue with pleural fluid microbiological studies in patients with evidence of infection is that most patients receive antibiotics before pleural fluid analysis. Another issue is that patients seldom undergo pleural biopsy with pleural tissue culture in an attempt to detect bacterial or fungal pathogens. A few methods have been recommended to increase the pleural fluid diagnostic yield. Collection of pleural fluid in blood culture bottles rather than in standard sterile tubes has been shown to increase the yield in cases where an infection is supected.11, 12 Nucleic acid amplification assays that target 16S ribosomal RNA sequence and biopsies are other techniques that show some promise in increasing the microbiological yield.13, 14 However, require validation and are subject to false positive results
In the AUDIO pilot study evaluating the role of ultrasound-guided pleural biopsies in pleural infection, ultrasound-guided pleural biopsies were found to be safe, and no adverse events were reported. 14Ultrasound-guided pleural biopsies were found to increase microbiological yield by 25% when added to pleural fluid and blood samples. The microbiological yield of ultrasound-guided biopsy (45%) was substantially higher than those of pleural fluid (20%) and blood culture (10%).14The study included only patients with proven empyema.
In our study, the incidence of true-positive pleural fluid microbiologic results was lower those reported by Walshe et al or Ferrer et al.4, 10 It is uncertain if this lower rate reflects differences in methods, antibiotic use, or patient populations. In our opinion, the most likely explanation is that the prevalence of pleural infections in cancer patients is lower than in the general population of patients who present with a pleural effusion. Most of our patients did have suspicion of malignancy at the time of thoracentesis however 542 patients did have clinical suspicion of pneumonia prior to thoracentesis and infection was in the differential. Interestingly 174 patients (10.6 %) were deemed as having a parapneumonic effusion, but none of these patients had positive pleural microbiological studies.
The limitations of our study include a narrow generalizability to other situations, especially situations in which patients do not have malignancy. We believe that in a different population, in which the pretest probability of infectious etiology is high, the diagnostic yield of pleural fluid microbiology studies on initial thoracentesis is likely to be higher than that found in this study. A second limitation is that we depended on evaluation of available clinical and imaging data. Only a small portion of our patients underwent pleural biopsies, so it is conceivable that some effusions recorded as negative for infections were actually false negatives.
In conclusion, in patients with a low pretest probability of pleural space infection who have a history of cancer and in whom there is a high suspicion of pleural malignancy, we recommend against routine microbiological studies because of their low diagnostic yield and high cost. However, if infection is suspected, we do recommend cultures to be sent and the use of blood culture bottles for pleural fluid collection. There is a convincing need for larger studies using other techniques, such as pleural biopsy and polymerase chain reaction, to increase the yield of microbiological studies and improve the management of suspected pleural infections.
FUNDING/SUPPORT:
This research is supported in part by the National Institutes of Health through MD Anderson’s Cancer Center Support Grant (CA016672).
Footnotes
CONFLICT OF INTEREST: none
References
- 1.Owings MF, Kozak LJ. Ambulatory and inpatient procedures in the United States, 1996. Vital and health statistics Series 13, Data from the National Health Survey. 1998: 1–119. [PubMed] [Google Scholar]
- 2.Kvale PA, Selecky PA, Prakash UB. Palliative care in lung cancer: ACCP evidence-based clinical practice guidelines (2nd edition). Chest. 2007; 132: 368s–403s. [DOI] [PubMed] [Google Scholar]
- 3.Maskell NA, Butland RJ. BTS guidelines for the investigation of a unilateral pleural effusion in adults. Thorax. 2003; 58 Suppl 2: ii8–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ferrer A, Osset J, Alegre J, Surinach JM, Crespo E, Fernandez de Sevilla T, Fernandez F. Prospective clinical and microbiological study of pleural effusions. European journal of clinical microbiology & infectious diseases : official publication of the European Society of Clinical Microbiology. 1999; 18: 237–41. [DOI] [PubMed] [Google Scholar]
- 5.Woolnough K, Hussain I. Late-breaking abstract: When is pleural fluid microbiology useful? European Respiratory Journal. 2011; 38: p2551. [Google Scholar]
- 6.Barnes TW, Olson EJ, Morgenthaler TI, Edson RS, Decker PA, Ryu JH. Low yield of microbiologic studies on pleural fluid specimens. Chest. 2005; 127: 916–21. [DOI] [PubMed] [Google Scholar]
- 7.Jimenez D, Diaz G, Garcia-Rull S, Vidal R, Sueiro A, Light RW. Routine use of pleural fluid cultures. Are they indicated? Limited yield, minimal impact on treatment decisions. Respiratory medicine. 2006; 100: 2048–52. [DOI] [PubMed] [Google Scholar]
- 8.Menzies SM, Rahman NM, Wrightson JM, Davies HE, Shorten R, Gillespie SH, Davies CW, Maskell NA, Jeffrey AA, Lee YC, Davies RJ. Blood culture bottle culture of pleural fluid in pleural infection. Thorax. 2011; 66: 658–62. [DOI] [PubMed] [Google Scholar]
- 9.Grosu HB, Kazzaz F, Vakil E, Molina S, Ost D. Sensitivity of Initial Thoracentesis for Malignant Pleural Effusion Stratified by Tumor Type in Patients with Strong Evidence of Metastatic Disease. Respiration; international review of thoracic diseases 2018; 96: 363–9. [DOI] [PubMed] [Google Scholar]
- 10.Walshe AD, Douglas JG, Kerr KM, McKean ME, Godden DJ. An audit of the clinical investigation of pleural effusion. Thorax. 1992; 47: 734–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Maskell NA, Davies CWH, Nunn AJ, Hedley EL, Gleeson FV, Miller R, Gabe R, Rees GL, Peto TEA, Woodhead MA, Lane DJ, Darbyshire JH, Davies RJO. U.K. Controlled Trial of Intrapleural Streptokinase for Pleural Infection. New England Journal of Medicine. 2005; 352: 865–74. [DOI] [PubMed] [Google Scholar]
- 12.Rahman NM, Maskell NA, West A, Teoh R, Arnold A, Mackinlay C, Peckham D, Davies CWH, Ali N, Kinnear W, Bentley A, Kahan BC, Wrightson JM, Davies HE, Hooper CE, Lee YCG, Hedley EL, Crosthwaite N, Choo L, Helm EJ, Gleeson FV, Nunn AJ, Davies RJO. Intrapleural Use of Tissue Plasminogen Activator and DNase in Pleural Infection. New England Journal of Medicine. 2011; 365: 518–26. [DOI] [PubMed] [Google Scholar]
- 13.Nielsen J, Meyer CN, Rosenlund S. Outcome and clinical characteristics in pleural empyema: A retrospective study. Scandinavian Journal of Infectious Diseases. 2011; 43: 430–5. [DOI] [PubMed] [Google Scholar]
- 14.Psallidas I, Kanellakis NI, Bhatnagar R, Ravindran R, Yousuf A, Edey AJ, Mercer RM, Corcoran JP, Hallifax RJ, Asciak R, Shetty P, Dong T, Piotrowska HEG, Clelland C, Maskell NA, Rahman NM. A Pilot Feasibility Study in Establishing the Role of Ultrasound-Guided Pleural Biopsies in Pleural Infection (The AUDIO Study). Chest. 2018; 154: 766–72. [DOI] [PubMed] [Google Scholar]