Journal Logo

Original Article

Predictors of Pleural Adhesion Formation and Success of Pleurodesis in Patients With Pleural Effusion

Wolff, Armand J. MD; Anderson, Eric D. MD, FCCP; Read, Charles A. MD, FCCP

Author Information
  • Free


One and a half to 2 million patients present each year in the United States with a pleural effusion. 1 One third of these result from congestive heart failure and approximately 20% are related to bacterial pneumonia. 1 Malignant effusions account for roughly 15% to 40%1–3 of pleural effusions, of which approximately 30% to 40% are associated with bronchogenic carcinoma and approximately 25% are the result of breast cancer. 1,4–6 When patients present with a malignant pleural effusion, the diagnostic yield of pleural fluid cytology is only 50% to 70%. 1,2,6–10 Blind pleural biopsy has a sensitivity of only 39% to 75% in the setting of malignant pleural effusion 1,2,11–14

Thoracoscopy has been advocated for undiagnosed exudative pleural effusions; by allowing visualization of the pleura, directed biopsies can be taken, increasing the diagnostic yield. 1,6,15–18 Thoracoscopy also allows for the insufflation of talc to promote pleurodesis. Such pleurodesis has been shown to be effective in preventing the recurrence of symptomatic pleural effusions resulting from diverse causes. 19–25

Thoracoscopy can be made more complicated by adhesions within the pleural space. 26,27 Adhesions can result in loculations that make dissection of the pleural space and complete evacuation of the pleural fluid difficult. 21 Furthermore, adhesions can make it difficult to evenly coat the parietal and visceral pleural surfaces with talc, possibly diminishing the effectiveness of pleurodesis.

Personal observation suggested that adhesions tended to form at the site of prior thoracentesis. Previous literature relates loculation of the pleural space to repeated thoracentesis. 4,10,19,21,28 Very recently, it was reported that repeated thoracentesis might result in the release of inflammatory cytokines that might lead to fibrin formation. 29 We undertook this study to determine whether the number of thoracenteses performed before thoracoscopy was related to the presence and extent of pleural adhesion formation. We also wanted to see if the presence and grade of adhesion would affect the success of talc pleurodesis.


Study Design and Definitions

We conducted a retrospective review of the charts of patients who underwent medical thoracoscopy for the purpose of diagnosis and/or treatment of a pleural effusion. Between August 1995 and March 2002, 121 patients underwent thoracoscopy performed by one of the authors (EDA or CAR). Indications for thoracoscopy included rapidly reaccumulated symptomatic malignant effusion in a patient whose life expectancy exceeded 1 month or else recurrent exudative effusion of unclear etiology. Contraindications for pleuroscopy included previous pleurodesis on that side or frank empyema. Data regarding patient demographics, number of thoracenteses before thoracoscopy, time from presentation with effusion to thoracoscopy (months from onset of symptoms thought to be related to effusion), malignancy/benignity of the pleural fluid (determined by pleural fluid cytology and/or pleural biopsy), history of thoracic irradiation, presence and extent of pleural adhesions, and success of the procedure were collected. Duration of effusion was analyzed as a continuous variable and a dichotomous variable was performed for several break points. Of the 121 patients, only 89 had complete charts for analysis.

For the purposes of this study, a grading system for pleural adhesions was modified from a system devised to grade abdominal and pelvic adhesions. 30 Grade 0 was defined as the absence of adhesions. Grade 1 was assigned if there were thin, filmy adhesions that were easily lysed. Grade 2 was defined as thick adhesions that were few in number and localized to one area of the hemithorax. Grade 3 encompassed thick adhesions that were seen extensively throughout the pleural space. Grade 3 also included patients with fibrothorax (trapped lung that failed to expand after surgery). Adhesion grade was assigned using the operative report as well as photographs of the pleural space taken intraoperatively. When multiple adhesions were seen, the grade was assigned based on the highest grade adhesion seen. Figures 1 through 3 show examples of adhesion grades 1–3, respectively.

Grade 1 adhesions (thin, easily lysed).
Grade 2 adhesions (thick, localized).
Grade 3 adhesions (thick, extensive).

The frequency of thoracentesis (number of thoracenteses/duration of effusion) is a measure of how quickly an effusion returns and becomes symptomatic again. It is our best available surrogate for the “severity” of the process underlying the pleural effusion.

We defined success of the procedure as reexpansion of the lung and the lack of need for another therapeutic thoracentesis on the operated side. Small pockets of fluid remaining after surgery did not connote failure of the procedure as long as they did not need drainage. Patients with fibrothorax in which pleurodesis was not attempted were excluded from the success analysis.

Data Analysis

Statistical analysis was done with SPSS software (SPSS, standard version 8.0, SPSS Inc., 1989–1997). Nominal data were compared using chi-squared testing. Numerical data were compared using independent samples t testing, analysis of variance, or Pearson's correlation when appropriate. Logistic regression was done to establish truly independent relationships. Significance was set at a P value of <0.05.


Table 1 shows the relationship of patient demographics, duration of the effusion before thoracoscopy, the mean number of thoracenteses, frequency of thoracentesis, history of thoracic irradiation, and malignancy/benignity of the pleural fluid to the presence or absence of adhesions. There were no statistically significant relationships between any of the demographic characteristics and the presence of adhesions. The number of thoracenteses and the frequency of thoracentesis were not related to adhesion formation. There was no association with mean duration of effusion. However, when the duration of effusion was categorized as a dichotomous variable (< or ≥5 months), there was an association with pleural adhesion formation (P = 0.032). Prior thoracic irradiation and malignancy of the fluid were not predictive of pleural adhesions.

Predictors of Pleural Adhesion Formation in Patients with Pleural Effusion (n = 89)

Table 2 shows the analysis limited to the 74 patients who proved to have malignant effusion. Again, there were no significant associations between pleural adhesion formation and demographic characteristics, the number of thoracenteses, frequency of thoracentesis, or prior thoracic irradiation. In this subgroup, however, the mean duration of effusion before thoracoscopy was significantly associated with the presence of adhesions (P = 0.045). Like in the overall population, duration of ≥5 months was predictive of adhesion formation (P = 0.020).

Predictors of Pleural Adhesion Formation in Patients with Malignant Pleural Effusion (n = 74)

Breast cancer is a frequent cause of malignant effusion. It is the most common indication for thoracoscopy and talc pleurodesis in our practice reflecting our referral base. We repeated our analysis, limiting it to the patients with breast cancer as the source of a malignant effusion (n = 39). Table 3 shows these results. No association was found between pleural adhesions and any of the demographic characteristics, number of thoracenteses, frequency of thoracentesis, or prior thoracic radiation. Mean duration of the effusion and duration ≥5 months were predictors of adhesion formation in the patients with effusion related to breast cancer (P = 0.037 and 0.008, respectively).

Predictors of Pleural Adhesion Formation in Patients with Malignant Pleural Effusion Related to Breast Cancer (n = 39)

Table 4 shows the distribution of adhesion grades within the 3 populations of patients. A significant relationship was noted between duration of effusion of at least 5 months and adhesion grade (P = 0.038, 0.037, and 0.011 in all patients, patients with malignant effusions, and patients with malignant effusions secondary to breast cancer, respectively). This primarily seems the result of an increased occurrence of grade 3 adhesions in patients with effusions of ≥5 months duration (P = 0.036, 0.036, and 0.032 in each of the subgroups, respectively; data not shown). In both the malignant effusion and breast cancer subgroups, mean duration of effusion was significantly associated with grade of adhesion (P = 0.018 and 0.035, respectively). No association was found between adhesion grade and number of thoracenteses, frequency of thoracentesis, prior thoracic irradiation, malignancy of the pleural fluid, or underlying breast cancer in any of the patient groups.

Adhesion Grade in Subgroups of Patients with Pleural Effusion

Logistic regression was performed with demographic data, duration of effusion, number of thoracenteses, frequency of thoracentesis, history of thoracic irradiation, malignancy/benignity of the pleural fluid, and presence of underlying breast cancer included in the regression model. Durations of 1 through 6 months were analyzed. Duration of effusion ≥5 months was the only significant predictor of adhesion formation in the entire patient population (P = 0.032; odds ratio, 13.7; 95% confidence interval, 1.3–150.2; data not shown).

Success of the pleurodesis procedure was a secondary end point. Of the 89 patients, 74 (83%) were included in the analysis of success of the pleurodesis. Of the 15 excluded patients, 3 had not actually had talc pleurodesis because of extensive pleural adhesions noted at thoracoscopy, 2 died too soon after the procedure to evaluate for success, and 10 had incomplete data or were lost to follow up. Overall, the success rate of pleurodesis was 94.6% (69 of 74 patients). There were no significant associations between demographic characteristics, duration of effusion, number of thoracenteses, frequency of thoracentesis, malignancy/benignity of the effusion, or prior thoracic irradiation with the likelihood of success. Also, the presence of adhesions and the grade of adhesions were not associated with failure of the procedure. This was true regardless of the patient subgroup analyzed (data not shown).

Of the 5 patients who had unsuccessful procedures (5.4%), the time to failure ranged from less than 1 week to almost 11 months. This group of patients was too small to attempt to determine differences between early and late failure.

In the entire patient population, the frequency of thoracentesis was lower in those patients with duration of effusion ≥5 months (data not shown, P < 0.001). This relationship was maintained when only patients with malignant effusion or effusion related to breast cancer were analyzed (data not shown, P < 0.001 and P = 0.016, respectively). Malignant effusions were more common in female patients than male patients (P = 0.004) as a result of our skewed referral basis. Malignancy of the fluid was not related to the number of thoracenteses or the duration of the effusion. The number of thoracenteses was strongly correlated to the duration of the effusion (Pearson correlation 0.471, P < 0.001).


The mean duration of the effusion before thoracoscopy was related to the presence and extent of pleural adhesions in patients with malignant effusion, especially those with underlying breast cancer. When duration of effusion was considered as a dichotomous variable (with a break point of ≥5 months), a significant tendency toward adhesion formation was noted regardless of the patient population analyzed. We also found that duration of effusion of at least 5 months was associated with more extensive adhesion formation (grade 3).

Contrary to the suggestion of the previous literature and anecdotal personal observation, we did not find an association between the number of thoracenteses performed before thoracoscopy and pleural adhesions. We were unable to find such a relationship even when the patient population was limited to those with malignant pleural effusion, so the lack of association is independent of the cytologic characteristics of the pleural fluid. We also failed to find an association between the number of thoracenteses and adhesion grade.

Our results in this regard seem to be in contrast to those that Chung et al. recently reported. 29 They demonstrated that after 3 daily thoracenteses, the level of inflammatory cytokines in the pleural fluid had increased and that almost half of the patients had fibrin stranding in the pleural space detected on ultrasonography. It would seem likely that the fibrin stranding they describe would correlate to our adhesion grade 1; we found no association between the number of thoracenteses or the frequency of thoracentesis with any adhesion grade. This was true even for our population of patients with malignant effusions, a group similar to that described by Chung et al. 29 Our study differs, however, in that the repeated thoracenteses took place over weeks to months rather than on 3 sequential days. Furthermore, our grading of adhesions was done by direct visualization at thoracoscopy rather than by ultrasonography. With direct visualization of the pleural space, it seems unlikely that we missed adhesions that would have altered our results.

The number of thoracenteses performed in most of our patients was fairly low. In fact, only 8 of our 92 (8.7%) patients had more than 3 thoracenteses. This could reflect our previous belief that the more thoracenteses a patient had, the more likely the thoracoscopy was to be complicated by pleural adhesions. In this regard, our hesitation to perform multiple thoracenteses before thoracoscopy could be a significant bias. Perhaps had a greater number of patients had more thoracenteses, we would have found a significant relationship between the number of thoracenteses and adhesion formation.

It is possible that adhesion formation is related to the severity of the underlying process. Effusions that reaccumulate more quickly could signify a more intense pleural reaction that could be more likely to result in adhesions. We attempted to investigate this possibility using the frequency of thoracentesis as a surrogate for the rapidity with which the pleural fluid returns after thoracentesis. This measure also helps to control for our bias toward performing fewer thoracenteses and taking patients with recurrent, symptomatic effusions to thoracoscopy fairly early. Although it might not be a perfect substitute, it is a useful one given that our practice is to perform repeat thoracentesis or thoracoscopy as soon as possible whenever a patient has recurrence of a symptomatic effusion.

We found no association between the frequency of thoracentesis and adhesion formation or grade. We did, however, show that the frequency of thoracentesis is lower in those patients who have duration of effusion ≥5 months. Patients who are followed expectantly for a longer period before thoracoscopy have slower reaccumulation of pleural fluid. This suggests that more indolent disease does not protect against developing significant pleural adhesions, further establishing the conclusion that pleural adhesion formation is independent of the characteristics of the pleural fluid and intensity of the pleural process leading to the effusion.

Not surprisingly, the number of thoracenteses is strongly correlated to the duration of the effusion. This could be the basis of our earlier belief that repeated thoracentesis led to a higher likelihood of pleural adhesion formation when, in fact, the important predictor is the duration of the effusion. Our finding that malignant pleural effusion was more common in women than men almost certainly reflects a referral bias; the most common reason for pleurodesis in our practice is recurrent effusion in the setting of breast cancer.

We were unable to find any significant predictors of the ultimate success of pleurodesis. We intentionally excluded 3 patients with fibrothorax from the analysis of success of the procedure because no talc was insufflated. Inclusion of these 3 patients as “failures” did not alter the results (data not shown). It is interesting that the presence and extent of adhesions was not related to failure. Although extensive pleural adhesions could make it technically more difficult to perform thoracoscopy and pleurodesis, they do not interfere with success. Pleurodesis works by forming adhesions within the pleural space; preexisting adhesions could augment this process. Because our success rate was high in all groups of patients, we were unable to associate pleural adhesion with either success or failure of pleurodesis.

Our results suggest that, regardless of the nature of the effusion fluid, a pleural space left without definitive treatment of at least 5 months is more likely to develop adhesions. This would suggest that for those practicing thoracoscopy, it would be an easier and more complete examination if it were accomplished within the first 5 months. However, for patients with rapidly reaccumulating symptomatic malignant effusion of more than 5 months duration in which the presence of grade 3 adhesions is suspected, they should still be considered candidates for thoracoscopic talc pleurodesis because these adhesions do not affect the success of pleurodesis.


1. Light RW. Pleural Diseases, 4th ed. Philadelphia: Lippincott Williams & Wilkins; 2001.
2. Prakash UB, Reiman HM. Comparison of needle biopsy with cytological analysis for the evaluation of pleural effusion: analysis of 414 cases. Mayo Clin Proc. 1985; 60:158–164.
3. Jimenez D, Perez-Rodriguez E, Diaz G, et al. Determining the optimal number of specimens to obtain with needle biopsy of the pleura. Respir Med. 2002; 96:14–17.
4. Anderson CB, Philpott GW, Ferguson TB. The treatment of malignant pleural effusions. Cancer. 1974; 33:916–922.
5. Prakash UB. Malignant pleural effusions. Postgrad Med. 1986; 80:201–209.
6. Marel M, Stastny B, Melinova L, et al. Diagnosis of pleural effusions: experience with clinical studies, 1986 to 1990. Chest. 1995; 107:1598–1603.
7. Hirsch A, Ruffie P, Nebut M, et al. Pleural effusion: laboratory tests in 300 cases. Thorax. 1979; 34:106–112.
8. Larsen K. Axelsen F. Closed pleural biopsy and fluid cytology in the diagnosis of suspected pulmonary cancer with pleural involvement. Scand J Thorac Cardiovasc Surg. 1982; 16:197–199.
9. Ong KC, Indumathi V, Poh WT, et al. The diagnostic yield of pleural fluid cytology in malignant pleural effusions. Singapore Med J. 2000; 41:19–23.
10. Sahn SA. Malignancy metastatic to the pleura. Clin Chest Med. 1998; 19:351–361.
11. Nance KV, Shermer RW, Askin FB. Diagnostic efficacy of pleural biopsy as compared with that of pleural fluid examination. Mod Pathol. 1991; 4:320–324.
12. Poe RH, Israel RH, Utell MJ, et al. Sensitivity, specificity, and predictive values of closed pleural biopsy. Arch Intern Med. 1984; 144:325–328.
13. Kirsch CM, Kroe DM, Azzi RL, et al. The optimal number of pleural biopsy specimens for a diagnosis of tuberculous pleurisy. Chest. 1997; 112:702–706.
14. Salyer WR, Eggleston JC, Erozan YS. Efficacy of pleural needle biopsy and pleural fluid cytopathology in the diagnosis of malignant neoplasm involving the pleura. Chest. 1975; 67:536–539.
15. Blanc FX, Atassi K, Bignon J, et al. Diagnostic value of medical thoracoscopy in pleural disease: a 6-year retrospective study. Chest. 2002; 121:1677–1683.
16. Menzies R, Charbonneau M. Thoracoscopy for the diagnosis of pleural disease. Ann Intern Med. 1991; 114:271–276.
17. Hucker J, Bhatnagar NK, al-Jilaihawi AN, et al. Thoracoscopy in the diagnosis and management of recurrent pleural effusions. Ann Thorac Surg. 1991; 52:1145–1147.
18. Harris RJ, Kavuru MS, Rice TW, et al. The diagnostic and therapeutic utility of thoracoscopy: a review. Chest. 1995; 108:828–841.
19. Cardillo G, Facciolo F, Carbone L, et al. Long-term follow-up of video-assisted talc pleurodesis in malignant recurrent pleural effusions. Eur J Cardiothorac Surg. 2002; 21:302–305; discussion 305–306.
20. de Campos JR, Vargas FS, de Campos Werebe E, et al. Thoracoscopy talc poudrage: a 15-year experience. Chest. 2001; 119:801–806.
21. Viallat JR, Rey F, Astoul P, et al. Thoracoscopic talc poudrage pleurodesis for malignant effusions: a review of 360 cases. Chest. 1996; 110:1387–1393.
22. Diacon AH, Wyser C, Bolliger CT, et al. Prospective randomized comparison of thoracoscopic talc poudrage under local anesthesia versus bleomycin instillation for pleurodesis in malignant pleural effusions. Am J Respir Crit Care Med. 2000; 162:1445–1449.
23. Aelony Y, King RR, Boutin C. Thoracoscopic talc poudrage in malignant pleural effusions: effective pleurodesis despite low pleural pH. Chest. 1998; 113:1007–1012.
24. Wohlrab JL, Read CA. Chylothorax associated with Noonan syndrome: management with medical thoracoscopic talc pleurodesis. Journal of Bronchology. 2002; 9:122–124.
25. Sanchez-Armengol A, Rodriguez-Panadero F. Survival and talc pleurodesis in metastatic pleural carcinoma, revisited: report of 125 cases. Chest. 1993; 104:1482–1485.
26. Janssen JP, Boutin C. Extended thoracoscopy: a biopsy method to be used in case of pleural adhesions. Eur Respir J. 1992; 5:763–766.
27. Waller DA, McConnell SA, Rajesh PB. Delayed referral reduces the success of video-assisted thoracoscopic surgery for spontaneous pneumothorax. Respir Med. 1998; 92:246–249.
28. Lambert CJ, Shah HH, Urshel Jr, HC et al. The treatment of malignant pleural effusions by closed trocar tube drainage. Ann Thorac Surg. 1967; 3:1–5.
29. Chung CL, Chen YC, Chang SC. Effect of repeated thoracenteses on fluid characteristics, cytokines, and fibrinolytic activity in malignant pleural effusion. Chest. 2003; 123:1188–1195.
30. Montz FJ, Fowler JM, Wolff AJ, et al. The ability of recombinant tissue plasminogen activator to inhibit post-radical pelvic surgery adhesions in the dog model. Am J Obstet Gynecol. 1991; 165:1539–1542.

pleural effusion; pleural adhesion; medical thoracoscopy; talc poudrage

© 2004 Lippincott Williams & Wilkins, Inc.