Malignant pleural effusions and the role of talc poudrage and talc slurry : a systematic review and meta-analysis

Malignant Pleural Effusion (MPE) is common with advanced Background: malignancy. Palliative care with minimal adverse events is the cornerstone of management. Although talc pleurodesis plays an important role in treatment, the best modality of talc application remains controversial. To compare rates of successful pleurodesis, rates of respiratory Objective: and non-respiratory complications between thoracoscopic talc insufflation/poudrage (TTI) and talc slurry (TS). MEDLINE (PubMed, OVID), EBM Data sources and study selection: Reviews (Cochrane database of Systematic Reviews, ACP Journal Club, DARE, Cochrane Central Register of Controlled Trials, Cochrane Methodology Register, Health Technology Assessment and NHS Economic Evaluation Database), EMBASE and Scopus. Randomized controlled trials published between 01/01/1980 10/1/2014 and comparing the two strategies were selected. Twenty-eight potential studies were identified of which 24 studies Results: were further excluded, leaving four studies. No statistically significant difference in the probability of successful pleurodesis was observed between TS and TTI groups (RR 1.06; 95 % CI 0.99-1.14; Q statistic, 4.84). There was a higher risk of post procedural respiratory complications in the TTI group compared to the TS group (RR 1.91, 95% CI= 1.24-2.93, Q statistic 3.15). No statistically significant difference in the incidence of non-respiratory complications between the TTI group and the TS group was observed (RR 0.88, 95% CI= 0.72-1.07, Q statistic 4.61). There is no difference in success rates of pleurodesis based on Conclusions: patient centered outcomes between talc poudrage and talc slurry treatments. Respiratory complications are more common with talc poudrage via thoracoscopy. 1,2 1,3

Palliation with minimal adverse events remains the cornerstone of management 3 . Talc pleurodesis and Indwelling Pleural Catheters (IPC) are the two most commonly used palliative approaches.
Talc pleurodesis can be achieved either by thoracoscopic instillation i.e.; talc insufflation/poudrage (TTI) or via a bedside chest tube i.e. talc slurry (TS). Existing systematic reviews concluded that thoracoscopic talc insufflation/poudrage was more efficacious in preventing recurrences when compared to bedside chest tube talc slurry 4,5 . New prospectively designed studies comparing TTI and TS have been published since then [6][7][8] . However the best initial approach for talc pleurodesis remains still unclear. To address the need for an update, a systematic review and meta-analysis of studies comparing thoracoscopic talc insufflation/poudrage and talc slurry in terms of patient centered outcomes was performed.

Data sources and search
We conducted a systematic review with meta-analysis of studies undertaken between 01/01/1980 and 10/1/2014 using MEDLINE (PubMed, OVID), EBM Reviews (Cochrane database of Systematic Reviews, ACP Journal Club, DARE, Cochrane Central Register of Controlled Trials, Cochrane Methodology Register, Health Technology Assessment and NHS Economic Evaluation Database), EMBASE and Scopus. Unpublished data sets such as conference abstracts and ClinicalTrials.gov were also included in the full review phase to reduce the effect of publication bias 9 .
In addition, a pearl growing strategy was employed using frequently cited reviews of malignant pleural effusion treatments. They were included to be analyzed in the full review phase of the study. Approval from the Institutional Review Board was unnecessary because this is a meta-analysis.

Study selection
Inclusion and exclusion criteria were framed prior to the implementation of the search strategy. To evaluate outcomes in adult malignant pleural effusion patients (18 + years) undergoing talc pleurodesis, we included studies based on the following criteria: 1) A randomized design was used in studying talc pleurodesis in patients with malignant pleural effusion between 01/01/1980 and 10/1/2014.
2) Patients undergoing bedside TS were compared with patients undergoing thoracoscopic talc insufflation/poudrage (TTI) in the above fashion.
3) Sufficient outcomes data were reported [Risk of recurrence of the pleural effusion, respiratory complications and nonrespiratory complications].
Non-English publications, case reports and series, pediatric studies, descriptive studies without a control group, retrospective studies and prospective controlled studies without randomization were excluded. Eligible articles were reviewed by two reviewers for inclusion; disagreements were resolved via discussion. An examination of the full-length articles was carried with the intent of eliminating duplicate studies or same patient cohorts.

Data extraction and outcome measures
Two reviewers independently extracted and rated the data from the selected full length articles using a standardized form. From each study, the data abstracted included study name/year, study design (prospective controlled, randomized controlled trial, retrospective etc.), cancer cell type, patient inclusion criteria, sample sizes for the bedside/surgical pleurodesis arms, technique employed in the bedside/surgical arms and the follow-up schedule.
Outcomes data pertaining to the risk of recurrent pleural effusions, respiratory complications, and non-respiratory complications were also extracted.
Talc pleurodesis for recurrent malignant pleural effusion is a palliative procedure and does not aim to have a mortality benefit. Therefore, measuring mortality outcomes was not the focus of the meta-analysis.
Various definitions for recurrent pleural effusions have been used in prior studies 6,10 . A clinically significant recurrent pleural effusion was defined a priori as accompanied by symptoms or a need for repeat pleural procedures. Where possible, asymptomatic radiological recurrences were not included in the meta-analysis.
Respiratory complications were defined as occurrence of respiratory conditions such as pneumonia, Acute Respiratory Distress Syndrome (ARDS), acute respiratory failure, re-expansion pulmonary edema, bronchospasm, empyema, pulmonary embolism, prolonged air leak, bronchopleural fistula, atelectasis requiring bronchoscopy and subcutaneous emphysema.
Immediate non-respiratory complications were tabulated from the complications mentioned in the full length articles. These included fever, wound infection, chest pain, tumor recurrence at site, myocardial infarction, need for blood transfusions, arrhythmias and immediate post procedural death.

Quality assessment criteria
The randomized controlled trials that met inclusion criteria were evaluated for quality using components of the modified Jadad scale 11 . The presence of the following features was appraised: A) A description of the study confirming the randomized nature.
B) Method of allocation to the study arms described and whether adequate/inadequate.

C) Description of withdrawals and dropouts.
Due to the nature of the comparison (surgical vs bedside procedure), we felt that other features of the scale (description of a double blind nature) could not be appraised during our quality assessment. Two raters independently determined the quality of the studies included. Disagreements were resolved by discussions and final consensus.

Statistical analysis
Outcomes data for recurrent pleural effusions, respiratory and nonrespiratory complications were summarized using descriptive statistics (simple count, proportion of the study sample). They were visually presented in Forest plots. The Mantel-Haenszel method 12 was used to combine data from individual studies and the results were reported as pooled relative risks (RR). Heterogeneity among the studies included was investigated by performing the I 2 test 13 . Meta-analyses were conducted using the fixed effects model when heterogeneity between studies was low (I 2 <40%) and the random effects model otherwise 9 .
To confirm the robust nature of the results, a sensitivity analysis was performed by removing one study at a time and determining the outcome.
Publication bias was examined by visually examining the filled funnel plots using trim and fill method. Other methods (Begg's correlation 14 and Egger's linear regression intercept) 15 were additionally used.
All analyses were performed using a statistical software package (Comprehensive Meta-Analysis, version 2.2.064; Biostat, Englewood, NJ).

Results
Based on initial search, 137 articles were obtained and reviewed independently by two reviewers. Pearl growing strategy was employed to seek additional articles and resulted in five articles.
A clinical trial registry (www.ClinicalTrials.gov) was also examined and resulted in one additional article. These 143 articles were reviewed and 115 articles were excluded based on title and abstract. A total of 28 potential studies were thus identified with our search strategy. Twenty-four studies were further excluded, leaving four studies 6,10,16,17 for the final analysis. The sequence describing the above process can be found in Figure 1. None of the studies restricted the study population to a single cancerous cell type.
None of the included studies employed thoracoscopic evacuation of the malignant pleural effusion prior to bedside TS insertion via a chest tube.
Follow-up periods varied through the studies (Range = 30-425 days). Where available, recurrence data for the most distal available time point were selected for the meta-analysis.
All of the studies included in the analysis underwent quality assessment. The average Jadad score 11 was 1.5 out of a maximum possible score of 4 (Ranges 1-2). Out of the possible seven ways to assess the quality 11 , only four questions could be answered due to the nature of the intervention. It was not possible reliably or ethically for the original investigators to have carried out efficient blinding in a surgical versus bedside clinical experiment.

Risk of recurrent pleural effusions
The results of the pooled RR are shown in Figure 2. The four studies included in this analysis enrolled a total of 454 patients with malignant pleural effusion (Table 1). There was no statistically significant difference in the risk of recurrent pleural effusions between the bedside TS (recurrences/patients who underwent pleurodesis, n/N = 51/218 pts) and the TTI groups (recurrences/patients who underwent pleurodesis, n/N = 39/236 pts, pooled RR 0.72; 95% CI 0.50-1.05; Q statistic, 3.58; I 2 statistic, 16.27%). There was no evidence of publication bias (P-value = 0.49 for the Begg's test, P-value = 0.54 for the Egger's regression intercept). After using the trim and fill methodology (Figure 3), these results did not change (RR-0.78, 95% CI = 0.54-1.12, Q statistic, 6.8).    The definitions of a recurrent pleural effusion in the included studies varied. One study defined the recurrence based on radiological data alone 6 . The remainder of the studies clearly mentioned the number of patients who were symptomatic and required further pleural procedures once a recurrent pleural effusion was diagnosed 10,16,17 .
A sensitivity analysis pooling data from studies reporting only clinically significant recurrences was performed leaving out one study 6 . This did not result in a different statistical outcome (pooled RR 0.49; 95% CI 0.11-2.22; Q statistic, 3.52; I 2 statistic 43.22%).
As follow-up periods varied widely, a sensitivity analysis pooling the data from studies reporting 30 day outcomes was carried out and did not result in a different statistical outcome.

Risk of respiratory complications
The results of the pooled RR are shown in Figure 4. The four studies included in this analysis reported outcomes on a total of 591 patients who underwent talc pleurodesis for palliation of malignant pleural effusion (Table 2).
There was a statistically significant higher risk of post procedural respiratory complications in the TTI group (Incidence of respiratory complications/Pts who underwent pleurodesis, n/N = 59/307 pts) compared to the TS group (Incidence of respiratory complications/ Pts who underwent pleurodesis, n/N = 28/284 pts, pooled RR 1.91, 95% CI = 1.24-2.93, Q statistic 3.15, I 2 statistic 4.79%). There was no evidence of publication bias (P-value = 1.0 for the Begg's test, P-value = 0.24 for the Egger's regression intercept). After using the    A sensitivity analysis pooling data from studies with ≥ 2 score on the Modified Jadad scale was performed leaving out one study with a score of 1 17 . This did not result in a different statistical outcome (pooled RR 1.99, 95% CI = 1.29-3.08, Q statistic 2.05, I 2 statistic 2.49).

Risk of non-respiratory complications
The results of the pooled RR are shown in Figure 6. The four studies included in this analysis reported outcomes on a total of 591 patients who underwent talc pleurodesis for palliation of malignant pleural effusion (Table 3).
A sensitivity analysis pooling data from studies with ≥2 score on the Modified Jadad scale was performed leaving out one study with a score of 1 17 . This did not result in a different statistical outcome (pooled RR 0.93, 95% CI = 0.76-1.14, Q statistic 0.06, I 2 statistic 0.0).

Discussion
Many experts believe that serial thoracentesis is not an ideal choice for treating the recurrent malignant pleural effusion 18,19 .
Talc pleurodesis was first performed in 1935 20 and is still commonly employed in the treatment of malignant pleural effusions. Although studies have shown talc to be the best chemical agent in terms of pleurodesis success and risk of recurrence 21,22 , the best method of applying talc remains controversial. Our meta-analysis demonstrates that both talc poudrage (TTI) and talc slurry (TS) offer similar protection against clinically significant recurrences. There was no difference in the risk of clinically significant recurrence (i.e., further need for pleural procedures or symptoms). TTI did have a greater risk of respiratory complications. There was, however, no difference in the rate of non-respiratory complications such as fever and need for blood transfusions.
Our results are in contrast to those of previous meta-analyses 5 , including the recently withdrawn Cochrane analysis which suggested     improved success rates of talc pleurodesis utilizing TTI. The conclusion of these analyses was that thoracoscopic pleurodesis with talc was the optimal method for pleurodesis in patients with malignant pleural effusions. However, several newer prospective studies have been published since 6,7 and have been incorporated into the present analysis.
Arguments in favor of TTI include the observation that there is more complete lung expansion after the procedure 18 . This is certainly understandable given that take-down of adhesions is typically performed during the procedure itself as opposed to TS. Interestingly, Terra et al. using CT scanning post-TTI and TS to assess degree of post procedure lung expansion did not find a correlation between clinical outcomes and initial degree of lung expansion 7 . These authors postulated that factors other than the degree of visceral and parietal pleura apposition were important in determining the success of pleurodesis. Similarly, Mager et al. using 99m Tc-labeled talc showed that rotation protocols did not affect the overall dispersion of talc suspensions after TS 23 . The degree of dispersion also did not affect pleurodesis success 23 .
In comparing TTI and TS, several difficulties arise. Pleurodesis success rates vary in the literature, due to the inconsistent definition of pleurodesis success and failure used in different studies. Failure or recurrence has been defined radiologically in some studies 6 but it has been argued that patient centered outcomes such as new symptoms and need for further pleural procedures are more pertinent outcomes 7 . In our meta-analysis, we determined recurrence a priori as the development of symptoms or the further need for pleural procedures and disregarded asymptomatic radiological recurrences where possible.
The technique of both TS and TTI vary significantly between centers and this is evident in the included studies. TS varied in regard to length of chest tube clamping, rotating or not-rotating the patient, size of chest tube, and timing of chest tube removal. With regard to TTI, in three of the four studies TTI was performed under general anesthesia and the ability to tolerate general anesthesia was in fact an entry criteria. Overall, 88% of patients in our analysis underwent general anesthesia for TTI. One could argue that the increased respiratory complications observed with TTI may be related to general anesthesia and single lung ventilation. Despite the concerns of ARDS with the use of ungraded talc, the studies included in our meta-analysis did not report specific cases of ARDS. Non-specific respiratory failure was, however, reported in patients in the study by Dresler 6 and Yim et al. reported a case of acute respiratory failure in the TS group 16 .
With the increasing numbers of interventional pulmonologists performing pleuroscopy (medical thoracoscopy) 24 under local and/or moderate sedation, the question of which procedure is the most optimal for talc pleurodesis is increasingly relevant. Whether talc poudrage performed during pleuroscopy with local or moderate sedation and dual lung ventilation is equivalent to surgical thoracoscopy (VATS) in terms of pleurodesis success and complications is unknown. Further studies are needed to compare talc poudrage performed with pleuroscopy versus TS.
One may wonder whether the question of TTI versus TS is still relevant in the era of indwelling pleural catheters (IPCs). Certainly, in the patient with trapped lung, both TS and TTI would likely be ineffective and indeed all of the studies in this meta-analysis excluded patients with possible trapped lung physiology. In the  Our study has several limitations. The results of meta-analyses are dependent on the quality of the studies included. The inclusion of only randomized controlled trials was necessary due to significant bias inherent in non-randomized prospective studies. Despite the lack of heterogeneity between studies, the individual studies varied substantially (technique of talc pleurodesis, varying definitions of recurrence and follow-up schedule). Sensitivity analyses performed leaving out one study at a time did not impact the results, suggesting robust data. Publication bias is an inherent limitation of meta-analyses. It is reassuring to see that accounting for it did not result in a statistically significant departure from the original point estimates.
In conclusion, our meta-analysis demonstrates that there is no difference in malignant pleural effusion recurrence based on patient centered outcomes between talc poudrage and talc slurry. Respiratory complications are more common with talc poudrage via thoracoscopy. Further studies are needed, however, to look at the role of talc pleurodesis via pleuroscopy. The decision of which procedure to perform needs to take into account also the patient preferences.
Author contributions SM co-conceived the study, participated in the design of the study, search strategy execution, performance of the statistical analysis, and writing of the manuscript.
AK participated in the search strategy execution, performance of the statistical analysis and writing of the manuscript.
PH co-conceived the study, participated in the design of the study, performance of the statistical analysis, and writing of the manuscript.
All authors read and approved the final manuscript.

Rationale 3
Describe the rationale for the review in the context of what is already known.

Page 3
Objectives 4 Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons, outcomes, and study design (PICOS). Page 3

Protocol and registration 5
Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide registration information including registration number.

Eligibility criteria 6
Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered, language, publication status) used as criteria for eligibility, giving rationale.

Page 3
Information sources 7 Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify additional studies) in the search and date last searched.

Page 3
Search 8 Present full electronic search strategy for at least one database, including any limits used, such that it could be repeated.

Appendix 1
Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable, included in the meta-analysis).

Page 3
Data collection process 10 Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes for obtaining and confirming data from investigators. Page 3

Data items 11
List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and simplifications made.

Risk of bias in individual studies 12
Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis.

Pages 3-4
Summary measures 13 State the principal summary measures (e.g., risk ratio, difference in means).

Pages 4-7
Synthesis of results 14 Describe the methods of handling data and combining results of studies, if done, including measures of consistency (e.g., I2) for each meta-analysis.

Pages 4-7
Risk of bias across studies 15 Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., publication bias, selective reporting within studies).

Pages 4-5
Section/topic # Checklist item Reported on page #

Study selection 17
Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions at each stage, ideally with a flow diagram. Figure 1, Pages 4-5 Study characteristics 18 For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up period) and provide the citations. Tables 1-3 Risk of bias within studies 19 Present data on risk of bias of each study and, if available, any outcome level assessment (see item 12).

Pages 4-7
Results of individual studies 20 For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for each intervention group (b) effect estimates and confidence intervals, ideally with a forest plot.

Figures 2,4,6
Synthesis of results 21 Present results of each meta-analysis done, including confidence intervals and measures of consistency.

Pages 4-7
Risk of bias across studies 22 Present results of any assessment of risk of bias across studies (see Item 15).

Summary of evidence 24
Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (e.g., healthcare providers, users, and policy makers).

Pages 7,9,10
Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of identified research, reporting bias).

Page 10
Conclusions

Overview:
The topic is important and clinically relevant. Talc pleurodesis is commonly practiced. The debate on the superiority of thoracoscopic poudrage (TTI) versus talc slurry (TS) has been ongoing for decades. A meta-analysis incorporating the most recent data is useful.
The methodology appears sound. The four selected articles are well described and acknowledged in most review articles as the only randomized studies directly addressing this topic.
The interpretation of the data is fair and balanced. The endpoints chosen are clinically relevant and important. Most other reviews or book chapters on the issue have only focused on the success/failure rate of the pleurodesis. The authors rightly highlighted the increased complication rates of TTI. The figures and graphs are clear and easy to understand.
The discussion is based on the analyses and the postulations for the observed results are sound and reasonable.
There are several few points that the authors may wish to consider. I believe that this article may be improved if the authors can address the following: The study by Manes (2000) et al.
The authors quite rightly included not only peer reviewed papers but also all published abstracts. This approach is according to standard meta-analysis practice, in order to avoid publication bias such that 'negative' studies (often not published in full) are not excluded. However, including data from published abstracts preclude scrutiny of the detailed methods, analyses and thus raises questions on the quality of the results. The situation in the meta-analysis of TTI vs TS highlights the pros vs cons of including data that were not peer reviewed.
Three of the four selected studies were published in respected journals and were subjected to peer-reviewing processes. The study by Manes however was published only as an abstract 14 years ago, and to date had not been published as a full paper. This by itself raises great concerns. The results of this particular study deviated significantly from all the other three. The authors have identified clear methodological concerns (especially 'recycling' of patients into randomization after failing pleurodesis). Unless the authors have obtained details from the primary research group, it is doubtful that a short abstract could provide adequate details for proper critique of the methods and results. Including this study without qualifying its many limitations may distort the interpretation of the readers.
I suggest that the authors should: Highlight the point that the Manes study was never published in full in the text/legends; Perform and show a separate analysis excluding the Manes study; Discuss the rationale of including/excluding this single study in the Discussion section. I believe the above measures are justified as the Manes study was not a 'negative' study and would never have been biased against if ever submitted for full publication.

An alternate way of presenting the data
Although it would not change the actual conclusion or the raw data, presenting the results as 'success' rather than 'failure' rates would quite significantly change the 'visual effects' of the graphs. Take for example the Dresler study. The RR for failure is 0.74 but if expressed as ratio of success rates it would become 1.11, the Yim study 1.04 and the Terres study 0.96. This probably presents a more useful interpretation for clinicians and patients -that the magnitude of superiority of TTI in any of the studies is at best 1.11 times over TS (excluding Manes ). et al.

Discussion
It would be useful to include that there are no scientific grounds why insufflation should be more advantageous than slurry. Talc does not work as a glue (otherwise we would have had major problems when talc was still included in baby nappy powders). Even distribution of talc over the pleural surface is not therefore critical. Radio-active isotope studies have shown that talc, even when applied as a slurry, can distribute around the pleural cavity via respiratory motions.

1.
I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
No competing interests were disclosed.

Competing Interests:
Reader Comment 17 Feb 2015 , Tuality Community Hospital/Oregon Health and Science University , USA

Srinivas Mummadi
Thank you very much your comments and a detailed critique of our manuscript.

Response to "The study by Manes (2000)" et al.
We agree with your assessment of the pros and cons of including the study by Manes et al. (2000) published only as an abstract. As our inclusion criteria were defined we a priori, included the study in our analysis.
We agree with your comments that it suffers from potential quality concerns. In addition to our stated concerns regarding the inappropriate allocation of randomization, we have made efforts to highlight additional quality concerns outlined below.
Footnotes to the tabular data have been added, highlighting the existing form of publication as an abstract. We also discussed in detail the lack of change in point estimates for all the predefined outcomes after removing the study as a part of the sensitivity analysis.
Due to its unique conclusion (TTI is superior to TS in terms of pleurodesis success rates) and a relatively large treatment effect, we believe that the inclusion of this study with the clear mentioning of its inherent limitations gives us a unique opportunity to present both sides of the talc poudrage versus talc slurry debate in a systematic review.
Inclusion of this study in the analysis allows meta-analysis to play the role of an adjudicator.
As mentioned earlier, removal of this study in the sensitivity analysis did not influence the results.
For the sake of clarity, we would like to report the results for all the studied outcomes after removing the above mentioned study in this reply.