We are grateful that scientists around the world are showing interest in our research, and delighted to reply to comments from Creaney et al. with regards to our paper the “Biological effect of tissue plasminogen activator (t-PA) and DNase intrapleural delivery in pleural infection patients”.[1]
Pleural infection is a significant clinical entity, has increasing incidence worldwide and is associated with high morbidity, mortality and burden to healthcare services. Effective treatment still relies upon effective pleural fluid drainage, and thus investigation of the pathological mechanisms behind fluid formation and treatment response is key to improving care.
The MIST2 study demonstrated that intrapleural delivery of tissue plasminogen activator (t-PA) alone or t-PA plus DNase increased volume of pleural fluid drained in humans with pleural infection, in a placebo controlled double blind randomised study.[2 ] Although the exact biological mechanisms via which t-PA induces the volume increment of drained pleural fluid are unknown, the design of the MIST2 study means that we are confident in the biological observation of increased fluid production in response to t-PA administration. Lansley et al. have demonstrated that the chemokine MCP-1 (also known as CCL-2) is the key protein that upon intrapleural t-PA administration induces fluid formation in healthy mice.[3]
We designed a study to directly assess their hypothesis in human pleural infection patients following intrapleural delivery of t-PA, using pleural fluid specimens collected for MIST2.
Our data suggest that MCP-1 is not the single modulator of the increment in pleural fluid drainage volume due to intrapleural administration of t-PA in humans. Furthermore, no correlation was detected between pleural fluid MCP-1 protein concentration and the volume of drained pleural fluid or the intrapleural drug given.
These data highlight the importance of the appropriate use of laboratory models when extrapolating preclinical findings to human biology.[4-6] Lansley et al. used a single mouse strain of the same sex for experiments (female CD1), and gave a single dose of drug to mice, while MIST2 participants received daily cumulative doses. It is unclear how the dose was adjusted to mouse body surface area. The mice were healthy, and the pleural cavity was not infected with bacteria in contrast to MIST2 samples which were derived from confirmed cased of pleural infection – it is well recognised that the presence of bacteria in the pleural space drastically affects the biological microenvironment.[7, 8]
These factors could explain the differences between our and their studies. Importantly, we highlight that our data suggest that MCP-1 plays a role but is not acting in isolation. More research is needed in order to fully understand the interplay between the pleura, bacteria and t-PA – however, the suggestion that MCP-1 is the sole driver of fluid formation in the human response to intrapleural t-PA based on a healthy mouse model is overly simplistic.
With regards to methodological comments from Creaney et al. we can clarify that total protein concentration was used in order to normalise the concentration of MCP-1. We did not use the total MCP-1 expression as suggested as this would not have allowed us to investigate the correlation between volume of pleural fluid drained and concentration of MCP-1 in pleural fluid. Their suggestion that adjusting for the total volume of pleural fluid present is required is not in our view rational, and makes the assumption that the concentration of MCP-1 is somehow different in the retained fluid in the thorax than that collected, and we see no justification for this assumption. The major pleural fluid parameters used in daily clinical practice (total protein, glucose, LDH) are all measured using concentrations – do they suggest that these parameters are not reliable unless the entire pleural space is drained, or unless total amount is measured? We are also somewhat surprised at their suggestion of the primacy of total (as opposed to concentration) of MCP-1, given that many papers routinely report the concentration of cytokines in pleural fluid (and not total amount) including papers published by the authors of the e-letter to which we here respond, including in their own studies of MCP-1.[9-13] Their suggestion that DNase may have interfered with the analysis is not based on any published data we can find - DNase is an enzyme that cleaves the phosphodiester linkages in DNA molecules and no interaction with chemokines has been reported. We acknowledge that the residual pleural fluid is a limitation of the study. However, even in animal models there is loss of fluid during aspiration.
There is no doubt that laboratory animal models of human disease are of importance and have significantly contributed to major scientific discoveries. However, preclinical disease models must be used with care to ensure that they faithfully and appropriately reflect the properties of any studied disease, and their purposes (which is to study the human disease condition) should not be forgotten. Creaney et al. conclude by suggesting that “Clinical observations are useful in generating hypotheses”. The MIST2 study was not a “clinical observation” and remains the only placebo controlled double blind study in humans with pleural infection treated with t-PA and DNase, and as such, translational discoveries from this dataset are likely to most closely represent the underlying pathobiology in human pleural infection.

References
1. Kanellakis NI, Wrightson JM, Hallifax R, et al. Biological effect of tissue plasminogen activator (t-PA) and DNase intrapleural delivery in pleural infection patients. BMJ Open Respir Res 2019;6(1):e000440. doi: 10.1136/bmjresp-2019-000440 [published Online First: 2019/11/02]
2. Rahman NM, Maskell NA, West A, et al. Intrapleural use of tissue plasminogen activator and DNase in pleural infection. N Engl J Med 2011;365(6):518-26. doi: 10.1056/NEJMoa1012740 [published Online First: 2011/08/13]
3. Lansley SM, Cheah HM, Varano Della Vergiliana JF, et al. Tissue plasminogen activator potently stimulates pleural effusion via a monocyte chemotactic protein-1-dependent mechanism. Am J Respir Cell Mol Biol 2015;53(1):105-12. doi: 10.1165/rcmb.2014-0017OC [published Online First: 2014/12/05]
4. Hackam DG, Redelmeier DA. Translation of research evidence from animals to humans. JAMA 2006;296(14):1731-2. doi: 10.1001/jama.296.14.1731 [published Online First: 2006/10/13]
5. Knight A. Systematic reviews of animal experiments demonstrate poor contributions toward human healthcare. Rev Recent Clin Trials 2008;3(2):89-96. doi: 10.2174/157488708784223844 [published Online First: 2008/05/14]
6. Roberts I, Kwan I, Evans P, et al. Does animal experimentation inform human healthcare? Observations from a systematic review of international animal experiments on fluid resuscitation. BMJ 2002;324(7335):474-6. doi: 10.1136/bmj.324.7335.474 [published Online First: 2002/02/23]
7. Corcoran JP, Wrightson JM, Belcher E, et al. Pleural infection: past, present, and future directions. Lancet Respir Med 2015;3(7):563-77. doi: 10.1016/S2213-2600(15)00185-X [published Online First: 2015/07/15]
8. Lee YC, Idell S, Stathopoulos GT. Translational Research in Pleural Infection and Beyond. Chest 2016;150(6):1361-70. doi: 10.1016/j.chest.2016.07.030 [published Online First: 2016/08/16]
9. Thomas R, Cheah HM, Creaney J, et al. Longitudinal Measurement of Pleural Fluid Biochemistry and Cytokines in Malignant Pleural Effusions. Chest 2016;149(6):1494-500. doi: 10.1016/j.chest.2016.01.001 [published Online First: 2016/02/03]
10. Bielsa S, Davies HE, Davies RJ, et al. Reproducibility and reliability of pleural fluid cytokine measurements. Eur Respir J 2009;34(4):1001-3. doi: 10.1183/09031936.00088409 [published Online First: 2009/10/03]
11. Jin HY, Lee KS, Jin SM, et al. Vascular endothelial growth factor correlates with matrix metalloproteinase-9 in the pleural effusion. Respir Med 2004;98(2):115-22. doi: 10.1016/j.rmed.2003.09.002 [published Online First: 2004/02/20]
12. Rovina N, Dima E, Psallidas I, et al. Interleukin-18 is up-regulated in infectious pleural effusions. Cytokine 2013;63(2):166-71. doi: 10.1016/j.cyto.2013.04.017 [published Online First: 2013/05/11]
13. Lansley SM, Cheah HM, Lee YC. Role of MCP-1 in pleural effusion development in a carrageenan-induced murine model of pleurisy. Respirology 2017;22(4):758-63. doi: 10.1111/resp.12951 [published Online First: 2016/11/24]

We read with interest the recent study by Morton et al. A “National COPD Collaborative” quality improvement (QI) initiative (The Collaborative) which is currently on-going in Ireland is also evaluating the efficacy of bundles, amongst other interventions, in improving COPD care. Running from September 2018 to December 2019, the Collaborative comprises 18 consultant-led teams in 19 hospitals across the country working to improve care for patients presenting with an acute exacerbation of COPD (AECOPD). The Collaborative is being run by the Royal College of Physicians of Ireland (RCPI) in conjunction with the Clinical Strategy and Programmes Division of the Irish Health Service Executive (HSE) and the National Clinical Programme for COPD within the HSE.
COPD is a major health burden in Ireland, as in the UK; based on the 2011 census (total population 4,588,252 [1]), it is estimated that at least 440,000 people in Ireland have COPD (of whom over 180,000 have moderate or severe disease) [2]. In 2015, Ireland had the highest rate of COPD hospital admissions out of all OECD countries [3]. The cost burden of COPD on the HSE is substantial; in 2014, the total cost of COPD hospitalisations was €70,813,040.00 [4]. According to the OECD, the average length of hospital stay (LOS) for COPD in Ireland in 2017 was eight days[5].
Prior to the initiation of the National COPD Collaborative, treatment of AECOPD within the acute Irish healthcare setting was highly varied; many aspects of the COPD hospital care pathway, including the diagnosis of COPD at presentation, LOS, treatment approaches and availability of respiratory consultant-led team(s) dedicated to in-patient care of AECOPD (as opposed to out-patient care), were not standardised between hospital sites. As Morton et al imply, improving care for patients with AECOPD relies on appropriately expert clinicians reliably following evidence based practice. Therefore, the National COPD Collaborative is attempting to improve and standardise the treatment of COPD/AECOPD patients across participating hospitals, at the following stages:
1) Presentation (e.g. time from registration to first medical review; time from registration to decision to admit)
2) Admission (e.g. time from registration to first respiratory specialist review; LOS; DECAF assessment [6], documented COPD diagnosis – spirometry; chest x-ray; steroids and route [PO/IV], antibiotics and route [PO/IV])
3) Discharge (e.g. inhaler technique; script review with patient; written self-management plan; follow-up)

In order to effect improvements in the care of AECOPD across participating hospital sites, the Institute for Healthcare Improvement “Breakthrough Series Collaborative” is being used to build QI capability amongst teams. Individual teams employ the “Model for Improvement” [7] , which utilises Plan, Do, Study, Act (PDSA) cycles, to bring about change. One of the primary aims of the National COPD Collaborative is to facilitate participating acute respiratory teams in the development their QI skills which will, therefore, allow them to apply QI methods to bring about treatment improvements at a local level.
Evaluation of the Collaborative focuses on assessment of monthly patient data which is submitted by each hospital for 22 measures related to the care of AECOPD at presentation, admission and discharge (the “bundles”); these will be analysed over time using run charts for QI purposes.
The hospital phase of the National COPD Collaborative will be completed by September 2019. Morton et al note that “Staff viewed bundles positively, believing they help standardise practice and facilitate communication between clinicians. However, they lacked skills in change management, leading to inconsistent implementation”. We speculate that should any improvement in patient care be demonstrated at the end of the National COPD Collaborative, this change may be due, at least in part, to the enthusiastic engagement of sites with QI methodology which allows them to adapt QI processes to the needs of their own hospital, while standardising care for patients with AECOPD nationally.

References
1. CSO. Census through History. 2019 30 July 2019]; Available from: https://www.cso.ie/en/census/censusthroughhistory/.
2. Jennings, S., Preventing chronic disease: defining the problem. 2014, Health Service Executive
3. NPSO, National Healthcare Qualtiy Reporting System Annual Report 2018, N.P.S. Office, Editor. 2018, Department of Health.
4. COPDSI, Manifesto for COPD, C.S. Ireland, Editor. 2015.
5. OECD. Hospital average length of stay by diagnostic categories: Chronic Obstructive Pulmonary Disease and bronchiectasis. 2017 30 July 2019]; Available from: https://stats.oecd.org/index.aspx?queryid=30165#.
6. Steer, J., J. Gibson, and S.C. Bourke, The DECAF Score: predicting hospital mortality in exacerbations of chronic obstructive pulmonary disease. Thorax, 2012. 67(11): p. 970-6.
7. Langley, G., et al., The Improvement Guide: A practical Approach to Enhancing Organizational Performance. 2nd ed. 2009, San Francisco: Jossey-Bass Publishers.

It was interesting to review the findings of the study by Barker and colleagues [1] regarding airway clearance techniques (ACTs) for patients with chronic obstructive pulmonary disease (COPD). The authors suggest there may be a relative ‘under-prescription’ of simple adjunctive therapy options such as oscillatory positive expiratory pressure (OPEP) devices for the care of this patient group and reflect on the ability for ‘actual use’ registry data to inform research questions in this area. The data certainly show OPEP devices to be used less frequently and at significantly lower costs than prescription medications such as Tiotropium and Carbocisteine. This is not surprising, and is a likely accurate reflection of clinical practice.

Identification of the most important clinical question(s) to arise from this data, however, appears quite challenging owing to significant limitations that may have been under-emphasised. The true incidence of OPEP device use amongst patients with COPD will certainly have been underestimated in the data obtained from the OpenPrescribing.net resource due to the ‘hidden’ market related to private or hospital-based device purchases. While this is acknowledged by the authorship team, we feel this may be more significant than that proposed. Our experience as clinical physiotherapists working in this area suggests a large proportion (if not the majority) of airway clearance therapy is the remit of physiotherapists working in cardiorespiratory medicine based primarily in (or outreached from) hospitals. This appears supported by the authors’ survey data responses, of which 72% of the 116 respondents stated they worked in hospitals. In these settings, ACTs are rarely formally ‘prescribed’ in medical registries, but rather implemented using existing resources, purchased privately or loaned / provided from the centre (as indicated in the survey instrument). Cost is a realistic barrier to OPEP device use, often resulting in education to substitute device use at home for cheaper, alternative means such as bottle-PEP (which could be performed at no cost). This would not appear on a prescription registry.

OPEP devices may also be considered clinically indicated for certain patients, but this does not necessarily mean they are prescribed. Previous data from physiotherapists across the world demonstrates considerable variability regarding ACT prescription for this patient group, particularly during acute exacerbations. The most commonly used ACTs include the active cycle of breathing technique and deep breathing exercises in the United Kingdom, Australia, New Zealand [2-4], ‘conventional chest physiotherapy’ techniques (e.g. postural drainage, percussion, vibration) in Canada [5] and even physical exercise in Australia [4], despite scarce research examining this precise effect [6]. PEP device use did not feature prominently in any of these surveys, however its use is known to be widespread in countries such as Sweden.

It is also unclear whether survey data (e.g. Table 2, Figure 1) represents opinion or actual practice patterns. Questions regarding ‘considerations’ or ‘recommendations’ may be quite different to treatment that a patient actually receives, particularly if the recommending therapist is different to the professional responsible for treatment ‘delivery’. And the survey responses, which represent only 12% of ACPRC members (and should therefore not be referred to as ‘clinical consensus’), may differ considerably from other members and/or the health professions represented in the prescription data. It would, of course, be interesting to know the extent of overlap between the two samples, however this is not likely to be possible. This uncertainty makes it challenging to comprehend the precise area in need of further research, particularly as it is possible the findings simply reflect differing views of two distinct health professions groups. For example, it is plausible the survey may primarily reflect opinions of physiotherapists but prescribing data that of general practitioners. If so, there remains a need to source more accurate information about OPEP prescription practices, potentially via closer examination of data originating from the therapists responsible for its implementation.

Finally, as stated by the authors, we currently have very little clinical data upon which to base decisions regarding long-term ACTs for patients with COPD and chronic sputum production [7]. Further research is indeed required, but it is crucial that good judgment underpins decisions regarding the most relevant clinical question(s) to investigate. Identifying patients with stable COPD and chronic sputum production (in the absence of co-existing bronchiectasis, which reportedly occurs in up to 50% of COPD cases [8]) is not without challenge, with one previous physiological investigation of PEP therapy for this patient group observing a 43% exclusion rate of those screened due to a lack of regular sputum production [9].

We feel some of the potential areas for future, clinically impactful research arising from this work may include identification of:
i) The reasons underpinning the high use of mucolytic therapy in this patient group despite a lack of consensus recommendations in international COPD guidelines [10-12];
ii) Patient acceptability regarding performance of long-term ACTs; and
iii) Distinct clinical phenotypes who may benefit most from future, high quality clinical trials of ACTs.

In conclusion, we congratulate Barker and colleagues for raising awareness of some of the challenging issues clinicians encounter when trying to deliver evidence-based care for patients with COPD and chronic sputum production. We encourage the issues to be further examined but implore a degree of caution when using inherently limited data as the foundation for future, clinically relevant research seeking to improve patient care.

1. Barker R, Laverty AA, Hopkinson NS. Adjuncts for sputum clearance in COPD: clinical consensus versus actual use. BMJ Open Respiratory Research 2017;4(1) doi: 10.1136/bmjresp-2017-000226.
2. Lee A, Button B, Denehy L. Current Australian and New Zealand physiotherapy practice in the management of patients with bronchiectasis and chronic obstructive pulmonary disease. New Zealand Journal of Physiotherapy 2008;36(2):49-58
3. Yohannes AM, Connolly MJ. A national survey: percussion, vibration, shaking and active cycle breathing techniques used in patients with acute exacerbations of chronic obstructive pulmonary disease. Physiotherapy 2007;93(2):110-13
4. Osadnik CR, McDonald CF, Holland AE. Airway clearance techniques in acute exacerbations of COPD: a survey of Australian physiotherapy practice. Physiotherapy 2013;99(2):101-6 doi: 10.1016/j.physio.2012.01.002.
5. Harth L, Stuart J, Montgomery C, et al. Physical therapy practice patterns in acute exacerbations of chronic obstructive pulmonary disease. Can Respir J 2009;16(3):86-92
6. Oldenburg FA, Jr., Dolovich MB, Montgomery JM, Newhouse MT. Effects of postural drainage, exercise, and cough on mucus clearance in chronic bronchitis. Am Rev Respir Dis 1979;120(4):739-45
7. Osadnik CR, McDonald CF, Jones AP, Holland AE. Airway clearance techniques for chronic obstructive pulmonary disease. Cochrane Database Syst Rev 2012(3):CD008328 doi: 10.1002/14651858.CD008328.pub2.
8. Patel IS, Vlahos I, Wilkinson TM, et al. Bronchiectasis, exacerbation indices, and inflammation in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2004;170(4):400-7 doi: 10.1164/rccm.200305-648OC200305-648OC.
9. Osadnik C, Stuart-Andrews C, Ellis S, Thompson B, McDonald CF, Holland AE. Positive expiratory pressure via mask does not improve ventilation inhomogeneity more than huffing and coughing in individuals with stable chronic obstructive pulmonary disease and chronic sputum expectoration. Respiration 2014;87(1):38-44 doi: 10.1159/000348546.
10. National Clinical Guideline Centre. Chronic obstructive pulmonary disease: management of chronic obstructive pulmonary disease in adults in primary and secondary care. 2010. http://guidance.nice.org.uk/CG101/Guidance/pdf/English (accessed 15/08/2017).
11. Global Initiative for Chronic Obstructive Lung Disease (GOLD). Global Strategy for the Diagnosis, Management and Prevention of COPD (2017 report). 2017. http://www.goldcopd.org/.
12. Yang IA, Dabscheck E, George J, et al. The COPD-X Plan: Australian and New Zealand Guidelines for the management of Chronic Obstructive Pulmonary Disease 2017. Version 2.50, June 2017. 2017. http://www.copdx.org.au/.