The National Institute for Health and Care Excellence (NICE) 2016 Tuberculosis (TB) guidelines no longer recommend screening contacts of adults with extra-pulmonary TB (ETB). However, no new evidence since the previous published guidelines was provided to support this policy change. Moreover, despite the guidance, some regional TB multidisciplinary teams and services continue to screen ETB contacts.(1)

In their original article in Thorax, Cavany et al estimated the cost-effectiveness of screening ETB contacts in London.(2) The authors’ findings suggest that screening of such contacts is unlikely to be cost-effective at the threshold of £30,000/QALY - the “willingness to pay” threshold commonly used by NICE.(3) The authors’ findings are tempered by the data being London-specific and not generalizable to the rest of England, and the lack of robust available evidence on either transmission rates or index cases’ pre-diagnosis symptom duration. Nevertheless, the authors recognise these limitations and their sensitivity analysis suggests that, even with assumptions of higher rates of transmission or prolonged symptom duration, their principal findings would not change.

The findings of this strong, well-designed study are important and provide much needed evidence for national debate around strategies for TB contact screening. Resources for TB services across England, especially those allocated to tracing contacts of TB patients, are becoming increasingly constrained. Therefore, it is vital that a pragmatic approach is taken to prioritise the most cost-effective, evidence-based strategies for identifying people most at risk of latent TB infection and progression to active TB disease. Cavany et al’s excellent paper raises certain key points relating to the wider policy implications of the cost-effectiveness of TB prevention strategies in low-burden settings.

First, as set out in the World Health Organization’s “End TB Strategy”(4) and PHE and NHS England’s “Collaborative Tuberculosis Strategy 2015-2020”,(5) the shared aim of the global TB community is to eliminate tuberculosis (defined as less than one case of active TB disease per 100,000 population). This goal is especially pertinent with regards to cost-effectiveness in low burden settings like Western Europe because, as incidence and prevalence of TB continue to decline, the cost per tracing episode will increase, probably non-linearly, and therefore meeting cost-effectiveness thresholds will become increasingly difficult.(6) For example, in North West England, declines in TB incidence have been associated with an increase in the proportion of TB cases who have complex social and clinical risk factors requiring enhanced case management (ECM) and hence greater per patient resource allocation.(7)

Second, cost-effectiveness estimates often assume homogeneity of clinical and social risk factors of TB patients and contacts when, in practice, such populations are profoundly diverse. We conducted an analysis of our North West England cohort of 2,139 TB cases and their 10,019 contacts. Our research showed that prevalence of active TB disease among contacts of patients with ETB was similar to the London cohort (0.44% versus 0.7% in Cavany et al, respectively) but varies widely according to the index patient’s social and clinical complexity as measured by ECM need.(1) Excluding contacts who would meet other criteria for routine contact tracing and/or active case finding (e.g. migrants, homeless people, or drug users), the rate of active TB disease in contacts of extrapulmonary TB patients with no ECM need versus the highest ECM need were 0.06% versus 1.5%, respectively.(8) Therefore, whilst we agree that contact tracing of all ETB contacts may not be cost-effective, focusing on a small number of high-risk ETB contacts (in whom approximately 1 in 66 will have active TB disease in our North West cohort) may still be an appropriate allocation of resources in the context of aiming towards regional elimination of TB.

Third, robust cost-effectiveness analyses such as Cavany et al’s should motivate the TB community in England to evaluate the predominant drivers of the costs of their TB detection and prevention services. If existing contact tracing strategies are, as in Cavany et al’s research, shown to lack cost-effectiveness, TB multi-disciplinary teams and service providers should consider designing and implementing innovative approaches that are more efficient and better value for money. Identification and characterisation of patients’ risk of an adverse treatment outcome and contacts’ risk of a positive screening event using ECM is one simple strategy,(7,8) but there are many other potential options including: electronic or virtual screening; community engagement and outreach clinics; patient and peer-led active case finding; and socioeconomic support, incentives and enablers. It is essential that future research designing and implementing such novel strategies follows Cavany et al’s lead and includes rigorous cost-effectiveness analysis in order to best inform and guide policy makers and public health commissioners.(9)

Cavany et al’s research highlights the importance of high-quality cost-effectiveness analysis to guide TB policy. In order to provide optimal TB detection, prevention, and care services tailored to individuals, especially those in underserved groups, future TB policy decisions should take into account heterogeneity within TB populations and the inevitable decrease in cost-effectiveness of interventions in line with declining TB rates. Without such considerations, we will be unlikely to meet the goal of eliminating TB in England.

References

1) Wingfield T, MacPherson P, Cleary P, Ormerod LP. High prevalence of TB disease in contacts of adults with extrapulmonary TB. Thorax. 2018 Aug;73(8):785-787. doi: 10.1136/thoraxjnl-2017-210202. Epub 2017 Nov 16.

2) Cavany SM, Sumner T, Vynnycky E, Flach C, White RG, Thomas HL, Maguire H, Anderson C. An evaluation of tuberculosis contact investigations against national standards. Thorax. 2017 Aug;72(8):736-745. doi: 10.1136/thoraxjnl-2016-209677. Epub 2017 Apr 7.

3) McCabe C, Claxton K, Culyer AJ. The NICE cost-effectiveness threshold: what it is and what that means. Pharmacoeconomics. 2008;26(9):733-44.

4) World Health Organisation’s End Tuberculosis Strategy, 2015.
http://www.who.int/tb/strategy/End_TB_Strategy.pdf?ua=1 WHO, Geneva. Last accessed 29th August 2018.

5) Collaborative Tuberculosis Strategy for England 2015 to 2020. NHS and Public Health England. https://assets.publishing.service.gov.uk/government/uploads/system/uploa... Last accessed 29th August 2018

6) Smit GS, Apers L, Arrazola de Onate W, Beutels P, Dorny P, Forier AM, Janssens K, Macq J, Mak R, Schol S, Wildemeersch D, Speybroeck N, Devleesschauwer B. Cost-effectiveness of screening for active cases of tuberculosis in Flanders, Belgium. Bull World Health Organ. 2017 Jan 1;95(1):27-35. doi: 10.2471/BLT.16.169383. Epub 2016 Nov 3.

7) Tucker A, Mithoo J, Cleary P, Woodhead M, MacPherson P, Wingfield T, Davies S, Wake C, McMaster P, Bertel Squire S. Quantifying the need for enhanced case management for TB patients as part of TB cohort audit in the North West of England: a descriptive study. BMC Public Health. 2017 Nov 15;17(1):881. doi: 10.1186/s12889-017-4892-5.

8) Wingfield T, MacPherson P, Sodha P, Tucker A, Mithoo J, Squire S B, Cleary P. The association of TB patients’ social risk factors and ethnicity with prevalence of TB infection and disease among their contacts: a cohort study in North West England. Under review with International Journal of Tuberculosis and Lung Disease, August 2018.

9) Lönnroth K, Migliori GB, Abubakar I, et al. Towards tuberculosis elimination: an action framework for low-incidence countries. Eur Respir J. 2015 Apr;45(4):928-52. doi: 10.1183/09031936.00214014

We support the view of Drs. Polkey and Ambrosino that recommendations for clinical practice should not be based on either positive or negative preoccupation concerning the potential effectiveness of a treatment but rather on an impartial evaluation of the available data. In their editorial entitled ‘Inspiratory Muscle Training in COPD: can data finally beat emotion’ they unfortunately provide a fairly one-sided evaluation of this treatment, based on an incomplete and largely outdated review of the available evidence1. It is unfortunate that they neglect a major part of available data, which could contribute to a more balanced and fair discussion about this intervention. We therefore deemed it necessary to add this missing evidence along with our own interpretation of recent findings to the discussion.
Complexity of studying add-on interventions to pulmonary rehabilitation
Based on the results from three recent multicentre trials2-4, Polkey and Ambrosino exclude a role for adjunctive IMT in the rehabilitation of patients with COPD. As emphasized in a previous opinion piece by Dr. Ambrosino5, it is important to distinguish between studies that evaluate the effects of inspiratory muscle training (IMT) as a standalone intervention (i.e. in comparison to no intervention or a sham control intervention) and studies on the effects of IMT added to a pulmonary rehabilitation program (PRP).
Concerning the first comparison, there is a large amount of data available supporting the effectiveness of IMT in COPD. The most recent meta-analysis identified 43 randomized controlled trials6. Based on data from more than 1200 patients randomized to either IMT or control interventions authors concluded that IMT improved symptoms of dyspnea (Baseline Dyspnea Index), Quality of life (Saint George's Respiratory Questionnaire), functional capacity (Six minute walking distance, 6MWD) and maximal inspiratory pressure (PImax) in patients with COPD6. A clinically relevant difference of 43m (95%CI: 17m to 69m) in improvements in 6MWD after IMT in comparison to control was reported. The claim that ‘the appeal of IMT in COPD endures in the absence of data’ seems inappropriate in the light of above-mentioned findings. Polkey and Ambrosino further support their conclusions by mentioning that IMT also failed to improve exercise capacity in healthy subjects. They support this statement by citing a single study performed in 20017. In doing so they neglect results from a systematic literature review from 2012 including results from 46 randomized controlled studies. From the meta-analysis it was concluded that ‘respiratory muscle training improves endurance exercise performance in healthy individuals with greater improvements in less fit individuals and in sports of longer durations’8.
In contrast to studies of IMT as standalone intervention, when evaluating IMT as adjunctive intervention to PRP the evidence base looks different. In these type of studies, the control group already receives a very effective treatment (PRP) that will result in clinically relevant improvements in functional capacity and quality of life9. This makes it challenging to demonstrate additional effects of the adjunctive intervention on functional outcomes. Consequently, study designs are more complex, the database is less comprehensive and conclusions less straightforward6. IMT is not the only intervention that has failed to result in additional improvements in the 6MWD in this context (i.e. studying add-on interventions to PRP). A recent systematic review of RCTs studying the effects of various adjunctive interventions applied during PRP (including non-invasive ventilation, lower limb strength training, upper limb exercise training, oxygen supplementation, and inspiratory muscle training) did not observe additional effects of any of these adjunctive interventions during PR on the 6MWD10. When comparing the results of studies on the adjunctive effects of either lower limb strength training or IMT during PRP the results are of striking similarity. In both cases, the adjunctive interventions result in additional specific improvements in muscle function of the trained muscle groups without resulting in further improvements in 6MWD or quality of life measures10. It is questionable whether additional improvements in health related quality of life are a realistic goal when studying effects of these adjunctive interventions. Control groups participating in comprehensive PRP without adjunctive interventions will already achieve large and clinically relevant improvements in quality of life. It has further been demonstrated that changes in functional exercise capacity correlate poorly with changes in quality of life measures11. The questions should rather be whether in the specific case of IMT the adjunct intervention can (1) further reduce specific symptoms of exertional breathlessness and (2) whether it can optimize the response in functional capacity to PRP in participants who would otherwise achieve a less optimal response. In addition, the most appropriate candidates (i.e. likely responders to the adjunct intervention) and most suitable outcome measures still need to be identified. In order to find out whether the intervention can result in additional effects on functional capacity it will be necessary to strictly control both adherence with and quality of the adjunctive intervention as well as the symptom-based progression of the general exercise training program in this type of studies. To the best of our knowledge only one of the available studies has so far undertaken this effort4. The absence of evidence from studies on add-on interventions at this point should therefore not be interpreted as evidence of absence. Along these lines, the value of lower limb strength training during PRP in patients with COPD is currently not put into question based on limted available data. It is surprising however that, based on very similar data, it is concluded that there should be no role for IMT in the rehabilitation care of patients with COPD anymore. Instead of interpreting the results of recent studies2-4 as ultimate proof to discourage the use of IMT in COPD, we would rather argue that these findings should be used to optimize the design of future studies in this area. Unfortunately, Polkey and Ambrosino neglect most of the key findings from our recent trial that could be helpful for the design of these future studies. We will discuss these aspects in the final part of our comment. Prior to that we would like to review some of the other arguments that are used by Polkey and Ambrosino to support their statements.
IMT and the diaphragm in patients with COPD
They start by citing Leith and Bradleys pioneer work on ‘ventilatory muscle strength and endurance training’ from 1976. According to Polkey and Ambrosino these data, obtained in a handful of young healthy volunteers, support the assumption that inspiratory muscle training improves ‘test performance’ rather than ‘true contractility’12. We are unsure what their definition of ‘true contractility’ is, but it is generally accepted that the principles of specificity of resistance training apply to both training of the peripheral muscles13, and respiratory muscle training14. Results of several randomized controlled studies in patients with COPD have taught us that endurance type training (high flow, low resistance breathing) will specifically improve endurance capacity of the respiratory muscles, while PImax will only improve when subjects perform their breathing training against an external load of sufficiently large magnitude15. We would further like to comment on their statement that the reduction in inspiratory muscle strength in COPD is ‘due to hyperinflation rather than weakness’. While static hyperinflation certainly affects diaphragm contractility, several other etiological factors and biological mechanisms are known to contribute to respiratory muscle dysfunction in COPD16 17. These additional factors, many of which are also involved in peripheral muscle dysfunction, probably also help to explain the large variability in PImax that has been observed for a given level of static lung hyperinflation18.
It is also noteworthy that, while limitations in diaphragm contribution in patients are acknowledged on the one hand, diaphragm strength evaluated by phrenic nerve stimulation is subsequently put forward as the most relevant measure of changes in inspiratory muscle function in patients with COPD. This seems to be a limited approach especially for patients with COPD in whom, with increasing disease severity, the role of the rib cage muscles becomes more and more important19. As a consequence it seems reasonable that both specific tests of diaphragm strength as well as measures of global inspiratory muscle strength (and endurance) should be used to evaluate the effect of interventions on inspiratory muscle function20. The fact that IMT did not increase diaphragm strength as judged by isolated (artificial) phrenic nerve stimulation (Twitch-Pdi) is used as one of the arguments to support the conclusion that ‘the story of IMT defied data at each stage’. We would like to comment on this. Firstly, the paper that is cited in support of this observation demonstrates a poor relationship between changes in Twitch-Pdi and PImax 21. Secondly, the data, again obtained in a handful of healthy subjects, reveal a high variability between Twitch-Pdi measurements performed before and after an intervention period in a control group that did not perform IMT21. We therefore believe that conclusions with regard to these observations should be formulated more cautiously.
High fatigue resistance of the diaphragm is subsequently put forward as another argument why respiratory muscles of patients with COPD should not be trained. Indeed, adaptations in the diaphragm towards more fatigue resistance, most likely in response to chronic increased loading during resting breathing, are present in patients16 17. We also agree with Polkey and Ambrosino that it is unclear whether diaphragmatic muscle fatigue typically occurs during exercise breathing in patients with COPD. We would however like to highlight that acute length changes induced by exercise breathing (due to dynamic hyperinflation) will further impair the ability of the diaphragm to contribute to ventilation and will place increasing demands on accessory and rib cage muscles19. It is reasonable to assume that this breathing pattern would promote non-diaphragmatic respiratory muscle fatigue rather than diaphragm fatigue. Notwithstanding the difficulties in reliably assessing non-diaphragmatic muscle fatigue there are indications that fatigue in these muscles can develop selectively and separate from diaphragm fatigue, depending on the breathing pattern that is maintained22-24. The reluctance of Polkey and Ambrosino to acknowledge that the stimulus of exercise hyperpnea, as a consequence of further acute shortening of respiratory muscles and increased elastic loads due to dynamic hyperinflation, can result in acute disturbances in the load – capacity balance of the respiratory muscles of patients with COPD is surprising to say the least25 26. Especially given the fact that in an earlier opinion piece Dr Polkey highlighted the ‘specific relevance of [breathing close to] end inspiratory lung volume’ in patients with COPD, supported ‘by the observation that augmenting inspiratory muscle action at high lung volumes […] can extend exercise duration and increase peak work load27.’ It remains his secret as to why he would not appreciate a role for inspiratory muscle training in augmenting inspiratory muscle performance at high lung volumes during exercise breathing. In this context we would like to refer to a recent study28, in which we were able to demonstrate dyspnea relief during a constant work rate cycling test in response to IMT in patients with COPD. The effects of IMT were observed in patients with static and dynamic hyperinflation. Dyspnea reduction was observed in conjunction with a reduced activation of the diaphragm relative to maximum and in the absence of significant changes in ventilation, breathing pattern and operating lung volumes during cycling exercise 28. These data support previous findings of reduced diaphragm fatigability after IMT in patients with COPD29.
Choosing adequate outcomes to evaluate the effects of adjunctive IMT
As mentioned earlier it is a pity that some important results from our recent trial (IMT as adjunctive intervention during PRP)4, along with suggestions to improve the design of future studies have not been addressed by Polkey and Ambrosino. We would therefore like to highlight some of our key findings. In the absence of an additional effect of IMT on the 6MWD, we were able to demonstrate a significant improvement in the ability to sustain a submaximal effort during a constant work rate endurance-cycling test, as well as a reduction in symptoms of dyspnea at a standardized timepoint during this test4. The absence of evaluating symptoms at standardized levels of exertion has been identified as one of the shortcomings in one of the studies published earlier this year30. In our opinion these iso-time evaluations should be performed with the multidimensional dyspnea profile (MDP) in addition to Borg-CR 10 scale dyspnea scores in future studies. The higher sensitivity of submaximal constant work rate endurance exercise tests in comparison with the 6MWD has consistently been documented in pharmacological and non-pharmacological interventions in patients with COPD 31-33. The constant work rate endurance test therefore seems to be an optimal choice as a primary outcome for evaluating the effects of adjunctive interventions during PRP on functional capacity in future studies.
Optimization of patient selection
Furthermore, our recent findings also offer some food for thought with regard to selecting appropriate candidates for the adjunct intervention (IMT). While for peripheral muscle strength training it is reasonable to assume that the subjects with most pronounced weakness will benefit the most from the intervention, this might not be the case when selecting patients with respiratory muscle weakness. In our study we selected patients with pronounced weakness. This resulted in a sample of severely hyperinflated subjects (average RV >200%pred). The mechanical disadvantage of the respiratory muscles in these severely hyperinflated participants is probably a major contributor to the reduction in respiratory muscle pressure generating capacity. It has also been demonstrated that some of these severely hyperinflated patients are unable to dynamically hyperinflate during exercise34. These patients would consequently not undergo additional functional weakening of their inspiratory muscles due to further acute shortening during exercise. These patients will also not experience a great deal of additional loading on their respiratory muscles during exercise breathing due to both their inability to achieve major increases in tidal volume expansion, minute ventilation and their inability to dynamically hyperinflate during exercise breathing. Both factors (the large impact of mechanical disadvantage on weakness and the absence of acute disturbances in the load-capacity balance during exercise hyperpnoea) might render IMT less effective in these patients. Measurements of dynamic hyperinflation during exercise (which were not available in our study) would be needed to explore these issues further in order to identify participants who are potentially less likely to benefit from adjunctive IMT. In a further step one might choose to either exclude these patients from participation or to stratify randomization for this criterion.
Improving control of interventions
In addition, we were the first to show a relationship between quality and quantity of the performed IMT sessions and improvements in PImax4. We were able to control training parameters by using an electronic training device that stored training parameters recorded during home based IMT sessions35. We also observed that changes in PImax were significantly related to increases in the symptom-based progression of the intensity of the general exercise-training program and to improvements in functional exercise capacity4. This highlights the importance of monitoring and controlling both adherence to and quality of the adjunctive intervention, as well as the symptom based progression of the general exercise training components of the PRP in future studies. Sufficient time also needs to be allowed to establish the effects of IMT on symptom perception during exercise. This will enable patients to achieve larger increases in symptom based progression of the general exercise program. These increases in training intensity are a prerequisite to achieve additional benefits in functional capacity. Of the three studies of adjunctive IMT in COPD published this year2-4, ours was the only one that made an attempt to control for these parameters and we strongly encourage the use of these measures in future work in this area.
Conclusions
In summary, we feel that formulating strong statements to either discourage or recommend the use of IMT for patients with COPD in clinical practice is inappropriate at this stage. The results of recently published studies should in our opinion rather be regarded as a call for further research on the effects of IMT as an adjunct to PRP in patients with COPD. These studies will be needed to identify the best candidates for this adjunctive treatment. Based on the existing evidence and our own experience we formulate the following recommendations for further research:
1. Select patients with respiratory muscle weakness and possibly exclude or stratify those patients who are unable to dynamically hyperinflate due to severe static lung hyperinflation.
2. Tightly control adherence with and quality of the adjunct IMT intervention.
3. Tightly control symptom based progression of general exercise training intensity during the PRP and preferably allow enough time (> 8 weeks) for effects of IMT on symptom perception to establish.
4. Have general exercise training intervention executed and monitored by highly trained professionals who are blinded to group allocation.
5. Select changes in symptoms of dyspnea (MDP scores or Borg CR-10 Scale dyspnea score) at standardized levels of exertion during constant work rate endurance exercise tests as the primary outcome of these studies30 36.
References
1. Polkey MI, Ambrosino N. Inspiratory muscle training in COPD: can data finally beat emotion? Thorax 2018 doi: 10.1136/thoraxjnl-2018-212070 [published Online First: 2018/06/28]
2. Beaumont M, Mialon P, Le Ber C, et al. Effects of inspiratory muscle training on dyspnoea in severe COPD patients during pulmonary rehabilitation: controlled randomised trial. Eur Respir J 2018;51(1) doi: 10.1183/13993003.01107-2017 [published Online First: 2018/01/27]
3. Schultz K, Jelusic D, Wittmann M, et al. Inspiratory muscle training does not improve clinical outcomes in 3-week COPD rehabilitation: results from a randomised controlled trial. Eur Respir J 2018;51(1) doi: 10.1183/13993003.02000-2017 [published Online First: 2018/01/27]
4. Charususin N, Gosselink R, Decramer M, et al. Randomised controlled trial of adjunctive inspiratory muscle training for patients with COPD. Thorax 2018 doi: 10.1136/thoraxjnl-2017-211417 [published Online First: 2018/06/20]
5. Ambrosino N. The case for inspiratory muscle training in COPD. Eur Respir J 2011;37(2):233-35.
6. Beaumont M, Forget P, Couturaud F, et al. Effects of inspiratory muscle training in COPD patients: A systematic review and meta-analysis. Clin Respir J 2018 doi: 10.1111/crj.12905 [published Online First: 2018/04/18]
7. Sonetti DA, Wetter TJ, Pegelow DF, et al. Effects of respiratory muscle training versus placebo on endurance exercise performance. Respir Physiol 2001;127(2-3):185-99. [published Online First: 2001/08/16]
8. Illi SK, Held U, Frank I, et al. Effect of respiratory muscle training on exercise performance in healthy individuals: a systematic review and meta-analysis. Sports Med 2012;42(8):707-24. doi: 10.2165/11631670-000000000-00000 [published Online First: 2012/07/07]
9. Spruit MA, Singh SJ, Garvey C, et al. An official American Thoracic Society/European Respiratory Society statement: key concepts and advances in pulmonary rehabilitation. Am J Respir Crit Care Med 2013;188(8):e13-e64.
10. Camillo CA, Osadnik CR, van Remoortel H, et al. Effect of "add-on" interventions on exercise training in individuals with COPD: a systematic review. ERJ Open Res 2016;2(1) doi: 10.1183/23120541.00078-2015
11. Puhan MA, Mador MJ, Held U, et al. Interpretation of treatment changes in 6-minute walk distance in patients with COPD. Eur Respir J 2008;32(3):637-43. doi: 10.1183/09031936.00140507 [published Online First: 2008/06/14]
12. Leith DE, Bradley M. Ventilatory muscle strength and endurance training. J Appl Physiol 1976;41(4):508-16.
13. Morrissey MC, Harman EA, Johnson MJ. Resistance training modes: specificity and effectiveness. Med Sci Sports Exerc 1995;27(5):648-60. [published Online First: 1995/05/01]
14. Romer LM, McConnell AK. Specificity and reversibility of inspiratory muscle training. Med Sci Sports Exerc 2003;35(2):237-44. doi: 10.1249/01.MSS.0000048642.58419.1E [published Online First: 2003/02/06]
15. Gosselink R, De Vos J, van den Heuvel SP, et al. Impact of inspiratory muscle training in patients with COPD: what is the evidence? Eur Respir J 2011;37(2):416-25.
16. Gea J, Pascual S, Casadevall C, et al. Muscle dysfunction in chronic obstructive pulmonary disease: update on causes and biological findings. J Thorac Dis 2015;7(10):E418-38. doi: 10.3978/j.issn.2072-1439.2015.08.04 [published Online First: 2015/12/02]
17. Barreiro E, Bustamante V, Cejudo P, et al. Guidelines for the evaluation and treatment of muscle dysfunction in patients with chronic obstructive pulmonary disease. Arch Bronconeumol 2015;51(8):384-95. doi: 10.1016/j.arbres.2015.04.011 [published Online First: 2015/06/15]
18. Langer D, Ciavaglia C, Webb K, et al. Inspiratory muscle weakness in mildly- to moderately-hyperinflated patients with COPD. Eur Respir J 2014;44:Suppl 58:3344.
19. Martinez FJ, Couser JI, Celli BR. Factors influencing ventilatory muscle recruitment in patients with chronic airflow obstruction. Am Rev Respir Dis 1990;142(2):276-82. doi: 10.1164/ajrccm/142.2.276 [published Online First: 1990/08/01]
20. Nava S, Ambrosino N, Crotti P, et al. Recruitment of some respiratory muscles during three maximal inspiratory manoeuvres. Thorax 1993;48(7):702-7. [published Online First: 1993/07/01]
21. Hart N, Sylvester K, Ward S, et al. Evaluation of an inspiratory muscle trainer in healthy humans. Respir Med 2001;95(6):526-31.
22. Fitting JW, Bradley TD, Easton PA, et al. Dissociation between diaphragmatic and rib cage muscle fatigue. J Appl Physiol (1985) 1988;64(3):959-65. doi: 10.1152/jappl.1988.64.3.959 [published Online First: 1988/03/01]
23. Hershenson MB, Kikuchi Y, Tzelepis GE, et al. Preferential fatigue of the rib cage muscles during inspiratory resistive loaded ventilation. J Appl Physiol (1985) 1989;66(2):750-4. doi: 10.1152/jappl.1989.66.2.750 [published Online First: 1989/02/01]
24. Segizbaeva MO, Donina Zh A, Timofeev NN, et al. EMG analysis of human inspiratory muscle resistance to fatigue during exercise. Adv Exp Med Biol 2013;788:197-205. doi: 10.1007/978-94-007-6627-3_29 [published Online First: 2013/07/10]
25. Dempsey JA, Romer L, Rodman J, et al. Consequences of exercise-induced respiratory muscle work. Respir Physiol Neurobiol 2006;151(2-3):242-50.
26. LeBlanc P, Summers E, Inman MD, et al. Inspiratory muscles during exercise: a problem of supply and demand. J Appl Physiol 1988;64(6):2482-89.
27. Luo YM, Hopkinson NS, Polkey MI. Tough at the top: must end-expiratory lung volume make way for end-inspiratory lung volume? Eur Respir J 2012;40(2):283-85.
28. Langer D, Ciavaglia CE, Faisal A, et al. Inspiratory muscle training reduces diaphragm activation and dyspnea during exercise in COPD. J Appl Physiol (1985) 2018 doi: 10.1152/japplphysiol.01078.2017 [published Online First: 2018/03/16]
29. Dekhuijzen PN, Folgering HT, van Herwaarden CL. Target-flow inspiratory muscle training during pulmonary rehabilitation in patients with COPD. Chest 1991;99(1):128-33.
30. Ekstrom M, Elmberg V, Lindow T, et al. Breathlessness measurement should be standardised for the level of exertion. Eur Respir J 2018;51(5) doi: 10.1183/13993003.00486-2018 [published Online First: 2018/06/01]
31. Puhan MA, Schunemann HJ, Frey M, et al. Value of supplemental interventions to enhance the effectiveness of physical exercise during respiratory rehabilitation in COPD patients. A systematic review. Respir Res 2004;5:25. doi: 10.1186/1465-9921-5-25 [published Online First: 2004/12/04]
32. Laviolette L, Bourbeau J, Bernard S, et al. Assessing the impact of pulmonary rehabilitation on functional status in COPD. Thorax 2008;63(2):115-21. doi: 10.1136/thx.2006.076844 [published Online First: 2007/09/29]
33. Pepin V, Brodeur J, Lacasse Y, et al. Six-minute walking versus shuttle walking: responsiveness to bronchodilation in chronic obstructive pulmonary disease. Thorax 2007;62(4):291-8. doi: 10.1136/thx.2006.065540 [published Online First: 2006/11/14]
34. Guenette JA, Webb KA, O'Donnell DE. Does dynamic hyperinflation contribute to dyspnoea during exercise in patients with COPD? Eur Respir J 2012;40(2):322-29.
35. Langer D, Charususin N, Jacome C, et al. Efficacy of a Novel Method for Inspiratory Muscle Training in People With Chronic Obstructive Pulmonary Disease. Phys Ther 2015;95(9):1264-73. doi: ptj.20140245 [pii];10.2522/ptj.20140245 [doi]
36. Beaumont M, Mialon P, Couturaud F. Breathlessness measurement should be standardised for the level of exertion. Eur Respir J 2018;51(5) doi: 10.1183/13993003.00820-2018 [published Online First: 2018/06/01]

We appreciate the points raised by the commentator about our study (Zeng et al.)[1] published in June 2018 issue of Thorax that (1) the prevalence of abnormal residual volume to total lung capacity ratio (RV:TLC) in our study of ever-smokers with preserved spirometry is substantially higher than that observed in the commentator’s past studies,[2-4] and (2) an assumption by the commentator that stringent exclusion of those with abnormally low TLC and those with diagnosis codes of interstitial lung diseases (ILD) in their electronic health records (EHR) may have resulted in overestimation of the prevalence of abnormally high RV:TLC among smokers with preserved spirometry.

We would like to draw the attention of commentator and readers to the following points:

1- The studies referenced by the commentator used pulmonary function tests (PFT) data collected from 708 patients in 2013 across 5 clinical sites associated with University of Minnesota Medical Center with inclusion criteria of patients 18 years of age or older with or without history of smoking.[2] They included about 50% women and 3 African Americans. Our study was performed on PFT data obtained from 1985 through 2017 through the United States Veterans Affairs (VA) nationwide EHR from 7,479 patients across 37 VA medical centers in the United States with inclusion criteria of patients 40 years of age or older with an EHR diagnosis code of smoking, which likely suggests heavy smoking for VA patients. Our study included only 7.5% women but 8.8% African Americans. The populations in our study and that of the commentator’s study are very different. Given the heavier smoking burden at the VA,[5] it is not unexpected to identify higher smoking-related pathology including higher prevalence of air trapping among the VA patients. Of note, through their deployments, veterans are also exposed to high levels of air pollutants,[6] which have been shown to contribute considerably to development of chronic obstructive pulmonary disease (COPD).[7]
2- As mentioned in our article, a limitation of our study was that spirometries, and specifically full PFT, were likely done due to clinical concerns for respiratory diseases, and given the inclusion of plethysmography, the likely existence of concerns for ILD. This selection bias may then contribute to a higher prevalence of receiving an ILD diagnosis in those patients’ records in our study.
3- Abnormally low TLC could occur for reasons other than ILD including obesity, which is a common health problem and thus may have resulted in exclusions of many patients from our study.
4- It is important to note that the assumption of stringent exclusion of those with ILD and abnormally low TLC should result in the remaining included patients to have higher than expected TLC, and should not result in the remaining patients to have high RV; that is, the remaining subjects should have the same level RV but higher TLC, and hence lower RV:TLC ratio. Thus, if the commentator’s assumption that “stringent exclusion criteria resulted in a very small denominator” were correct, then the algorithm of our study indeed would have been biased towards generating lower than expected prevalence of abnormally high RV:TLC, leading to an underestimation, not overestimation, of the prevalence of abnormally high RV:TLC.
5- Most remarkably, a finding that we would like to emphasize and should not be overlooked is that regardless of whether the RV:TLC ratio was normal or abnormal, it was linearly and directly associated with respiratory morbidity and progression to spirometric COPD. No threshold level was observed.
6- It is now estimated that approximately 20% of COPD is caused by exposure to air pollutants (including pollutants other than biomass smoke such as the daily air pollution experienced by all across the globe).[8] Given above, the use of “normal” or “abnormal” values and reference equations should be approached cautiously as it is potentially extremely difficult if not impossible to develop a “normal” cohort when all earth inhabitants are being exposed to air pollutants. This is consistent with our related findings in a previous article showing that in never-smokers with history of exposure to air pollutants, all with preserved spirometry AND normal range of RV:TLC, higher RV:TLC ratio was associated with worse physiologic performance.[9]
7- Finally, the commentator’s statement about lesser reliability and lack of rationale for use of lung volumes in smokers at risk for COPD is fascinating. Spirometry is an extremely important tool in assessment of respiratory diseases including COPD. It is easier to use than plethysmography and more widely available. However, COPD is a complex disease and reducing it to one or two dimensions is unhelpful.[10] As demonstrated in our article, lung volumes, regardless of whether in the normal or abnormal range, provide an additional dimension to help prognosticate in early obstructive lung disease in those at risk for COPD. In their study of 708 patients with normal spirometry,[2] the commentator and colleagues did conclude that their “findings point out the importance of measuring lung volumes in symptomatic patients who have normal spirometry.” We fully agree with them.

References:

1. Zeng, S., et al., Lung volume indices predict morbidity in smokers with preserved spirometry. Thorax, 2018.
2. Fortis, S., E.O. Corazalla, and H.J. Kim, Does normal spirometry rule out an obstructive or restrictive ventilatory defect? Respir Investig, 2017. 55(1): p. 55-57.
3. Fortis, S., et al., The difference between slow and forced vital capacity increases with increasing body mass index: a paradoxical difference in low and normal body mass indices. Respir Care, 2015. 60(1): p. 113-8.
4. Fortis, S., et al., Persistent Empiric COPD Diagnosis and Treatment After Pulmonary Function Test Showed No Obstruction. Respir Care, 2016. 61(9): p. 1192-200.
5. Odani, S., et al., Tobacco Product Use Among Military Veterans - United States, 2010-2015. MMWR Morb Mortal Wkly Rep, 2018. 67(1): p. 7-12.
6. Falvo, M.J., et al., Airborne hazards exposure and respiratory health of Iraq and Afghanistan veterans. Epidemiol Rev, 2015. 37: p. 116-30.
7. van Koeverden, I., et al., Secondhand Tobacco Smoke and COPD Risk in Smokers: A COPDGene Study Cohort Subgroup Analysis. COPD, 2015. 12(2): p. 182-9.
8. Eisner, M.D., et al., An official American Thoracic Society public policy statement: Novel risk factors and the global burden of chronic obstructive pulmonary disease. Am J Respir Crit Care Med, 2010. 182(5): p. 693-718.
9. Arjomandi, M., et al., Lung volumes identify an at-risk group in persons with prolonged secondhand tobacco smoke exposure but without overt airflow obstruction. BMJ Open Respir Res, 2018. 5(1): p. e000284.
10. Agusti, A., B. Celli, and R. Faner, What does endotyping mean for treatment in chronic obstructive pulmonary disease? Lancet, 2017. 390(10098): p. 980-987.