Comparison between breath stacking technique associated with expiratory muscle training and breath stacking technique in amyotrophic lateral sclerosis patients: Protocol for randomized single blind trial

Introduction Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease affecting both upper and lower motor neurons, and lead to respiratory failure. Strategies are suggested to respiratory management in ALS patients, as the breath stacking and Expiratory muscle training (EMT), which have been used as aid to assist cough in neuromuscular disorders. However, the randomized controlled trials performed in ALS patients have not investigated the addiction of EMT together breath stacking in this population. This trial aims to determine if breath stacking plus EMT is more effective than breath stacking alone to decrease the decline rate on the inspiratory/expiratory muscle strength, FVC and voluntary PCF in ALS patients. Methods This parallel-group, assessor-blinded randomized controlled trial, powered for superiority, aims to assess pulmonary function, respiratory muscle strength, peak cough flow as primary outcomes. Forty-two participants are being recruited referral neuromuscular disease center at Brasilia, Brazil. Following baseline testing, participants are randomized using concealed allocation, to receive either: a) breath stacking technique alone or b) breath stacking technique plus EMT. Conclusion There is a lack of evidence regarding the benefit of EMT plus breath stacking in ALS patients. This trial will contribute to evidence currently being generated in national and international trials by implementing and evaluating a respiratory therapy program including two components not yet combined in previous research, for people with ALS involving longer-term follow-up of outcomes. This trial is ongoing and currently recruiting. Trial registration This trial was prospectively registered on the Clinical Trials Registry NCT04226144.


. Introduction
Amy otrophic lat eral scle ro sis (ALS) is a pro gres sive neu rode gen era tive dis ease af fect ing both up per and lower mo tor neu rons, which leads to a loss of vol un tary mus cle con trol [ 1 ]. As the dis ease progresses, mus cles of res pi ra tion are af fected, lead ing to chronic res pi ratory fail ure [ 2 ]. Res pi ra tory fail ure is pri mar ily due to res pi ra tory mus cle weak ness, lead ing ALS pa tients to cough im pair ment, one of the main causes of hos pi tal iza tion [ 3 , 4 ]. Pro duc ing an ef fec tive cough ini tially re quires tak ing a deep breath with the in spi ra tory mus cles, fol lowed by max i mum con trac tion of the ex pi ra tory mus cles, with glot tis clo sure and sub se quent open ing us ing the oropha ryn geal muscles, gen er at ing an ex pi ra tory flow due to con trac tion of ex pi ra tory mus cles ca pa ble of elim i nat ing se cre tions [ 5 , 6 ]. This rep re sents the in spi ra tory, com pres sion and ex pi ra tory phases of cough pro duc tion. Thus, strate gies to en hance ef fec tive cough and lung re cruit ment should be rec om mended in ALS pa tients [ 7 ].
Among these strate gies sug gested for res pi ra tory man age ment in ALS pa tients, breath stack ing has been used to as sist cough in neu romus cu lar dis or ders [ 8 ]. The breath -stacking tech nique uses a man ual re sus ci ta tion bag with one -way valve to de liver large breath vol umes to the pa tient via a suit able in ter face [ 9 ]. Pe ri odic lung ex pan sion by breath stack ing de creases basal at elec ta sis, and main tains com pli ance of the lungs and chest wall, with in creases of Peak cough flow (PCF) and cough ef fec tive ness [ 5 ]. Rafiq et al. have demon strated that breath stack ing should be rec om mended as a first line in ter ven tion for lung re cruit ment and cough aug men ta tion in ALS pa tients, with no dif fer ences be tween breath stack ing and me chan i cal in suf fla torexsufflator for sur vival, qual ity of life and hos pi tal iza tion [ 9 ].
Ex pi ra tory mus cle train ing (EMT) has been in ves ti gated as a new ef fec tive tool to en hance max i mal PCF in in di vid u als with neu ro muscu lar dis eases. This tech nique aims to im prove sub glot tic air pres sure gen er a tion, swal low ing and air way clear ance [ 10 ]. Reyes et al. have shown that EMT plus breath stack ing im proved vol un tary and re flex cough in Parkin son dis ease [ 11 ]. Plow man et al. have demon strated that 5 -week EMT train ing was fea si ble and led to im prove ments in res pi ra tory and bul bar func tion in ALS pa tients [ 10 , 12 ]. How ever, both stud ies have not used the tech niques to gether, which may improve the clin i cal re sults. Of note, the ran dom ized con trolled tri als per formed in ALS pa tients have not in ves ti gated the ad dic tion of EMT to gether breath stack ing in this pop u la tion.
This trial aims to de ter mine if breath stack ing plus EMT is more effec tive than breath stack ing alone to de crease the de cline rate on the in spi ra tory/ ex pi ra tory mus cle strength, pul monary func tion test ing and vol un tary PCF in ALS pa tients. We hy poth e size that those re ceiving breath stack ing plus EMT will have a slower de cline in these pa ra -me ters from base line to 6 months, com pared to pa tients that re ceiv ing breath stack ing alone. The key sec ondary out comes are to as sess whether EMT plus breath stack ing is su pe rior to breath stack ing alone to swal low ing func tion and abil ity to speech. This pro to col is re ported ac cord ing to the Stan dard Pro to col Items: Rec om men da tions for in terven tional Tri als (SPIRIT guid ance) [ 13 ].

1 . Methods/ design
This two -arm, par al lel (1:1), su pe ri or ity, blinded -assessor, ran domized con trolled trial is be ing con ducted at re fer ral neu ro mus cu lar disease cen ter in Brasilia, Brazil. This trial has been prospec tively reg istered on Clin i cal Tri als Reg istry (NC T04226144). Re cruit ment commenced in Au gust 2019.

2 . Participants
Fig. 1 out lines par tic i pants' flow through the study. El i gi ble par tici pants are iden ti fied through screen ing at the neu ro mus cu lar dis ease cen ter. To be el i gi ble, par tic i pants must have a di ag no sis of neu romus cu lar dis ease con firmed by neu rol o gists at the re fer ral cen ter for neu ro mus cu lar dis eases in Brasília prior to screen ing for re cruit ment. Other el i gi bil ity re quire ments are: age over 18 years; pre served cog nition, as ev i denced by a score greater than or equal to 24 points in the Mini -Mental Sta tus Exam; no bar ium al ler gies; with out tra cheostomy or in va sive me chan i cal ven ti la tion; no di aphrag matic pace maker; and no as so ci ated res pi ra tory dis ease. Par tic i pants are ex cluded if they are preg nant; had pre vi ous kid ney dis ease or other con comi tant dis eases; had res pi ra tory dis eases; or are hos pi tal ized in in ten sive care units (ICUs) dur ing the study.
El i gi ble par tic i pants are con tacted by the trial co or di na tor who explains the study aims and re quire ments, and those ex press ing in ter est in the study are pro vided with a Pa tient In for ma tion and Con sent Form. All par tic i pants pro vide writ ten in formed con sent prior to complet ing base line out come mea sures. Re cruited par tic i pants may choose to with draw from the study at any stage. Data col lected prior to the time of study with drawal will be in cluded in data analy ses.

3 . Randomization and allocation
Fol low ing in formed con sent and all base line as sess ments, par tic ipants will be ran dom ized 1:1 to ei ther the in ter ven tion group (EMT plus breath stack ing) or the con trol group (breath stack ing aloneusual care). A strat i fied block per mu ta tion ran dom iza tion will be used with bul bar and non -bulbar on set symp toms as the strat i fi ca tion factors, to en sure bal ance be tween the groups. The pa tients will be dichotomized into nor mal to mod er ate bul bar im pair ment (score 7 -12) and se vere bul bar im pair ment (score 0 -6).
The ran dom iza tion sched ule was pre pared by a re searcher in depen dent of the trial.
The trial is be ing con ducted in ac cor dance with theDe c la ra tion of Helsinki and was ap proved by the Fun dação de En sino e Pesquisa em Ciên cias da Saúde (FEPECS) Ethics Com mit tee and re ceived ap proval 6/ 18/ 2019 (Doc u ment num ber 3400412).

4. 1 . Control group
This group will just per form the breath stack ing tech nique. For this, the pa tients will have a choice be tween us ing an in flated oronasal mask (Lu miar Health Care, Sao Paulo, Brazil) or a mouthpiece (Respiron ics, Utah, USA). The care givers of the pa tients will be trained by the re search -coordinators (VM, AD and SM) to per form the tech nique at least two ses sions per day, one in the morn ing and one at bed time. The care givers will be mon i tored con tin uedly, with monthly vis its to the re fer ral cen ter to en sure that the tech nique is per formed well. All re search co or di na tors have ex pe ri ence with the tech niques and worked with res pi ra tory man age ment in ALS pa tients for at least 3 years.
To per formed the breath stack ing tech nique, we uti lize the pro tocol pre vi ously de scribed by Bach et al. [ 5 ]. The lungs are in flated as fully as pos si ble by stack ing suc ces sive breaths with out ex pi ra tion until the pa tients' lung in suf fla tion ca pac ity (LIC). The par tic i pant will be in structed to sus tain the air in the lung, clos ing the glot tis. Once the lungs are max i mally in flated, the com pressed air vol ume is released un der ex pi ra tory mus cle force, thus gen er at ing a cough with lung and chest wall re coil. They will per form 5 -8 cy cles of breath stack ing per ses sion, stack ing 3 -5 breaths per cy cle.

4. 2 . Experimental group
This group will per form the breath stack ing tech nique in ad dic tion with EMT. The valve will be changed to the one -way valve called VUP (Lu miar, Sao Paulo, Brazil), a one -way valve ( Fig. 2 ) that al lows patients blow out air with a counter re sis tance dur ing all ex pi ra tory phase.
. The ini tial ex pi ra tory pres sure will be 8 cmH 2 O and can be changed at each visit de pend ing on par tic i pants' tol er ance, whether it is easy or dif fi cult to ex hale as sessed by re search co or di na tors. The par tic i pants will be en cour aged to blow out as slowly that they can. Both pro to cols will be per formed for 24 weeks, and out come mea surements will be per formed every 8 weeks.

4. 3 . Safety and adverse events reporting
A se ri ous ad verse event is de fined in this study as any event oc curring ei ther dur ing or up to 60 min fol low ing the trial in ter ven tion or out come as sess ment that is life threat en ing or re sults in death, hos pital iza tion or pro lon ga tion of ex ist ing hos pi tal iza tion, dis abil ity or inca pac ity. Mi nor ad verse events di rectly re lat ing to in ter ven tion or outcome mea sure ses sions can in clude: aeropha gia, se vere breath less ness, new or pro gres sive pain, al tered men tal sta tus, pal pi ta tions and progres sive fa tigue [ 4 ]. Fol low ing each in ter ven tion and out come measure ses sion trial staff are re quired to com plete data en try forms in dicat ing if a se ri ous or mi nor ad verse event has oc curred. In the case of se ri ous ad verse events the study chief in ves ti ga tor is no ti fied im me diately, par tic i pants are man aged ap pro pri ately, and the in ci dent will be re ported to the rel e vant hos pi tal ethics com mit tee.

4. 4 . Outcomes
Out comes will be as sessed dur ing in di vid ual ap point ments at baseline, 8 weeks, 16 weeks and 24 weeks by a re search as sis tant blinded to group al lo ca tion. At base line, de mo graphic and clin i cal de tails will be recorded, in clud ing age, sex, lo ca tion of on set of symp toms, time of di ag no sis, treat ment de tails and so cial his tory. Sur vival data will be col lected af ter 2 years post study re cruit ment.
Treat ment ef fi cacy will be de ter mined by changes in de cline rate of the pri mary out comes (Max i mal In spi ra tory Pres sure [

. Pul monary func tion and res pi ra tory mus cle strength.
The PFTs will be mea sured us ing a Vi ta lo graph 6800 PFT spirom e ter (Vi ta lograph Inc, USA) and the de ci sion whether to use the mouth piece with nasal clip or the mask will be based on the eval u a tion of each sin gle pa tient, specif i cally on the pres ence of fa cial mus cle weak ness. The PFTs that will be in cluded are the FVC and SVC, which will be performed in both and supine po si tions; they are ex pressed both as absolute value, in FVC% and SVC%, re spec tively) [ 14 ]. The pa tient is guided to in hale deeply and, sub se quently, to ex pire as fast as far they could.
Max i mal in spi ra tory (MIP), and max i mal ex pi ra tory pres sure (MEP) were mea sured as pre vi ously de scribed by Black & Hy att [ 15 ]. Sniff nasal in spi ra tory pres sure (SNIP) will be mea sured at func tional resid ual ca pac ity as pre vi ously de scribed [ 16 ]. The high est value from three or more at tempts will be se lected for analy ses and stan dard ized to pre dicted per cent ages of Brazil ian pop u la tion [ 17 ]. 1.4.4.2 . Peak cough flow (PCF) and as sisted PCF. The ma neu ver, expressed in L/ min, will be per formed with a mask of fit ting size, with a seated pa tient who is asked to in hale as much air as pos si ble and then cough it out. At least three ac cept able and re peat able tri als were performed, and the best out of these ma neu vers will be cho sen.
PCF and man u ally as sisted PCF by LIC will be mea sured us ing a sealed oronasal mask with a hand held portable peak flow me ter (Respiron ics INC. USA), which re cent guide lines rec om mend for measur ing ex pi ra tory flow rates dur ing vol un tary cough [ 18 ].

. Bul bar func tion.
Bul bar func tion will be as sessed us ing the three com po nents of the re vised ALS func tional rat ing scale (ALS FRS -R) rel e vant to bul bar func tion (i.e. speech, sali va tion and swal low ing, each scored on a scale of 0 -4).

. Sec ondary out comes.
Data on blood gases will be col lected to as sess PaCO 2 lev els at base line. The Amy otrophic Lat eral Scle ro sis Func tional Rat ing Scale re vised score (ALSFR -r), site of on set [bul bar (B) ver sus non -bulbar (NB)], time elapsed (months) be tween symp tom on set and ALS di ag no sis, months elapsed from di ag no sis to non in vasive ven ti la tion (NIV) pre scrip tion will also be col lected. Voice pa rame ters will be as sessed by Max i mum Phona tion Time and its acoustic analy sis.
The Func tional Oral In take Scale (FOIS) [ 19 ] will be used as an index of daily oral in take. This val i dated 7 -point or di nal scale mea sures what foods an in di vid ual con sumes to meet their daily nu tri tional and hy dra tion re quire ments and ranges from a 1 (noth ing by mouth) to 7 (full oral diet with no re stric tions). The Eat ing As sess ment Tool -10 (EAT -10) was ad min is tered as a pa tient re ported out come (PRO) of swal low ing func tion [ 20 ]. The EAT -10 is a val i dated 10 -item pa tient rated ques tion naire with each ques tion rated on a 5 -point or di nal scale. A to tal EAT -10 score ranges from 0 (in dica tive of no selfperceived swal low ing im pair ments) to 40 (in dica tive of se vere swallow ing im pair ments).

5 . Sample size calculation
This study is pow ered to de tect a clin i cal mean ing ful dif fer ence in de cline rate of pul monary func tion (% pre dicted FVC). Based on metaanalysis con ducted by Silva et al. [ 21 ], the min i mal clin i cal im por tant dif fer ence (MCID) from base line to 1 year is 2 -6%, 16 par tic i pants for each arm will need to be re cruited with 80% power at two -tailed 5% level of sig nif i cance. Af ter take into ac count 30% at tri tion, a to tal sam ple size of 42 par tic i pants is re quired.

6 . Statistical analysis
Base line char ac ter is tics will be sum ma rized, in clud ing age, sex, time of di ag no sis, on set of symp toms lo cal, drugs, pri mary and secondary out comes, and im bal ances will be in ves ti gated.
For each vari able, the Shapiro Wilk test will be used to eval u ate the nor mal ity of the dis tri b u tion and the Lev ene test will be performed to eval u ate the ho mo gene ity of vari ance. The dif fer ences between groups will be an a lyzed by the un paired t -test and for con tin uous vari ables and chi -square test for cat e gor i cal vari ables. The de cline rate will be cal cu lated as the dif fer ence be tween base line and fi nal obser va tion, di vid ing the re sult by the num ber of months be tween baseline and last eval u a tion. Spear man test will be used to eval u ate the lon gi tu di nal de cline cor re la tions be tween pul monary func tion tests and both bul bar im pair ment and dis ease pro gres sion rate.
Ex ploratory analy sis of 2 -year sur vival will in clude de scrip tive Kaplan -Mayer sur vival curves and cox re gres sion with treat ment in tent.

. Discussion
This study will as sess the im pact of ex pi ra tory mus cle train ing asso ci ated with breath stack ing tech nique on pul monary func tion, res pira tory mus cle strength, peak cough flow and func tional as sess ment in ALS pa tients. We chose the pul monary func tion pa ra me ters and cough ef fi cacy as pri mary out comes be cause they plays a crit i cal role in predict ing the prog no sis of these pa tients, both in terms of sur vival and func tional abil ity [ 22 ].
A re cent study sug gests that ex pi ra tory mus cle train ing was fea sible and well -tolerated in ALS pa tients, and led to im prove ments of expi ra tory mus cle strength and swal low ing kine mat ics [ 10 ]. How ever, no dif fer ences were noted for FVC. The breath stack ing tech nique can in crease in spired vol umes that lead to greater peak cough flow, al lowing for im prove ments in mu cus clear ance and re duc tion in at elec ta sis (6). We be lieve that the in crease of in spired vol ume may in crease the ef fi cacy of ex pi ra tory mus cle train ing in the ALS pop u la tion.
The Amer i can Acad emy of Neu rol ogy, Prac tice Pa ra me ters has recom mended that breath stack ing tech nique as a res pi ra tory aid in patients with neu ro mus cu lar dis eases. This trial aims to de ter mine the im pact of an in spi ra tory and ex pi ra tory syn er gist pro gram in ALS patients us ing the same re sus ci ta tion bag with an ad dic tion of a new valve (VUP) that al lows for in creased lung ca pac ity and the ad dic tion of a sub max i mal load to ex pi ra tory mus cles dur ing the ex pi ra tion phase, which may pro vide mus cu lar con di tion ing in ALS pa tients and con tribute to a main tained bul bar func tion.

. Conclusion
ALS pa tients have a poor sur vival and demon strate high bur den of dis ease when their bul bar and res pi ra tory func tion de cline. This random ized trial will as sess the ef fects of home -based pro gram in pulmonary and bul bar func tion. If ben e fi cial, the in ter ven tion is de signed in way to en able easy trans la tion into prac ti cal guide lines for this pop u la tion.

Author statement
Alessan dra Dorça: Con cep tu al iza tion, Method ol ogy, Livia A. Al cântara: In ves ti ga tion, Re sources, Denise Sis terolli Di niz: Writ ing -orig inal draft, Max Sarmet: In ves ti ga tion, Method ol ogy,Sér gio Ri cardo Menezes Ma teus: Writ ing -orig i nal draft, Luis Vi cente Franco Oliveira: Writ ing -orig i nal draft, Hamil ton Franco: In ves ti ga tion, Resources, Vini cius Mal daner: Con cep tu al iza tion, Method ol ogy, Writ ing -orig i nal draft

Funding
No fund ing or spon sor ship was re ceived for this study or pub li cation of this ar ti cle.

Declaration of competing interest
All au thors de clare that have no con flict of in ter est.