Abstract
Long-term outcome of patients with BOS after HSCT with home spirometry is encouraging despite suboptimal adherence http://ow.ly/mKfM30jDknz
To the Editor:
Bronchiolitis obliterans syndrome (BOS) is a well-characterised late-onset noninfectious pulmonary complication of allogeneic haematopoietic stem cell transplantation (HSCT), occurring in 2% to 26% of recipients [1]. It is considered to be a pulmonary manifestation of graft versus host disease (GVHD) [2] and is associated with a mortality rate varying from 14% to 100% in historical series [3, 4]. Although immunosuppressive drugs are modestly effective, early diagnosis and treatment could improve its outcome [5]. After lung transplantation, home spirometry monitoring of pulmonary function allows early detection of BOS [6]; it is associated with a better response to steroids [7] and consequently is considered as a standard of care [8]. Our group was the first to report the use of home spirometry in 37 HSCT recipients [9]. More recently, Cheng et al. [10] showed a good correlation between home spirometry and classical laboratory spirometry in this population. We present here the incidence and long-term outcome of BOS occurring after HSCT in an extended cohort of 110 patients monitored with home spirometry.
This single-centre study was conducted in 110 patients who received an allogeneic HSCT for haematological malignancy from June 2001 to November 2008. Inclusion criteria were: 1) living in the Paris area, 2) having a landline telephone and 3) consenting to the study. BOS was defined as a persistent non-reversible obstructive ventilatory impairment, with a drop of forced expiratory volume in 1 s (FEV1) above 20% compared to baseline value before the graft, after exclusion of other causes, using the International Society of Heart and Lung Transplantation criteria [11]. No histological confirmation was required due to the high risk of the procedure in this population.
Before HSCT, complete pulmonary functional tests (PFT) were performed and patients were trained to perform flow–volume curves twice a week with home spirometry (individual portable spirometer, Spirotel, M-Elect, France, and its modem). home spirometry monitoring was scheduled to start three months after transplantation for 18 months. home spirometry results were centrally transmitted in real time using a landline telephone for analysis and monitoring. For each patient, correlation between data from home spirometry and from classical laboratory spirometry was checked before transplant. When home spirometry deterioration was detected, with a FEV1 drop superior to 20%, patients were referred to pulmonologists for BOS diagnosis by PFT confirmation and performance of a high-resolution thoracic computed-tomography scan with expiratory phases and an infectious workup including a fiberoptic bronchoscopy with bronchoalveolar lavage to rule out a pulmonary infection. The trial was conducted in accordance with the declaration of Helsinki.
Main characteristics of the patients are described in table 1. The median time from transplant to home spirometry monitoring onset was 4.4 months (IQR 3–7.1) and the monitoring was continued for a median time of 16.2 months (IQR 7.2–21.2). According to the criteria published by Kugler et al. [12], good adherence (>80% scheduled home spirometry effectively performed) was observed in 35.5% patients, moderate adherence (50–80% home spirometry performed) in 31.8% and poor adherence (<50% home spirometry performed) in 32.7%. During the follow-up time, 49 drops ≥20% in FEV1 in home spirometry were reported in 34 patients, and 34/49 (69%) in 26 patients were confirmed by standard spirometry. 17 drops were related to BOS diagnosis (n=13) or worsening (n=4), but home spirometry anomalies also allowed the detection of other pulmonary complications: infections (n=6), pulmonary embolism (n=1), cardiac failure (n=3) or others (n=7).
26 BOS were diagnosed at a median time of 14 months (IQR 10.5–21.1) post-HSCT, leading to a 5-year cumulative incidence of 27.7% (95% CI 19.9–38.8%). In 13 of the cases, BOS detection was made by home spirometry. Reasons for lack of BOS diagnosis by home spirometry were non-adherence (n=6), BOS occurring before (n=2) or after (n=4) monitoring or non-specified (n=1). Finally, home spirometry detected 13 (65%) out of 20 BOS occurring during the effective period of equipment: nine (82%) out of 11 in good/moderate adherers versus four (44%) out of nine in poor adherers. Thus, home spirometry was a useful tool for BOS detection in good adherers.
At BOS diagnosis, three patients were asymptomatic. The median initial drop in FEV1, compared with pre-transplant data, was 39.9% (IQR 27.3–45.8). All patients with BOS except one presented extra pulmonary manifestations of chronic GVHD. The 5-year cumulative incidence of BOS was 36.9% (95% CI 25.2–48.5%) in patients with chronic GVHD, compared to 2.4% (95% CI 0–7%) in those who did not (p=0.006 Fine and Gray test). This confirmed that patients with extrapulmonary chronic GVHD have the highest risk of developing BOS, and could represent a group of patients that could benefit from home spirometry.
BOS features were followed for a median of 9.4 years (IQR 7.4–10.7). The median maximal drop in FEV1, compared with pre-transplant data, was 45.8% (IQR 36.4–58.5). Treatment of BOS consisted in systemic steroids less than 1 mg per kg per day in 65.4% patients, associated with other immunosuppressive therapies depending on those previously administrated and on GVHD extrapulmonary organ involvement and specific pulmonary treatment (inhaled corticosteroids and bronchodilatators, and azithromycin) as well as infectious prophylaxis. At last evaluation, worsening of FEV1 (decrease >15% compared with diagnosis) was observed in seven (26.9%) patients whereas improvement (increase >15%) occurred in eight (30.8%) and stabilisation (variation ≤15%) in 11 (42.3%). From the 26 patients with BOS, two died of evolution of BOS associated with respiratory infection leading to an 8-year mortality due to BOS of 8.1% (95% CI 0–19.1%).
Aiming to indirectly evaluate the home spirometry efficacy, we compared BOS outcome in good/moderate or poor adherers and failed to detect any difference between these patients (table 1). Although this could argue against the home spirometry utility after HSCT, no definitive conclusions can be drawn and only a randomised trial would definitely conclude about the home spirometry usefulness.
We report here with a long follow-up time the features of BOS in a population of HSCT recipients monitored by home spirometry, remarkable by the absence of pejorative effect of BOS occurrence. Indeed, only two of the 26 BOS patients died from respiratory failure, while 73% patients showed a favourable evolution (FEV1 stable or improved) which has been correlated to a better outcome in post-HSCT BOS [13]. It is noteworthy that BOS incidence, 27.7%, was similar to other studies [14] and FEV1 drop at diagnosis was close to that reported in the NIH series [15] making unlikely the good outcome of BOS patients being related to a lower gravity of BOS.
We showed in this large cohort that home spirometry was feasible in HSCT patients. However, adherence rate was 35.5%, which is lower than the 68% observed after pulmonary transplantation [12]. This low adherence rate could be a limiting factor for home spirometry use in HSCT patients. In further studies, specific interventions are warranted to improve adherence of these patients less aware of the severity of pulmonary complications than lung transplant patients. These include 1) a better education of patients with dedicated consultations, requiring a close collaboration between haematologists, pulmonologists and patients, 2) development of systems with internet based transmission of home spirometry data and 3) providing reminders to patients through an automated reminding system.
Our study shows that home spirometry allows BOS detection in most adherer patients and raises the issue of a potential favourable impact of home spirometry on post-HSCT BOS outcome. A randomised trial, comparing home spirometry versus absence of home spirometry monitoring in HSCT patients, with specific interventions to improve adherence, would definitively address the interest of this non-invasive procedure on BOS outcome.
Acknowledgement
We thank LVL Medical Company for technical assistance with home spirometry monitoring devices and Polly Gobin for English language revision.
Footnotes
Conflict of interest: E. Catherinot has received funding for meeting attendance from LVL Medical, CSL Behring and LFB, outside the submitted work.
Conflict of interest: H. Salvator has received funding for meeting attendance from GSK and Oxyvie, outside the submitted work.
- Received April 27, 2017.
- Accepted April 15, 2018.
- Copyright ©ERS 2018