Article

Reduced mortality in cystic fibrosis patients treated with tobramycin inhalation solution

Division of Respiratory Diseases, Children's Hospital Boston, Harvard Medical School, Boston, Massachusetts 02115, USA. .
Pediatric Pulmonology (Impact Factor: 2.7). 01/2012; 47(1):44-52. DOI: 10.1002/ppul.21521
Source: PubMed
ABSTRACT
Though tobramycin inhalation solution has been used for over a decade to improve lung function and reduce exacerbations in patients with cystic fibrosis (CF), its effects on mortality have not been well-described. This study aimed to assess the association between use of tobramycin inhaled solution and mortality in patients with CF and chronic Pseudomonas aeruginosa (PA) infection.
Longitudinal logistic regression was used to assess the association between current-year reported use of tobramycin inhalation solution and subsequent-year mortality of patients meeting recommended criteria for tobramycin inhalation solution use in the United States Cystic Fibrosis Foundation's Patient Registry (1996-2008).
Among 12,740 patients meeting inclusion criteria, 2,538 deaths were observed during a median follow-up of 6 years. After regression adjustment, use of tobramycin inhaled solution was associated with a 21% reduction in the odds of subsequent year mortality (odds ratio (95% CI): 0.79 (0.72-0.88), P < 0.001). In our model, use of dornase alfa was also associated with a 15% reduction in the odds of subsequent year mortality (odds ratio (95% CI): 0.85 (0.76-0.95), P = 0.005). Underweight for age, CF-related diabetes, female gender, worse lung function and cultures positive for Pseudomonas aeruginosa or Burkholderia cepacia complex, among multiple other patient characteristics, were associated with significantly increased mortality. Adjusted mortality rates for patients reporting tobramycin inhalation solution use in all versus none of the follow-up years were 1.3% versus 2.1% at 2 years, 5.2% versus 8.0% at 5 years, and 9.9% versus 15.0% at 10 years.
After adjustment for multiple patient characteristics and known risk factors, use of tobramycin inhalation solution was associated with significantly reduced mortality among patients with CF.

Full-text

Available from: Gregory S Sawicki, Jan 21, 2014
ISSN 8755-6863
PED I AT RIC
PULMONOLOGY
ONLINE SUBMISSION AND PEER REVIEW
http://mc.manuscriptcentral.com/ppul
Reduced Mortality in Cystic Fibrosis Patients
Treated With Tobramycin Inhalation Solution
Gregory S. Sawicki, James E. Signorovitch, Jie Zhang, Dominick Latremouille-Viau,
Markus von Wartburg, Eric Q. Wu, and Lizheng Shi
DIGITAL COPY
Reproduced from
Pediatric Pulmonology 2012; 47(1): 44-52
This PDF is supplied to Novartis (Spain) on 14/05/2012:
•  A maximum of 1 copy is allowed per physician (100 copies in total).
•  A maximum total of 100 downloads are permitted.
•  This PDF is not to be forwarded or further disseminated elsewhere,
without prior written permission from Wiley-Blackwell.
•  This article is protected by copyright. All rights reserved.
Page 1
Pediatric Pulmonology 47:44–52 (2012)
Reduced Mortality in Cystic Fibrosis Patients Treated
With Tobramycin Inhalation Solution
Gregory S. Sawicki, MD, MPH,
1
* James E. Signorovitch, PhD,
2
Jie Zhang, PhD,
3
Dominick Latremouille-Viau, MS,
2
Markus von Wartburg, PhD,
2
Eric Q. Wu, PhD,
2
and Lizheng Shi, PhD
4
Summary. Background: Though tobramycin inhalation solution has been used for over a de-
cade to improve lung function and reduce exacerbations in patients with cystic fibrosis (CF), its
effects on mortality have not been well-described. This study aimed to assess the association
between use of tobramycin inhaled solution and mortality in patients with CF and chronic Pseu-
domonas aeruginosa (PA) infection. Methods: Longitudinal logistic regression was used to as-
sess the association between current-year reported use of tobramycin inhalation solution and
subsequent-year mortality of patients meeting recommended criteria for tobramycin inhalation
solution use in the United States Cystic Fibrosis Foundation’s Patient Registry (1996–2008).
Results: Among 12,740 patients meeting inclusion criteria, 2,538 deaths were observed during
a median follow-up of 6 years. After regression adjustment, use of tobramycin inhaled solution
was associated with a 21% reduction in the odds of subsequent year mortality (odds ratio
(95% CI): 0.79 (0.72–0.88), P < 0.001). In our model, use of dornase alfa was also associated
with a 15% reduction in the odds of subsequent year mortality (odds ratio (95% CI): 0.85
(0.76–0.95), P ¼ 0.005). Underweight for age, CF-related diabetes, female gender, worse lung
function and cultures positive for Pseudomonas aeruginosa or Burkholderia cepacia complex,
among multiple other patient characteristics, were associated with significantly increased mor-
tality. Adjusted mortality rates for patients reporting tobramycin inhalation solution use in all
versus none of the follow-up years were 1.3% versus 2.1% at 2 years, 5.2% versus 8.0%
at 5 years, and 9.9% versus 15.0% at 10 years. Conclusion: After adjustment for multiple
patient characteristics and known risk factors, use of tobramycin inhalation solution was asso-
ciated with significantly reduced mortality among patients with CF. Pediatr Pulmonol. 2012;
47:44–52.
ß 2011 Wiley Periodicals, Inc.
Key words: Pseudomonas aeruginosa; anti-bacterial agents; survival; registries;
medication therapy management.
Funding source: Novartis Pharmaceuticals Corporation, East Hanover, NJ.
INTRODUCTION
Predicted survival among patients with cystic fibrosis
(CF) has improved substantially during the past decade
due in part to the introduction of new chronic
therapies.
1
However, lung function decrease associated
with chronic lung infection and obstructive lung disease
remains the primary cause of death,
1
and Pseudomonas
aeruginosa (PA) remains the most common airway
pathogen contributing to shortened survival in CF.
2–4
Additional supporting information may be found in the online version of
this article.
1
Division of Respiratory Diseases, Children’s Hospital Boston, Harvard
Medical School, Boston, Massachusetts.
2
Analysis Group, Inc., Boston, Massachusetts.
3
Evidence Based Medicine, Novartis Pharmaceuticals Corporation, East
Hanover, New Jersey.
4
Health Systems Management, Tulane University, New Orleans, Louisiana.
Publication of an abstract: Elements of this study were presented at the
24th North American Cystic Fibrosis Conference, October 21–23, 2010,
Baltimore Convention Center, Baltimore, MD.
*Correspondence to: Gregory S. Sawicki, MD, MPH, Division of
Respiratory Diseases, Children’s Hospital Boston, 300 Longwood Ave,
Boston, MA 02115. E-mail: gregory.sawicki@childrens.harvard.edu
Received 14 February 2011; Accepted 25 May 2011.
DOI 10.1002/ppul.21521
Published online 3 August 2011 in Wiley Online Library
(wileyonlinelibrary.com).
ß 2011 Wiley Periodicals, Inc.
Page 2
Evidence-based guidelines for chronic CF pulmonary
therapy recommend agents targeting airway infection,
inflammation, and mucociliary clearance to improve
lung function and reduce risk of exacerbations.
5
The
impact of these recommended therapies on mortality in
CF remains unclear.
Tobramycin inhalation solution (TOBI
1
; Novartis
Pharmaceuticals Corporation, East Hanover, NJ [TIS]),
approved by the US Food and Drug Administration
(FDA) in 1997, has been used for over a decade in
patients with CF and infection with PA. In CF clinical
trials, TIS has resulted in improved lung function and
reduced risk of airway exacerbations.
6–8
Although bet-
ter lung function and lower rates of exacerbation have
been associated with longer survival in epidemiologic
studies,
2,9–11
the direct effect of TIS on survival has not
been thoroughly assessed. A secondary analysis of a
clinical trial, presented as a conference abstract,
12
found
that patients randomized to TIS had about half the mor-
tality of those randomized to placebo, but the original
trial was not powered to evaluate mortality.
6
A previous
retrospective study of the Cystic Fibrosis Foundation
Patient Registry (CFFPR) described an association
between TIS use and higher odds of mortality in the
first 2 years following its FDA approval, but this result
was presented by the authors as an example of strong
confounding by indication, rather than as a realistic
finding.
13
The CFFPR now contains 11 years of patient data
following TIS approval and a wealth of clinical meas-
ures, making it a valuable resource for comparative
effectiveness research in CF. This study utilized CFFPR
data up to 2009 to assess the association between
TIS use in accordance with recommended criteria and
mortality in patients with CF.
METHODS
Sample Selection and Study Measures
The CFFPR contains annual and encounter-based
questionnaires from over 36,000 people receiving care
at CFF-accredited centers across the US. Data from
1996 to 2008 were utilized in this analysis. Inclusion in
the study cohort required confirmed CF diagnosis and
at least 1 year of follow-up meeting recommended cri-
teria for chronic TIS use,
5
detected as (1) age 6 years,
(2) a history of four or more airway cultures positive
for PA since entry into the CFFPR and (3) forced expi-
ratory volume in 1 sec (FEV
1
) %-predicted between
25% and 75% within the last 2 years. A patient’s first
year meeting these criteria served as the index year, and
was identified using all airway cultures and FEV
1
assessments recorded on encounter-based questionnaires
(Fig. 1). Patients with index years occurring in their
first year of registry data or prior to TIS approval were
excluded. Included patients were retrospectively followed
from their index year to the end of continuous annual
follow-up or death. Reported TIS use during a study
year (available only as a yes/no variable assessed annu-
ally in the CFFPR) was the primary exposure variable.
Death from any cause was the primary outcome variable.
Statistics
Patient characteristics (see Table 1) were summarized
in the index year, and compared between follow-up
ABBREVIATIONS:
BCC Burkholderia cepacia complex
CF Cystic fibrosis
CFFPR Cystic Fibrosis Foundation Patient Registry
FDA US Food and Drug Administration
FEV
1
Forced expiratory volume in 1 sec
GEEs Generalized estimating equations
MRSA Methicillin-resistant Staphylococcus aureus
N Number
OR Odds ratio
PP-value
PA Pseudomonas aeruginosa
TIS Tobramycin inhalation solution
Fig. 1. Sample selection. CF, cystic fibrosis; CFF, cystic fibro-
sis foundation; PA, Pseudomonas aeruginosa; TIS, tobramycin
inhalation solution; FEV
1
, forced expiratory volume in 1 sec.
Reduced Mortality With Inhaled Tobramycin 45
Pediatric Pulmonology
Page 3
years with versus without reported use of TIS, using
generalized estimating equations (GEEs) to account for
correlation among multiple follow-up years from the
same patient.
14
The association between TIS use in any study year
and mortality in the subsequent year was assessed using
longitudinal logistic regression, with adjustment for
pre-selected patient characteristics at each year of fol-
low-up.
15
Each patient’s reported TIS use was assessed
annually (yes vs. no), and could vary from year to year,
but could not reflect treatment intensity within a year.
Analyses included the index year and all following
years, and used GEEs to account for within-patient cor-
relation. Absolute rates of mortality in the subsequent
year, and mortality risk differences, were estimated for
a hypothetical average patient with and without
reported TIS use in the current year based on the fitted
longitudinal logistic regression model. Sensitivity anal-
yses investigated the impacts of (1) requiring three or
ve airway cultures positive for PA during sample se-
lection, (2) excluding patients awaiting or receiving
lung transplant, or with Burkholderia cepacia complex
(BCC) infection, as of the index year, or (3) including
random effects of CFF care center and TIS-by-care cen-
ter in a logistic mixed effects model.
Patients reporting TIS use in all versus none of the
follow-up years were also compared in terms of base-
line characteristics, and in terms of survival using lon-
gitudinal Cox proportional hazards models
16
adjusted
for the same patient characteristics as the longitudinal
logistic models. In this analysis, patient follow-up was
censored upon switching of TIS status: patients
TABLE 1— Patient Characteristics
Patient characteristics
1
Index year
All follow-up years
With TIS Without TIS
Number of annual assessments 12,740 49,642 24,863
Number of patients represented 12,740 11,140 8,452
Demographics
Age, mean SD 21.64 10.13 23.29 9.54 26.73 11.21
5
Female 6,267 (49.2) 24,561 (49.5) 12,245 (49.2)
5
Underweight for age
2
2,013 (15.9) 8,701 (17.6) 3,915 (16.5)
Comorbidities
Asthma 927 (7.3) 5,979 (12.0) 2,762 (11.1)
CF-related diabetes 1,023 (8.0) 10,083 (20.3) 5,033 (20.2)
Liver disease 1,000 (7.8) 5,059 (10.2) 2,165 (8.7)
5
Sinus disease 829 (6.5) 5,805 (11.7) 2,828 (11.4)
Microbiology
Cumulative % of PA positive cultures, mean SD
3
78.26 27.24 82.32 23.44 79.96 24.73
5
Positive culture for Burkholderia cepacia 484 (3.8) 1,741 (3.5) 1,067 (4.3)
Positive culture for MRSA 1,414 (11.1) 8,874 (17.9) 3,111 (12.5)
5
Treatments
Dornase alfa 9,132 (71.7) 41,711 (84.0) 13,008 (52.3)
5
High-dose ibuprofen (e.g., 25–30 mg/kg) 693 (5.4) 2,382 (4.8) 730 (2.9)
5
Pancreatic enzyme supplements 11,951 (93.8) 47,186 (95.1) 21,697 (87.3)
5
Waiting for transplantation 126 (1.0) 1,517 (3.1) 622 (2.5)
5
Clinical measures
Average FEV
1
%-predicted, mean SD
4
61.84 20.44 57.39 21.20 59.81 22.38
5
History of TIS intolerance 35 (0.3) 232 (0.5) 656 (2.6)
5
Resource use (number of events)
Hospitalizations, mean SD 1.17 1.60 1.38 1.76 1.02 1.59
5
Home IV antibiotic treatments, mean SD 0.61 1.14 0.77 1.28 0.51 1.06
5
Acute exacerbations, mean SD 1.54 2.13 1.69 2.11 1.14 1.78
5
Outpatient visits, mean SD 6.09 2.62 6.13 3.29 4.88 3.17
5
TIS, tobramycin inhalation solution; SD, standard deviation; CF, cystic fibrosis; PA, Pseudomonas aeruginosa; MRSA, methicillin-resistant
Staphylococcus aureus; FEV
1
, forced expiratory volume in 1 sec; IV, intravenous.
1
All characteristics are defined in the present year unless otherwise noted. Comparisons of additional characteristics are available in the
Online Data e-Table 1.
2
Body mass index <18.5 when aged 20 years; body mass index <5th percentile when aged <20 years. World Health Organization. Global
Database on Body Mass Index 2005 Dec (cited 2010 June 14). Available from: http://apps.who.int/bmi/index.jsp?introPage¼intro_3.html.
3
Defined using all available data in the current and prior years.
4
Average of FEV
1
%-predicted across all reported assessments in the present year.
5
P < 0.05 for a test comparing years with versus without TIS using generalized estimating equations to account for repeated assessments
from individual patients.
46 Sawicki et al.
Pediatric Pulmonology
Page 4
initiating without TIS were censored at the first year of
reported TIS use; patients initiating with TIS were cen-
sored at the first year without report TIS use. Based on
the fitted Cox model, adjusted survival curves were pre-
dicted for two hypothetical patients, each with a covari-
ate history equal to the population average, but one
reporting TIS use in all years and the other reporting
TIS use in none of the years. A two-sided P-value of
0.05 or less was considered statistically significant.
Analyses were performed using SAS
1
9.2 (Cary, NC).
RESULTS
Among 36,239 patients included in the CFFPR from
1996 to 2008, 35,012 had a confirmed diagnosis of CF.
Of these, 12,740 met recommended criteria for TIS use
in 1998 or later and had pre-index annual data (Fig. 1).
Among patients in their index years, mean age was
22 years, mean FEV
1
%-predicted was 62%, 16% were
underweight for age, and 64% reported TIS use
(Table 1). During a median follow-up of 6 years, TIS
use was reported in 67% of patient-years. Compared to
patient-years without TIS use (N ¼ 24,863), those with
TIS use (N ¼ 49,642) had a significantly younger pa-
tient age, lower FEV
1
%-predicted, greater reported use
of other treatments (dornase alfa [Pulmozyme
1
; Gen-
entech, Inc., San Francisco, CA], high-dose ibuprofen),
greater proportions with positive cultures for methicil-
lin-resistant Staphylococcus aureus (MRSA), more prior
positive cultures for PA, and a greater proportion with
episodes of acute care (hospitalizations, acute exacerba-
tions, and home IV antibiotics) and outpatient visits.
Among years without TIS use, patients were more likely
to have a history of intolerance to TIS or positive culture
results for BCC, but the absolute rates of these events
were low (2.6% and 4.3%, respectively) (Table 1).
A total of 2,538 patients, or 20% of the study popula-
tion, died during follow-up. In unadjusted analyses, TIS
use in a given year was not associated with mortality in
the subsequent year (odds ratio [OR] (95% CI) ¼ 0.99
(0.90–1.08), P ¼ 0.8). In unadjusted analyses stratified
by index year, and assessing outcomes only in the first
year following the index year, TIS use was associated
with odds ratios for subsequent-year mortality of 1.58
(95% CI ¼ 1.12–2.23) among patients with index year
1998, 1.05 (95% CI ¼ 0.63–1.75) among patients with
index year 1999, and 0.73 (95%CI ¼ 0.51–1.04) among
patients with index year 2000 or later. Using all avail-
able data and after adjustment for multiple patient char-
acteristics (Table 2), TIS use was associated with a
statistically significant 21% reduction in the odds of
subsequent year mortality (OR: 0.79, 95% CI: 0.72–
0.88, P < 0.001). Based on this model, estimated abso-
lute risks of mortality in the subsequent year for the
average patient derived from entire cohort were 1.26%
(95% CI ¼ 1.17–1.40) with TIS versus 1.58% (95%
CI ¼ 1.43–1.80) without TIS (risk difference ¼0.32;
P < 0.001). The use of dornase alfa was similarly asso-
ciated with reduced subsequent year mortality (OR:
0.85, CI: 0.76–0.95, P ¼ 0.005). In the same model,
other factors associated with increased mortality
(P < 0.05), in order of decreasing OR, were cancer, re-
nal dysfunction, BCC infection, hypertension, liver dis-
ease, transplantation, underweight, lack of insurance
coverage, depression, waiting for transplantation, CF-
related diabetes, non-white race, prior hospitalizations,
lack of age-appropriate education, female gender, prior
home intravenous antibiotics, earlier calendar year, low-
er FEV
1
%-predicted, and additional cultures with PA
(Table 2). In sensitivity analyses, the association be-
tween TIS use and reduced mortality was robust to ex-
clusion of patients waiting for transplantation or with a
history of BCC infection at the index year, requiring
three or five PA cultures or inclusion of random effects
for care center and the interaction between TIS and
care center (Table 3). The effect of TIS did not exhibit
statistically significant heterogeneity across care centers
in the mixed effects model.
Within the total sample, 34% of patients (N ¼ 4,288)
reported TIS use in all study years and 13%
(N ¼ 1,600) reported TIS use in none of the study
years following their index year. Patients without
reported TIS use were more likely to be older, male,
have an older age at CF diagnosis, and had less
reported use of dornase alfa and lower rates of acute
care utilization (Table 4). During follow-up, death was
reported for 849 (19.8%) of patients reporting TIS use
in all follow-up years and 282 (17.6%) of patients
reporting TIS use in none of the follow-up years. After
multivariable adjustment in a longitudinal Cox propor-
tional hazards model, however, reporting TIS use in all
versus none of the follow-up years was associated with
a 36% reduction in the hazard of mortality (hazard ratio
(95% CI) ¼ 0.64 (0.53–0.78), P < .001). Based on this
model, using TIS in all versus none of the years follow-
ing the index year increased the predicted percentage of
patients surviving to 2 years by 0.8%, to 5 years by
2.8%, and to 10 years by 5.1% after the first year meet-
ing recommended criteria for TIS use (Fig. 2).
DISCUSSION
This retrospective analysis of the CFFPR included a
majority of CF patients in the US who met recom-
mended criteria for TIS use in the decade following its
FDA approval. Median follow-up of 6 years exceeded
that of randomized trials for inhaled antibiotics in CF.
1
SAS Institute, Inc., 100 SAS Campus Dr, Cary, NC 27513.
Reduced Mortality With Inhaled Tobramycin 47
Pediatric Pulmonology
Page 5
After adjusting for multiple patient characteristics,
including demographics, lung function, comorbidities,
microbiology, medical resource use, and other CF treat-
ments, use of TIS in a study year was associated with
significantly reduced mortality in the subsequent year.
Reduced mortality was especially pronounced among
patients reporting TIS use in all versus none of the
follow-up years.
TABLE 3— Sensitivity Analyses
Sensitivity analyses
1
Number
of patients
Effect of current year TIS use on
subsequent year mortality
Odds ratio
2
P-value
1. Requiring 3 PA positive cultures 13,124 0.81 <0.001
2. Requiring 5 PA positive cultures 12,094 0.79 <0.001
3. Excluding patients with transplantation, waiting for transplantation,
or with culture results of Burkholderia cepacia at baseline
12,174 0.81 <0.001
4. Mixed effects model, with random effects for CFF care center,
and TIS-by-care center interaction
12,740 0.78 <0.001
TIS, tobramycin inhalation solution; PA, Pseudomonas aeruginosa; CFF, Cystic Fibrosis Foundation.
1
Described as changes or additions to the core analysis.
2
Odds ratios <1 indicate an association with reduced mortality in the subsequent year; estimated from multivariable logistic regression
models with generalized estimating equations to account for repeated assessments from individual patients.
TABLE 2— Multivariable Logistic Regression Model for Mortality (N ¼ 12,740)
Patient characteristics
1
Odds ratio
2
95% Confidence interval P-value
Use of tobramycin inhalation solution 0.79 (0.72, 0.88) <0.001
Demographics
Female 1.15 (1.05, 1.26) 0.002
White race 0.79 (0.64, 0.99) 0.037
Underweight for age
3
1.42 (1.28, 1.58) <0.001
Age-appropriate education level 0.80 (0.67, 0.95) 0.013
Insurance coverage 0.59 (0.38, 0.92) 0.020
Calendar year 0.93 (0.90, 0.95) <0.001
Comorbidities
Cancer 4.69 (2.82, 7.77) <0.001
CF-related diabetes 1.32 (1.17, 1.49) <0.001
Depression 1.37 (1.21, 1.56) <0.001
Hypertension 1.68 (1.29, 2.18) <0.001
Liver disease 1.67 (1.47, 1.90) <0.001
Renal dysfunction 3.42 (2.32, 5.04) <0.001
Microbiology
Cumulative number of PA positive cultures
4
1.01 (1.00, 1.01) 0.034
Positive culture for Burkholderia cepacia 2.02 (1.70, 2.41) <0.001
Other treatments
Dornase alfa 0.85 (0.76, 0.95) 0.005
Transplantation 1.58 (1.28, 1.94) <0.001
Waiting for transplantation 1.33 (1.12, 1.58) 0.001
Clinical measures and resource use
Average FEV
1
%-predicted
5
0.95 (0.94, 0.95) <0.001
Number of hospitalizations 1.25 (1.20, 1.30) <0.001
Number of home IV treatments 1.07 (1.03, 1.12) <0.001
CF, cystic fibrosis; PA, Pseudomonas aeruginosa; FEV
1
, forced expiratory volume in 1 sec; IV, intravenous.
1
All characteristics are defined in the present year unless otherwise noted. Age, aged 18 years at first CF diagnosis, Hispanic origin, asthma,
arthritis, gastrointestinal complications, sinus disease, use of pancreatic enzyme supplements, and use of high-dose ibuprofen were also
included in the model, but did not have statistically significant associations with mortality. The full fitted model is available from the Online
Data e-Table 2.
2
Odds ratios <1 indicate an association with reduced mortality in the subsequent year.
3
Body mass index <18.5 when aged 20 years; body mass index <5th percentile when aged <20 years. World Health Organization. Global
Database on Body Mass Index 2005 Dec (cited 2010 June 14). Available from: http://apps.who.int/bmi/index.jsp?introPage¼intro_3.html.
4
Defined using all available data in the current and prior years.
5
Average of FEV
1
%-predicted across all reported assessments in the present year.
48 Sawicki et al.
Pediatric Pulmonology
Page 6
TABLE 4— Characteristics of Patients Reporting TIS Use in All Versus None of the Follow-Up Years
Patient characteristics
1
All follow-up years
Patients with TIS reported
in all follow-up years
Patients with TIS reported
in none of the follow-up years
Number of annual assessments 20,216 5,233
Number of patients represented 4,288 1,600
Demographics
Age, mean SD 21.90 8.98 28.01 11.75
5
Aged 18 years at first CF diagnosis 637 (3.2) 505 (9.7)
Female 9,946 (49.2) 2,391 (45.7)
5
Underweight for age
2
3,579 (17.7) 868 (16.6)
5
Insurance
Public insurance 9,730 (48.1) 2,193 (41.9)
5
Private insurance 12,844 (63.5) 3,284 (62.8)
No insurance coverage 137 (0.7) 173 (3.3)
5
Microbiology
Cumulative % of PA positive cultures, mean SD
3
79.96 25.44 73.68 28.14
5
Positive culture for Burkholderia cepacia 657 (3.2) 262 (5.0)
5
Positive culture for MRSA 3,794 (18.8) 600 (11.5)
5
Treatments
Dornase alfa 17,989 (89.0) 2,522 (48.2)
5
High-dose ibuprofen (e.g., 25–30 mg/kg) 983 (4.9) 143 (2.7)
5
Pancreatic enzyme supplements 19,361 (95.8) 4,528 (86.5)
5
Waiting for transplantation 558 (2.8) 111 (2.1)
Clinical measures
Average FEV
1
%-predicted, mean SD
4
58.72 21.45 61.79 22.35
5
Resource use (number of events)
Hospitalizations, mean SD 1.36 1.73 0.91 1.54
5
Home IV antibiotic treatments, mean SD 0.76 1.25 0.44 0.97
5
Acute exacerbations, mean SD 1.67 2.07 1.02 1.75
5
TIS, tobramycin inhalation solution; SD, standard deviation; CF, cystic fibrosis; PA, Pseudomonas aeruginosa; MRSA, methicillin-resistant
Staphylococcus aureus; FEV
1
, forced expiratory volume in 1 sec; IV, intravenous.
1
All characteristics are defined in the present year unless otherwise noted. Comparisons of additional characteristics are available in the
Online Data e-Table 3.
2
Body mass index <18.5 when aged 20 years; body mass index <5th percentile when aged <20 years. World Health Organization. Global
Database on Body Mass Index 2005 Dec (cited 2010 June 14). Available from: http://apps.who.int/bmi/index.jsp?introPage¼intro_3.html.
3
Defined using all available data in the current and prior years.
4
Average of FEV
1
%-predicted across all reported assessments in the present year.
5
P < 0.05 for a test comparing years with versus without TIS using generalized estimating equations to account for repeated assessments
from individual patients.
Fig. 2. Survival curves for patients reporting TIS use in all versus none of the follow-up years.
TIS, tobramycin inhalation solution.
Reduced Mortality With Inhaled Tobramycin 49
Pediatric Pulmonology
Page 7
The present study included patients meeting recom-
mended criteria for TIS use. Current evidence-based
guidelines and FDA labeling support chronic use of TIS
in CF patients aged 6 years with moderate–severe
lung disease (FEV
1
25–75% predicted) and persistent
airway infection with PA.
5,17
While age and lung func-
tion are unambiguously assessed in the CFFPR, chronic
PA infection lacks a consensus definition. Definitions of
chronic, as opposed to transient, PA infection vary,
18–21
as does the frequency and method of culture assessment
for PA across clinics.
19
The present study identified
chronic PA infection as a history of at least four cul-
tures positive for PA, which resulted in the average
included patient having six positive cultures and 78%
of their total cultures positive for PA. Changing the re-
quirement to a history of three or ve cultures positive
for PA had little impact on the association between TIS
use and reduced mortality.
TIS use was reported during only two-thirds of fol-
low-up years in the present study, though all included
patients met recommended criteria for chronic TIS use.
This level of non-use is consistent with prior analyses
of the CFFPR
1
and prior studies of CF medication ad-
herence.
22
Possible medical reasons for non-use, includ-
ing TIS intolerability, infection with BC or lung
transplantation, were reported for only small propor-
tions of non-users in the present study. Non-use more
likely reflects clinician variation or patient non-adherence
to recommended therapy. Continued efforts to under-
stand factors leading to underuse of TIS and to improve
overall medication adherence in CF will be valuable.
As in any non-randomized assessment of treatment
effects, a key challenge in the present study was to
adjust for confounding by indication, or the tendency
for TIS to be used in sicker patients who already have a
greater risk of death.
13
In our analysis, patients report-
ing TIS use do appear sicker, with more exacerbations
and hospitalizations, and worse baseline lung function.
The wealth of clinical measures and long-term follow-
up in the CFFPR allowed adjustment for almost all
characteristics that have been previously associated
with mortality, and adjustment for these and multiple
other patient characteristics resulted in a positive asso-
ciation between TIS use and reduced mortality. In
particular, female gender, lower FEV
1
, an increased
number of cultures positive for PA, cultures positive for
BCC and the presence of CF-related diabetes were
included in our adjusted model and were associated
with increased mortality, in accordance with prior
studies.
2,23,24
Importantly, use of dornase alfa was also
associated with significantly reduced mortality in the
present study. Like use of TIS, use of dornase alfa has
been previously associated with improved FEV
1
25,26
and reduced risk of pulmonary exacerbations or hospi-
talization,
27,28
but has not been previously associated
with reduced mortality. As patients using TIS were also
more likely to be using dornase alfa, our data
suggest that more aggressive compliance with chronic
therapy guidelines can improve survival with CF.
Depression was also associated with increased mortality
in the present study, and has been previously associated
with negative impacts on treatment adherence and
health-related quality of life in CF.
29
Associations
between depression and increased mortality in CF
warrant further study, particularly since depression may
be under-reported in the CFFPR.
Earlier epidemiologic studies have identified associa-
tions between more aggressive practice patterns with
higher median lung function among care centers.
30
Variation in practice patterns across CFF care centers
may have contributed to the association between TIS
use and reduced mortality in the present study. How-
ever, care center effects are unlikely to fully explain the
association between TIS use and reduced mortality in
our study, since this association remained at a similar
magnitude even after such care center effects were
allowed in a mixed effects model.
Despite adjustment for multiple risk factors and care
center effects, confounding could arise from unobserved
factors, as in any non-randomized study. Only a ran-
domized trial could attempt to avoid this limitation.
The analyses based on multivariate logistic regression
and Cox proportional hazards models also make the
assumption that drop-out from the CFFPR is non-
informative, that is, unrelated to prognosis, given the
history of covariate values included in the model. Bias
could also arise in the present study due to mismeasure-
ment of risk factors or outcomes. It should also be
pointed out that while TIS is approved and recom-
mended for chronic use in cycles of 28 days on treat-
ment followed by 28 days off treatment, the present
study could only assess the presence or absence of any
reported TIS use in each year. Thus, patients recorded
as using TIS in the present study may not have been
fully compliant with the recommended dosing regimen.
The association between TIS use and reduced mortal-
ity in the present study differs from an earlier analysis
of the CFFPR
13
which found the opposite association.
This earlier study used only the first 2 years of data
following FDA approval of TIS and was presented as
an illustration of confounding by indication. In particu-
lar, Rothman and Wentworth
13
suggest that ‘such a
relation is expected because TSI is used for those who
are close to death, resulting in strong confounding by
indication. In the first year after its approval, many
patients initiating TIS may have done so late in their
disease course, rather than in a timely fashion early in
the disease course, resulting in particularly strong con-
founding by indication and perhaps also limited benefits
of TIS due to late use. This interpretation is supported
50 Sawicki et al.
Pediatric Pulmonology
Page 8
by the change in the unadjusted association between
TIS use and subsequent-year mortality from a positive
association in years 1998 and 1999 to a negative asso-
ciation in years 2000 or later. The present study
employed a similar design to that of Rothman and
Wentworth,
13
assessing the effect of current year TIS
use on subsequent year mortality using multivariate
logistic regression, but leveraged an additional 9 years
of CFFPR data, included only those patients meeting
recommended criteria for chronic TIS use, and adjusted
for a larger number of risk factors. By finding a positive
association between TIS use and reduced mortality, as
might be expected given the established effects of TIS
on the improvement of lung function and reduced exac-
erbations, the present study suggests that incorporation
of more data can help overcome the limitations
discussed in the earlier analysis.
13
Clinical guidelines have recommended chronic use
of both TIS and dornase alfa in select populations of
individuals with CF.
5
In particular, a strong recommen-
dation is made in the guidelines for chronic TIS use for
patients over the age of 6 years with positive
PA cultures and moderate to severe lung disease. This
recommendation is based on an evidence-based review
identifying substantial benefits of TIS on improving
lung function and reducing the frequency of CF pulmo-
nary exacerbations. The clinical implications of our
analysis include further support of this strong recom-
mendation. Additionally, our analysis suggests that TIS
use may improve survival, both in the acute and chronic
setting. Although the absolute risk of mortality in any
given year is quite low among the patient population,
we have found that TIS use is associated with reduction
in subsequent year mortality, suggesting that treatment
with TIS in patients at high risk of mortality may be
of benefit. Moreover, the large difference in overall
survival between individuals with consistent use of TIS
compared to those with no evidence of TIS use further
buttresses the strategy promoted by treatment guidelines
for chronic TIS treatment in appropriate CF populations.
In summary, a retrospective analysis of the CFFPR
found that reported TIS use was associated with signifi-
cantly reduced mortality in the CF population, especial-
ly among patients reporting TIS use in all versus none
of the study years following their first eligibility for
TIS. This finding further supports chronic use of TIS
for patients with CF, moderate–severe lung disease and
chronic PA infection. The extent of potential underuse
of TIS and the reasons for underuse warrant further
examination.
ACKNOWLEDGMENTS
This study was funded by Novartis Pharmaceuticals
Corporation, East Hanover, NJ. J.E.S., D.L-V., M.V.W.,
and E.Q.Y. are employees of Analysis Group, Inc.,
which received research funding from Novartis Pharma-
ceuticals. J.Z. is an employee of Novartis Pharmaceuti-
cals Corporation. G.S.S. and L.S. received research
support from Novartis Pharmaceuticals Corporation.
Author contributions: Conception and design: G.S.S.,
J.E.S., J.Z., E.Q.W., L.S.; acquisition of data: J.E.S.,
J.Z., L.S.; analysis and interpretation: G.S.S., J.E.S.,
J.Z., D.L-V., M.W.; drafting the manuscript for impor-
tant intellectual content: J.E.S., D.L-V., M.W.; revision
of the manuscript for important intellectual content:
G.S.S., J.E.S., D.L-V., M.W., J.Z., E.Q.W., L.S. provid-
ed final approval of the version to be published: G.S.S.,
J.E.S., D.L-V., M.W., J.Z., E.Q.W., L.S. Other contribu-
tions: The authors are grateful to the CF Foundation for
access to the Patient Registry.
REFERENCES
1. Cystic Fibrosis Foundation. Patient Registry: Annual Data
Report 2009. Bethesda, Maryland: Cystic Fibrosis Foundation;
2011. Available from: http://www.cff.org/LivingWithCF/Quality
Improvement/PatientRegistryReport/.
2. Liou TG, Adler FR, Fitzsimmons SC, Cahill BC, Hibbs JR,
Marshall BC. Predictive 5-year survivorship model of cystic
fibrosis. Am J Epidemiol 2001;153:345–352.
3. Hoiby N, Flensborg EW, Beck B, Friis B, Jacobsen SV,
Jacobsen L. Pseudomonas aeruginosa infection in cystic
fibrosis. Diagnostic and prognostic significance of Pseudomonas
aeruginosa precipitins determined by means of crossed immu-
noelectrophoresis. Scand J Respir Dis 1977;58:65–79.
4. Henry RL, Mellis CM, Petrovic L. Mucoid Pseudomonas aeru-
ginosa is a marker of poor survival in cystic fibrosis. Pediatr
Pulmonol 1992;12:158–161.
5. Flume PA, O’Sullivan BP, Robinson KA, Goss CH, Mogayzel
PJ Jr, Willey-Courand DB, Bujan J, Finder J, Lester M, Quittell
L, et al. Cystic Fibrosis Pulmonary Guidelines: chronic medica-
tions for maintenance of lung health. Am J Respir Crit Care
Med 2007;176:957–969.
6. Ramsey BW, Pepe MS, Quan JM, Otto KL, Montgomery AB,
Williams-Warren J, Vasiljev-K M, Borowitz D, Bowman CM,
Marshall BC, et al. Intermittent administration of inhaled
tobramycin in patients with cystic fibrosis. N Engl J Med 1999;
340:23–30.
7. Moss RB. Administration of aerosolized antibiotics in cystic fi-
brosis patients. Chest 2001;120:s107–s113.
8. Murphy TD, Anbar RD, Lester LA, Nasr SZ, Nickerson B, Van-
Devanter DR, Colin AA. Treatment with tobramycin solution
for inhalation reduces hospitalizations in young CF subjects
with mild lung disease. Pediatr Pulmonol 2004;38:314–320.
9. Kerem E, Reisman J, Corey M, Canny GJ, Levison H. Predic-
tion of mortality in patients with cystic fibrosis. N Engl J Med
1992;326:1187–1191.
10. Corey M, Edwards L, Levison H, Knowles M. Longitudinal
analysis of pulmonary function decline in patients with cystic
fibrosis. J Pediatr 1997;131:809–814.
11. Hayllar KM, Williams SG, Wise AE, Pouria S, Lombard M,
Hodson ME, Westaby D. A prognostic model for the prediction
of survival in cystic fibrosis. Thorax 1997;52:313–317.
12. Enger C, Rothman KJ, Kylstra JW. Mortality rates during two
years of treatment with intermittent inhaled tobramycin (TOBI)
in cystic fibrosis. Pediatr Pulmonol 1999;28:339–340.
Reduced Mortality With Inhaled Tobramycin 51
Pediatric Pulmonology
Page 9
13. Rothman KJ, Wentworth CE III. Mortality of cystic fibrosis
patients treated with tobramycin solution for inhalation. Epide-
miology 2003;14:55–59.
14. Zeger SL, Liang KY, Albert PS. Models for longitudinal data: a
generalized estimating equation approach. Biometrics 1988;44:
1049–1060.
15. Wooldridge JM. Discrete response models. In: Econometric
analysis of cross section and panel data. Cambridge: The MIT
Press; 2002. pp 247–496.
16. Lin DY, Wei LJ. The robust inference for the proportional haz-
ards model. J Am Stat Assoc 1989;84:1074–1078.
17. NDA 50-753 TOBI (tobramycin inhalation solution, USP)
Nebulizer Solution. FDA approved drug products. Washington,
DC: FDA/Center for Drug Evaluation and Research. Available
at: http://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?id¼14345
(Last accessed 20 September 2010).
18. Do
¨
ring G, Conway SP, Heijerman HG, Hodson ME, Høiby N,
Smyth A, Touw DJ. Antibiotic therapy against Pseudomonas
aeruginosa in cystic fibrosis: a European consensus. Eur Respir
J 2000;16:749–767.
19. Pedersen SS, Espersen F, Hoiby N. Diagnosis of chronic Pseu-
domonas aeruginosa infection in cystic fibrosis by enzyme-
linked immunosorbent assay. J Clin Microbiol 1987;25:1830–
1836.
20. Lee TW, Brownlee KG, Conway SP, Denton M, Littlewood JM.
Evaluation of a new definition for chronic Pseudomonas aerugi-
nosa infection in cystic fibrosis patients. J Cyst Fibros 2003;2:
29–34.
21. Proesmans M, Balinska-Miskiewicz W, Dupont L, Bossuyt X.
Evaluating the ‘Leeds criteria’ for Pseudomonas aeruginosa in-
fection in a cystic fibrosis centre. Eur Respir J 2006;27:937–943.
22. Zindani GN, Streetman DD, Streetman DS, Nasr SZ. Adherence
to treatment in children and adolescent patients with cystic
fibrosis. J Adolesc Health 2006;38:13–17.
23. O’Connor GT, Quinton HB, Kahn R, Robichaud P, Maddock J,
Lever T, Detzer M, Brooks JG; Northern New England
Cystic Fibrosis Consortium. Case-mix adjustment for evaluation
of mortality in cystic fibrosis. Pediatr Pulmonol 2002;33:99–
105.
24. Rosenfeld M, Davis R, FitzSimmons S. Gender gap in cystic
fibrosis mortality. Am J Epidemiol 1997;145:794–803.
25. Johnson CA, Butler SM, Konstan MW, Breen TJ, Morgan WJ.
Estimating effectiveness in an observational study: a case study
of dornase alfa in cystic fibrosis. The Investigators and Coordi-
nators of the Epidemiologic Study of Cystic Fibrosis. J Pediatr
1999;134:734–739.
26. Fuchs HJ, Borowitz DS, Christiansen DH, Morris EM, Nash
ML, Ramsey BW, Rosenstein BJ, Smith AL, Wohl ME. Effect
of aerosolized recombinant human DNase on exacerbations of
respiratory symptoms and on pulmonary function in patients
with cystic fibrosis. The Pulmozyme Study Group. N Engl J
Med 1994;331:637–642.
27. Hodson ME, McKenzie S, Harms HK, Koch C, Mastella G,
Navarro J, Strandvik B; Investigators of the Epidemiologic
Registry of Cystic Fibrosis (ERCF). Dornase alfa in the treat-
ment of cystic fibrosis in Europe: a report from the Epidemio-
logic Registry of Cystic Fibrosis. Pediatr Pulmonol 2003;36:
427–432.
28. Oster G, Huse DM, Lacey MJ, Regan MM, Fuchs HJ. Effects
of recombinant human DNase therapy on health care use and
cost in patients with cystic fibrosis. Ann Pharmacother 1995;29:
459–464.
29. Quittner AL, Barker DH, Snell C, Grimley ME, Marciel K,
Cruz I. Prevalence and impact of depression in cystic fibrosis.
Curr Opin Pulm Med 2008;14:582–588.
30. Johnson C, Butler SM, Konstan MW, Morgan W, Wohl ME.
Factors influencing outcomes in cystic fibrosis: a center-based
analysis. Chest 2003;123:20–27.
52 Sawicki et al.
Pediatric Pulmonology
SG 05/2012/1769
1205017426
Page 10
  • Source
    • "Следует подчеркнуть, что для препарата, наряду с муколитическими, также характерны противовоспалительные и антибактериальные свойства, которые обеспечиваются за счет снижения концентрации эластазы и интерлейкина 8 в мокроте, уменьшения процентного содержания нейтрофилов, снижения концентрации нейтрофильной эластазы и интерлейкина 8 в жидкости бронхоальвеолярного лаважа, влияния на биофильм мукоидной синегнойной палочки, снижения содержания матричных металлопротеиназ в жидкости бронхоальвеолярного лаважа, т. е. уменьшения вклада деструктивного компонента легочной ткани при воспа- лении [23, 24] . Анализ данных 12 740 пациентов из регистра США (1996–2008 гг.), а также 2538 случаев смерти пациентов в течение 6 лет наблюдений позволил сделать вывод о том, что ингаляции с дорназой альфа сокращают смертность на 15% [26]. Развития существенных нежелательных явлений при лечении дорназой альфа не отмечено. "
    Full-text · Article · Dec 2014
  • Source
    • "Ironically, the potentially most heretical use of antibiotics in CF, the chronic suppression of bacterial opportunists in the airway without their eradication, is the one with the greatest empirical evidence for clinical benefit. Chronic or chronic intermittent administration of the inhaled antipseudomonal antibiotics colistimethate, tobramycin and aztreonam has been associated with improved quality of life [110,111], decreased risk of exacerbation [34,112], improved pulmonary function [34,111], and decreased mortality [97]. This practice, which has expanded substantially in recent decades [47], was predated by scheduled, periodic treatment of patients with intravenous antibiotics [113,114]. "
    [Show abstract] [Hide abstract] ABSTRACT: Since the discovery of the gene associated with cystic fibrosis (CF), there has been tremendous progress in the care of patients with this disease. New therapies have entered the market and are part of the standard treatment of patients with CF, and have been associated with marked improvement in survival. Now there are even more promising therapies directed at different components of the pathophysiology of this disease. In this review, our current knowledge of the pathophysiology of lung disease in patients with CF is described, along with the current treatment of CF lung disease, and the therapies in development that offer great promise to our patients.
    Full-text · Article · Aug 2012 · BMC Medicine
  • Source
    • "Tobramycin inhalation solution (TIS) has been introduced for the long-term management of chronic P. aeruginosa infection, with a Cochrane review [20] suggesting some benefit from TIS in terms of lung function and pulmonary exacerbation rate but also concern regarding an increase in antibiotic resistance. A recent registry study examining data from the Cystic Fibrosis Foundation's Patient Registry has suggested that TIS use is associated with reduced mortality [21]. TIS has also been demonstrated to be effective in delaying re-infection in those with early P. aeruginosa infection [22]. "
    [Show abstract] [Hide abstract] ABSTRACT: Pseudomonas aeruginosa chronically infects patients with cystic fibrosis and is associated with greater morbidity. There has been limited progress on the clinical development of new antibiotics with novel modes of action. This review addresses some of the latest research developments on the exploitation of candidate adjuvant therapeutic agents that may act alongside conventional antibiotics as an alternative therapeutic strategy. After considering key mechanisms this opportunistic pathogen employs to control virulence, the progress of various strategies including the inhibition of quorum sensing, efflux pumps and lectins, and the use of iron chelators, bacteriophages, immunisation and immunotherapy is reviewed. Both therapeutic approaches in early development and clinical phase are discussed.
    Full-text · Article · Jun 2012 · European Respiratory Journal
Show more