ERJ Express. Published on December 23, 2014 as doi: 10.1183/09031936.

00148613

ORIGINAL ARTICLE
IN PRESS | CORRECTED PROOF

Nonspecific interstitial pneumonia:

survival is influenced by the
underlying cause
Hilario Nunes1,2, Kirsten Schubel2, Diane Piver3, Eline Magois4,
Séverine Feuillet5, Yurdagul Uzunhan1,2, Zohra Carton2, Abdellatif Tazi5,
Pierre Levy6, Pierre-Yves Brillet3, Andrew G. Nicholson7,
Marianne Kambouchner8 and Dominique Valeyre1,2

Affiliations: 1Université Paris 13, Sorbonne Paris Cité, EA2363 "Réponses cellulaires et fonctionnelles à
l’hypoxie", Bobigny, France. 2AP-HP, Service de Pneumologie, Hôpital Avicenne, Bobigny, France. 3AP-HP,
Service de Radiologie, Hôpital Avicenne, Bobigny, France. 4Service de Pneumologie, Hôpital d’Amiens,
Université de Picardie Jules Verne, Amiens, France. 5Université Paris Diderot, Sorbonne Paris Cité, AP-HP,
Service de Pneumologie, Hôpital Saint Louis, Paris, France. 6AP-HP, Département de Santé Publique, Hôpital
Tenon, INSERM, U707, Université Paris 6 Pierre et Marie Curie, UMR-S 707, Paris, France. 7Dept of
Histopathology, Royal Brompton and Harefield NHS Foundation Trust and NHLI Division, Imperial College,
London, UK. 8AP-HP, Service d’Anatomie Pathologique, Hôpital Avicenne, Bobigny, France.

Correspondence: Hilario Nunes, Service de Pneumologie, Hôpital Avicenne, 125 rue de Stalingrad, 93009
Bobigny, France. E-mail: hilario.nunes@avc.aphp.fr

ABSTRACT Idiopathic, nonspecific interstitial pneumonia (NSIP) is most often associated with various
clinical disorders, including connective tissue diseases (CTDs) and chronic hypersensitivity pneumonitis
(cHP). Emerging evidence also suggests that “idiopathic” NSIP may be the lung manifestation of
undifferentiated CTD (UCTD). However, whether or not NSIP outcome is influenced by the underlying
cause remains uncertain.
This retrospective study included 127 biopsy-proven NSIP patients (65 women, mean±SD age
55±12 years). Survivals were estimated using a Kaplan–Meier curve and compared using the log-rank test.
Multivariate analyses were based on a Cox model.
15 (11.8%) patients had cHP, 29 (22.8%) had CTD, 32 (25.2%) satisfied the Kinder criteria for UCTD
and 51 (40.1%) had idiopathic NSIP. At the end of follow-up (mean±SD 64±54 months), a difference in
survival was observed between aetiological groups ( p=0.002). Survival was better for UCTD than for
idiopathic NSIP ( p=0.020) and similar to that observed for CTD. cHP survival tended to be poorer than
that of idiopathic NSIP ( p=0.087) and was an independent predictor of mortality (hazard ratio 2.17, 95%
CI 1.05–4.47; p=0.035).
NSIP outcome is influenced by its cause. cHP exhibits the highest mortality. UCTD does not differ from
CTD supporting the concept of autoimmune NSIP, with a prognosis that is better than that of idiopathic NSIP.

@ERSpublications
NSIP patients should be investigated for the presence of an underlying cause, which significantly
impacts survival http://ow.ly/F1lO2

Received: Aug 25 2014 | Accepted after revision: Nov 21 2014

Conflict of interest: Disclosures can be found alongside the online version of this article at erj.ersjournals.com
Copyright ©ERS 2014

Eur Respir J 2014; in press | DOI: 10.1183/09031936.00148613 1

Copyright 2014 by the European Respiratory Society.
INTERSTITIAL LUNG DISEASES | H. NUNES ET AL.

Introduction
Nonspecific interstitial pneumonia (NSIP) remains an area of uncertainty that requires further research
[1, 2]. Although most often idiopathic, the histologic pattern of NSIP is also observed in a wide variety of
clinical settings, including connective tissue diseases (CTDs), chronic hypersensitivity pneumonitis (cHP),
drug toxicity and slowly resolving diffuse alveolar damage [1]. NSIP is the most common histological
pattern in interstitial lung diseases (ILDs) associated with CTDs, including their forme fruste variants.
Several authors have suggested that a subset of patients previously classified as “idiopathic” NSIP meet the
criteria for undifferentiated CTD (UCTD) [3–5]. UCTD is characterised by the presence of features
reflecting a systemic autoimmune process that do not fulfil the accepted diagnostic criteria for
differentiated CTDs, i.e. rheumatoid arthritis (RA), Sjögren’s syndrome (SjS), systemic scleroderma (SSc),
polymyositis and dermatomyositis, systemic lupus erythematosus and mixed CTD [5, 6]. Whether or not
patients with CTD-associated NSIP have a better outcome than those with idiopathic NSIP remains
controversial [7]. Similarly, the prognostic impact of UCTD in NSIP patients has not been fully elucidated
[8–11]. Although it is now well known that NSIP can represent the sole histological expression of a
percentage of patients with cHP [12–16], no studies have compared the survival of these patients with that
of patients with idiopathic NSIP.
The aim of the study was to compare the prognosis of NSIP patients stratified according to the underlying
cause (idiopathic, UCTD, CTDs and cHP) in terms of survival, response to therapy and long-term
functional outcome.

Methods
Patient selection and data collection
This retrospective study received institutional review board approval (Comité de Protection des Personnes
Ile-de-France X, No. 2012-12-01). All consecutive patients with a histological pattern of NSIP on surgical
lung biopsy (SLB) examined at the Avicenne University Hospital Pathology department (Bobigny, France)
were selected. The patient cohort has already been the subject of a previous study focusing on histology
[17]. The diagnosis of NSIP was based on a consensus by two pathologists (M. Kambouchner and
A.G. Nicholson) [1]. Clinical and laboratory characteristics, pulmonary function tests (PFTs) and
bronchoalveolar lavage (BAL) findings at the time of SLB were collected from medical records. The battery
of individual serological tests and serum precipitins were ordered as part of the initial workup or during
follow-up on clinical grounds. Two radiologists (D. Piver and P.Y. Brillet) reviewed in consensus
high-resolution computed tomography (HRCT) scans that were available within 6 months of SLB. Patients
were classified as presenting a HRCT pattern either “suggestive or consistent with NSIP” or “suggestive of
usual interstitial pneumonia (UIP)” [1, 2].

Diagnostic criteria
Standard diagnostic criteria were applied for individual differentiated CTDs [18–24]. Patients were
considered to have UCTD when they had at least one symptom suggestive of CTDs and evidence of
systemic inflammation in the absence of infection, as defined by KINDER et al. [5]. A narrower definition of
UCTD was also applied, as proposed by CORTE et al. [8, 25]. The diagnosis of cHP was established using
the criteria of RICHERSON et al [26]. In addition to clinical and radiological evidence of ILD, patients were
required to have a history of exposure to an inhaled antigen known to cause cHP and either confirmatory
serum precipitins or a lymphocytic BAL [26].

Patient outcome
Therapeutic response was recorded within 3–6 months of treatment initiation. Long-term functional
outcome was evaluated for patients with available PFTs at least 12 months after their initial assessment.
Improvement was defined as a ⩾10% increase in forced vital capacity (FVC) % predicted or a ⩾15%
increase in diffusing capacity of the lung for carbon monoxide (DLCO) % predicted from initial values, and
worsening was defined as >10% decrease in FVC or >15% decrease in DLCO.

Statistical analysis
All results are expressed as percentages or mean±SD. The various aetiological groups were compared using
a Chi-squared test or Fisher’s exact test for categorical variables, and Kruskal–Wallis test for continuous
variables. Survival was calculated from the date of inclusion, which corresponded to the date of SLB and
ranged from July 1987 to November 2011, until the end of the follow-up period. Patients were followed
until death, lung transplantation or September 1, 2012, with complete follow-up for 125 out of 127 cases.
Information regarding vital status and cause of death was obtained by reviewing the patient’s medical
charts and by contacting the referring physician and general practitioner. Transplanted patients were
censored at the time of transplantation. The survival probability of aetiological groups was estimated using

2 DOI: 10.1183/09031936.00148613
 INTERSTITIAL LUNG DISEASES | H. NUNES ET AL.

the Kaplan–Meier method and compared by a log-rank test. Univariate analysis was based on a log-rank
test. For continuous variables, patients were classified into two groups on either side of the median value.
All parameters with a p-value <0.20 were then entered into the multivariate Cox proportional hazards
model. Results are reported as hazard ratios (HRs) and 95% confidence intervals.

Results
Aetiological groups
136 cases with NSIP were listed in the Pathology department. Nine cases were excluded because the
medical charts were not available. The study population comprised 127 patients (table 1). SLB were
obtained from a single lobe in 29 (22.8%) cases, two lobes in 75 (59.1%) cases and three lobes in four
(3.1%) cases. The site of SLB was not available for 19 (15.0%) patients, and for the other 108 patients
corresponded to: upper lobes n=74 (68.5%); lower lobes n=90 (83.3%); middle lobe n=18 (16.7%); or
lingula: n=5 (4.6%).
The study population consisted of 62 men with a mean±SD age of 55±12 years at the time of SLB. 15
(11.8%) patients had a diagnosis of cHP induced by the following antigens: birds n=12, domestic moulds
n=1, hay n=1 and textile dusts n=1. 29 (22.8%) patients had a differentiated CTD (RA n=7, SjS n=7, SSc
n=4, polymyositis or dermatomyositis n=7, mixed CTD n=2, systemic lupus erythematosus plus SjS n=1,
and SSc plus SjS n=1). NSIP occurred during the course of a previously known CTD in nine cases, with a
median time to onset of 60 months (range: 5–192 months). The two conditions were diagnosed
concomitantly in 16 cases. NSIP preceded the diagnosis of CTD in four cases by 8, 48, 120 and
120 months, respectively. While one of these four patients with RA initially had no features suggestive of a

TABLE 1 Patients’ characteristics at the time of surgical lung biopsy (SLB) according to aetiological groups

Idiopathic NSIP UCTD# CTDs cHP p-value Whole population

Subjects 51 32 29 15 127
Demographics
Males/Females 32/19 11/21¶ 10/19+ 9/6 0.020 62/65
Non-/current or ex-smokers 21/30 23/9¶ 17/12 7/8 0.044 68/59
Age years 55.8±11.5 56.5±12.8 51.6±14.4 52.7±10.3 0.206 54.7±12.5
Clinical symptoms
NYHA functional class I–II/III–IV 41/10 25/7 24/5 11/4 0.871 101/26
Squawks 4 5 3 3 0.560 15
Digital clubbing 18 9 5 4 0.402 36
Duration of symptoms months 35.4±49.2 48.1±47.7 32.4±31.0 32.5±44.0 0.445 37.7±44.5
Site of SLB 0.963
At least two lobes 32 19 19 9 79
A single lobe or unknown 19 13 10 6 48
HRCT pattern, n 0.375
Subjects 37 23 23 11 94
Suggestive or consistent with NSIP 36 22 20 10 88
Suggestive of UIP 1 1 3 1 6
Pulmonary function tests
Subjects 49 31 27 15 122
FEV1 % predicted 70.0±19.0 75.7±22.2 66.1±20.8 60.2±17.4 0.115 69.4±20.4
FEV1/FVC % 85.2±9.6 83.5±8.2 85.1±9.9 82.9±8.5 0.830 84.5±9.2
FVC % predicted 67.4±17.3 73.6±20.3 63.3±20.0 59.0±14.3 0.071 67.1±18.8
TLC % predicted 70.6±17.5 73.8±15.7 66.6±20.2 61.4±13.8 0.081 69.4±17.6
Subjects 46 28 24 13 110
DLCO % predicted 50.0±13.8 54.9±17.3 47.2±14.4 44.0±18.9 0.238 50.0±15.7
Bronchoalveolar lavage
Subjects 42 28 22 14 106
Lymphocytes % 15.6±13.3 17.1±17.2 20.1±16.7 20.6±19.7 0.861 17.6±15.9
Neutrophils % 10.7±11.5 12.3±14.4 15.5±15.2 7.0±9.2 0.198 11.7±13.1
Eosinophils % 3.5±5.3 6.9±13.2 3.7±4.7 3.0±3.0 0.835 4.4±8.1

Data are presented as n, n/n or mean±SD, unless otherwise stated. NSIP: nonspecific interstitial pneumonia; UCTD: undifferentiated connective
tissue disease; CTD: connective tissue disease; cHP: chronic hypersensitivity pneumonitis; NYHA: New York Heart Association; HRCT:
high-resolution computed tomography; UIP: usual interstitial pneumonia; FEV1: forced expiratory volume in 1 s; FVC: forced vital capacity; TLC:
total lung capacity; DLCO: diffusing capacity of the lung for carbon monoxide. #: UCTD as defined by Kinder’s criteria [5]; ¶: p<0.05 for
comparison between idiopathic and UCTD groups; +: p<0.05 for comparison between idiopathic and CTDs groups.

DOI: 10.1183/09031936.00148613 3
INTERSTITIAL LUNG DISEASES | H. NUNES ET AL.

systemic autoimmune disorder, the other three patients with dermatomyositis, RA and SjS met the criteria
for UCTD prior to the onset of differentiated CTD. Among the remaining 83 patients, 32 (38.5%, i.e. 25.2%
of the whole population) satisfied the criteria for UCTD and did not develop differentiated CTD during
follow-up, and 51 (61.4%, i.e. 40.1% of the whole population) had idiopathic NSIP with no UCTD. 19 cases
(22.9%, i.e. 15% of the whole population) met the definition for UCTD proposed by CORTE et al. [8].

Clinical characteristics, PFTs, BAL and HRCT findings

Patient characteristics at the time of SLB are summarised in table 1. Overall, aetiological groups were
significantly different in terms of sex ratio ( p=0.020) and smoking status ( p=0.048). More specifically,
patients with UCTD and CTDs were more likely to be females than those with idiopathic NSIP (UCTD
versus idiopathic NSIP: 65.6% versus 37.2%, p=0.014; and CTDs versus idiopathic NSIP: 65.5% versus
37.2%, p=0.020). The UCTD group comprised a significantly higher percentage of nonsmokers than the
idiopathic NSIP group (UCTD versus idiopathic NSIP: 71.8% versus 41.2%; p=0.007). Conversely, no
significant difference was found between the UCTD and CTDs groups, or between the cHP and idiopathic
NSIP groups. HRCT was available for 94 patients, and showed a pattern “suggestive or consistent with
NSIP” in 88 (93.6%) cases and “suggestive of UIP” in six (6.4%) cases. All of these six cases had a biopsy
taken from two lobes, providing evidence of a concordant histological pattern of NSIP.

Systemic autoimmune symptoms and laboratory findings

The results are shown in tables 2 and 3. As expected, the aetiological groups differed in terms of the
presence of systemic autoimmune symptoms ( p<0.0001) and auto-antibodies ( p<0.0001). All patients with
UCTD and CTDs had at least one autoimmune symptom, a much higher rate than that observed in
patients with idiopathic NSIP (100% versus 43.1%; p<0.0001). By contrast, no significant difference was
observed between the cHP and idiopathic NSIP groups. The proportion of patients with at least two
symptoms was higher in the CTDs group than in the UCTD group (96.5% versus 78.1%; p=0.033).

Therapeutic response and long-term functional outcome

14 (11%) patients did not receive any specific treatment for NSIP during the study period, 106 (83.5%)
patients were given corticosteroids (n=99) and/or at least one immunosuppressive agent (azathioprine:
n=52, cyclophosphamide: n=33, mycophenolate mofetil: n=24, methotrexate: n=4, rituximab: n=3,
cyclosporine: n=1, leflunomide: n=1 and plasmapheresis: n=1), and information was not available in seven

TABLE 2 Distribution of autoimmune symptoms according to aetiological groups

Idiopathic NSIP UCTD# CTDs cHP p-value

Subjects 51 32 29 15
Symptoms associated with CTD
Raynaud’s phenomenon 2 5 15¶,+ 1 <0.0001
Arthralgias/multiple joint swelling 5 19§ 20¶ 3 <0.0001
Morning stiffness 1 6§ 14¶,+ 1 <0.0001
Proximal muscle weakness 0 5§ 13¶,+ 1 <0.0001
Sicca features 2 15§ 15¶ 2 <0.0001
Dysphagia 2 2 2 0 0.780
Gastro-oesophageal reflux 13 8 12 1 0.096
Photosensitivity 0 2 1 0 0.226
Skin changes (rash) 1 3 7¶ 0 0.006
Oral ulceration 0 0 1 1 0.118
Nonandrogenic alopecia 0 0 1 0 0.346
Recurrent unexplained fever 0 3 5 0 0.005
Unintentional weight loss 6 9 8 1 0.096
Number of symptoms
⩾1 22 32§ 29¶ 7 <0.0001
⩾2 8 25§ 28¶,+ 3 <0.0001
⩾3 0 12§ 21¶,+ 1 <0.0001

Data are presented as n, unless otherwise stated. NSIP: nonspecific interstitial pneumonia; UCTD:
undifferentiated connective tissue disease; CTD: connective tissue disease; cHP: chronic hypersensitivity
pneumonitis. #: UCTD as defined by Kinder’s criteria [5]; ¶: p<0.05 for comparison between idiopathic and
CTDs groups; +: p<0.05 for comparison between UCTD and CTDs groups; §: p<0.05 for comparison between
idiopathic and UCTD groups.

4 DOI: 10.1183/09031936.00148613
 INTERSTITIAL LUNG DISEASES | H. NUNES ET AL.

TABLE 3 Distribution of autoimmune laboratory signs according to aetiological groups

Idiopathic NSIP UCTD# CTDs cHP p-value

Subjects 51 32 29 15
Auto-antibody
Rheumatoid factor 2/47 13/31¶ 13/27+ 2/15 <0.0001
Anti-CCP antibody 0/17 3/17 7/14+ 0/6 0.002
Antinuclear antibody 8/46 22/31¶ 24/28+ 3/15 <0.0001
Antinuclear antibody titer 120 (80–320) 160 (80–1600) 320 (80–2650) + 80 (80–320) 0.05
Anti-dsDNA antibody 1/8 0/21 3/24 0/3 0.356
Anti-Sm antibody 0/40 0/29 3/24+ 0/14 0.012
Anti-SSA antibody 0/40 0/29 8/25+ 0/14 <0.0001
Anti-SSB antibody 0/40 0/29 3/25+ 0/14 0.024
Anti-RNP antibody 0/40 1/29 2/24 0/14 0.065
Anti-Scl 70 antibody 0/29 3/21 2/24 0/8 0.153
Anti-Jo-1 antibody 0/30 2/24 2/25 0/10 0.337
Anti-PL7 antibody 0/10 0/4 0/8 0/4 1
Anti-PL12 antibody 0/10 0/5 0/8 0/4 1
Anti-PM-ScL antibody 0/10 0/7 0/9 0/4 1
Anti-Mi2 antibody 0/10 0/3 0/9 0/4 1
Anti-Ku antibody 0/9 0/3 1/9 0/4 1
Anti-SRP antibody 0/9 0/4 0/9 0/4 1
ANCA 4/33 7/27 1/22 1/11 0.194
Elevated erythrocyte sedimentation rate 7 19¶ 20+ 3 <0.0001
(⩾2 times normal) or CRP in the absence of infection
At least one auto-antibody 12 29¶ 28+ 4 <0.0001

Data are presented as n positive/N tested, median (range) or n, unless otherwise stated. NSIP: nonspecific interstitial pneumonia; UCTD:
undifferentiated connective tissue disease; CTD: connective tissue disease; cHP: chronic hypersensitivity pneumonitis; CCP: cyclic citrullinated
peptide; dsDNA: double stranded DNA; RNP: ribonucleoprotein; SRP: signal recognition particle; ANCA: anti-neutrophil cytoplasmic antibodies;
CRP: C-reactive protein. #: UCTD as defined by Kinder’s criteria [5]; ¶: p<0.05 for comparison between idiopathic and UCTD groups; +: p<0.05
for comparison between idiopathic and CTDs groups.

cases. All cHP patients had been removed from antigen exposure. 101 patients were evaluated for
therapeutic response based on PFTs. Overall, responses were not different between groups (p=0.219) (table 4).
However, when patients with UCTD and CTDs were pooled and compared with the rest of the
population, they presented a significantly higher rate of response (43.7% versus 24.5%; p=0.041).
Long-term functional follow-up (⩾12 months) was available for 105 patients. At the last follow-up, the
distribution of patients according to functional evolution outcome and the annual decline in FVC and
DLCO were similar between groups (table 4).

Survival and causes of death

Mean follow-up after SLB was 63.7±54.2 months. At the end of the study, 75 patients were alive, 40
patients had died, 10 had undergone transplant and two were lost to follow-up. 2-, 5- and 10-year overall
survival was 89.0%, 65.6% and 49.2%, respectively. Survival was significantly different between groups
( p=0.002) (fig. 1a), with an even more marked difference when patients with UCTD and CTDs were
pooled ( p=0.0006) (fig. 1b). Patients with cHP exhibited the poorest survival (2-, 5- and 10-year survival
of 73.3%, 41.9% and 27.9%, respectively), followed by patients with idiopathic NSIP (2-, 5- and 10-year
survival of 87.7%, 60.7% and 35.9%, respectively) and patients with autoimmune NSIP (2-, 5- and 10-year
survival of 94.5%, 77.1% and 72.5%, respectively). A side-by-side comparison showed that UCTD survival
was significantly better than that of idiopathic NSIP ( p=0.020), but similar to that of CTD ( p=0.583). cHP
patients tended to have a worse survival than patients with idiopathic NSIP ( p=0.087). Similar results were
observed according to the criteria proposed by CORTE et al. [8], with a tendency towards better survival in
the UCTD group than in idiopathic NSIP group ( p=0.090) and no significant difference between the
UCTD and CTDs groups ( p=0.614).
The causes of death were as follows: end-stage respiratory failure n=13; acute exacerbation n=11;
respiratory tract infection n=2; pulmonary hypertension n=1; sudden death n=1; pneumothorax n=1;
neoplasia n=5 (lung n=2, pancreas n=1, nasopharynx n=1 and colon n=1); upper gastrointestinal tract
haemorrhage n=1; and unknown n=5.

DOI: 10.1183/09031936.00148613 5
INTERSTITIAL LUNG DISEASES | H. NUNES ET AL.

TABLE 4 Patient outcomes according to aetiological groups

Idiopathic NSIP UCTD# CTDs cHP p-value

Subjects 51 32 29 15
Duration of follow-up months 66.7±58.0 64.3±49.3 67.1±61.7 45.8±31.5 0.716
Treatment at some point during follow-up
Corticosteroids 39 26 21 13 0.849
Azathioprine 20 14 12 6 0.966
Cyclophosphamide 12 6 11 4 0.352
Mycophenolate mofetil 8 8 7 1 0.341
Response to first or second-line treatment n/N 10/39 10/25 11/23 3/14 0.219
Long-term evolution of PFTs
Subjects 43 26 23 13
Interval between PFTs years 4.6±3.9 5.2±4.0 5.6±3.8 3.7±2.4 0.485
Annual FVC decline % −3.2±7.2 −1.5±3.9 −2.0±3.5 −2.0±4.6 0.562
Annual DLCO decline % −4.5±13.2 −1.5±5.2 −0.8±1.9 −3.1±7.2 0.304
Improvement/stability/worsening 6/11/26 5/8/13 1/12/10 2/4/7 0.396

Data are presented as n or mean±SD, unless otherwise stated. NSIP: nonspecific interstitial pneumonia; UCTD: undifferentiated connective
tissue disease; CTD: connective tissue disease; cHP: chronic hypersensitivity pneumonitis; PFTs: pulmonary function tests; FVC: forced vital
capacity; DLCO: diffusing capacity of the lung for carbon monoxide. #: UCTD as defined by Kinder’s criteria [5].

Predictive factors of mortality

Results of the univariate analysis are presented in table 5. In multivariate analyses independent predictors
of mortality were absence of response to therapy (HR 10.38, 95% CI 3.1–34.2; p=0.0001) (fig. 2) and a
diagnosis of cHP (HR 2.17, 95% CI 1.05–4.47; p=0.035) (fig. 3).

Discussion
Only limited data are available in the literature on whether patients with secondary NSIP have a different
outcome to those with idiopathic NSIP. The present study demonstrates that, despite similar baseline
functional impairment and long-term functional decline, the prognosis of NSIP is influenced by the
underlying cause of the disease. Patients with cHP appear to have a poorer outcome, while a diagnosis of
cHP is independently associated with a higher mortality. Although patients with autoimmune NSIP, i.e.
NSIP associated with CTDs or UCTD, have a better survival than those with cHP and idiopathic NSIP,
the impact of a diagnosis of autoimmune NSIP is no longer significant on multivariate analysis. The
absence of response to therapy is the strongest independent determinant of mortality.
A NSIP pattern is observed on histology in 16–50% of cases of cHP [12–16]. Hypersensitivity pneumonitis
is traditionally thought to be associated with a good prognosis, but a wide range of mortality rates are

a) 1.0 b) 1.0

0.8 0.8
CTD
UCTD+CTD

0.6 UCTD 0.6

Survival

Survival

Idiopathic
0.4 Idiopathic 0.4

0.2 cHP 0.2 cHP

0.0 0.0
0 20 40 60 80 100 120 0 20 40 60 80 100 120
Time months Time months
At risk n 127 104 77 46 31 27 19 At risk n 127 104 77 46 31 27 19

FIGURE 1 Patient survival according to aetiological group. a) Patients with undifferentiated connective tissue disease (UCTD) (as defined using Kinder’s
criteria [5]) and connective tissue diseases (CTDs) are separated. b) Patients with UCTD and CTDs are pooled (autoimmune). cHP: chronic hypersensitivity
pneumonitis.

6 DOI: 10.1183/09031936.00148613
 INTERSTITIAL LUNG DISEASES | H. NUNES ET AL.

TABLE 5 Univariate analysis of factors associated with mortality

Variable p-value

Males 0.910
Current or ex-smoker 0.162
Age >56.1 years 0.139
Absence of systemic autoimmune symptoms 0.017
Absence of auto-antibody 0.013
No diagnosis of UCTD# or CTDs 0.001
Diagnosis of cHP 0.004
Duration of symptoms >22.5 months 0.839
Digital clubbing 0.067
FVC <65.2% predicted 0.125
DLCO <50.0% predicted 0.062
Alveolar lymphocytosis >20% 0.019
Alveolar neutrophilia >5% 0.090
No response to therapy <0.0001

UCTD: undifferentiated connective tissue disease; CTDs: connective tissue diseases; cHP: chronic
hypersensitivity pneumonitis; FVC: forced vital capacity; DLCO: diffusing capacity of the lung for carbon
monoxide. #: UCTD as defined by Kinder’s criteria [5].

actually reported [27]. Most recent studies have estimated the 5-year survival to be between 25% and 55%
in the presence of histological signs of fibrosis [12, 15, 28, 29]. A similarly poor outcome is observed in
the present series, with a 5-year survival of 41.9%. This may partly reflect the large number of patients
with bird exposure (12 (80%) out of 15), which has been suggested to be linked with a more severe course
[27]. It is remarkable that the disease progressed in the majority of our patients, all of whom presented
with avian cHP, despite exposure avoidance, which raises the hypothesis that, once NSIP is established, the
disease may become independent of continuing exposure [15].
There is emerging evidence that so-called “idiopathic” NSIP is frequently associated with an autoimmune
“flavour” [3–5]. KINDER et al. [5] first applied a set of diagnostic criteria for UCTD to an American cohort
with idiopathic interstitial pneumonias (IIPs), and found that 88% of cases with NSIP had UCTD. In two
other Japanese [10] and British series [8] the proportion of UCTD patients was 47% and 71%,
respectively. CORTE et al. [8] emphasised the highly nonspecific nature of Kinder’s criteria, which were met
by one third of patients with idiopathic pulmonary fibrosis (IPF). Using a more stringent definition, the
frequency of UCTD dropped from 71% to 31% [8]. In our series, the proportion of UCTD patients was
38.5% and 22.9% using the Kinder criteria and the Corte criteria, respectively. No clearly validated criteria
for the diagnosis of UCTD are available at the present time.
As in previous studies, our patients with UCTD were more likely to be females [5, 8, 30] and nonsmokers
[5, 30] than those with idiopathic NSIP. UCTD patients were similar to CTDs patients, including in terms
of demographic characteristics and survival. In line with previous observations [10], three cases with
UCTD developed differentiated CTD over time. Taken together, these observations support the grouping
of these patients into a subset of “autoimmune NSIP”.

1.0 Response to therapy

0.8

0.6
Survival

No response to therapy
0.4

0.2

0.0
FIGURE 2 Patient survival according
to the response to first- and/or 0 20 40 60 80 100 120
second-line therapy. Time months

DOI: 10.1183/09031936.00148613 7
INTERSTITIAL LUNG DISEASES | H. NUNES ET AL.

1.0

0.8

0.6 No cHP
Survival

0.4
cHP

0.2

0.0
FIGURE 3 Patient survival according
0 20 40 60 80 100 120 to the presence or absence of chronic
Time months hypersensitivity pneumonitis (cHP).

Most studies have shown that individuals with CTDs-ILD live significantly longer than patients with IIPs
[7, 31], but in the study of PARK et al. [7] this difference disappeared in cases diagnosed by histology as
NSIP. However, the absence of distinction of UCTD in this study may have led to overestimation of the
survival of patients with “idiopathic” NSIP. In fact, the prognostic significance of a concomitant diagnosis
of UCTD has been questioned in a small number of studies [8–11]. CORTE et al. [8] showed that the
presence of UCTD was not associated with a survival advantage in IIP patients taken as a whole, but this
association was not tested in the NSIP subset. In the study by VIJ et al. [11], no difference was observed
between IIPs with autoimmune features (with a similar definition to Kinder’s definition of UCTD) and
IPF. It is, however, noteworthy that 80.6% of cases of IIPs with autoimmune features who underwent SLB
displayed an UIP pattern [11]. By contrast with these papers, SUDA et al. [10] demonstrated that UCTD
patients had a significantly lower mortality compared with patients with NSIP not fulfilling the criteria for
UCTD. Similarly, in our study, UCTD patients had a significantly better survival than those with
idiopathic NSIP ( p=0.002) and this survival difference persisted when patients with UCTD and CTD were
pooled ( p=0.0006). Comparable results were observed when the definition of UCTD was restricted to the
criteria proposed by CORTE et al [8]. Univariate analysis revealed that the presence of auto-antibodies as
well as the presence of systemic autoimmune symptoms was significantly correlated with a lower risk of
death. We failed to establish the role of a diagnosis of autoimmune NSIP on multivariate analysis, possibly
because of a link with the response to therapy.
This is the first study to show that the response to therapy is the most robust prognostic predictor. The
response rate of 25.6% observed in idiopathic NSIP was similar to that reported in several previous studies
(between 25% and 33.3%) [32–34], but lower than that estimated by PARK et al. [35] (53%). Importantly,
none of these studies had distinguished UCTD. The response rate observed in this study for patients with
UCTD was comparable to that observed by KINDER et al. [9] (40% versus 38%, respectively). Interestingly,
when patients with UCTD and CTD were pooled, i.e. patients with autoimmune NSIP, they responded
more frequently to therapy than those with other forms of NSIP (43.7% versus 24.5%; p=0.041).
Our group has previously published a study based on the same cohort concerning the histological
findings, in which we clearly demonstrated that NSIP subdivision into histological subgroups was clinically
relevant for prognostic and aetiological purposes [17]. In addition to the widely accepted NSIP criteria,
several histological subgroups could be identified according to superimposed minor histological features,
which were associated with significantly different survivals. NSIP/organising pneumonia overlap was
significantly associated with CTDs and NSIP/cHP overlap with a clinical diagnosis of cHP. Interestingly,
only five out of the 15 patients with a clinical diagnosis of cHP presented a histological pattern of NSIP/
cHP overlap [17]. These two studies, therefore, provide different and complementary information and
support the role of thorough clinical investigation of the underlying cause of NSIP as well as a detailed
examination of SLB.
Our study encompasses several limitations. First, we cannot exclude that some cases with UCTD or cHP
have escaped diagnosis. Nevertheless, in our routine practice, we systematically question patients with ILDs
about the presence of symptoms of CTD and exposures, and almost all patients were evaluated for at least
antinuclear antibody and rheumatoid factor. One of the strengths of our study is the fairly long duration
of follow-up, which decreases the risk of missing patients with late development of CTD. Secondly,
interpretation of survival is limited by the small number of patients with cHP. Thirdly, SLB is rarely
performed in the context of CTD and cHP, creating a potential selection bias. This issue has already been
extensively discussed in CTD. In our patients with identifiable exposures, the decision to perform SLB

8 DOI: 10.1183/09031936.00148613
 INTERSTITIAL LUNG DISEASES | H. NUNES ET AL.

may have been dictated by a particular clinical presentation or disease course, so that our findings may not
be extrapolated to all cHP patients. Lastly, it is recognised that a histological pattern of NSIP and UIP may
coexist in the same patient and, in the presence of interlobar variability, the outcome is that of IPF [36].
As SLBs were performed in various centres using different procedures over a long period of time,
specimens were not taken from two lobes in all of our patients. However, the overall survival of our cohort
is much better than expected for IPF, and 93.6% of our patients presented a pattern of NSIP on HRCT.
In conclusion, the outcome of NSIP is influenced by the underlying aetiology, with a poorer survival for cHP
patients and better survival for patients with autoimmune NSIP. In addition to autoimmune signs and the
recognition of forme fruste variants of CTD and UCTD, NSIP compels the clinician to be vigilant in
questioning patients about environmental or occupational exposures. As the antigenic source frequently
remains undetected and as new causative agents continue to be identified, so-called “idiopathic” NSIP may
simply conceal a forme fruste of cHP. The role of detailed evaluation based on a systematic questionnaire for
home and workplace exposures and a panel of serum precipitins needs to be investigated by further studies.

References
1 American Thoracic Society/European Respiratory Society International Multidisciplinary Consensus Classification
of the Idiopathic Interstitial Pneumonias. This joint statement of the American Thoracic Society (ATS), and the
European Respiratory Society (ERS) was adopted by the ATS board of directors, June 2001 and by the ERS
Executive Committee, June 2001. Am J Respir Crit Care Med 2002; 165: 277–304.
2 Travis WD, Costabel U, Hansell DM, et al. An official American Thoracic Society/European Respiratory Society
statement: update of the international multidisciplinary classification of the idiopathic interstitial pneumonias. Am
J Respir Crit Care Med 2013; 188: 733–748.
3 Fischer A, West SG, Swigris JJ, et al. Connective tissue disease-associated interstitial lung disease: a call for
clarification. Chest 2010; 138: 251–256.
4 Fujita J, Ohtsuki Y, Yoshinouchi T, et al. Idiopathic non-specific interstitial pneumonia: as an “autoimmune
interstitial pneumonia”. Respir Med 2005; 99: 234–240.
5 Kinder BW, Collard HR, Koth L, et al. Idiopathic nonspecific interstitial pneumonia: lung manifestation of
undifferentiated connective tissue disease? Am J Respir Crit Care Med 2007; 176: 691–697.
6 Mosca M, Neri R, Bombardieri S. Undifferentiated connective tissue diseases (UCTD): a review of the literature
and a proposal for preliminary classification criteria. Clin Exp Rheumatol 1999; 17: 615–620.
7 Park JH, Kim DS, Park IN, et al. Prognosis of fibrotic interstitial pneumonia: idiopathic versus collagen vascular
disease-related subtypes. Am J Respir Crit Care Med 2007; 175: 705–711.
8 Corte TJ, Copley SJ, Desai SR, et al. Significance of connective tissue disease features in idiopathic interstitial
pneumonia. Eur Respir J 2012; 39: 661–668.
9 Kinder BW, Shariat C, Collard HR, et al. Undifferentiated connective tissue disease-associated interstitial lung
disease: changes in lung function. Lung 2010; 188: 143–149.
10 Suda T, Kaida Y, Nakamura Y, et al. Acute exacerbation of interstitial pneumonia associated with collagen
vascular diseases. Respir Med 2009; 103: 846–853.
11 Vij R, Noth I, Strek ME. Autoimmune-featured interstitial lung disease: a distinct entity. Chest 2011; 140:
1292–1299.
12 Churg A, Sin DD, Everett D, et al. Pathologic patterns and survival in chronic hypersensitivity pneumonitis. Am J
Surg Pathol 2009; 33: 1765–1770.
13 Gaxiola M, Buendía-Roldán I, Mejía M, et al. Morphologic diversity of chronic pigeon breeder’s disease: clinical
features and survival. Respir Med 2011; 105: 608–614.
14 Lima MS, Coletta EN, Ferreira RG, et al. Subacute and chronic hypersensitivity pneumonitis: histopathological
patterns and survival. Respir Med 2009; 103: 508–515.
15 Ohtani Y, Saiki S, Kitaichi M, et al. Chronic bird fancier’s lung: histopathological and clinical correlation. An
application of the 2002 ATS/ERS consensus classification of the idiopathic interstitial pneumonias. Thorax 2005;
60: 665–671.
16 Vourlekis JS, Schwarz MI, Cool CD, et al. Nonspecific interstitial pneumonitis as the sole histologic expression of
hypersensitivity pneumonitis. Am J Med 2002; 112: 490–493.
17 Kambouchner M, Levy P, Nicholson AG, et al. Prognostic relevance of histological variants in nonspecific
interstitial pneumonia. Histopathology 2014; 65: 549–560.
18 Preliminary criteria for the classification of systemic sclerosis (scleroderma). Subcommittee for scleroderma criteria
of the American Rheumatism Association Diagnostic and Therapeutic Criteria Committee. Arthritis Rheum 1980;
23: 581–590.
19 Alarcón-Segovia D, Villarreal M. Classification and diagnostic criteria for mixed connective tissue disease.
In: Kasukawa R, Sharp GC, eds. Mixed Connective Tissue Disease and Anti-Nuclear Antibodies. Amsterdam,
Elsevier Science, 1987; pp. 33–40.
20 Arnett FC, Edworthy SM, Bloch DA, et al. The American Rheumatism Association 1987 revised criteria for the
classification of rheumatoid arthritis. Arthritis Rheum 1988; 31: 315–324.
21 Bohan A, Peter JB. Polymyositis and dermatomyositis (first of two parts). New Engl J Med 1975; 292: 344–347.
22 Bohan A, Peter JB. Polymyositis and dermatomyositis (second of two parts). New Engl J Med 1975; 292: 403–407.
23 Hochberg MC. Updating the American College of Rheumatology revised criteria for the classification of systemic
lupus erythematosus. Arthritis Rheum 1997; 40: 1725.
24 Vitali C, Bombardieri S, Jonsson R, et al. Classification criteria for Sjögren’s syndrome: a revised version of the
European criteria proposed by the American-European Consensus Group. Ann Rheum Dis 2002; 61: 554–558.
25 Cottin V. Significance of connective tissue diseases features in pulmonary fibrosis. Eur Respir Rev 2013; 22:
273–280.

DOI: 10.1183/09031936.00148613 9
INTERSTITIAL LUNG DISEASES | H. NUNES ET AL.

26 Richerson HB, Bernstein IL, Fink JN, et al. Guidelines for the clinical evaluation of hypersensitivity pneumonitis.
Report of the Subcommittee on Hypersensitivity Pneumonitis. J Allergy Clin Immunol 1989; 84: 839–844.
27 Lacasse Y, Girard M, Cormier Y. Recent advances in hypersensitivity pneumonitis. Chest 2012; 142: 208–217.
28 Sahin H, Brown KK, Curran-Everett D, et al. Chronic hypersensitivity pneumonitis: CT features comparison with
pathologic evidence of fibrosis and survival. Radiology 2007; 244: 591–598.
29 Vourlekis JS, Schwarz MI, Cherniack RM, et al. The effect of pulmonary fibrosis on survival in patients with
hypersensitivity pneumonitis. Am J Med 2004; 116: 662–668.
30 Romagnoli M, Nannini C, Piciucchi S, et al. Idiopathic nonspecific interstitial pneumonia: an interstitial lung
disease associated with autoimmune disorders? Eur Respir J 2011; 38: 384–391.
31 Bouros D, Wells AU, Nicholson AG, et al. Histopathologic subsets of fibrosing alveolitis in patients with systemic
sclerosis and their relationship to outcome. Am J Respir Crit Care Med 2002; 165: 1581–1586.
32 Nicholson AG, Colby TV, du Bois RM, et al. The prognostic significance of the histologic pattern of interstitial
pneumonia in patients presenting with the clinical entity of cryptogenic fibrosing alveolitis. Am J Respir Crit Care
Med 2000; 162: 2213–2217.
33 Daniil ZD, Gilchrist FC, Nicholson AG, et al. A histologic pattern of nonspecific interstitial pneumonia is
associated with a better prognosis than usual interstitial pneumonia in patients with cryptogenic fibrosing
alveolitis. Am J Respir Crit Care Med 1999; 160: 899–905.
34 Kondoh Y, Taniguchi H, Yokoi T, et al. Cyclophosphamide and low-dose prednisolone in idiopathic pulmonary
fibrosis and fibrosing nonspecific interstitial pneumonia. Eur Respir J 2005; 25: 528–533.
35 Park IN, Jegal Y, Kim DS, et al. Clinical course and lung function change of idiopathic nonspecific interstitial
pneumonia. Eur Respir J 2009; 33: 68–76.
36 Flaherty KR, Travis WD, Colby TV, et al. Histopathologic variability in usual and nonspecific interstitial
pneumonias. Am J Respir Crit Care Med 2001; 164: 1722–1727.

10 DOI: 10.1183/09031936.00148613