Lung (2014) 192:141–149 DOI 10.1007/s00408-013-9530-0
Staging of Acute Exacerbation in Patients with Idiopathic Pulmonary Fibrosis Tomoo Kishaba • Hitoshi Tamaki • Yousuke Shimaoka Hajime Fukuyama • Shin Yamashiro
•
Received: 27 February 2013 / Accepted: 22 October 2013 / Published online: 13 November 2013 Ó Springer Science+Business Media New York 2013
Abstract Background The purpose of this study was to evaluate the predictors of a 3-month mortality rate of acute exacerbation of idiopathic pulmonary fibrosis (IPF) and provide a new staging system. Methods A total of 594 patients with IPF were included in this retrospective, observational study conducted from January 2001 to December 2010 at Okinawa Chubu Hospital. Results Among the 594 patients, 58 (9.8 %) developed acute exacerbation (AE) of IPF during the 10-year observation period. The median follow-up period for AE was 10.4 months. In-hospital mortality was 56.9 % and the 3-month mortality rate was 63.8 %. We identified the following four parameters in a multivariate analysis as: serum lactate dehydrogenase, sialylated carbohydrate antigen (KL-6), ratio of partial pressure of oxygen and fraction of inspiratory oxygen, and total extent of abnormal findings on high-resolution computed tomography of the chest. Patients were divided into two groups on the basis of the four composite parameters. Patients in the extensive disease-stage group required more mechanical ventilation and intensive therapy than those in the limited disease-stage group. The 3-month mortality rate was higher in patients in the extensive disease-stage group than in patients in the limited disease-stage group (80.6 vs. 54.5 %, respectively; p = 0.007).
Conclusions Staging of AE in patients with IPF provided useful information regarding disease severity and shortterm outcome. Keywords Acute exacerbation Extensive Idiopathic pulmonary fibrosis Limited Staging
Introduction Idiopathic pulmonary fibrosis (IPF) is a progressive interstitial lung disease (ILD) of unknown etiology [1]. The clinical course of individual patients is highly variable [2, 3]. Acute exacerbation (AE) of IPF, which is a sudden deterioration of respiratory condition superimposed on chronic fibrotic lung disease, can sometimes develop during the clinical course [4, 5]. Collard et al. [6] published diagnostic criteria for AE of IPF, and Song et al. [7] have reported the incidence, risk factors, and outcomes of AE in patients with IPF. In addition, Kondoh et al. [8] have described the risk factors for AE. However, variation in the severity of AE in patients with IPF remains unknown. Therefore, the purpose of this study was to use composite parameters to assess the severity of AE in patients with IPF before mechanical ventilation (MV) and provide a new staging system for predicting the 3-month mortality rate.
Materials and Methods T. Kishaba (&) Y. Shimaoka H. Fukuyama S. Yamashiro Department of Respiratory Medicine, Okinawa Chubu Hospital, Okinawa, Japan e-mail: [email protected] H. Tamaki Sunagawa Naika Clinic, Okinawa, Japan
Study Subjects This was a retrospective and observational study. The study population comprised of 594 patients with IPF diagnosed according to the American Thoracic Society (ATS)/European
123
142
Respiratory Society (ERS) consensus statement [1] at Okinawa Chubu Hospital, Okinawa, Japan from January 2001 to December 2010. Data Collection We collected medical records of the patients during the same period. Patients with secondary ILD, such as connective-tissue disease, due to drug, environmental, or occupational factors, were excluded. Baseline variables, including medical treatment, were evaluated at the time of IPF diagnosis. Among the remaining patients, 58 developed AE. In this study, AE was defined as a sudden aggravation of dyspnea within 30 days with new bilateral infiltration accompanying known IPF or evidence of honeycombing on high-resolution computed tomography (HRCT) of the chest [9–11]. Five patients underwent surgical lung biopsy and demonstrated pathological findings of usual interstitial pneumonia (UIP). All 58 patients were diagnosed on the basis of a multidisciplinary discussion. Three patients developed two episodes of AE. Therefore, only the first episodes were included in the analysis. All laboratory data were collected on the date of admission due to AE. We performed bronchoalveolar lavage (BAL) in all patients with IPF to exclude secondary causes such as pneumonia, sepsis, and aspiration and to diagnose the cause of AE [12]. BAL cultures of all patients revealed \103 colony-forming units. CT angiogram was negative and an echocardiogram showed no evidence of systolic or diastolic heart failure, which excluded pulmonary embolism and congestive heart failure. Modified Medical Research Council (mMRC) dyspnea scale [13] and dyspnea duration defined as the interval between symptom onset and HRCT date also were reviewed. In addition, we reviewed laboratory findings and baseline pulmonary function test results. The median follow-up period was defined as the period from the date of diagnosis of AE in patients with IPF until death or final follow-up.
Lung (2014) 192:141–149
the area above the main carina, the lower was defined as the area between the area of the lung below the level of the inferior pulmonary vein and the level immediately above the right hemi-diaphragm, and the middle was defined as the area between the two according to the protocol of Sumikawa et al. [17]. Both chest physicians and radiologists read whole images and reached a consensus through multidisciplinary discussion. The extent of traction bronchiectasis, ground-glass opacity, consolidation, and honeycombing was defined as follows: 0, none; 1, involving 1–25 %; 2, involving 26–50 %; and 3, involving ]50 %. Among these four abnormal findings, we considered that both ground-glass opacity and consolidation contributed to the severity of AE in patients with IPF. Therefore, each zone’s radiological score of both groundglass opacity and consolidation were summed and was defined as the total extension score. Therefore, the minimum score was zero and the highest was 36. Calculation of Staging Points Receiver operating characteristic (ROC) curve analysis was used to define the staging points, which were later summed to determine staging. We defined the points used to determine the staging as follows: 0, serum lactate dehydrogenase (LDH) level \280, and 1, serum LDH level ]280; 0, serum sialylated carbohydrate antigen (KL-6) level\1,000, and 1, KL-6 level ]1,000; 0, partial pressure of oxygen and fraction of inspiratory oxygen (P/F) ratio ]100, and 1, P/F ratio \100; 0,overall HRCT score \20, and 1, overall HRCT score ]20. These values were obtained at the time of AE diagnosis. Each parameter predicted the 3-month mortality rate. However, a combination of these parameters more actively predicted the 3-month mortality rate with the Cox proportion hazard model and c-statistic. On the basis of this information, we defined staging based on the total score obtained as follows: limited disease stage 0–2 and extensive disease stage ]3.
Radiological Evaluation Treatment Protocol With patients in the supine position, thin-section CT scans were obtained at the end-inspiratory phase. Images were obtained at window settings appropriate for viewing the lung parenchyma. We evaluated the presence, distribution, and extent of ground-glass opacity, consolidation, traction bronchiectasis, and honeycombing on admission [14–16]. In addition, we compared HRCT findings during the acute phase with that during the stable phase for the diagnosis of IPF. Lungs were divided into six zones (upper, middle, and lower of both sides). The upper lung zone was defined as
123
Our standardized uniformed treatment protocol was as follows: we always initiated methylprednisolone (mPSL) pulse therapy (1 g/day for 3 consecutive days) when we treated AE in patients with IPF and maintained PSL monotherapy for patients responding to steroid pulse therapy. However, if a patient showed poor response after the first cycle of steroid pulse therapy, we initiated immunosuppressants, such as cyclosporine A or azathioprine, and repeated the steroid pulse therapy weekly for a maximum of 4 weeks.
Lung (2014) 192:141–149
Mortality In-hospital mortality was defined as death during admission. AE onset was defined as that on the date of admission. The 3-month mortality rate was defined as death within 3 months from AE onset. The Ethics Committee of Okinawa Chubu Hospital approved the study protocol. Informed consent was waived because this study was a retrospective and anonymous review of medical records. Our Ethics Committee did not require informed consent, because this study was epidemiological rather than a prospective interventional clinical study.
143
characteristics of all patients with IPF are shown in Table 1. We evaluated another cohort of patients with AE of IPF for validation. The baseline clinical parameters and prior treatments of these two cohorts are shown in Table 2. Median survival rate was 9.3 months. Laboratory Findings Mean serum white blood cell count and levels of C-reactive protein, LDH, and KL-6 were 11,357 mm3, 8.4 mg/dL, 396 IU/L, and 1,889 IU/L, respectively. All patients required oxygen, and the mean ratio of partial pressure of oxygen and fraction of inspiratory oxygen concentration (P/F ratio) was 158 (Table 3).
Statistical Analysis Physiological Parameters Continuous variables are presented as means ± standard deviations, and categorical variables are presented as percentages. The v2 and Fisher’s exact tests were used for categorical data, and the unpaired t test and Mann–Whitney U test were used for continuous data. The j statistic was used to analyze the interobserver variations in the findings. Cox regression analysis was used to identify significant variables capable of predicting AE severity or those that could act as prognostic factors. Model discrimination was quantified based on C statistic, which is the probability that among two randomly selected patients the patient with the highest predicted risk of an event will be the first to experience the event. C-statistic value ranges from 0.5 (model discrimination is no better than chance) to 1 (model discrimination is perfect). C-statistic values ranging between 0.7 and 0.8 is typically considered ‘‘acceptable,’’ whereas a value [0.8 is typically considered ‘‘excellent.’’ The Kaplan–Meier survival curve and the log-rank test were used to evaluate survival. The level of statistical significance was set at P \ 0.05. STATA software V.11.0 (Stata Corp., College Station, TX, USA) was used to perform all clinical data analyses.
The mean baseline forced vital capacity (FVC), percent predicted FVC, and percent predicted diffusion of carbon mono oxide (DLco) values were 1.63 L, 58.2 %, and 38.3 %, respectively. MV and Hospital Stay Half of the patients required MV, and the length of MV was 8.5 ± 12.9 days. The length of hospital stay was 21.1 ± 13.7 days. Table 1 Baseline characteristics of patients without AE of IPF, and patients with AE of IPF Without AE of IPF (n = 536)
AE of IPF (n = 58)
P value
Age (year)
72 ± 9.2
75.0 ± 9.6
0.738
Male/female
386/150
38/20
0.704
Pack-year Smoking status (active/ex/never)
42.4 ± 40.9 (232/272/32)
35.4 ± 41.6 (22/36/0)
0.692 0.541
mMRC
2.4 ± 1.0
2.8 ± 1
0.428
Dyspnea duration (months)
12.8 ± 9.6
6.7 ± 6
0.036
Results
Percent predicted FVC (%)
69.5 ± 24.6
58.2 ± 21.2
0.025
Clinical Characteristics
Percent predicted DLco (%)
61.3 ± 22.8
52.3 ± 19.7
0.042
We identified 594 patients with IPF during the 10-year study period. Among these patients, 58 developed AE. Mean age of all patients with AE was 75 years, and 66 % were men. The median follow-up period from the date of diagnosis of AE of IPF until death or the final follow-up was 10.2 (range 0.1–112) months. Among patients with AE of IPF, smokers approximately had a 35 pack-year history. The mean mMRC dyspnea scale score was 2.8, and the mean dyspnea duration was 6.7 days. Clinical
Anticoagulants use
41
4
0.751
Prednisolone use
38
4
0.683
Immunosuppressive therapy
16
2
0.582
Anti proliferative therapy
2
0
0.864
PCP-prophylaxis
28
3
0.795
mMRC modified Medical Research Council dyspnea scale, FVC forced vital capacity, DLco diffusion of carbon mono oxide, PCP Pneumocystis jirovecii pneumonia
123
144
Lung (2014) 192:141–149
Table 2 Baseline characteristics of patients with an AE of IPF between different periods AE of IPF (n = 52) 1995–2000
AE of IPF (n = 58) 2001–2010
P value 0.868
Age (year)
72.0 ± 8.2
75.0 ± 9.6
Male/female
36/16
38/20
0.796
Pack-year
42.8 ± 33.2
35.4 ± 41.6
0.692
Smoking status (active/ex)
24/28
22/36
0.573
mMRC
3.0 ± 1.4
2.8 ± 1.0
0.726
Dyspnea duration (days)
10.2 ± 5.8
6.7 ± 6.0
0.462
Percent predicted FVC (%)
62.6 ± 29.5
58.2 ± 21.2
0.628
Percent predicted DLco (%)
51.9 ± 24.6
52.3 ± 19.7
0.795
LDH (IU/L)
326.8 ± 128.7
396.4 ± 147.8
0.483
KL-6 (IU/L)
1,581.6 ± 1,225.3
1,889.2 ± 1,735.6
0.693
P/F Ground-glass opacity ? consolidation score
166.2 ± 79.5 21.9 ± 0.8
158.4 ± 94.3 23.2 ± 0.7
0.759 0.695
3-month mortality rate (%)
75.0
70.7
0.685
Survival time (months)
7.6 ± 28.2
9.3 ± 24.6
0.824
Anticoagulants use
3
4
0.826
Prednisolone use
6
4
0.496
Immunosuppressive therapy
3
2
0.672
Antiproliferative therapy
0
0
0.998
PCP prophylaxis
4
3
0.704
mMRC modified Medical Research Council dyspnea scale, FVC forced vital capacity, DLco diffusion of carbon mono oxide, KL-6 sialylated carbohydrate antigen, P/F ratio ratio of partial pressure of oxygen and fraction of inspiratory oxygen, PCP Pneumocystis jirovecii pneumonia
Table 3 Clinical data on the date of admission due to AE cohort AE in IPF (n = 58) Age (year)
75 ± 9.6
WBC (mm3)
11,356.9 ± 5,544
CRP (mg/dl)
8.4 ± 7.5
LDH (IU/L)
396.4 ± 147.8
KL-6 (IU/L)
1,889.2 ± 1,735.6
pH
7.42 ± 0.1
PaO2 (mmHg)
62.1 ± 17.6
PaCO2 (mmHg)
43.5 ± 16.3
P/F
158.4 ± 94.3
MV (%)
50
Length of MV (days) Duration of hospital stay (days)
8.5 ± 12.9 21.1 ± 13.7
Survival time (months)
9.3 ± 24.6
LDH lactate dehydrogenase, KL-6 sialylated carbohydrate antigen, P/F ratio ratio of partial pressure of oxygen and fraction of inspiratory oxygen concentration
AE staging Univariate analysis was performed to assess the useful parameters for predicting the 3-month mortality rate of AE (Table 4). Results of the multivariate analysis are shown in Table 5. Serum LDH, and KL-6 levels, P/F ratio, and total
123
extent of abnormal findings on HRCT of the chest score were identified as predictors of the 3-month mortality rate. The j value of four chest physicians and two radiologists was 0.64. We assigned staging levels of 0–3 for traction bronchiectasis, honeycombing, ground-glass opacity, and consolidation. Among them, both ground-glass opacity and consolidation scores were closely associated for predicting the 3-month mortality rate (r = 0.846). Therefore, we used the overall score of these two HRCT findings for staging. The ROC curve values for serum LDH level, KL-6 level, P/F ratio, and total extent of abnormal findings on HRCT of the chest score were 0.618, 0.626, 0.658, and 0.72, respectively, and the cutoff values were 280, 1,000, 100, and 20, respectively. We assigned 0 or 1 for serum LDH level, serum KL-6 level, P/F ratio, and total extent of abnormal findings on HRCT of the chest score (Table 6). The Cox-proportional hazard model was used to investigate useful ROC-curve cutoff points for staging. The four parameters had different hazard ratios (HRs) for predicting the 3-month mortality rate: each parameter predicted the 3-month mortality rate. However, a combination of these parameters more actively predicted the 3-month mortality rate using the Cox proportion hazard model. In addition, we evaluated the discrimination rate for different clinical models using C statistic (Table 7). Therefore, we defined staging on the
Lung (2014) 192:141–149
145
Table 4 Univariate analysis of 3-month mortality rate due to AE
Table 6 Point assignment for AE staging
HR
P value
Definition
Point
Age
0.997
0.833
LDH
Gender
0.898
0.76
\280
0
Pack-year
1.009
0.027
]280
1
mMRC
1.161
0.386
Dyspnea duration
1.013
0.649
\1,000
0
Finger clubbing
1.45
0.299
]1,000
1
WBC
1
0.465
P/F ratio
LDH
1.003
0.009
]100
0
KL-6
2.01
0.001
\100
1
CRP
0.984
0.47
P/F Traction bronchiectasis ? honeycombing score
1.999 1.324
0.048 0.061
Ground-glass opacity ? consolidation score
1.853
0.033
BALF neutrophil
0.944
0.327
BALF lymphocyte
1.002
0.974
PSL alone
0.436
0.033
PSL ? immunosuppressants
2.294
0.033
MV
1.921
0.026
mMRC modified Medical Research Council dyspnea scale, LDH lactate dehydrogenase, KL-6 sialylated carbohydrate antigen KL-6, P/ F ratio ratio of partial pressure of oxygen and fraction of inspiratory oxygen concentration, BALF bronchoalveolar lavage fluid, PSL prednisolone
KL-6
Ground-glass opacity ? consolidation score \20 ]20
LDH lactate dehydrogenase, KL-6 sialylated carbohydrate antigen KL-6, P/F ratio ratio of partial pressure of oxygen and fraction of inspiratory oxygen concentration
Table 7 Different clinical model specifications Model
Independent variables
C-statistic (95 % CI)
Comprehensive
LDH, KL-6, P/F, ground-glass opacity ? consolidation score
0.76 (0.7–0.82)
Model 1
LDH
0.56 (0.52–0.6)
Model 2 Model 3
KL-6 P/F
0.62 (0.580.66) 0.6 (0.56–0.64)
Model 4
Ground-glass opacity ? consolidation score
0.66 (0.62–0.7)
Model 5
LDH and P/F
0.62 (0.58–0.66)
Model 6
LDH, ground-glass opacity ? consolidation score
0.68 (0.62–0.72) 0.66 (0.62–0.7)
Table 5 Multivariate analysis of 3-month mortality rate due to AE HR
P value
0 1
LDH KL-6
2.024 2.909
0.047 0.038
Model 7
KL-6 and P/F
P/F
2.42
0.041
Model 8
Ground-glass opacity ? consolidation score
2.289
0.030
KL-6, ground-glass opacity ? consolidation score
LDH lactate dehydrogenase, KL-6 sialylated carbohydrate antigen KL-6, P/F ratio ratio of partial pressure of oxygen and fraction of inspiratory oxygen concentration
0.7 (0.66–0.74)
LDH lactate dehydrogenase, KL-6 sialylated carbohydrate antigen KL-6, P/F ratio ratio of partial pressure of oxygen and fraction of inspiratory oxygen concentration
Table 8 Staging system for patients with AE of IPF patients
basis of the four clinical parameter scores obtained as follows: limited stage 0–2, and extensive stage ]3 (Table 8). Clinical characteristics of each stage are shown in Table 9. Base line treatment and HRCT score for each stage were similar. The extensive disease stage group had higher KL-6 levels and HRCT scores and required more MV than the limited disease-stage group (2,303 vs. 1,164, 8.4 vs. 5.6, and 61 vs. 32 %, respectively). Among all patients, 29 (50 %) required MV, and only 6 patients survived. Fifteen patients underwent MV 48 h after admission. Representative images of the two groups are shown in Figs. 1 and 2.
Points Limited exacerbation (n = 22)
0–2
Extensive exacerbation (n = 36)
]3
Treatment Response and Prognosis In the limited disease stage group, 68 % patients received PSL alone. Moreover, 69 % of patients in the extensive disease stage group received combination therapy, such as PSL and another immunosuppressant. We did not use
123
146
Lung (2014) 192:141–149
Table 9 Clinical characteristics of patients in each stage Limited disease (n = 22)
Extensive disease (n = 36)
P value
Age
74.68 ± 11.13
75.17 ± 8.64
0.93
Pack-year
26.95 ± 31.66
40.56 ± 46.28
0.425
Smoking status (active/ex/never)
(10/17/0)
(12/19/0)
0.486
mMRC
2.55 ± 1.06
3.03 ± 0.94
0.049
Dyspnea duration (months)
5.59 ± 4.92
7.33 ± 6.53
0.274
WBC (mm3)
12,259 ± 6,887
10,806 ± 4,558
0.496
CRP (mg/dl)
11.15 ± 9.19
6.85 ± 5.88
0.141
LDH (IU/L)
372.10 ± 122.66
410.5 ± 160.59
0.519
KL-6 (IU/L)
1,164.35 ± 838.60
2,303.46 ± 1976.22
0.0007
PaO2 (mmHg)
56.62 ± 16.01
65.44 ± 17.84
0.041
P/F ratio Traction bronchiectasis ? honeycombing score
170.58 ± 104.72 19
138.55 ± 72.09 25
0.309 \0.0001
Ground-glass opacity ? consolidation score
17
27
\0.0001
MV (%)
31.8
61.1
0.032
PSL alone (%)
68.2
30.6
0.006
Combination therapy (%)
31.8
69.4
0.006
Length of hospital stay (days)
24.36 ± 13.98
19.17 ± 13.31
0.123
90-day mortality rate (%)
54.5
80.6
0.007
Baseline traction bronchiectasis ? honeycombing score
15
17
0.491
Baseline ground-glass opacity ? consolidation score
6
5
0.726
Baseline anticoagulants use
2
2
0.374
Baseline prednisolone use
3
1
0.286
Baseline immunosuppressive therapy
1
1
0.793
Baseline antiproliferative therapy
0
0
0.902
Baseline PCP-prophylaxis
2
1
0.684
mMRC modified Medical Research Council dyspnea scale, LDH lactate dehydrogenase, KL-6 sialylated carbohydrate antigen KL-6, P/F ratio ratio of partial pressure of oxygen and fraction of inspiratory oxygen concentration, PCP Pneumocystis jirovecii pneumonia
Fig. 1 Representative image of the limited disease stage
extracorporeal membrane oxygenation or lung transplantation. The 3-month mortality rate was worse in the extensive disease-stage group than in the limited diseasestage group (80.6 vs. 54.5 %; p = 0.007; Fig. 3). We evaluated another cohort of patients with AE of IPF,
123
Fig. 2 Representative image of the extensive disease stage
comprising 52 patients from January 1995 to December 2000. Our staging was applied to these patients and showed a similar result (82.9 vs. 58.8 %; p = 0.005).
Lung (2014) 192:141–149
Fig. 3 3-month survival curve according to disease stage. The extensive disease stage group showed high mortality compared with that of the limited disease stage group
Discussion We found that 58 of 594 patients with IPF developed AE during the 10-year study period. Song et al. [7] reported 1- and 3-year mortality rate of 14.2 and 20.7 %, respectively. The mean age of the patients who developed AE was 75 years, and 66 % of them were men. Many reports on the cause of death or occurrence of AE in patients with IPF have shown that older age, AE, decreased functional impairment, and immunosuppressive therapy are significant predictors of poor overall prognosis [5, 18]. Although many reports are available regarding the incidence and cause of AE in patients with IPF, limited studies are available on AE severity in patients with IPF. Song et al. reported that the immediate outcome was poor and that median survival time of patients was 2.2 months. In addition, these patients usually develop severe respiratory failure and required MV. The in-hospital mortality rate was 95.7 % in 22 of 23 patients [19]. Pe´rez et al. showed that high positive end-expiratory pressure (PEEP) and Acute Physiology and Chronic Health Evaluation III scores predicted mortality. In addition, low P/F ratio is an independent determinant of survival in patients with ILD [20]. Among our cohort of patients with AE of IPF, 29 (50 %) required MV, and a lung protective strategy for profound hypoxemic patients was used. However, MV provided little benefit for AE of IPF. Saydain et al. [21] reported that the actual in-hospital mortality rate was 61 and 92 % of the hospital survivors died at a median of 2 months after discharge. MV itself may constitute one of the most important outcome determinants in patients with IPF. Based on these findings, we recognized the importance of staging for determining AE mortality in patients with IPF. Other possibilities of acute respiratory failure, such as infection, heart failure, pneumothorax, and pulmonary embolism, should be excluded for AE diagnosis. Thus, we
147
performed BAL in all patients. We believe the main purpose of performing BAL is to exclude infection. In addition, a detailed history, physical examination, medication review, environmental exposure, and family history are very important [1]. We used univariate and multivariate analyses to predict the 3-month mortality rate and the Cox proportional hazards model to investigate useful variables. Percent predicted FVC and DLco are crucial parameters for predicting IPF mortality [22, 23]. Pulmonary function may be important for predicting prognosis; however, many patients cannot perform the test in an AE state because of severe dyspnea. We identified the following parameters that affected mortality and severity of AE in patients with IPF: serum LDH level, KL-6 level, P/F ratio, and total extent of abnormal findings on HRCT of the chest score. MV also was an important factor. However, in our cohort, 15 of 29 patients (52 %) required MV after 48 h after admission. Our main purpose was immediate severity evaluation. In addition, the indication of MV includes subjective elements and is an ethical issue rather than only conducting a blood test and evaluating HRCT findings. Therefore, we considered that it was unreasonable to add this to the staging system. Akira et al. [24] showed that the HRCT pattern and overall CT extent are associated with mortality; they divided patients into the following three patterns: peripheral, multifocal, and diffuse. The strongest correlations were observed between CT patterns and survival (odds ratio 4.629; 95 % confidence interval (CI) 1.9–11.278; P \ 0.001). Sumikawa et al. reported that traction bronchiectasis and fibrosis scores are significant predictors of outcome in patients with pathological UIP patients (HR 1.3 and 1.1, respectively; 95 % CI 1.18–14.2 and 1.03–1.19, respectively) [17]. We set our HRCT findings scores based on these reports. We chose four useful parameters for staging after multivariate analysis. Each parameter predicted the 3-month mortality rate but individual HRs differed. Among them, HRCT scores were variable. Traction bronchiectasis and honeycombing signify a chronic process and are useful predictors of long-term mortality because of IPF. Fibrotic findings were not different between the two groups at baseline. In contrast, we consider that both ground-glass opacity and consolidation usually indicate new findings that contribute to increased severity of AE in patients with IPF. In addition, a minority of patients received, PSL, immunosuppressants, and anticoagulants before AE. Anticoagulants were used to treat atrial fibrillation not IPF. Thus, we thought that history of prior treatment had little effect on severity and HRCT findings of AE. Therefore, we chose a mixed score of these two HRCT findings for creating the 3-month mortality rate staging system. The most
123
148
accurate and discriminating HRCT score value was 20 in the ROC curve analysis. We thought that a simple staging system would be practical and easy to use. Therefore, we assigned only two points for each parameter based on the ROC curve analysis. We also investigated the difference in treatment strategy for staging. There have been many reports on high-dose corticosteroids for treating AE of IPF [25–27]. We always initiate methylprednisolone pulse therapy (1 g/day for 3 consecutive days) when we treat AE of IPF and maintain PSL monotherapy for patients responding to steroid pulse therapy, such as those with limited-disease stage. Patients in the extensive disease stage group received combination therapy, such as PSL with immunosuppressants. We usually added an immunosuppressant such as cyclosporine A or azathioprine based on previous reports [28, 29]. In addition, we repeated methylprednisolone pulse therapy every week for either 2 or 4 weeks. If these patients survived, we maintained low-dose PSL and a cytotoxic agent. Finally for non-responders, such as patients in the extensive disease stage group, we repeated steroid pulse therapy and high-dose cyclophosphamide or another immunosuppressant. However, these patients often were entirely resistant to such intensive therapy and showed progressive respiratory failure. Therefore, supportive care may be the mainstay approach for this patient sub group. We described the variable clinical courses of patients with AE of IPF in our cohort. We plan to use a different management strategy in the future based on our staging system. Our study had several limitations. First, this was a retrospective and observational study. Second, the long study period (10 years) in combination with the relatively small sample size and the sophisticated statistical approach was an important limitation of our study. Third, most of our patients with IPF were diagnosed according to clinical criteria. Only five patients underwent surgical lung biopsy. Therefore, patients with AE and pathological UIP may show a different presentation. However, the clinical characteristics, physiological impairment, and clinical course of our cohorts were entirely compatible with IPF. Therefore, misclassification of other patients with ILD was not high. Fourth, most patients received broadspectrum antibiotics: however, we evaluated their sputum smear and performed BAL for all patients. The BALF culture showed no evidence of active infection, including clinical parameters, such as C-reactive protein. Therefore, we excluded pneumonia. Finally, this was a single-center study and the number of patients with AE patients in our cohort was small. Therefore, the results are not generalizable to all patients with AE of IPF. However, the clinical presentation and HRCT findings in our patients were comparable to those of previous reports. In addition, we
123
Lung (2014) 192:141–149
evaluated another cohort during another period with the same staging system and observed similar results. Therefore, our staging system is valid and applicable.
Conclusions We propose a new staging system for patients with AE of IPF. Patients in each stage showed different clinical characteristics and short-term outcomes. Different management strategies will be required to rescue patients with AE of IPF patients according to our staging system. Our composite staging system, including clinical parameters and HRCT findings, provides useful information for physicians. A multicenter study is warranted to validate our results. Acknowledgments The authors thank all of our residents and Drs. Yasutani and Takara for their interpretation of the medical records and imaging results. Conflict of interest
All the authors declare none.
References 1. American Thoracic Society, European Respiratory Society (2000) Idiopathic pulmonary fibrosis: diagnosis and treatment. International consensus statement. Am J Respir Crit Care Med 161:646–664 2. Kim DS, Collard HR, King TE Jr (2006) Classification and natural history of the idiopathic interstitial pneumonias. Proc Am Thorac Soc 3:285–292 3. Ley B, Collard HR, King TE Jr (2011) Clinical course and prediction of survival in idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 183:431–440 4. Kondoh Y, Taniguchi H, Kawabata Y, Yokoi T, Suzuki K, Takagi K (1993) Acute exacerbation in idiopathic pulmonary fibrosis. Analysis of clinical and pathologic findings in three cases. Chest 103(6):1808–1812 5. Panos RJ, Mortenson RL, Niccoli SA et al (1990) Clinical deterioration in patients with idiopathic pulmonary fibrosis: causes and assessment. Am J Med 88:396–404 6. Collard HR, Moore BB, Flaherty KR et al (2007) Acute exacerbations of idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 176:636–643 7. Song JW, Hong S-B, Lim C-M, Koh Y, Kim DS (2011) Acute exacerbation of idiopathic pulmonary fibrosis: incidence, risk factors and outcome. Eur Respir J 37:356–363 8. Kondoh Y, Taniguchi H, Katsuta T, Kataoka K, Kimura T, Nishiyama O, Sakamoto K, Johkoh T, Nishimura M, Ono K, Kitaichi M (2010) Risk factors of acute exacerbation of idiopathic pulmonary fibrosis. Sarcoidosis Vasc Diffuse Lung Dis 27(2):103–110 9. Nishimura K, Kitaichi M, Izumi T, Nagai S, Kanaoka M, Itoh H (1992) Usual interstitial pneumonia: histologic correlation with high-resolution CT. Radiology 182:337–342 10. Johkoh T, Muller NL, Cartier Y, Kavanagh PV, Hartman TE, Akira M, Ichikado K, Ando M, Nakamura H (1999) Idiopathic interstitial pneumonias: diagnostic accuracy of thin-section CT in 129 patients. Radiology 211:555–560
Lung (2014) 192:141–149 11. Hansell DM, Bankier AA, Macmahon H, McLoud TC, Muller NL, Remy J (2008) Fleischner society: glossary of terms for thoracic imaging. Radiology 246:697–722 12. Meyer KC, Raghu G, Baughman RP, Brown KK, Costabel U, du Bois RM, Drent M, Haslam PL, Kim DS, Nagai S, Rottoli P, Saltini C, Selman M, Strange C, Wood B (2012) An Official American Thoracic Society Clinical Practice Guideline: the clinical utility of bronchoalveolar lavage cellular analysis in interstitial lung disease. Am J Respir Crit Care Med 185:1004–1014 13. Bestall JC, Paul EA, Garrod R, Garnham R, Jones PW, Wedzicha JA (1999) Usefulness of the Medical Research Council (MRC) dyspnoea scale as a measure of disability in patients with chronic obstructive pulmonary disease. Thorax 54:581–586 14. Akira M, Hamada H, Sakatani M, Kobayashi C, Nishioka M, Yamamoto S (1997) CT findings during phase of accelerated deterioration in patients with idiopathic pulmonary fibrosis. Am J Roentgenol 168:79–83 15. Kim DS, Park JH, Park BK, Lee JS, Nicholson AG, Colby T (2006) Acute exacerbation of idiopathic pulmonary fibrosis: frequency and clinical features. Eur Respir J 27:143–150 16. Akira M, Sakatani M, Ueda E (1993) Idiopathic pulmonary fibrosis: progression of honeycombing at thin-section CT. Radiology 189:687–691 17. Sumikawa H, Johkoh T, Colby TV, Ichikado K, Suga M, Taniguchi H, Kondoh Y, Ogura T, Arakawa H, Fujimoto K, Inoue A, Mihara N, Honda O, Tomiyama N, Nakamura H, Mu¨ller ML (2008) Computed tomography findings in pathological usual interstitial pneumonia. Relationship to survival. Am J Respir Crit Care Med 177:433–439 18. Nagai S, Kitaichi M, Hamada Knagao T, Hoshino Y, Miki H, Izumi T (1999) Hospital-based historical cohort study of 234 histologically proven Japanese patients with IPF. Sarcoidosis Vasc Diffuse Lung Dis 16:209–214 19. Stern J-B, Mal H, Groussard O, Brugie`re O, Marceau A, Jebrak G, Fournier M (2001) Prognosis of patients with advanced idiopathic pulmonary fibrosis requiring mechanical ventilation for acute respiratory failure. Chest 120(1):213–219 20. Fernandez-Perez ER, Yilmaz M, Jenad H, Daniels CE, Ryu JH, Hubmayr RD, Gajic O (2008) Ventilator settings and outcome of
149
21.
22.
23.
24.
25.
26.
27.
28.
29.
respiratory failure in chronic interstitial lung disease. Chest 133(5):1113–1119 Saydain G, Islam A, Afessa B, Ryu JH, Scott JP, Peters SG (2002) Outcome of patients with idiopathic pulmonary fibrosis admitted to the intensive care unit. Am J Respir Crit Care Med 166:839–842 Wells AU, Desai SR, Rubens MB, Goh NS, Cramer D, Nicholson AG, Colby TV, du Bois RM, Hansell DM (2003) Idiopathic pulmonary fibrosis: a composite physiologic index derived from disease extent observed by computed tomography. Am J Respir Crit Care Med 167(7):962–969 du Bois RM, Weycker D, Albera C, Bradford WZ, Costabel U, Kartashov A, King TE Jr, Lancaster L, Noble PW, Sahn SA, Thomeer M, Valeyre D, Wells AU (2011) Forced vital capacity in patients with idiopathic pulmonary fibrosis. Test properties and minimal clinically important difference. Am J Respir Crit Care Med 184:1382–1389 Akira M, Kozuka T, Yamamoto S, Sakatani M (2008) Computed tomography findings in acute exacerbation of idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 178:372–378 Kondoh Y, Taniguchi H, Yokoi T, Nishiyama O, Ohishi T, Kato T, Suzuki K, Suzuki R (2005) Cyclophosphamide and low-dose prednisolone in idiopathic pulmonary fibrosis and fibrosing nonspecific interstitial pneumonia. Eur Respir J 25:528–533 Parambil JG, Myers JL, Ryu JH (2005) Histopathologic features and outcome of patients with acute exacerbation of idiopathic pulmonary fibrosis undergoing surgical lung biopsy. Chest 128:3310–3315 Kondo A, Saiki S (1989) Acute exacerbation in idiopathic interstitial pneumonia (IIP). In: Harasawa M, Fukuchi Y, Morinari H (eds) Interstitial pneumonia of unknown etiology. University of Tokyo Press, Tokyo, pp 33–42 Homma S, Sakamoto S, Kawabata M, Kishi K, Tsuboi E, Motoi N, Yoshimura K (2005) Cyclosporin treatment in steroid-resistant and acutely exacerbated interstitial pneumonia. Intern Med 44:1144–1150 Pereira CAC, Malheiros T, Coletta EM, Ferreira RG, Rubin AS, Otta JS, Rocha NS (2006) Survival in idiopathic pulmonary fibrosis cytotoxic agents compared to corticosteroids. Respir Med 100:340–347
123
Staging of Acute Exacerbation in Patients with Idiopathic Pulmonary Fibrosis Tomoo Kishaba • Hitoshi Tamaki • Yousuke Shimaoka Hajime Fukuyama • Shin Yamashiro
•
Received: 27 February 2013 / Accepted: 22 October 2013 / Published online: 13 November 2013 Ó Springer Science+Business Media New York 2013
Abstract Background The purpose of this study was to evaluate the predictors of a 3-month mortality rate of acute exacerbation of idiopathic pulmonary fibrosis (IPF) and provide a new staging system. Methods A total of 594 patients with IPF were included in this retrospective, observational study conducted from January 2001 to December 2010 at Okinawa Chubu Hospital. Results Among the 594 patients, 58 (9.8 %) developed acute exacerbation (AE) of IPF during the 10-year observation period. The median follow-up period for AE was 10.4 months. In-hospital mortality was 56.9 % and the 3-month mortality rate was 63.8 %. We identified the following four parameters in a multivariate analysis as: serum lactate dehydrogenase, sialylated carbohydrate antigen (KL-6), ratio of partial pressure of oxygen and fraction of inspiratory oxygen, and total extent of abnormal findings on high-resolution computed tomography of the chest. Patients were divided into two groups on the basis of the four composite parameters. Patients in the extensive disease-stage group required more mechanical ventilation and intensive therapy than those in the limited disease-stage group. The 3-month mortality rate was higher in patients in the extensive disease-stage group than in patients in the limited disease-stage group (80.6 vs. 54.5 %, respectively; p = 0.007).
Conclusions Staging of AE in patients with IPF provided useful information regarding disease severity and shortterm outcome. Keywords Acute exacerbation Extensive Idiopathic pulmonary fibrosis Limited Staging
Introduction Idiopathic pulmonary fibrosis (IPF) is a progressive interstitial lung disease (ILD) of unknown etiology [1]. The clinical course of individual patients is highly variable [2, 3]. Acute exacerbation (AE) of IPF, which is a sudden deterioration of respiratory condition superimposed on chronic fibrotic lung disease, can sometimes develop during the clinical course [4, 5]. Collard et al. [6] published diagnostic criteria for AE of IPF, and Song et al. [7] have reported the incidence, risk factors, and outcomes of AE in patients with IPF. In addition, Kondoh et al. [8] have described the risk factors for AE. However, variation in the severity of AE in patients with IPF remains unknown. Therefore, the purpose of this study was to use composite parameters to assess the severity of AE in patients with IPF before mechanical ventilation (MV) and provide a new staging system for predicting the 3-month mortality rate.
Materials and Methods T. Kishaba (&) Y. Shimaoka H. Fukuyama S. Yamashiro Department of Respiratory Medicine, Okinawa Chubu Hospital, Okinawa, Japan e-mail: [email protected] H. Tamaki Sunagawa Naika Clinic, Okinawa, Japan
Study Subjects This was a retrospective and observational study. The study population comprised of 594 patients with IPF diagnosed according to the American Thoracic Society (ATS)/European
123
142
Respiratory Society (ERS) consensus statement [1] at Okinawa Chubu Hospital, Okinawa, Japan from January 2001 to December 2010. Data Collection We collected medical records of the patients during the same period. Patients with secondary ILD, such as connective-tissue disease, due to drug, environmental, or occupational factors, were excluded. Baseline variables, including medical treatment, were evaluated at the time of IPF diagnosis. Among the remaining patients, 58 developed AE. In this study, AE was defined as a sudden aggravation of dyspnea within 30 days with new bilateral infiltration accompanying known IPF or evidence of honeycombing on high-resolution computed tomography (HRCT) of the chest [9–11]. Five patients underwent surgical lung biopsy and demonstrated pathological findings of usual interstitial pneumonia (UIP). All 58 patients were diagnosed on the basis of a multidisciplinary discussion. Three patients developed two episodes of AE. Therefore, only the first episodes were included in the analysis. All laboratory data were collected on the date of admission due to AE. We performed bronchoalveolar lavage (BAL) in all patients with IPF to exclude secondary causes such as pneumonia, sepsis, and aspiration and to diagnose the cause of AE [12]. BAL cultures of all patients revealed \103 colony-forming units. CT angiogram was negative and an echocardiogram showed no evidence of systolic or diastolic heart failure, which excluded pulmonary embolism and congestive heart failure. Modified Medical Research Council (mMRC) dyspnea scale [13] and dyspnea duration defined as the interval between symptom onset and HRCT date also were reviewed. In addition, we reviewed laboratory findings and baseline pulmonary function test results. The median follow-up period was defined as the period from the date of diagnosis of AE in patients with IPF until death or final follow-up.
Lung (2014) 192:141–149
the area above the main carina, the lower was defined as the area between the area of the lung below the level of the inferior pulmonary vein and the level immediately above the right hemi-diaphragm, and the middle was defined as the area between the two according to the protocol of Sumikawa et al. [17]. Both chest physicians and radiologists read whole images and reached a consensus through multidisciplinary discussion. The extent of traction bronchiectasis, ground-glass opacity, consolidation, and honeycombing was defined as follows: 0, none; 1, involving 1–25 %; 2, involving 26–50 %; and 3, involving ]50 %. Among these four abnormal findings, we considered that both ground-glass opacity and consolidation contributed to the severity of AE in patients with IPF. Therefore, each zone’s radiological score of both groundglass opacity and consolidation were summed and was defined as the total extension score. Therefore, the minimum score was zero and the highest was 36. Calculation of Staging Points Receiver operating characteristic (ROC) curve analysis was used to define the staging points, which were later summed to determine staging. We defined the points used to determine the staging as follows: 0, serum lactate dehydrogenase (LDH) level \280, and 1, serum LDH level ]280; 0, serum sialylated carbohydrate antigen (KL-6) level\1,000, and 1, KL-6 level ]1,000; 0, partial pressure of oxygen and fraction of inspiratory oxygen (P/F) ratio ]100, and 1, P/F ratio \100; 0,overall HRCT score \20, and 1, overall HRCT score ]20. These values were obtained at the time of AE diagnosis. Each parameter predicted the 3-month mortality rate. However, a combination of these parameters more actively predicted the 3-month mortality rate with the Cox proportion hazard model and c-statistic. On the basis of this information, we defined staging based on the total score obtained as follows: limited disease stage 0–2 and extensive disease stage ]3.
Radiological Evaluation Treatment Protocol With patients in the supine position, thin-section CT scans were obtained at the end-inspiratory phase. Images were obtained at window settings appropriate for viewing the lung parenchyma. We evaluated the presence, distribution, and extent of ground-glass opacity, consolidation, traction bronchiectasis, and honeycombing on admission [14–16]. In addition, we compared HRCT findings during the acute phase with that during the stable phase for the diagnosis of IPF. Lungs were divided into six zones (upper, middle, and lower of both sides). The upper lung zone was defined as
123
Our standardized uniformed treatment protocol was as follows: we always initiated methylprednisolone (mPSL) pulse therapy (1 g/day for 3 consecutive days) when we treated AE in patients with IPF and maintained PSL monotherapy for patients responding to steroid pulse therapy. However, if a patient showed poor response after the first cycle of steroid pulse therapy, we initiated immunosuppressants, such as cyclosporine A or azathioprine, and repeated the steroid pulse therapy weekly for a maximum of 4 weeks.
Lung (2014) 192:141–149
Mortality In-hospital mortality was defined as death during admission. AE onset was defined as that on the date of admission. The 3-month mortality rate was defined as death within 3 months from AE onset. The Ethics Committee of Okinawa Chubu Hospital approved the study protocol. Informed consent was waived because this study was a retrospective and anonymous review of medical records. Our Ethics Committee did not require informed consent, because this study was epidemiological rather than a prospective interventional clinical study.
143
characteristics of all patients with IPF are shown in Table 1. We evaluated another cohort of patients with AE of IPF for validation. The baseline clinical parameters and prior treatments of these two cohorts are shown in Table 2. Median survival rate was 9.3 months. Laboratory Findings Mean serum white blood cell count and levels of C-reactive protein, LDH, and KL-6 were 11,357 mm3, 8.4 mg/dL, 396 IU/L, and 1,889 IU/L, respectively. All patients required oxygen, and the mean ratio of partial pressure of oxygen and fraction of inspiratory oxygen concentration (P/F ratio) was 158 (Table 3).
Statistical Analysis Physiological Parameters Continuous variables are presented as means ± standard deviations, and categorical variables are presented as percentages. The v2 and Fisher’s exact tests were used for categorical data, and the unpaired t test and Mann–Whitney U test were used for continuous data. The j statistic was used to analyze the interobserver variations in the findings. Cox regression analysis was used to identify significant variables capable of predicting AE severity or those that could act as prognostic factors. Model discrimination was quantified based on C statistic, which is the probability that among two randomly selected patients the patient with the highest predicted risk of an event will be the first to experience the event. C-statistic value ranges from 0.5 (model discrimination is no better than chance) to 1 (model discrimination is perfect). C-statistic values ranging between 0.7 and 0.8 is typically considered ‘‘acceptable,’’ whereas a value [0.8 is typically considered ‘‘excellent.’’ The Kaplan–Meier survival curve and the log-rank test were used to evaluate survival. The level of statistical significance was set at P \ 0.05. STATA software V.11.0 (Stata Corp., College Station, TX, USA) was used to perform all clinical data analyses.
The mean baseline forced vital capacity (FVC), percent predicted FVC, and percent predicted diffusion of carbon mono oxide (DLco) values were 1.63 L, 58.2 %, and 38.3 %, respectively. MV and Hospital Stay Half of the patients required MV, and the length of MV was 8.5 ± 12.9 days. The length of hospital stay was 21.1 ± 13.7 days. Table 1 Baseline characteristics of patients without AE of IPF, and patients with AE of IPF Without AE of IPF (n = 536)
AE of IPF (n = 58)
P value
Age (year)
72 ± 9.2
75.0 ± 9.6
0.738
Male/female
386/150
38/20
0.704
Pack-year Smoking status (active/ex/never)
42.4 ± 40.9 (232/272/32)
35.4 ± 41.6 (22/36/0)
0.692 0.541
mMRC
2.4 ± 1.0
2.8 ± 1
0.428
Dyspnea duration (months)
12.8 ± 9.6
6.7 ± 6
0.036
Results
Percent predicted FVC (%)
69.5 ± 24.6
58.2 ± 21.2
0.025
Clinical Characteristics
Percent predicted DLco (%)
61.3 ± 22.8
52.3 ± 19.7
0.042
We identified 594 patients with IPF during the 10-year study period. Among these patients, 58 developed AE. Mean age of all patients with AE was 75 years, and 66 % were men. The median follow-up period from the date of diagnosis of AE of IPF until death or the final follow-up was 10.2 (range 0.1–112) months. Among patients with AE of IPF, smokers approximately had a 35 pack-year history. The mean mMRC dyspnea scale score was 2.8, and the mean dyspnea duration was 6.7 days. Clinical
Anticoagulants use
41
4
0.751
Prednisolone use
38
4
0.683
Immunosuppressive therapy
16
2
0.582
Anti proliferative therapy
2
0
0.864
PCP-prophylaxis
28
3
0.795
mMRC modified Medical Research Council dyspnea scale, FVC forced vital capacity, DLco diffusion of carbon mono oxide, PCP Pneumocystis jirovecii pneumonia
123
144
Lung (2014) 192:141–149
Table 2 Baseline characteristics of patients with an AE of IPF between different periods AE of IPF (n = 52) 1995–2000
AE of IPF (n = 58) 2001–2010
P value 0.868
Age (year)
72.0 ± 8.2
75.0 ± 9.6
Male/female
36/16
38/20
0.796
Pack-year
42.8 ± 33.2
35.4 ± 41.6
0.692
Smoking status (active/ex)
24/28
22/36
0.573
mMRC
3.0 ± 1.4
2.8 ± 1.0
0.726
Dyspnea duration (days)
10.2 ± 5.8
6.7 ± 6.0
0.462
Percent predicted FVC (%)
62.6 ± 29.5
58.2 ± 21.2
0.628
Percent predicted DLco (%)
51.9 ± 24.6
52.3 ± 19.7
0.795
LDH (IU/L)
326.8 ± 128.7
396.4 ± 147.8
0.483
KL-6 (IU/L)
1,581.6 ± 1,225.3
1,889.2 ± 1,735.6
0.693
P/F Ground-glass opacity ? consolidation score
166.2 ± 79.5 21.9 ± 0.8
158.4 ± 94.3 23.2 ± 0.7
0.759 0.695
3-month mortality rate (%)
75.0
70.7
0.685
Survival time (months)
7.6 ± 28.2
9.3 ± 24.6
0.824
Anticoagulants use
3
4
0.826
Prednisolone use
6
4
0.496
Immunosuppressive therapy
3
2
0.672
Antiproliferative therapy
0
0
0.998
PCP prophylaxis
4
3
0.704
mMRC modified Medical Research Council dyspnea scale, FVC forced vital capacity, DLco diffusion of carbon mono oxide, KL-6 sialylated carbohydrate antigen, P/F ratio ratio of partial pressure of oxygen and fraction of inspiratory oxygen, PCP Pneumocystis jirovecii pneumonia
Table 3 Clinical data on the date of admission due to AE cohort AE in IPF (n = 58) Age (year)
75 ± 9.6
WBC (mm3)
11,356.9 ± 5,544
CRP (mg/dl)
8.4 ± 7.5
LDH (IU/L)
396.4 ± 147.8
KL-6 (IU/L)
1,889.2 ± 1,735.6
pH
7.42 ± 0.1
PaO2 (mmHg)
62.1 ± 17.6
PaCO2 (mmHg)
43.5 ± 16.3
P/F
158.4 ± 94.3
MV (%)
50
Length of MV (days) Duration of hospital stay (days)
8.5 ± 12.9 21.1 ± 13.7
Survival time (months)
9.3 ± 24.6
LDH lactate dehydrogenase, KL-6 sialylated carbohydrate antigen, P/F ratio ratio of partial pressure of oxygen and fraction of inspiratory oxygen concentration
AE staging Univariate analysis was performed to assess the useful parameters for predicting the 3-month mortality rate of AE (Table 4). Results of the multivariate analysis are shown in Table 5. Serum LDH, and KL-6 levels, P/F ratio, and total
123
extent of abnormal findings on HRCT of the chest score were identified as predictors of the 3-month mortality rate. The j value of four chest physicians and two radiologists was 0.64. We assigned staging levels of 0–3 for traction bronchiectasis, honeycombing, ground-glass opacity, and consolidation. Among them, both ground-glass opacity and consolidation scores were closely associated for predicting the 3-month mortality rate (r = 0.846). Therefore, we used the overall score of these two HRCT findings for staging. The ROC curve values for serum LDH level, KL-6 level, P/F ratio, and total extent of abnormal findings on HRCT of the chest score were 0.618, 0.626, 0.658, and 0.72, respectively, and the cutoff values were 280, 1,000, 100, and 20, respectively. We assigned 0 or 1 for serum LDH level, serum KL-6 level, P/F ratio, and total extent of abnormal findings on HRCT of the chest score (Table 6). The Cox-proportional hazard model was used to investigate useful ROC-curve cutoff points for staging. The four parameters had different hazard ratios (HRs) for predicting the 3-month mortality rate: each parameter predicted the 3-month mortality rate. However, a combination of these parameters more actively predicted the 3-month mortality rate using the Cox proportion hazard model. In addition, we evaluated the discrimination rate for different clinical models using C statistic (Table 7). Therefore, we defined staging on the
Lung (2014) 192:141–149
145
Table 4 Univariate analysis of 3-month mortality rate due to AE
Table 6 Point assignment for AE staging
HR
P value
Definition
Point
Age
0.997
0.833
LDH
Gender
0.898
0.76
\280
0
Pack-year
1.009
0.027
]280
1
mMRC
1.161
0.386
Dyspnea duration
1.013
0.649
\1,000
0
Finger clubbing
1.45
0.299
]1,000
1
WBC
1
0.465
P/F ratio
LDH
1.003
0.009
]100
0
KL-6
2.01
0.001
\100
1
CRP
0.984
0.47
P/F Traction bronchiectasis ? honeycombing score
1.999 1.324
0.048 0.061
Ground-glass opacity ? consolidation score
1.853
0.033
BALF neutrophil
0.944
0.327
BALF lymphocyte
1.002
0.974
PSL alone
0.436
0.033
PSL ? immunosuppressants
2.294
0.033
MV
1.921
0.026
mMRC modified Medical Research Council dyspnea scale, LDH lactate dehydrogenase, KL-6 sialylated carbohydrate antigen KL-6, P/ F ratio ratio of partial pressure of oxygen and fraction of inspiratory oxygen concentration, BALF bronchoalveolar lavage fluid, PSL prednisolone
KL-6
Ground-glass opacity ? consolidation score \20 ]20
LDH lactate dehydrogenase, KL-6 sialylated carbohydrate antigen KL-6, P/F ratio ratio of partial pressure of oxygen and fraction of inspiratory oxygen concentration
Table 7 Different clinical model specifications Model
Independent variables
C-statistic (95 % CI)
Comprehensive
LDH, KL-6, P/F, ground-glass opacity ? consolidation score
0.76 (0.7–0.82)
Model 1
LDH
0.56 (0.52–0.6)
Model 2 Model 3
KL-6 P/F
0.62 (0.580.66) 0.6 (0.56–0.64)
Model 4
Ground-glass opacity ? consolidation score
0.66 (0.62–0.7)
Model 5
LDH and P/F
0.62 (0.58–0.66)
Model 6
LDH, ground-glass opacity ? consolidation score
0.68 (0.62–0.72) 0.66 (0.62–0.7)
Table 5 Multivariate analysis of 3-month mortality rate due to AE HR
P value
0 1
LDH KL-6
2.024 2.909
0.047 0.038
Model 7
KL-6 and P/F
P/F
2.42
0.041
Model 8
Ground-glass opacity ? consolidation score
2.289
0.030
KL-6, ground-glass opacity ? consolidation score
LDH lactate dehydrogenase, KL-6 sialylated carbohydrate antigen KL-6, P/F ratio ratio of partial pressure of oxygen and fraction of inspiratory oxygen concentration
0.7 (0.66–0.74)
LDH lactate dehydrogenase, KL-6 sialylated carbohydrate antigen KL-6, P/F ratio ratio of partial pressure of oxygen and fraction of inspiratory oxygen concentration
Table 8 Staging system for patients with AE of IPF patients
basis of the four clinical parameter scores obtained as follows: limited stage 0–2, and extensive stage ]3 (Table 8). Clinical characteristics of each stage are shown in Table 9. Base line treatment and HRCT score for each stage were similar. The extensive disease stage group had higher KL-6 levels and HRCT scores and required more MV than the limited disease-stage group (2,303 vs. 1,164, 8.4 vs. 5.6, and 61 vs. 32 %, respectively). Among all patients, 29 (50 %) required MV, and only 6 patients survived. Fifteen patients underwent MV 48 h after admission. Representative images of the two groups are shown in Figs. 1 and 2.
Points Limited exacerbation (n = 22)
0–2
Extensive exacerbation (n = 36)
]3
Treatment Response and Prognosis In the limited disease stage group, 68 % patients received PSL alone. Moreover, 69 % of patients in the extensive disease stage group received combination therapy, such as PSL and another immunosuppressant. We did not use
123
146
Lung (2014) 192:141–149
Table 9 Clinical characteristics of patients in each stage Limited disease (n = 22)
Extensive disease (n = 36)
P value
Age
74.68 ± 11.13
75.17 ± 8.64
0.93
Pack-year
26.95 ± 31.66
40.56 ± 46.28
0.425
Smoking status (active/ex/never)
(10/17/0)
(12/19/0)
0.486
mMRC
2.55 ± 1.06
3.03 ± 0.94
0.049
Dyspnea duration (months)
5.59 ± 4.92
7.33 ± 6.53
0.274
WBC (mm3)
12,259 ± 6,887
10,806 ± 4,558
0.496
CRP (mg/dl)
11.15 ± 9.19
6.85 ± 5.88
0.141
LDH (IU/L)
372.10 ± 122.66
410.5 ± 160.59
0.519
KL-6 (IU/L)
1,164.35 ± 838.60
2,303.46 ± 1976.22
0.0007
PaO2 (mmHg)
56.62 ± 16.01
65.44 ± 17.84
0.041
P/F ratio Traction bronchiectasis ? honeycombing score
170.58 ± 104.72 19
138.55 ± 72.09 25
0.309 \0.0001
Ground-glass opacity ? consolidation score
17
27
\0.0001
MV (%)
31.8
61.1
0.032
PSL alone (%)
68.2
30.6
0.006
Combination therapy (%)
31.8
69.4
0.006
Length of hospital stay (days)
24.36 ± 13.98
19.17 ± 13.31
0.123
90-day mortality rate (%)
54.5
80.6
0.007
Baseline traction bronchiectasis ? honeycombing score
15
17
0.491
Baseline ground-glass opacity ? consolidation score
6
5
0.726
Baseline anticoagulants use
2
2
0.374
Baseline prednisolone use
3
1
0.286
Baseline immunosuppressive therapy
1
1
0.793
Baseline antiproliferative therapy
0
0
0.902
Baseline PCP-prophylaxis
2
1
0.684
mMRC modified Medical Research Council dyspnea scale, LDH lactate dehydrogenase, KL-6 sialylated carbohydrate antigen KL-6, P/F ratio ratio of partial pressure of oxygen and fraction of inspiratory oxygen concentration, PCP Pneumocystis jirovecii pneumonia
Fig. 1 Representative image of the limited disease stage
extracorporeal membrane oxygenation or lung transplantation. The 3-month mortality rate was worse in the extensive disease-stage group than in the limited diseasestage group (80.6 vs. 54.5 %; p = 0.007; Fig. 3). We evaluated another cohort of patients with AE of IPF,
123
Fig. 2 Representative image of the extensive disease stage
comprising 52 patients from January 1995 to December 2000. Our staging was applied to these patients and showed a similar result (82.9 vs. 58.8 %; p = 0.005).
Lung (2014) 192:141–149
Fig. 3 3-month survival curve according to disease stage. The extensive disease stage group showed high mortality compared with that of the limited disease stage group
Discussion We found that 58 of 594 patients with IPF developed AE during the 10-year study period. Song et al. [7] reported 1- and 3-year mortality rate of 14.2 and 20.7 %, respectively. The mean age of the patients who developed AE was 75 years, and 66 % of them were men. Many reports on the cause of death or occurrence of AE in patients with IPF have shown that older age, AE, decreased functional impairment, and immunosuppressive therapy are significant predictors of poor overall prognosis [5, 18]. Although many reports are available regarding the incidence and cause of AE in patients with IPF, limited studies are available on AE severity in patients with IPF. Song et al. reported that the immediate outcome was poor and that median survival time of patients was 2.2 months. In addition, these patients usually develop severe respiratory failure and required MV. The in-hospital mortality rate was 95.7 % in 22 of 23 patients [19]. Pe´rez et al. showed that high positive end-expiratory pressure (PEEP) and Acute Physiology and Chronic Health Evaluation III scores predicted mortality. In addition, low P/F ratio is an independent determinant of survival in patients with ILD [20]. Among our cohort of patients with AE of IPF, 29 (50 %) required MV, and a lung protective strategy for profound hypoxemic patients was used. However, MV provided little benefit for AE of IPF. Saydain et al. [21] reported that the actual in-hospital mortality rate was 61 and 92 % of the hospital survivors died at a median of 2 months after discharge. MV itself may constitute one of the most important outcome determinants in patients with IPF. Based on these findings, we recognized the importance of staging for determining AE mortality in patients with IPF. Other possibilities of acute respiratory failure, such as infection, heart failure, pneumothorax, and pulmonary embolism, should be excluded for AE diagnosis. Thus, we
147
performed BAL in all patients. We believe the main purpose of performing BAL is to exclude infection. In addition, a detailed history, physical examination, medication review, environmental exposure, and family history are very important [1]. We used univariate and multivariate analyses to predict the 3-month mortality rate and the Cox proportional hazards model to investigate useful variables. Percent predicted FVC and DLco are crucial parameters for predicting IPF mortality [22, 23]. Pulmonary function may be important for predicting prognosis; however, many patients cannot perform the test in an AE state because of severe dyspnea. We identified the following parameters that affected mortality and severity of AE in patients with IPF: serum LDH level, KL-6 level, P/F ratio, and total extent of abnormal findings on HRCT of the chest score. MV also was an important factor. However, in our cohort, 15 of 29 patients (52 %) required MV after 48 h after admission. Our main purpose was immediate severity evaluation. In addition, the indication of MV includes subjective elements and is an ethical issue rather than only conducting a blood test and evaluating HRCT findings. Therefore, we considered that it was unreasonable to add this to the staging system. Akira et al. [24] showed that the HRCT pattern and overall CT extent are associated with mortality; they divided patients into the following three patterns: peripheral, multifocal, and diffuse. The strongest correlations were observed between CT patterns and survival (odds ratio 4.629; 95 % confidence interval (CI) 1.9–11.278; P \ 0.001). Sumikawa et al. reported that traction bronchiectasis and fibrosis scores are significant predictors of outcome in patients with pathological UIP patients (HR 1.3 and 1.1, respectively; 95 % CI 1.18–14.2 and 1.03–1.19, respectively) [17]. We set our HRCT findings scores based on these reports. We chose four useful parameters for staging after multivariate analysis. Each parameter predicted the 3-month mortality rate but individual HRs differed. Among them, HRCT scores were variable. Traction bronchiectasis and honeycombing signify a chronic process and are useful predictors of long-term mortality because of IPF. Fibrotic findings were not different between the two groups at baseline. In contrast, we consider that both ground-glass opacity and consolidation usually indicate new findings that contribute to increased severity of AE in patients with IPF. In addition, a minority of patients received, PSL, immunosuppressants, and anticoagulants before AE. Anticoagulants were used to treat atrial fibrillation not IPF. Thus, we thought that history of prior treatment had little effect on severity and HRCT findings of AE. Therefore, we chose a mixed score of these two HRCT findings for creating the 3-month mortality rate staging system. The most
123
148
accurate and discriminating HRCT score value was 20 in the ROC curve analysis. We thought that a simple staging system would be practical and easy to use. Therefore, we assigned only two points for each parameter based on the ROC curve analysis. We also investigated the difference in treatment strategy for staging. There have been many reports on high-dose corticosteroids for treating AE of IPF [25–27]. We always initiate methylprednisolone pulse therapy (1 g/day for 3 consecutive days) when we treat AE of IPF and maintain PSL monotherapy for patients responding to steroid pulse therapy, such as those with limited-disease stage. Patients in the extensive disease stage group received combination therapy, such as PSL with immunosuppressants. We usually added an immunosuppressant such as cyclosporine A or azathioprine based on previous reports [28, 29]. In addition, we repeated methylprednisolone pulse therapy every week for either 2 or 4 weeks. If these patients survived, we maintained low-dose PSL and a cytotoxic agent. Finally for non-responders, such as patients in the extensive disease stage group, we repeated steroid pulse therapy and high-dose cyclophosphamide or another immunosuppressant. However, these patients often were entirely resistant to such intensive therapy and showed progressive respiratory failure. Therefore, supportive care may be the mainstay approach for this patient sub group. We described the variable clinical courses of patients with AE of IPF in our cohort. We plan to use a different management strategy in the future based on our staging system. Our study had several limitations. First, this was a retrospective and observational study. Second, the long study period (10 years) in combination with the relatively small sample size and the sophisticated statistical approach was an important limitation of our study. Third, most of our patients with IPF were diagnosed according to clinical criteria. Only five patients underwent surgical lung biopsy. Therefore, patients with AE and pathological UIP may show a different presentation. However, the clinical characteristics, physiological impairment, and clinical course of our cohorts were entirely compatible with IPF. Therefore, misclassification of other patients with ILD was not high. Fourth, most patients received broadspectrum antibiotics: however, we evaluated their sputum smear and performed BAL for all patients. The BALF culture showed no evidence of active infection, including clinical parameters, such as C-reactive protein. Therefore, we excluded pneumonia. Finally, this was a single-center study and the number of patients with AE patients in our cohort was small. Therefore, the results are not generalizable to all patients with AE of IPF. However, the clinical presentation and HRCT findings in our patients were comparable to those of previous reports. In addition, we
123
Lung (2014) 192:141–149
evaluated another cohort during another period with the same staging system and observed similar results. Therefore, our staging system is valid and applicable.
Conclusions We propose a new staging system for patients with AE of IPF. Patients in each stage showed different clinical characteristics and short-term outcomes. Different management strategies will be required to rescue patients with AE of IPF patients according to our staging system. Our composite staging system, including clinical parameters and HRCT findings, provides useful information for physicians. A multicenter study is warranted to validate our results. Acknowledgments The authors thank all of our residents and Drs. Yasutani and Takara for their interpretation of the medical records and imaging results. Conflict of interest
All the authors declare none.
References 1. American Thoracic Society, European Respiratory Society (2000) Idiopathic pulmonary fibrosis: diagnosis and treatment. International consensus statement. Am J Respir Crit Care Med 161:646–664 2. Kim DS, Collard HR, King TE Jr (2006) Classification and natural history of the idiopathic interstitial pneumonias. Proc Am Thorac Soc 3:285–292 3. Ley B, Collard HR, King TE Jr (2011) Clinical course and prediction of survival in idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 183:431–440 4. Kondoh Y, Taniguchi H, Kawabata Y, Yokoi T, Suzuki K, Takagi K (1993) Acute exacerbation in idiopathic pulmonary fibrosis. Analysis of clinical and pathologic findings in three cases. Chest 103(6):1808–1812 5. Panos RJ, Mortenson RL, Niccoli SA et al (1990) Clinical deterioration in patients with idiopathic pulmonary fibrosis: causes and assessment. Am J Med 88:396–404 6. Collard HR, Moore BB, Flaherty KR et al (2007) Acute exacerbations of idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 176:636–643 7. Song JW, Hong S-B, Lim C-M, Koh Y, Kim DS (2011) Acute exacerbation of idiopathic pulmonary fibrosis: incidence, risk factors and outcome. Eur Respir J 37:356–363 8. Kondoh Y, Taniguchi H, Katsuta T, Kataoka K, Kimura T, Nishiyama O, Sakamoto K, Johkoh T, Nishimura M, Ono K, Kitaichi M (2010) Risk factors of acute exacerbation of idiopathic pulmonary fibrosis. Sarcoidosis Vasc Diffuse Lung Dis 27(2):103–110 9. Nishimura K, Kitaichi M, Izumi T, Nagai S, Kanaoka M, Itoh H (1992) Usual interstitial pneumonia: histologic correlation with high-resolution CT. Radiology 182:337–342 10. Johkoh T, Muller NL, Cartier Y, Kavanagh PV, Hartman TE, Akira M, Ichikado K, Ando M, Nakamura H (1999) Idiopathic interstitial pneumonias: diagnostic accuracy of thin-section CT in 129 patients. Radiology 211:555–560
Lung (2014) 192:141–149 11. Hansell DM, Bankier AA, Macmahon H, McLoud TC, Muller NL, Remy J (2008) Fleischner society: glossary of terms for thoracic imaging. Radiology 246:697–722 12. Meyer KC, Raghu G, Baughman RP, Brown KK, Costabel U, du Bois RM, Drent M, Haslam PL, Kim DS, Nagai S, Rottoli P, Saltini C, Selman M, Strange C, Wood B (2012) An Official American Thoracic Society Clinical Practice Guideline: the clinical utility of bronchoalveolar lavage cellular analysis in interstitial lung disease. Am J Respir Crit Care Med 185:1004–1014 13. Bestall JC, Paul EA, Garrod R, Garnham R, Jones PW, Wedzicha JA (1999) Usefulness of the Medical Research Council (MRC) dyspnoea scale as a measure of disability in patients with chronic obstructive pulmonary disease. Thorax 54:581–586 14. Akira M, Hamada H, Sakatani M, Kobayashi C, Nishioka M, Yamamoto S (1997) CT findings during phase of accelerated deterioration in patients with idiopathic pulmonary fibrosis. Am J Roentgenol 168:79–83 15. Kim DS, Park JH, Park BK, Lee JS, Nicholson AG, Colby T (2006) Acute exacerbation of idiopathic pulmonary fibrosis: frequency and clinical features. Eur Respir J 27:143–150 16. Akira M, Sakatani M, Ueda E (1993) Idiopathic pulmonary fibrosis: progression of honeycombing at thin-section CT. Radiology 189:687–691 17. Sumikawa H, Johkoh T, Colby TV, Ichikado K, Suga M, Taniguchi H, Kondoh Y, Ogura T, Arakawa H, Fujimoto K, Inoue A, Mihara N, Honda O, Tomiyama N, Nakamura H, Mu¨ller ML (2008) Computed tomography findings in pathological usual interstitial pneumonia. Relationship to survival. Am J Respir Crit Care Med 177:433–439 18. Nagai S, Kitaichi M, Hamada Knagao T, Hoshino Y, Miki H, Izumi T (1999) Hospital-based historical cohort study of 234 histologically proven Japanese patients with IPF. Sarcoidosis Vasc Diffuse Lung Dis 16:209–214 19. Stern J-B, Mal H, Groussard O, Brugie`re O, Marceau A, Jebrak G, Fournier M (2001) Prognosis of patients with advanced idiopathic pulmonary fibrosis requiring mechanical ventilation for acute respiratory failure. Chest 120(1):213–219 20. Fernandez-Perez ER, Yilmaz M, Jenad H, Daniels CE, Ryu JH, Hubmayr RD, Gajic O (2008) Ventilator settings and outcome of
149
21.
22.
23.
24.
25.
26.
27.
28.
29.
respiratory failure in chronic interstitial lung disease. Chest 133(5):1113–1119 Saydain G, Islam A, Afessa B, Ryu JH, Scott JP, Peters SG (2002) Outcome of patients with idiopathic pulmonary fibrosis admitted to the intensive care unit. Am J Respir Crit Care Med 166:839–842 Wells AU, Desai SR, Rubens MB, Goh NS, Cramer D, Nicholson AG, Colby TV, du Bois RM, Hansell DM (2003) Idiopathic pulmonary fibrosis: a composite physiologic index derived from disease extent observed by computed tomography. Am J Respir Crit Care Med 167(7):962–969 du Bois RM, Weycker D, Albera C, Bradford WZ, Costabel U, Kartashov A, King TE Jr, Lancaster L, Noble PW, Sahn SA, Thomeer M, Valeyre D, Wells AU (2011) Forced vital capacity in patients with idiopathic pulmonary fibrosis. Test properties and minimal clinically important difference. Am J Respir Crit Care Med 184:1382–1389 Akira M, Kozuka T, Yamamoto S, Sakatani M (2008) Computed tomography findings in acute exacerbation of idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 178:372–378 Kondoh Y, Taniguchi H, Yokoi T, Nishiyama O, Ohishi T, Kato T, Suzuki K, Suzuki R (2005) Cyclophosphamide and low-dose prednisolone in idiopathic pulmonary fibrosis and fibrosing nonspecific interstitial pneumonia. Eur Respir J 25:528–533 Parambil JG, Myers JL, Ryu JH (2005) Histopathologic features and outcome of patients with acute exacerbation of idiopathic pulmonary fibrosis undergoing surgical lung biopsy. Chest 128:3310–3315 Kondo A, Saiki S (1989) Acute exacerbation in idiopathic interstitial pneumonia (IIP). In: Harasawa M, Fukuchi Y, Morinari H (eds) Interstitial pneumonia of unknown etiology. University of Tokyo Press, Tokyo, pp 33–42 Homma S, Sakamoto S, Kawabata M, Kishi K, Tsuboi E, Motoi N, Yoshimura K (2005) Cyclosporin treatment in steroid-resistant and acutely exacerbated interstitial pneumonia. Intern Med 44:1144–1150 Pereira CAC, Malheiros T, Coletta EM, Ferreira RG, Rubin AS, Otta JS, Rocha NS (2006) Survival in idiopathic pulmonary fibrosis cytotoxic agents compared to corticosteroids. Respir Med 100:340–347
123
