|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Division of Pulmonary, Critical Care, Allergy, Immunology and Sleep Medicine (Drs. Quefatieh and DiGiovine), and the Department of Pathology (Dr. Stone), Henry Ford Health Sciences Center, Detroit; and the Division of Pulmonary and Critical Care Medicine (Drs. Toews and Hyzy), University of Michigan Medical Center, Ann Arbor, MI.
Correspondence to: Robert C. Hyzy, MD, FCCP, 3916 Taubman Center, University of Michigan Medical Center, 1500 East Medical Center Dr, Ann Arbor, MI 48109; e-mail: rhyzy{at}umich.edu
| Abstract |
|---|
|
|
|---|
Design: Retrospective chart review.
Setting: A large, urban, teaching hospital.
Interventions: Patients were classified into idiopathic (AIP group) and secondary causes of DAD (ARDS group) according to available clinical and microbiology data. AIP and ARDS cases were compared, and ARDS cases were analyzed for long-term outcome.
Measurements and results: Twenty patients with pathologic diagnosis of DAD were identified. Four cases were excluded; eight cases of ARDS due to known etiologies were identified. These etiologies included pneumonia and sepsis (n = 6), cocaine use (n = 1), and carmustine chemotherapy (n = 1). Eight cases of AIP were found. When compared with the ARDS group, patients in the AIP group had a longer time from the onset of symptoms until hospital admission (16.8 ± 15.7 days vs 2.2 ± 1.0 days, p = 0.0015) and a shorter time from hospital admission to open-lung biopsy (8.3 ± 3.0 days vs 15.5 ± 9.5 days, p = 0.02) [mean ± SD]. Seven of eight patients with AIP and four of eight patients with ARDS survived to hospital discharge (p = not significant). The 12.5% mortality rate for patients with AIP reported in this series was significantly lower than the previously reported cumulative rate of 69.5% (p = 0.0025). Follow-up in five AIP survivors for a mean of 7.6 ± 3.5 years (range, 3 to 11 years) showed all to be without shortness of breath or relapse despite mild residual fibrosis on chest radiograph and mild-to-moderate restriction on pulmonary function tests (mean total lung capacity, 68.5 ± 6.2% predicted).
Conclusions: Our data support a favorable hospital and long-term outcome for patients with AIP, with no evidence of recurrence or progression to chronic interstitial lung disease.
Key Words: acute interstitial pneumonia ARDS diffuse alveolar damage outcome
| Introduction |
|---|
|
|
|---|
Typically, patients present with respiratory failure following a viral illness such as prodrome of cough and dyspnea. The histologic pattern is identical to that seen in established ARDS2 ; however, AIP differs from ARDS in the absence of known inciting events, and the lack of multiorgan involvement.3 Since having been first described by Hamman and Rich4 in 1935, few additional series have been reported in the literature.5
While mortality in all patients with ARDS has decreased to 31%,6 7 there have been no analogous reports of an improvement in mortality in patients with AIP, wherein every reported series the hospital mortality has been > 50%.5 Although only a few hospital survivors have been described in each series, the functional recovery among these survivors has been good, with improvement in lung function and resolution of pulmonary infiltrates. The good outcome among hospital survivors has been recently challenged by Vourlekis et al,8 who noted the development of recurrence and/or progression to end-stage lung fibrosis in four of seven patients who survived their initial hospitalization.
We examined our 10-year experience with AIP at a large, urban, teaching hospital, evaluating in-hospital mortality and the functional recovery of survivors. In addition, we compared patients with AIP to patients with ARDS who had open-lung biopsy specimens showing DAD.
| Materials and Methods |
|---|
|
|
|---|
Following the review of the medical record from each case, three patients were excluded from further analysis because the chest radiograph had shown only a localized infiltrate: one patient underwent open-lung biopsy in the context of a previously diagnosed pulmonary embolism with a persistent pulmonary infiltrate; one patient had undergone thoracotomy for a pneumothorax with persistent air leak, during which time a pulmonary nodule had been sampled; and a third patient had undergone a biopsy of a pulmonary nodule following a liver transplant. One additional patient with DAD and a history of dermatomyositis underwent open-lung biopsy for the abrupt worsening of chronic pulmonary infiltrates and was excluded from further analysis.
Among the remaining 16 cases with diffuse lung infiltrates and acute respiratory failure, eight patients had known causes for ARDS and were classified as non-AIP ARDS (ARDS group). Causes of ARDS in this group included the following: pneumonia and sepsis (n = 6), cocaine use (n = 1), and carmustine chemotherapy (n = 1). The eight remaining cases had organizing DAD plus no identifiable cause for lung injury, and were classified as AIP (AIP group). In the AIP group, all bacterial, fungal, and viral culture findings as well as serologies had been negative; each patient had failed to respond to antimicrobial therapy.
The medical records from the 16 cases above were analyzed for the following: demographics; time from symptom onset to hospital admission; PaO2/fraction of inspired oxygen (FIO2) ratio at presentation to the ICU; time from hospital admission to open-lung biopsy; steroid use and dose; sequential organ failure assessment (SOFA) score of organs other than the lungs9 ; hospital mortality; and hospital length of stay. The outpatient medical records from our hospital of survivors in the AIP group were retrospectively analyzed for the presence of dyspnea, limitation in activities of daily living, and pulmonary function data recorded during any follow-up visits.
Given the small sample sizes and the fact that most of the continuous data were skewed, we did not feel that the data were normally distributed. As such, we compared all continuous outcomes, such as mean time from onset of symptoms to hospital admission and mean time from hospital admission to open-lung biopsy, using the Mann-Whitney U test. To compare the mortality in the AIP group with the ARDS group, we used the Fisher exact test.
We also compared the mortality of our patients with AIP to the published literature. This comparison was done using the Fisher exact test to compare our mortality rate to the previously published rate. Information regarding 82 patients with AIP from previous reports tabulated by Voulekis et al8
was combined with information on 13 additional acutely hospitalized patients contained in that report to yield a total of 29 survivors among 95 patients (mortality rate of 69.5%). Information from a recently published series by Ichikado et al10
was not included in this analysis, as 7 of 31 patients contained in this series had been reported previously,11
and were included in the analysis by Vourlekis et al.8
In addition, the analysis by Ichikado et al10
did not specify which patients had been reported earlier. All analyses were done using SAS software (SAS Institute; Cary, NC). For every analysis, the
level was set at 0.05.
| Results |
|---|
|
|
|---|
|
|
All patients in both groups required mechanical ventilation for respiratory failure prior to the performance of the open-lung biopsy, except for two patients, one in each group (patient 1 in the AIP group and patient 7 in the ARDS group). More patients in the AIP group received high-dose steroid treatment after open-lung biopsy (n = 8; mean methylprednisolone dose, 280 ± 143 mg/d) than in the ARDS group (n = 2; mean methylprednisolone dose, 180 ± 84 mg/d).
Organ failure other than respiratory was uncommon in the patients with AIP. Four of eight patients with AIP had an extrapulmonary SOFA score of 0 during their ICU stay. The remaining four patients all had a highest extrapulmonary SOFA score
4 during their ICU stay: one patient had a SOFA score of 1 (renal SOFA score of 1); one patient had SOFA score of 2 (renal SOFA score of 1, and hepatic SOFA score of 1); one patient had a SOFA score of 3 (renal SOFA score of 2, and hepatic SOFA score of 1); and one patient had a SOFA score of 4 (coagulation SOFA score of 4).
The mortality in the AIP group (12.5%) was lower than the mortality in the ARDS group (50%). This difference was not statistically significant using the Fisher exact test (p = 0.28). The mortality rate of our population (12.5%) was significantly different than the published rate of 69.5% (p = 0.0025). There was a trend toward a shorter hospital length of stay in AIP survivors than in ARDS survivors (25 ± 13.6 days vs 40.5 ± 15.6 days, p = 0.11).
Two of seven AIP survivors were unavailable for follow-up. The remaining five patients were followed up for a mean of 7.6 ± 3.5 years (range, 3 to 11 years). All five patients reported significant clinical improvement following discharge, and none experienced a relapse or had significant limitations in activities of daily living resulting from dyspnea. Available chest radiographs were clear except for mild, basilar scarring. Follow-up pulmonary function studies in four patients showed a mild restrictive defect, with total lung capacity of 68 ± 6.2% predicted, performed at a mean of 21.5 ± 26.0 months following discharge. Mean FEV1 was found to be 1.82 L (range, 1.58 to 2.06 L), with a mean FEV1/FVC ratio of 68% (range, 61 to 76%) in the two patients in whom this information was available.
Information indicative of patient quality of life was available on five of seven survivors with AIP: patient 1 was without limitations, enjoying biking and skiing at 5 years of follow-up; patient 2 was working in her original occupation at 8 years of follow-up; patient 4 had retired at age 58 years, was not dyspneic but had acquired end-stage renal failure requiring dialysis at the 5-year follow-up; patient 6 reported no shortness of breath but had lost his job as a painter at 2 years of follow-up; and patient 8 was determined to be asymptomatic from a pulmonary standpoint at 11 years of follow-up from several emergency department visits during that period for sexually transmitted diseases.
| Discussion |
|---|
|
|
|---|
Organizing DAD represents the fibrinoproliferative phase of acute lung injury and is commonly seen in patients with ARDS during the second and third weeks of their illness.12 DAD is characteristically seen throughout both lungs in patients with AIP and ARDS. Three of 20 patients in our series had had only localized infiltrates. Histopathologically, such regional alveolar damage is an occasional finding that is believed to represent a nonspecific response to an array of insults.12 13 Some patients with underlying interstitial lung disease acutely deteriorate and are found to have DAD superimposed on chronic fibrotic changes at open-lung biopsy.14 This was the case in one patient in our series with a history of interstitial lung disease due to polymyositis, who then acquired ARDS.15
AIP involves onset of the new, diffuse lung infiltrates of no identifiable cause with progressive, hypoxemic acute respiratory failure in a previously healthy patient. The histopathology, organizing DAD, is identical to that is seen in fibrinoproliferative ARDS. Nevertheless, AIP and fibrinoproliferative ARDS are believed to represent two distinct entities.3 5 8 The ARDS group seen in this series presented in a time frame typical of patients with ARDS, developing acute respiratory failure 2.2 ± 1.0 days following the onset of risk factors, mostly pneumonia and sepsis.16
In contrast, patients with AIP present to the hospital after a longer prodromal period, consisting of nonspecific symptoms suggestive of a viral illness, including cough, fever, and dyspnea. The mean time from the onset of these symptoms to hospital admission of 16.8 ± 15.7 days in our AIP group is similar to values reported elsewhere, as is the fact that approximately one half of the patients in our series presented with
7 days of symptoms.5
The statistically significant difference between ARDS and AIP groups in the time to hospitalization from the onset of disease, while typical of each condition, has not been reported previously in series of patients with AIP, and is consistent with the concept that AIP and ARDS constitute two separate diseases.
In our series, patients in the AIP group underwent open-lung biopsy sooner than patients with ARDS, probably due to the lack of a presumptive risk factor for the development of ARDS, and sooner than the only other series that reported this interval.2 Some patients with AIP can become too ill to tolerate an open-lung biopsy and go undiagnosed; hence, the incidence of AIP is probably underestimated, as the diagnosis is applied only to those in whom an open-lung biopsy has been performed. The term idiopathic ARDS has been applied to patients who meet the diagnostic criteria for AIP yet lack a tissue diagnosis.17
The mortality from AIP has generally been reported to be in excess of 50%. We used a mortality rate of 69.5% for comparison with our AIP group, as it reflects all previous series of uniquely reported patients who were acutely hospitalized for AIP.8 The mortality rate seen in our series is significantly lower than that reported in the combined previous series. The absence of significant extrapulmonary organ failure, as evidenced by low extrapulmonary SOFA scores, contributed to the low mortality rate. The mortality from ARDS has declined to 31 to 40% in the last decade.6 7 The low mortality seen in our AIP group might reflect this overall decline. However, most of our patients with AIP were seen in the early to mid-1990s and did not receive low tidal volume ventilation, one factor known to have contributed to the decreased mortality rate.7
In contrast with some series,8 10 11 none of our cases were diagnosed at autopsy. Some cases of idiopathic ARDS may have gone undiagnosed as AIP because an autopsy was not performed. Other series reporting high mortality rates had either no or few autopsy diagnoses.2 Thus, it seems unlikely that our low mortality rate could have been accounted for by this fact alone.
Alternatively, the low hospital mortality in our series may be a sampling error related to the small numbers in our AIP group. Yet, our series is similar in size to those in many previous reports. The low mortality could have been influenced by more severe cases of idiopathic ARDS having gone undiagnosed as AIP. Recently, findings on high-resolution CT of the chest, such as extensive ground-glass attenuation or airspace consolidation with traction bronchiectasis, have been associated with more severe disease and a higher mortality.10 High-resolution CT scans were not performed on our patients. Nevertheless, our patients with AIP were severely ill on presentation to the ICU, as evidenced by a mean PaO2/FIO2 ratio of 113.5, values similar to those seen in series where the mortality rate was high.8 Hence, it is unlikely the low mortality rate seen in our series was significantly influenced by the lack of a tissue diagnosis in severely ill patients with idiopathic ARDS.
The selection bias introduced by the need to obtain an open-lung biopsy in severely ill patients with idiopathic ARDS is a confounding variable in all reported series. It is unknown, for example, if series reporting a high AIP mortality rate preferentially selected patients for open-lung biopsy who were more likely to die, or do not select patients for biopsy who are believed to be responding clinically and hence more likely to live.
The management of patients with AIP is supportive. The role of high-dose steroids in this condition has not been prospectively evaluated and remains unclear.18 Corticosteroids have been shown to have a beneficial impact on the outcome of patients with ARDS and fibrinoproliferation.19 The earlier and more frequent use of corticosteroids in the AIP group, as compared with the ARDS group and other AIP series, may have favorably influenced the outcome seen in our patients with AIP when compared with other series.
Recently, a mortality rate of 14.3% was seen among seven patients with AIP in Korea,20 confirming the validity of our observation. Some investigators have suggested that genetic factors may play a role in survivorship in AIP. Although our series of patients with AIP had a disproportionate number of African Americans (37%) compared to the overall population of the United States, this was probably insufficient to have played a major role in the favorable mortality rate observed.
Previous studies18 21 22 reported a good overall prognosis in hospital survivors, with improvements in lung function, near-resolution of pulmonary infiltrates, and a lack of disease recurrence. Nonetheless, recurrence of disease and evolution into chronic interstitial lung disease has been reported.8 23 In our series, there was no evidence of disease recurrence among the five patients in whom follow-up was available.
In addition, follow-up in five of seven survivors for a mean of 7.6 years (range, 3 to 11 years) indicated the absence of significant dyspnea, a mild restrictive defect on pulmonary function testing (mean total lung capacity, 68 ± 6.2%) and only minimal residual fibrosis on chest radiography. No patient went on to acquire chronic, progressive interstitial lung disease, and most were found to be leading active lives.
In summary, our series describes a group of patients with AIP with a low hospital mortality and no evidence of either recurrent or progressive disease. Our 12.5% mortality rate contrasts favorably with earlier series in the medical literature. Because the diagnosis of AIP requires the performance of an open-lung biopsy in a group of patients who are critically ill, selection bias may be present in this and in previous series. It is unknown how many patients with idiopathic ARDS go undiagnosed as AIP. The determination of the actual mortality rate of AIP awaits further study, and would likely necessitate the acceptance of a clinical diagnosis of idiopathic ARDS as indicative of AIP.
| Footnotes |
|---|
This study was supported by ALI SCOR P50HL60289.
Received for publication August 6, 2002. Accepted for publication January 29, 2003.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. Hyzy, S. Huang, J. Myers, K. Flaherty, and F. Martinez Acute Exacerbation of Idiopathic Pulmonary Fibrosis Chest, November 1, 2007; 132(5): 1652 - 1658. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. G. Parambil, J. L. Myers, M.-C. Aubry, and J. H. Ryu Causes and Prognosis of Diffuse Alveolar Damage Diagnosed on Surgical Lung Biopsy Chest, July 1, 2007; 132(1): 50 - 57. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Y. Suh, E. H. Kang, M. P. Chung, K. S. Lee, J. Han, M. Kitaichi, and O J. Kwon Early Intervention Can Improve Clinical Outcome of Acute Interstitial Pneumonia Chest, March 1, 2006; 129(3): 753 - 761. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Ichikado, M. Suga, H. Muranaka, Y. Gushima, H. Miyakawa, M. Tsubamoto, T. Johkoh, N. Hirata, T. Yoshinaga, Y. Kinoshita, et al. Prediction of Prognosis for Acute Respiratory Distress Syndrome with Thin-Section CT: Validation in 44 Cases Radiology, December 1, 2005; 238(1): 321 - 329. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. I. Schwarz and R. K. Albert "Imitators" of the ARDS: Implications for Diagnosis and Treatment Chest, April 1, 2004; 125(4): 1530 - 1535. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |