Study population was composed of 66 patients having open lung biopsy (OLB) for the differential diagnosis of interstitial lung diseases (ILD) from 1984) to 1995. After clinical evaluations including pulmonay function tests (PET) and HRCT were taken, OLB was performed at the determined sites in consideration of chest PA and/or HRCT findings. For pathologic proof, all pathologic specimens were reviewed by two pathologists in our hospital. The cases of interstitial pneumonia were consulted outside, mainly with Professor Kitaichi in Kyoto University, when they were not compatible with UIP, DIP or AIP. Six cases of NIP were pathologically confirmed by Professor Kitaichi and one cases of NIP by Dr. Katzenstein in the State University of New York, New York.

NIP is characterized by a variable degree of intersitial inflammation with or without fibrosis. However, it appeared temporally uniform within each case and areas of honeycombing were not found. Histologic features for a specific category of interstitial pneumonia were not present. These cases were divided into three groups according to Katzenstein’s criteria⁵⁾. Group I was characterized by a cellular interstitial pneumonia with littly fibrosis. The interstitial infiltrate was minimal to mild and consisted of small lymphocytes (Fig. 1). There were patch involvement and accentuation around bronchioles. Group II was characterized by a cellular interstitial pneumonia with significant fibrosis (Fig. 2). The alveolar septa were widened by cellular infiltrations, which consisted of lymphocytes, plasma cells and a few eosinophils. An area of germinal center formation was noted. There was a marked hyperplasia of alveolar lining cells. The fibroblasts were composed of mature collagen and proliferating fibroblasts. Large numbers of intraalveolar macrophages and fibroblast foci were found. Group III was characterized by architectural derangement due to dense interstitial collagenous fibrosis and less degree of inflammation (Fig. 3). This process was diffuse in one case and patchy in the other case. The alveolar septa were markedly thickened, resulting in narrowing of air spaces. The alveolar lining cells were hyperplastic and alveolar spaces were filled with alveolar macrophages. Areas of patchy interstitial infiltrate were noted, but their degree was mininal compared to that of group II.

Chart of seven patients were reviewed. A standardized history taking and physical examinations were performed by the charged physicians. The standardized history included duration of respiratory symtoms, past medical history, family history, smoking history, medication history and occupational history with exposures to organic or inorganic dusts or toxic fumes. Clinical histories were reviewed to identify any possible underlying diseases or other potential causes of pulmonary abnormalities. Respiratory questionnaire included dyspnea, cough, sputum, chest tightness and chest pain Rales, rhonchi and wheezing were also identified as respiratory signs.

Routine blood tests included CBC, liver function tests and kidney function tests. Serologic and immunologic tests included ESR, C-reactive proten (CRP), anti-nuclear antibody (ANA), rheumtoid factor (RF), double stranded-DNA (ds-DNA) antibody, LE cell test, immunoglobulin (Ig) G, A, M and complements (C3, C4).

Prior to OLB, forced vital capacity (FVC), forced expiratory volume at one second (FEVI), total lung capacity (TLC) and DLCO were measured with pulmonary function analyzer ERS-1500 (Fukuda-Sankyo, Tokyo). TLC was measured with a heliumdilution method. All data were expressed as percentages of predicted values. Diffusion capacity for carbon monoxide (DLCO) was measured by the single-breath method and expressed as a percentage of the predicted values.

Six Patients with NIP underwent bronchoalveolar lavage (BAL). One patient was excluded due to severe dyspnea. BAL was done as usual. The sites of BAL were decided in consideration of HRCT findings. Four times’ instillation of 50 ml sterile normal saline and gentle suction with negative pressure below 50 mmHg was done. Total cell count were calculated with a hemocytometer (American opics, USA). Cell pellets were sparated by centirifugation at 500 g for 5 min. Differential count of 500 cells was done on the slides prepared by cytocentrifuge and Diff-Quik stain.

PET and HRCT performed again at the 2nd week after OLB and then steroid a were begun to be administered if the patients did not need urgently tereatment. Prednisolone was given orally in a single dose with 1 mg/kg of body weight for 2 weeks to 4 weeks and then was tapered 10 mg every one or two weeks, if patients showed clinical response of improvement. At 4th week, 3rd, 6th, and 12th month after treatment with steroids, follow-up measurements of PET and HRCT were performed.

The clinical response was analysed by a subjective improvement of dyspnea and by the objective evaluation of PET and HRCT. The physiologic improvement was defined when the increment of FVC was more than 10%, or that of DLCO or TLC was greater than 20% compared to those of initial values. The stationary state was defined when the changes of FVC was less than 10%, or that DLCO or TLC was less than 20%.

SPSS/PC program (Chicago, Illinois, USA) was used. The differences between groups BAL cell analysis were compared with the nonparametric Mann-Whitney U test. The difference was considered significant when p value was less than 0.05. All results were expressed as mean ±SE.

Among the sixty-six patients having OLB for the differential diagnosis of interstitial lung disease (ILD) from 1984 to 1994, thirty-five patients were proved as IPF and 10 patients as collagen vascular disease-associated pulmonary fibrosis (CVD-PF). Among IPF, seven patients were diegnosed as NIP. The first case was diagnosed in June, 1992 and the lat one in May, 1994.

Female was the predominant gender of the patients with NIP: six patients were female and one patient was male. The mean age of the patients was 56 years and ranged from 43 to 69 years. One male patient was a current smoker, and the others were never-smokers. As for the undrlying and associated medical conditions, one patient had diabetes mellitus for about six years, an other one had mild bronchial asthma for three years and another patient had hypertension for three years.

As an initial symptom, five patients had systemic flu-like discomfort such as fever, myalgia and fatigue for a few days. Two patients had elevated body temperature (37.5°C and 37.6°C). Mean duration of symptoms was 4 months before diagnosis. They were followed up for more than 12 months, except one patient who was followed up for 1 month after discharge. However, she has been contacted by telephone till now at 16 months after discharge (Table 1).

The most common respiratory symptom was dyspnea, which all of the seven patients complained of The median grade of dyspnea was about 10 according to CRP classifications⁷⁾. Four patients complained of cough and chest tightness, three patients complained of sputum, not purulent but mucoid, and two patients had chest pain. On physical examinations, rale was audible in all patients. One patient had wheeze on expiation (Table 2).

On complete blood counts, two patients had a mild degree of anemia: 10.2% and 10.5% Hb. Peripheral blood morphology, RBC index, the level of serum iron. TIBC and ferritin indicated that their anemias were due to a complication of chronic diseases. In four cases, WBC elevated more than 10,000/mm³ (22,500/mm³, 19,400/mm³, 12,300/mm³). Differeential counts were within normal limit in all patients, except one patient who had eosinophilia with 9%. Platelet count was normal in all patients. One patient had and elevation of SGT (45.61 IU/ml), SGPT (36.81 IU/ml) and alkaline phosphatase (122.61 IU/ml). Kidney function tests were normal in all patient (Table 3).

ESR was elevated significantly in all patients with a mean of 41±4 mm/hr rage from 24 mm to 57 mm/hr. All of four measured cases had more than one positive value of one of six patients. LE cell was negative in all cases (Table 3). One patient had a positive ds-DNA antibody, but she did not have the other manifestations of SLE. The levels of IgG (1883±193 mg/dl), IgA (321±93 mg/dl), IgM (339±56 mg/dl), C3 (116.3±35 mg/dl) and C4 (31.2±18.0mg/dl) were within normal ranges in the three measured patients.

FVC decreased in six patients. Three patients showed a severe degree of restriction less than 50% of predicted value, two showed a moderated degree of restriction betwee 51% and 65% of predicted value and one showed a mild degree of restriction between 6% and 75% of predicted value. FEV1/FVC was normal in all cases. TLC ws measured in 6 cases. One male patient had normal ranges. of TLC. In five cases, TLC was decreased from mild to severe degree. DLCO decreased markedly in two of three measured patients, and the other patients could not perform PET adequately because of difficulty in breath holding (Table 4).

BAL was performed in 6 cases. Profiles of BAL cells were compared with those of normal controls which were published previously⁸⁾. The mean recovery rate was a mean of 43.6±4.6%, which was lower than that of the normal controls (p<0.05).

Total cell count was a mean number of 57.5±17.53×10⁶. which was about three times of normal value (p<0.01). On differential counts, the percentage of lymphocytes were 36.5±9.10% and eosinophils 4.8±2.59%. In the patients with NIP, there was a marked increase in the percentages of lymphocytes (p<0.01), neutrophils (p<0.01) and eosinophils (p<0.05) when compared to those of normal controls (Table 5).

A prominent finding of chest radiology was a patchy opacification of parenchyme in six patients. It bilaterally involved the middle (n=6) and lower (n=6) lung zones more than the upper lung zone (n=2) (Fig. 4). In one patient, the parenchymal abnormality was combined with opacifications and reticular densites. In the remaining one patient, no abnormalities were noted on chest radiograhp, but HRCT, obtained two days later, depicted pathy areas of ground-glass attenuation.

On initial HRCT scan, the most common findings were bilaterally patchy areas of ground-glass attenuation alone and/or areas of alveolar consolidation in five patients, without central or subpleural predominancy (Fig. 5). Irregular lines were present for two patients (29%) without honey-hcombing or traction bronchioloectasia on initial or follow-up HRCT scans. Enlarged lymph node was observed in two patients. Pleural effusin was not present in any patient. Follow-up CT scan revealed that the initial abnormalities had resolved completely in three patients within 13 months. In the other three patients, the overall extent of parenchymal abnormalities had improved.

Our cases were divided into three groups a ccording to Katzenstein’s criteria⁵⁾. Salient pathologic findings were summarized in Table 6. Only one case blonged to Group I, four cases were catergorized as Group II and two cases were in Group III.

Corticosteroid was trid on 6 patients. Because one female patient showed subjective improvement during two weeks after OLB, she did not receive steroids and was followed up till now without symptomatic aggravation. The average period from diagnosis to the last follow-up was 24.8 months. Six patients were regularly followed up form 12 months to 36 months. One patient was not followed up at our hospital after first month of treatment with steroids. By correspondence, dyspnea was revealed to the be makedly improved, but she was excluded from the data analysis.

All patients felt improvement dyspnea within 2 weeks to 8 weeks spontaneously or by trial with steroids. Five patients completely recovered to the previously well-being state before the illness within one to five months later. One female patient felt dyspnea to be gradually improved during 9 months’ trial with steroids, but not to disappear completely. The recovery of subjective symptoms took a mean duration of 4.2 months (Fig. 6). All of he six patients showed improvement of PET during follow-up. FVC and TLC were recovered to normal values spontaneously in one patient and by treatment with steroids in five patients (Fig. 7). The recovery of FVC took a mean duration of 6.3 months. However, DLCO has remained lower than the predicted value during follow-up. In two cases dyspnea had relapsed 4 months and 6 months after the discontinuation of steroids. Three patients remained asymtomatic and had normal FVC and TLC at 4th, 12th and 15th months after the discontinuation of steoids. One case, who needed corticosteroid, still remained with sub-normal lung functions at 30th month of follow-up period. She had a remarkable foci of BOOP on the pathologic findings and was classified as Group 3.