Serum Lactate Dehydrogenase Level As a Predictor Of Survival For Patients With Acute Exacerbation of Idiopathic Interstitial Pneumonias Undergoing Polymyxin B-Immobilized Fiber Column-Direct Hemoperfusion


 BackgroundAcute exacerbation of chronic fibrosing idiopathic interstitial pneumonias (AE-IIPs) is known to be associated with a poor prognosis. In Japan, polymyxin B-immobilized fiber column-direct hemoperfusion (PMX-DHP) therapy is often used along with corticosteroids for the treatment of patients with AE-IIPs. However, the serum marker that predicts treatment response to PMX-DHP is still unknown. Our aim was to evaluate the serum marker that predicts the outcome in patients with AE-IIPs undergoing PMX-DHP therapy.MethodsWe retrospectively collected the medical records of 104 patients who developed AE-IIPs for the first time and visited Juntendo University Hospital between April 2009 and July 2020. Among these patients, 33 patients who received PMX-DHP were identified. Among the patients who received PMX-DHP, 18 patients who survived for over 30 days since the initiation of PMX-DHP therapy were classified into the “survival group,” and 15 patients who succumbed to the condition in less than 30 days were categorized into the “non-survival group.” We evaluated data on the patients’ background, survival, and differences in the serum markers associated with AE-IIPs and in oxygenation markers between the groups.ResultsAmong patient characteristics, median forced vital capacity was significantly higher in the survival group than in the non-survival group. Among serum markers, 7 days after the initiation of PMX-DHP therapy, serum lactate dehydrogenase (LDH) levels, white blood cell counts, and D-dimer levels were significantly higher and serum lymphocyte count was significantly lower in the non-survival group than in the survival group. PaO2/FiO2 ratio (P/F ratio) and alveolar-arterial oxygen difference (Aa-DO2) were significantly different between the two groups. The changes in LDH level (ΔLDH), P/F ratio, and Aa-DO2 were significantly different from day 1 to 7 days after the initiation of PMX-DHP between the two groups. Multivariate analysis revealed that ΔLDH was the only risk factor associated with poor prognosis. The cut-off value of ΔLDH was calculated as −8. The median survival time (MST) was significantly longer in patients with low ΔLDH values (<−8) than in those with high ΔLDH values (>−8).ConclusionsChanges in serum LDH levels are reasonable markers for evaluating the prognosis of PMX-DHP therapy.

the rst-line treatment for AE-IPF [4]. Although almost all patients who developed AE-IPF receive steroid pulse therapy, many patients do not recover because of respiratory failure. Thus, in Japan, we often try to combine steroids with immunosuppressants, including cyclophosphamide, cyclosporine, and tacrolimus; recombinant thrombomodulin; and polymyxin B-immobilized ber column-direct hemoperfusion (PMX-DHP) when treatment of AE-IIPs with corticosteroid alone does not yield satisfactory results [4]. However, recently, intravenous administration of cyclophosphamide and recombinant thrombomodulin for the treatment of AE-IPF was reported to have detrimental effects by a prospective study and a propensity score much analysis [5,6]. Therefore, we will not use both intravenous cyclophosphamide and recombinant thrombomodulin for the treatment of AE-IIPs.
PMX-DHP is usually used for the treatment of gram-positive or gram-negative sepsis in Japan [7]. PMX-DHP treatment has been reported to improve respiratory failure in patients with acute respiratory distress syndrome (ARDS) [8,9]. Because both ARDS and AE-IPF are known to have similar pathogenesis, that is, diffuse alveolar damage, PMX-DHP was considered to treat AE-IPF. Several clinical retrospective studies have reported the effectiveness of PMX-DHP against AE-IPF [10][11][12]. However, the effects or prognostic factors after the initiation of PMX-DHP therapy are still unknown.
Our aim was to determine the prognostic factors in patients with AE undergoing PMX-DHP treatment. We focused on AE in patients with chronic brosing idiopathic interstitial pneumonias (CF-IIPs) including IPF, idiopathic nonspeci c interstitial pneumonia (INSIP), and unclassi ed IIPs (UCIP) according to the American Thoracic Society (ATS) and European Respiratory Society (ERS) IIP classi cation statement published in 2013 [13] and our previous report [14].

Patient selection and evaluations
We collected data from patients with CF-IIPs who developed AE (AE-IIPs) for the rst time and visited Juntendo University Hospital between April 2009 and July 2020.
AE-IIPs was diagnosed on the basis of the criteria de ned by ATS in 2011 and our previous report: previous or concurrent diagnosis of any of the CF-IIPs, acute (within 1 month) worsening or development of dyspnea that typically lasted for one month, and identi cation of new bilateral ground glass opacity and/or consolidation superimposed on a background pattern associated with CF-IIPs on high resolution computed tomography (HRCT), and deterioration that is not fully explained by cardiac/renal failure or uid overload [15].
Among the patients who developed AE-IIPs for the rst time, the patients who received PMX-DHP were selected and were divided into two groups depending on the survival time from the initiation of PMX-DHP: the "survival group" and "non-survival group." Patients who survived for > 30 days were included in the former group, and the patients who survived for < 30 days were categorized into the latter group. Data on patient characteristics, treatment for AE-IIPs, serum markers associated with AE-IIPs (including Krebs von den Moreover, among the patients who received PMX-DHP, 18 patients survived for > 30 days after the initiation of PMX-DHP treatment (long survival group) and 15 patients survived for < 30 days (non-long survival group) ( Figure. 1). There were no signi cant differences in the patient characteristics, including age, sex, smoking history, prior treatment for baseline IIPs, gender, age, and physiology (GAP) index for the severity of interstitial pneumonia 1-6 months before the development of AE, and diffusing capacity.
In contrast, median forced vital capacity (FVC) was signi cantly higher in the survival group than in the non-survival group.

Characteristics of acute exacerbation
We previously reported differences in survival among the clinical AE-IIP types, including idiopathic, infection-triggered, and non-infection-triggered AE-IIPs, and evaluated the difference in survival among these categories in the current study. Twelve patients were considered to have idiopathic AE-IIPs, 11 were suspected with infection-triggered AE-IIPs, and ve were diagnosed with non-infection-triggered AE-IIPs.
Proportions of patients with these three categories of AE-IIPs were signi cantly different between the survival and non-survival groups.
Subsequently, we compared the treatment for AE-IIPs between the two groups. All patients received methylprednisolone and antibiotics. Eighteen patients (54.55%) received immunosuppressants, including cyclophosphamide, tacrolimus, or cyclosporine. However, there was no signi cant difference in the frequency of receiving immunosuppressants between the survival and non-survival groups. Then, the duration since the development of AE-IIPs to the initiation of PMX-DHP treatment was evaluated. PMX-DHP treatment was initiated within 48 h of the development of AE-IIPs in 21 patients (63.64%). The proportion of patients who received PMX-DHP treatment within 48 h tended to be higher in the survival group than in the non-survival group (p = 0.0627).

Serum markers and oxygenation associated with AE-IIPs
We evaluated the serum markers associated with IIPs, including serum levels of KL-6, SP-D, LDH, CRP, and D-dimer and counts of WBCs, neutrophils, lymphocytes, and platelet, prior to the development of AE-IIPs, on the day of initiation of PMX-DHP treatment, and 7 days after the initiation of PMX-DHP treatment and compared these markers between the survival and non-survival groups (Table. 3 and Supplementary   Table 1). Then, the changes in clinical parameters differences, particularly LDH and lymphocyte, between rst day and 7 were shown in Fig. 2A and 2B. There were no signi cant differences in all marker levels/counts prior to the development of AE-IIPs and on the rst day of the PMX-DHP treatment. In contrast, 7 days after the initiation of PMX-DHP treatment, serum LDH levels, WBC, neutrophil, and lymphocyte counts, and D-dimer levels differed signi cantly between the non-survival and survival groups. Serum KL-6 and SP-D levels tended to be higher in the non-survival group than in the survival group. Moreover, we analyzed the difference in the changes in serum marker levels/counts between baseline and the rst day of PMX-DHP treatment and between the rst day of PMX-DHP treatment and 7 days after the initiation of PMX-DHP. Although there were no signi cant differences in the serum marker levels/counts between baseline and the rst day of PMX-DHP treatment, the changes in LDH levels (ΔLDH) and lymphocyte counts (ΔLymphocyte) differed signi cantly between the two groups.
In addition, we evaluated the PaO 2 /FiO 2 ratio (P/F ratio) and alveolar-to-arterial oxygen difference (Aa-DO 2 ) on the rst day of PMX-DHP and 7 days after PMX-DHP initiation (Table 4). Then, the change in P/F ratio between rst day and 7 was shown in Fig. 2C. At the initiation, although the P/F ratio tended to be higher in the non-survival group than in the survival group, there was no signi cant difference in Aa-DO 2 between the two groups. The P/F ratio was signi cantly higher in the survival group than in the nonsurvival group at 7 days after the initiation of PMX-DHP (p = 0.0019). Aa-DO 2 was also signi cantly higher in the non-survival group than in the survival group (p = 0.0032). Finally, we evaluated the differences in the two oxygenation markers (ΔP/F ratio and ΔAa-DO 2 ) between 7 days after the initiation of PMX-DHP treatment and the rst day. The ΔP/F ratio was signi cantly higher in the survival group than in the non-survival group at 7 days after the initiation of PMX-DHP treatment (p = 0.0235). ΔAa-DO 2 value was signi cantly higher in the non-survival group than in the survival group (p = 0.0409). Table 4 The difference in the change in PaO 2 /FiO 2 ratio and alveolar-to-arterial difference for oxygen between survivor and non-survivor

Discussion
To our knowledge, this study is the rst to evaluate prognostic factors after the initiation of PMX-DHP treatment in patients with AE-IIPs. Our main results are as follows: (1)  Many serum markers associated with interstitial pneumonia have been reported. Among them, serum KL-6 and SP-D levels were speci cally related to interstitial pneumonia progression [15,16]. Serum LDH levels, WBC counts, and CRP levels are also associated with interstitial pneumonia progression [17][18][19]; however, there have been no reports regarding the association between serum markers and development or progression of AE-IIP. We have previously reported that the elevation of serum SP-D levels was signi cantly related to drug-induced interstitial lung disease, particularly the development of diffuse alveolar damage and poor prognosis in patients with advanced non-small cell lung carcinoma [20]. In a previous study, SP-D change was de ned as the difference in SP-D level between onset of the development of drug-induced interstitial lung disease and baseline. SP-D was also associated with lung injury in a rat brosis model [16]. However, there was no signi cant association between the prognosis of AE-IIPs and SP-D changes in our study. Changes in serum LDH levels were reported to be signi cantly associated with poor prognosis after the development of AE-IPF [18,21,22]. Kishaba et. al. demonstrated that the cut-off level of ΔLDH level of > 80 IU/L, as measured within 2 weeks, was related to poor prognosis. Furthermore, changes in LDH were reportedly associated with AE-related poor prognosis in patients with connective tissue disease associated with interstitial pneumonia [23]. Our study suggests that LDH change was useful as a prognostic factor. Although there is a difference in the two measurement points between our study and other studies, these studies focused on the same marker, LDH; therefore, these publications might support our results. Therefore, the evaluation of changes in serum LDH levels may be a reasonable and convenient method to evaluate the development or progression of AE-IIPs worldwide. LDH is present in the cytoplasm in any cell in any organ. Damage to alveolar cells is known as the rst step in the pathogenesis of AE-IIPs. This is the reason why serum LDH levels are elevated in the early phase of AE-IIPs. We particularly focused on changes in the serum marker levels/counts associated with AE-IIPs at the initiation of PMX-DHP and 7 days after PMX-DHP initiation. In this study, there was a signi cant difference in the patients' background, especially baseline FVC between the two groups. However, we did not evaluate the prognostic factors using baseline FVC levels because several patients who developed AE-IPF had not undergone pulmonary function tests, including measurement of FVC and diffusing capacity prior to the development of AE-IIPs in real time. To predict the outcome after the development of AE-IIPs in all patients with AE-IIPs, we determined prognostic factors using from serum markers after the development of AE-IIPs.
Subsequently, we also focused on the time of initiation of PMX-DHP since the development of AE-IIPs. In a previous retrospective study, the survival time was signi cantly longer in patients who received PMX-DHP treatment within 48 h than in those who received PMX-DHP after 48 h [24]. Similar ndings were obtained in the current study. Administration of PMX-DHP therapy may be recommended as soon as possible from the onset of AE-IIPs. Although the role of PMX-DHP is to remove endotoxins of gramnegative bacteria from patients with sepsis, it lowers the serum levels of cytokines, including interleukin-12, vascular endothelial growth factor (VEGF) [25], and high mobility group box-1 protein (HMGB-1) [26] in patients with AE-IIPs. Thus, we need to perform PMX-DHP therapy as soon as possible because PMX-DHP cannot purify serum completely after progression of AE-IIPs. The pathogenesis of AE is known to have two phases: the early (in ammation) and late ( brosis) phases. It is important to initiate PMX-DHP before transition to the brosis phase [3].
Several studies have reported the e cacy of PMX-DHP therapy for respiratory failure, including oxygenation, Aa-DO 2 , and P/F ratio. In this study, both P/F ratio and Aa-DO 2 at 7 days after the initiation of PMX-DHP were related to poor prognosis. We performed multivariate analysis using two variables, either of oxygenation markers (P/F ratio or Aa-DO 2 ) and change in the serum LDH level however, both the analyses showed that LDH was still a signi cant risk factor associated with poor prognosis.
This study had several limitations. First, the study was retrospective and small. The incidence of AE-IIPs was reported to be 10-20% in patient with CF-IIPs; thus, the development of AE-IIPs is rare. [3]. Second, almost all patients were diagnosed and classi ed into each type of CF-IIPs based on HRCT ndings, serological ndings, and medical examination via an interview, but not based on histological ndings.
Therefore, the patients' backgrounds were heterogeneous, which may have introduced selection bias.
Third, LDH is present in the cytoplasm in every cell in almost all organs. Thus, serum LDH can be released from the cytoplasm by any cell damage in any organs other than the lungs, particularly the liver. We need to distinguish between lung disease and diseases of any other organs when serum LDH levels are elevated. Fourth, in this study, changes in serum lymphocyte counts were signi cantly associated with the prognosis of AE-IIPs. However, steroid therapy was initiated before the initiation of PMX-DHP therapy in several cases. Therefore, changes in WBC counts, including neutrophil and lymphocyte counts, may be in uenced by steroid treatment.

Conclusion
We focused on the serum brotic and oxygenation markers associated with AE-IIPs for evaluating the effect of PMX-DHP on long survival. Measurement of serum LDH levels is an easier and more reasonable methods for the evaluation of prognosis after the initiation of PMX-DHP therapy than is that of other brotic markers, including serum KL-6 and SP-D levels. We may carefully observe serum LDH changes after the development of AE-IIPs and initiation of PMX-DHP. The ethics committee waived the requirement for informed consent because of the retrospective nature of this study.