Disease activity of lung cancer at the time of acute exacerbation of interstitial lung disease during cytotoxic chemotherapy

Abstract Background The prognosis of lung cancer patients with interstitial lung disease (ILD) is poor, and acute exacerbation (AE) of ILD can occur during chemotherapy as a fatal adverse event. Although AE‐ILD development is correlated with various factors, no reports are investigating the disease activity of lung cancer at the time of AE‐ILD development. Methods All consecutive lung cancer patients with ILD who developed chemotherapy‐related AE‐ILD within 28 days after the last administration of cytotoxic chemotherapy between 2011 and 2020 were retrospectively reviewed. Results Among 206 lung cancer patients with ILD who were treated with cytotoxic chemotherapy, 30 patients were included. The median age was 72 years and all patients were men with smoking history. Usual interstitial pneumonia (UIP) and non‐UIP patterns of ILD was observed in 17 and 13 patients. Most of AE‐ILD occurred during second‐ or later‐line (22/30, 73.3%) and developed within first or second courses during chemotherapy (19/30, 63.3%). Regarding tumor response to chemotherapy at AE‐ILD development, majority of patients (18 patients, 60.0%) experienced progressive disease and only one patient (3.3%) experienced a partial response. Notably, 27 patients (90.0%) did not exhibit any tumor shrinkage of the thoracic lesions. Conclusion Lung cancer was uncontrolled with cytotoxic chemotherapy at the time of AE‐ILD development. Although AE‐ILD during chemotherapy has been generally discussed in terms of drug‐specific adverse effects, uncontrolled lung cancer may be also correlated with AE‐ILD development.


INTRODUCTION
Interstitial lung disease (ILD) is characterized by various degrees of inflammation and fibrosis, 1 and the presence of ILD is widely accepted as a risk factor for lung cancer development. [2][3][4] Lung cancer patients with ILD reportedly present with a very poor prognosis and acute exacerbation (AE) of ILD can sometimes develop during chemotherapy as a fatal adverse event. [5][6][7][8] As a result, multiple studies have investigated the safety and efficacy of therapeutic regimens for lung cancer patients with ILD. [9][10][11][12][13][14][15][16] Although AE-ILD development is reported to be correlated with various factors, including low vital capacity, radiologically usual interstitial pneumonia (UIP) pattern, and non-small cell lung cancer (NSCLC), [17][18][19] no reports have investigated the disease activity of lung cancer at the time of AE-ILD development. We speculated that the tumor response itself is inversely correlated with AE-ILD development.
Therefore, this retrospective study verifies the abovementioned hypothesis.

Patients
We retrospectively investigated all lung cancer patients with ILD who were treated with systemic cytotoxic chemotherapy between January 2011 and December 2020. ILD was diagnosed based on medical history, physical examination, and radiological abnormalities compatible with bilateral lung fibrosis including ground-glass opacity, consolidation, and/or reticular shadow. Patients who developed cytotoxic chemotherapyrelated AE-ILD were included. As per the methods in previous studies, this study included UIP pattern ILD and non-UIP pattern ILD. UIP or probable UIP was defined as UIP pattern ILD, while the other type of ILD was defined as non-UIP pattern ILD based on the International Consensus Statement. 20 The radiological diagnosis of ILD and AE-ILD was made based on the consensus of at least two board-certified chest physicians (A.S., T.B., O.T., R.O.) and chest radiologist (T.E.). Patient characteristics including sex, age, smoking history, performance status, tumor histology, disease stage, radiological classification, treatment history and results of laboratory data including C-reactive protein (CRP), Krebs von den Lungen-6 (KL-6) and surfactant protein-D (SP-D) at the time of initiation of last course of chemotherapy were investigated.

Chemotherapy-related AE-ILD and evaluation of disease activity
Chemotherapy-related AE-ILD was confirmed if all of the following four criteria were met according to previous reports 17,211 : acute worsening or development of dyspnea 2 ; high-resolution computed tomography (HRCT) findings indicating new bilateral ground-glass attenuations with/without nonsegmental consolidation superimposed on pre-existing interstitial shadows 3 ; deterioration not fully explained by cardiac failure or fluid overload, on the basis of the results of biochemical tests and echocardiography and subsequent clinical course; and 4 <4 weeks interval between the last administration of chemotherapeutic drugs and the onset of AE-ILD. In addition, patients who developed AE-ILD immediately after or during radiotherapy or immune-checkpoint inhibitors were excluded because radiation pneumonitis and pseudoprogression could not be completely denied. 22 At the diagnosis of AE-ILD, the percent change in intrathoracic lesion size from baseline or best response was evaluated using Response Evaluation Criteria in Solid Tumor (RECIST) version 1.1. 23

Overall survival
Overall survival (OS) after AE-ILD diagnosis and after initiating first line chemotherapy was investigated in all patients who developed chemotherapy-related AE-ILD. In addition, OS was also evaluated in all patients who did not develop any AE-ILD during their clinical courses. Death of chemotherapy-related AE-ILD was defined as death within 28 days of AE-ILD development because a previous report showed that about 60% of patients with idiopathic pulmonary fibrosis (IPF) died within a month of AE development. 24

Statistical analysis
Descriptive statistics were expressed as n (%) or median and range. Fisher's exact test or chi-squire test was used to compare categorical variables and the Mann-Whitney U test was performed to compare continuous variables. Overall survival was defined as the period from the diagnosis of chemotherapy-related AE-ILD or the initiation of first-line chemotherapy to the day of death of any cause, using the Kaplan-Meier method. The outcome was censored if a patient had not died at the time of the last follow-up. JMP 10 software (SAS Institute) was used for all statistical analysis. This study was approved by the Institutional Ethical Review Board (IRB) of Kanagawa Cardiovascular Respiratory Center, Yokohama, Japan (IRB: KCRC-19-0020).

RESULTS
Patient characteristics at the time of last chemotherapy immediately before AE-ILD Figure 1 shows the study flow. A total of 206 patients with ILD were treated with cytotoxic chemotherapy. During their clinical courses, 54 patients developed AE-ILD and 152 patients did not. Among 54 patients who developed AE-ILD, 30 patients were confirmed as developing cytotoxic chemotherapy-related AE-ILD. Table 1 shows the patient characteristics at the time of initiating last course of chemotherapy. All patients were male and smokers with a median age of 72 years. In most patients (25/30), performance status was 0-1. The UIP pattern of ILD was observed in 20 patients (66.6%), whereas non-UIP was present in 10 patients (33.3%). The histological subtypes of the tumors were adenocarcinoma in 10 patients, squamous cell carcinoma in seven patients, neuroendocrine cell cancer in seven patients, and other nonspecified tumors in six patients. The disease stage was stage 3 in 13 patients, stage 4 in 14 patients, and three patients experienced postoperative recurrence. With regards to cause of ILD, 29 patients had idiopathic ILD. More than half of patients had comorbidity such as diabetes mellitus and hypertension. The median level of KL-6, SP-D and CRP were 938 U/ml, 177.1 ng/ml and 2.85 mg/dl, respectively. Regarding previous treatment, three patients received immune-checkpoint inhibitors and one patient underwent thoracic radiotherapy.
Treatment regimen and disease activity of lung cancer at the time of AE diagnosis AE-ILD occurred during first-line treatment in eight patients, second-line treatment in 11 patients, third-line treatment in seven patients and fourth-line treatment or later in four patients, as shown in Table 2. Thus, AE-ILD mostly developed during second-or later-line chemotherapy (   showed that almost all patients (27/30, 90.0%) did not experience any tumor shrinkage, as shown in Figure 2.  Table S1.

DISCUSSION
The present study showed the following two findings.  cytotoxic chemotherapy (22/30, 73.3%), and within second courses of treatment (19/30, 63.3%). To the best of our knowledge, this is the first study investigating the disease activity of lung cancer at the time of AE-ILD development.
In the present study, among 30 patients who developed AE-ILD, majority of the patients (18 patients, 60.0%) presented with progressive disease according to the RECIST criteria, and almost all patients (27 patients, 90.0%) did not exhibit any tumor shrinkage of the thoracic lesions. These results may indicate that uncontrolled lung cancer is correlated with AE-ILD development. To our knowledge, there have been eight prospective studies which have investigated the safety and efficacy of first-line cytotoxic chemotherapy for lung cancer patients with ILD, as shown in Table 3. [9][10][11][12][13][25][26][27] These studies employed varying therapeutic regimens and slightly different definitions of AE-ILD in terms of the need for decreased partial pressure of oxygen in arterial blood; all studies showed a low rate of progressive disease (3.0%-19.0%) and good disease control rate (66%-98.2%) with acceptable rates of AE-ILD (range, 1.6%-12.1%). In contrast, there were four retrospective studies in total. These studies showed a high progressive disease rate (33.3%-69.6%) [14][15][16]26 and low response rate (8.6%-33%). [14][15][16]28 Of note, all four studies showed a relatively higher rate of development of AE-ILD (12.0-27%). Although AE-ILD has been generally discussed in terms of drug-specific adverse effects, [9][10][11][12][13][14][15][16]25 these reports and the results of our study indicate that uncontrolled lung cancer is correlated with the development of AE-ILD.
It remains unclear why lung cancer was uncontrolled in most of our patients at AE-ILD development. However, uncontrolled lung cancer may have the following two adverse effect on ILD. First, immune activations during disease progression of lung cancer may trigger AE-ILD. Generally, patients with advanced lung cancer are reported to be in a state of immune activation. [29][30][31] Further, lung cancer patients with ILD and high level of serum C-reactive protein before chemotherapy were more likely to develop AE-ILD. 32,33 In fact, our study showed that serum CRP level was relatively high at initiating last course of chemotherapy. Second, decreased FVC accompanied by uncontrolled lung cancer may also be a risk factor for AE-ILD. Uncontrolled lung cancer potentially causes bronchial obstruction, the emergence of pleural effusion, and a reduction in performance status, which lead to decreased FVC. A retrospective study showed that low FVC was a risk factor for AE-ILD in lung cancer patients with ILD. 17 Although this study contains some important findings, there were three limitations. First, this was a small-sized retrospective study at a single institution, which potentially leads to selection bias. The current study employed multiple chemotherapeutic agents of various treatment lines which showed diverse response rates and the incidence rate of AE-ILD. These factors make it difficult to perform statistical analysis. Therefore, future large-scale studies with matching conditions should be performed to confirm our results. Second, other diseases mimicking ILD or AE-ILD, such as carcinomatous lymphangitis, pulmonary infection or cardiac failure, could not be completely excluded because of the retrospective nature of this study. However, it would be difficult to perform invasive examination due to deteriorated oxygenation. Finally, pseudoprogression could not be completely ruled out, although the present study included AE-ILD during cytotoxic chemotherapy, not immune-checkpoint inhibitors.
In conclusion, lung cancer was mostly uncontrolled with cytotoxic chemotherapy at the time of AE-ILD development. Although AE-ILD during chemotherapy has been generally discussed in terms of drug-specific adverse effects, uncontrolled lung cancer may be also correlated with AE-ILD development.