Survival after repeated surgery for lung cancer with idiopathic pulmonary fibrosis: a retrospective study

The incidence of lung cancer is higher in patients with idiopathic pulmonary fibrosis (IPF) than in the general population; the relative risk of lung cancer in such patients ranges from 6 to 17% [1, 2]. In the general population, the likelihood of a new primary lung cancer developing after complete resection for an initial lung cancer has been reported to be 1% to 2% per patient per year [3, 4]. On the other hand, in patients with IPF, the cumulative rate of developing lung cancer has been reported to increase as the duration of follow-up increased (3.3%, 15.4%, and 54.7% at 1, 5, and 10 years, respectively) [5].

Thoracic surgeons, as well as medical and radiation oncologists, must often make difficult decisions in treating lung cancer patients with IPF because of the poor prognosis of IPF itself and the complications, such as acute exacerbation (AE), which arise after each intervention [5–9]. Several previous studies demonstrated a median survival time of 2 to 3 years after diagnosis in patients with IPF [10–14]. Sato and colleagues [15] reported that 9.3% of lung cancer patients with IPF developed AE after pulmonary resection. To estimate the risk of surgery, they proposed a risk score using clinical characteristics and surgical procedures. However, repeated surgical intervention was not included as a risk factor. With regard to the outcome of surgical intervention, several studies [9, 16, 17] reported 5-year survival rates of about 40% to 60% for stage I patients; therefore, surgical treatment for lung cancer with concomitant IPF might not be an absolute contraindication, as long as the patients are carefully selected. For lung cancer patients with IPF who undergo surgical interventions, many challenging problems, such as AE of IPF and high rates of second and third primary cancers, have been cited; however, to the best of our knowledge, there have been no studies focusing on the incidence of postoperative AE of IPF and the long-term outcome after a second pulmonary resection. Thus, the purpose of this study was to evaluate the outcomes and risks after a second pulmonary resection and to elucidate the implications of surgical interventions in lung cancer patients with IPF.

The medical records of all lung cancer patients admitted from 2001 to 2015 to the Division of Thoracic and Cardiovascular Surgery at Niigata University Hospital and the Department of Thoracic Surgery at Nishi-Niigata Chuo National Hospital were retrospectively reviewed; patients diagnosed with IPF before surgical treatment for lung cancer were identified. The eligibility criteria for surgical resection of lung cancer with IPF were: a resting partial pressure of arterial oxygen > 60 mmHg; predicted postoperative forced expiratory volume in 1 s > 1.0 L; clinically stable and symptomless IPF; and complete resection possible. Of a total of 108 patients enrolled in this study, 17 (15.7%) with IPF developed second primary lung cancers. Thirteen of these patients underwent a second pulmonary resection, 2 patients received radiotherapy, 1 patient received chemotherapy, and 1 patient had best supportive care. The institutional review board approved this study (Niigata University, 2302) and waived the requirement for informed consent because the study was a retrospective review.

Radiologic assessment of the preoperative conventional chest computed tomography (CT) or high-resolution CT (HRCT) of all patients was performed to confirm the following criteria for IPF: 1) CT patterns compatible with IPF, as proposed by the American Thoracic Society and the European Respiratory Society [18], with bilateral reticular opacities and/or honeycombing predominant in the peripheral, subpleural, and basal locations; and 2) absence of known causes of pulmonary fibrosis, such as hypersensitivity pneumonitis, pneumoconiosis, sarcoidosis, eosinophilic pneumonia, lymphangioleiomyomatosis, drug-induced lung disease, and collagen vascular disease. One thoracic radiologist (HI) and one thoracic surgeon (SS) who were blinded to the clinical data evaluated the preoperative chest CT scans.

The medical records were reviewed to obtain the: demographic and clinical characteristics; chest CT scan findings; pulmonary function test results, including percent vital capacity (%VC) and percent forced expiratory volume in one second (FEV1%); surgical procedure; histologic findings; morbidity within 30 days of surgery; postoperative AE of IPF; and survival. AE was defined as: 1) onset within 30 days after pulmonary resection; 2) increasing respiratory distress; 3) newly developed fibrosis, ground glass opacity, or infiltrates on chest radiograph; 4) decrease in the resting partial pressure of arterial oxygen > 10 mmHg; and 5) the absence of heart failure or infectious lung disease [19]. A recently reported scoring system was used to assess the 30-day risk of AE onset after pulmonary resection in lung cancer patients with interstitial lung disease (ILD) based on seven risk factors, including a history of AE of ILD, preoperative steroid use, elevated serum sialylated carbohydrate antigen, KL-6 level, surgical procedure, usual interstitial pneumonia (UIP) pattern on CT scan, male sex, and low %VC [15].

Pathologic cancer stage was determined using the 7th edition of the International Union Against Cancer tumor-node-metastasis staging system [20]. Information was obtained for all survivors, either during office visits or by telephone interviews with the patient or a relative. The criteria for the diagnosis of multiple primary lung cancers were those described by Martini and Melamed [21] in 1975: 1) different histology or 2) same histology, if the disease-free interval between the two lesions was at least 2 years or development of a new neoplasm from an in situ carcinoma and occurrence of the second tumor in a different lobe or lung, provided that extrapulmonary metastases and lymphatic involvement common to both tumors were excluded. Tumors were designated as ‘synchronous’ when detected or resected simultaneously and as ‘metachronous’ when the second tumor was found some time later.

The incidence of lung cancer is higher in patients with IPF than in those without IPF. Notably, Fujimoto and colleagues [22] reported a high incidence of second primary lung cancer in patients with IPF.

Lung cancer patients with IPF are more likely to develop severe morbidity and have poor outcomes after pulmonary resection. Considering the high recurrence rate and the poor prognosis for this patient population, operative indicators for major pulmonary resection in patients with lung cancer with IPF remain unclear. Kushibe and colleagues [23] reported that patients with IPF who had postoperative acute lung injury/acute respiratory distress syndrome had a significantly lower preoperative percent forced VC (%FVC) than those without such complications. IPF patients with a preoperative %FVC < 80% may not have an operative indication for lung cancer, and those with a preoperative %FVC ≥90% could have a good operative indication. Fujimoto and colleagues [22] suggested that patients with lung cancer invading the chest wall were excluded from surgery because chest wall resection is associated with major morbidity [24]. Pneumonectomy should be avoided for the same reason.

There have been several reports on the surgical outcomes of lung cancer patients with IPF [1, 7, 9, 15–17, 22, 23, 25]; however, to the best of our knowledge, the implications of repeated surgery in such patients have not been reported. The present study found a poor prognosis for patients with IPF even after complete repeated resection of lung cancer, with 3-year DFS and OS rates of 8.7% and 34.6%, respectively.

Recently, a large cohort study by Sato and colleagues [15] reported their derived scoring system for the 30-day risk of developing AE of IPF after pulmonary resection and classified patients into three risk groups (i.e., low, intermediate, and high). In the present study, higher rates of developing postoperative AE were found than cited in the previous report. According to Sato and colleagues, the predicted AE incidence was < 10% [95% confidence interval (CI): 0–10] in the low risk group and 10–25% (95% CI: 8.8–29.7) in the intermediate risk group, though 2 of 11 (18.2%) patients at low risk and 1 of 2 (50%) patients at intermediate risk developed and died of AE in the present study. The patients who developed postoperative AE at the initial surgery were compared with those who developed postoperative AE at the second surgery, and AE in the second surgery cohort developed with a significantly lower risk score than in the initial surgery cohort. The patients who did not undergo surgical treatment for second primary lung cancers were also examined, and no patients developed AE of IPF. Although the present study sample was very small, repeated surgery for lung cancer in patients with IPF could be a risk factor for AE.

The etiologic agents of AE of IPF after pulmonary resection remain unclear. Sakamoto and colleagues [26] reported the possible factors contributing to AE after surgery in patients with IPF: 1) high activity of the disease prior to the surgery; 2) oxygen supplementation at a high concentration during surgery; 3) surgical stress including mechanical ventilation-related lung injury; 4) complicating respiratory infections; 5) postoperative reduction of steroid dose; and 6) medications (anesthesia, anticancer drugs, and so on). Misthos and colleagues [27] investigated the possibility of an association between oxygen radical toxicity and the occurrence of AE of IPF. They found the following: 1) lung re-expansion after one-lung ventilation (OLV) provoked severe oxidative stress; 2) the degree of oxygen-derived free radicals generated was associated with the duration of OLV; 3) patients with lung cancer had higher production of oxygen-derived free radicals than the normal population; 4) tumor resection removes a large oxidative burden from the organism; 5) mechanical ventilation and surgical trauma are weak free radical generators; and 6) manipulated lung tissue is also a source of oxygen-derived free radicals, not only intraoperatively, but also for several hours later.

The present analysis showed that the rate of %VC decrease was significantly higher in patients with AE than in those without AE. %VC has been considered a reliable marker of fibrotic change [11, 13], and some previous studies [23, 25] reported that %VC had a significant and independent association with the development of AE. In the present study, the %VC of 1 patient with AE was not lower than 80%, but the rate of %VC decrease was 24.3%. Therefore, the AE incidence at the second surgery for lung cancer in patients with IPF is associated not only with a low %VC, but also with a high rate of %VC decrease. Also, among the pulmonary function tests, %DLCO has been considered a survival predictor [7, 28] and a reliable indicator of fibrotic change [12, 13, 29]. However, it was not included in the present study because the values of only 3 of 13 patients were available.

In the present study, the reason for the poor prognosis was a high rate of cancer recurrence, aside from development of AE. Although all patients were clinical stage IA or IB at the second surgery, 8 of 10 patients had cancer recurrence, except those who died of AE. Saito and colleagues [16] reported that the 5-year survival of lung cancer stage IA patients with IPF was 54.2%. Watanabe and colleagues [9] reported a 5-year survival of 61.6% after pulmonary resection for patients with stage I. Sato and colleagues [17] reported that the 5-year survival rates were 59% and 42% for pathologic stages IA and IB, respectively. Watanabe and colleagues [9] and Okamoto and colleagues [7] suggested that the frequency of cancer recurrence was higher in patients with IPF than in those without. Sato and colleagues [17] reported that recurrence was the main cause of death and posed a risk that was about twice as high as that of respiratory disorders; they underscored the importance of oncologic control for survival. In the present study, sublobar resection was performed in all patients, and lymph node dissection was performed in only 1 patient at the second surgery. Sato and colleagues [17] noted that stage IA patients who underwent wedge resection had a prolonged survival and were less likely to develop AE of IPF, but they had a higher cancer recurrence rate than patients who underwent lobectomy. Furthermore, patients who underwent segmentectomy had less favorable oncologic outcomes than patients who underwent lobectomy. At the second surgery, accurate pathologic staging might not be possible because lymph node dissection was not performed in almost all patients. Needless to say, it was necessary to consider the influence of lung cancer recurrence not only in the second surgery, but also at the initial surgery. Actually, of the 5 of 8 patients who developed recurrence, 3 were stage II and 2 were stage III at the time of the initial surgery.

With regard to distinguishing multiple primary lung cancers from primary lung cancer with intrapulmonary metastasis, the possible effects of multiple lung tumors, as defined by Martini and Melamed [21], were considered. Although it is important to distinguish a second primary cancer from local recurrence or metastatic disease, this is sometimes difficult and even impossible. Girard and colleagues [30] considered that biologic examinations could be performed, assuming that the independent tumor clones harbor distinct mutations. In the present study, the same histologic diagnoses at the initial and second surgeries were seen in 4 patients with squamous cell carcinoma and in 1 patient with adenocarcinoma. Among them, 2 patients with squamous cell carcinoma and the patient with adenocarcinoma were investigated, but they were negative for epidermal growth factor receptor mutation at both the initial and second surgeries.

Squamous cell carcinoma had a higher prevalence than the other histopathological types among the patients with IPF in the present study, consistent with the previous reports. The cause of the high prevalence of squamous cell carcinoma in IPF patients remains unclear. Song and colleagues [31] found many foci of squamous metaplasia in honeycombing epithelium, and Hironaka and Fukuyama [32] reported that IPF patients with lung cancer showed more frequent foci of squamous metaplasia than IPF patients without lung cancer. Calabrese and colleagues [33] reported the overexpression of squamous cell antigen, a serine protease inhibitor typically expressed by dysplastic and neoplastic cells of epithelial origin, more often in squamous cell tumors, in IPF. These pathological findings support the notion that IPF may be a precursor to the development of squamous cell carcinoma.

Postoperative development of AE and the long-term survival of patients with a second primary lung cancer with IPF who underwent repeated surgery were investigated. The main cause of their poor prognosis was cancer death, possibly related to sublobar resection. Repeated surgery for patients with lung cancer and concomitant IPF could increase the risk of AE development despite limited surgery. The rate of %VC decrease might be correlated with the incidence of AE. Although these results demonstrated that surgical intervention for multiple primary lung cancers might be contraindicated in patients with IPF, selection of patients who may benefit from such treatment is very important. To confirm these findings, a large, long-term, multi-center surveillance study will be required.