The prognostic value of PET/CT in clinical stage I lung cancer patients: A propensity-match analysis.

Introduction The application of PET/CT directly improved treatment choice and management in 25% of non-small cell lung cancer patients and 29% of small cell lung cancer patients. However, the long-term outcome of altering the management of these patients remains unclear. The aim of this study was to compare the 5-year overall survival rates of two groups of clinical stage I lung cancer patients: those who received PET/CT and those who did not. Methods Data were obtained from the Taiwan Society of Cancer Registry. There were 6,587 clinical stage I lung cancer patients analyzed between 2009 and 2014 in this retrospective study. We performed propensity matching to reduce the bias; it resulted in both groups having 2,649 patients. We measured the 1, 3, and 5-year survival rates of all clinical stage I lung cancer patients and the survival rates of pathological I, II and III lung cancer patients and compared the survival rates between clinical stage I lung cancer patients with PET/CT scans and patients without PET/CT scans. Results The 1, 3, and 5-year survival rates of all clinical stage I lung cancer patients are 97.2%, 88.2% and 79.0%, respectively. The 1, 3, and 5-year survival rates are 97.0%, 88.2% and 79.8% in the PET/CT group and 97.5%, 88.1% and 78.2% in the no PET/CT group; there was no statistical difference (p= 0.6528). Conclusion Although stage I lung cancer patients who received PET/CT had their management strategies modified and avoided any unnecessary thoracotomies, our data showed that there was no 5-year survival benefit for these patients.


Abstract
Introduction The application of PET/CT directly improved treatment choice and management in 25% of non-small cell lung cancer patients and 29% of small cell lung cancer patients. However, the long-term outcome of altering the management of these patients remains unclear. The aim of this study was to compare the 5-year overall survival rates of two groups of clinical stage I lung cancer patients: those who received PET/CT and those who did not. Conclusion Although stage I lung cancer patients who received PET/CT had their management strategies modified and avoided any unnecessary thoracotomies, our data showed that there was no 5-year survival benefit for these patients.

Background
Lung cancer is identified as the leading cause of cancer death worldwide. 1 Around threequarters of lung cancer cases were diagnosed at a late stage and less than 20% of lung cancer cases were diagnosed at stage I. 2 Through the popularity of screening nowadays, we increasingly detect earlier stage nodules and the algorithms grow more complex. Both planning and prognosis are dependent on the precise staging. The pretreatment evaluation of stage I lung cancer includes pulmonary function tests, a bronchoscopy, mediastinal lymph node evaluation, brain MRI with contrast, and positron emission tomography/computed tomography (PET/CT). 3 Computed tomography (CT) with contrast provides excellent tumor information. However, it is hard to differentiate between benign and malignant lesions. 4 The combination of PET and CT scanners was introduced into clinical practice in 1998. 5 Previous studies showed PET/CT has high sensitivity (96.8%) and intermediate specificity (77.8%) for malignancy. 6 The application of PET/CT directly improved treatment choice and management in 25% of non-small cell lung cancer patients and 29% of small cell lung cancer patients. 7,8 Furthermore, PET/CT helps to differentiate between benign and malignant pulmonary nodules. 9 Though there are so many benefits after PET/CT evaluation, the long-term survival rates remained unclear.
For this study, we obtained data from the Taiwan Society of Cancer Registry (TSCR) over a 5-year period. We aimed to analyze if the introduction of PET/CT can provide better survival rate to clinical stage I lung cancer patients.

Database
This study was approved in our hospital's institutional review board. The population data was obtained from the TSCR. This data includes the entire population of 23 million people in Taiwan. This database includes registration files and original claims data for each patient. All the patients were strictly confirmed by tissue diagnosis. The following items were included in the study: age, gender, smoking status, cell type, operative method, clinical stage, pathologic stage and treatment.

Study Sample
This study searched data from the TSCR between January 2009 and December 2014. We identified patients who were diagnosed with lung cancer by the diagnostic codes C34.0, C34.1, C34.2, C34.3, C34.8, and C34.9. We identified a total of 64,918 patients with malignant lung neoplasm who received surgical treatment. (Fig 1.) There were 8,566 patients who were diagnosed at clinical stage I.
A total of 1,979 patients were excluded from the study. Among these, 1,210 patients had a missing follow-up 3 months post-operation; 625 had a missing pathological stage; 47 patients had a missing smoking status; 9 patients had a missing tumor size and 88 patients had missing lymph node data. Therefore, a total of 6,587 patients were enrolled into the study. There were 2,727 patients who received PET/CT, and the other 3,860 patients did not receive PET/CT. In order to reduce the bias, we used propensity matching of age, gender, smoking status, cell type and clinical stage. There were 2,649 patients in both groups after the propensity match.

Statistical Analysis
We used SAS software (SAS System for Windows, version 9.2; SAS Institute, Cary, NC) to perform the statistical analysis for this study.
The outcome measures for our study were 1, 3, and 5-year survival rates of all clinical stage I lung cancer patients; and survival rates of pathological I, II and III lung cancer patients who was diagnosed as clinical stage I and received surgery. We compared the survival rates between clinical stage I lung cancer patients with PET/CT scans and patients without PET/CT scans. Survival curves were plotted by the Kaplan-Meier method, and the difference in survival was calculated by the log-rank test. Univariate and multivariate analyses were performed with the Cox proportional hazards model using SAS software. Statistical analysis with a p value less than 0.05 was considered statistically significant.

Database
This study was approved in our hospital's institutional review board. The population data was obtained from the TSCR. This data includes the entire population of 23 million people in Taiwan. This database includes registration files and original claims data for each patient. All the patients were strictly confirmed by tissue diagnosis. The following items were included in the study: age, gender, smoking status, cell type, operative method, clinical stage, pathologic stage and treatment.

Study Sample
This study searched data from the TSCR between January 2009 and December 2014. We patients had missing lymph node data. Therefore, a total of 6,587 patients were enrolled into the study. There were 2,727 patients who received PET/CT, and the other 3,860 patients did not receive PET/CT. In order to reduce the bias, we used propensity matching of age, gender, smoking status, cell type and clinical stage. There were 2,649 patients in both groups after the propensity match.

Statistical Analysis
We used SAS software (SAS System for Windows, version 9.2; SAS Institute, Cary, NC) to perform the statistical analysis for this study.
The outcome measures for our study were 1, 3, and 5-year survival rates of all clinical stage I lung cancer patients; and survival rates of pathological I, II and III lung cancer patients who was diagnosed as clinical stage I and received surgery. We compared the survival rates between clinical stage I lung cancer patients with PET/CT scans and patients without PET/CT scans.
Survival curves were plotted by the Kaplan-Meier method, and the difference in survival was calculated by the log-rank test. Univariate and multivariate analyses were performed with the Cox proportional hazards model using SAS software. Statistical analysis with a p value less than 0.05 was considered statistically significant.

Discussion
Our study was a retrospective study investigating the prognosis of clinical stage I lung cancer patients in Taiwan after receiving PET/CT. Previous studies showed lots of prognostic factors in lung cancer such as age, gender, stage, performance status, tumor differentiation and lactate dehydrogenase. 10,11 Our study revealed that older age, male gender, current smoking status, SqCC cell type, clinical stage Ib and advanced pathological stage were independent factors of a poor 5-year survival rate. However, PET/CT was not independent factors. 63.2%). 12 Another study reported that the sensitivity and specificity of malignant solitary pulmonary nodules were 82% and 66% for CT, 88% and 71% for PET, and 88% and 77% for PET/CT. 13 Both studies indicated that the combination of PET and CT was effective for lung cancer screening. Therefore, PET/CT is commonly suggested after a CT examination. Not showed that PET/CT could contribute to 71.6% of the modifications of management strategies in lung cancer patients. 23 Our data indicated that clinical stage I lung cancer patients who received PET/CT had a higher lobectomy rate and a lower wedge resection rate. Moreover, patients were more likely to receive chemotherapy after receiving PET/CT.
We assumed that most people in our study were upstaged after receiving a PET/CT scan and that the management strategies altered as well.
In the past, we merely used history taking and CT with contrast to determine the clinical stage. While PET/CT was introduced for pretreatment evaluation, a high proportion of patients receiving PET/CT had stage migration. Theoretically, PET/CT could lead to better prognosis of these patients. However, our data showed that a higher 5-year survival rate is not a benefit. There were few studies reporting the effects of PET/CT on the prognosis of lung cancer patients. Although management strategies were altered after lung cancer patients received PET/CT, their pathological stage did not change. We only varied the clinical stage and considered the different managements of the disease, but the 5-year survival rate did not change even when management strategies altered after PET/CT. Previous studies only mentioned that PET/CT lead to a more accurate clinical stage and changed the management. Our study pointed out that PET/CT did not alter the prognosis of stage I lung cancer patients. Further prospective study is needed to confirm this result.
There are some limitations of our study. First, this is a retrospective study. Although there was a huge amount of data, a prospective study is more convincing. Second, there might be a selection bias due to PET/CT scans being expensive in Taiwan. Patients who could afford the cost of a PET/CT scan might be able to afford other precise medication and have a better outcome.

Conclusions
Although stage I lung cancer patients receiving PET/CT resulted in the modification of management strategies and the avoidance of unnecessary thoracotomy, our data showed that there was no 5-year survival benefit for these patients.

Clinical Practice Points
Previous studies claims that PET/CT has high sensitivity (96.8%) and intermediate specificity (77.8%) for malignancy. In non-small cell lung cancer patients, it improved treatment choice in 25%. However, in our study the application of PET/CT provides no 5year survival benefit in clinical stage I lung cancer patients.

Declarations
Ethics approval and consent to participate: Not applicable Consent for publication: Not applicable Availability of data and material: The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Competing interests: The authors declare that they have no competing interests.
Funding: The authors declare that they have no funding for this study.

Authors' contributions:
Overall conception and design: YF and BY. Technical design and analysis of data: WH, JY and HC. Contribution to acquisition of clinical data and interpretation and analysis of clinical data: CH and SH.
All authors critically reviewed the manuscript and revised it. All authors approved the final version of the manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Acknowledgment:
A special gratitude I give to the chairman of our department, Mr. Cheng, whose stimulating suggestions and encouragement helped me to coordinate my project, especially when writing this report.    Figure 1 Flow chart of patients through study The survival rates of clinical stage I lung cancer patients The survival rates of pathologic stage I, II and III lung cancer patients who was diagnosed as clinical stage I and received surgery