Over the last decade, the diagnosis and staging of non-small cell lung cancer (NSCLC) has significantly improved primarily with the use of fluoro-2-deoxyglucose positron-emission tomography (FDG-PET/CT). Previous randomized controlled trials have demonstrated the utility of FDG-PET/CT (typically in combination with CT) with respect to avoidance of futile surgery in patients with both early stage and more advanced NSCLC (1–3). The current guidelines from the National Comprehensive Cancer Network (NCCN) recommend FDG-PET/CT scans for staging of all newly diagnosed patients with NSCLC, and also as a potential tool to evaluate (4, 5) response to treatment as well as for ongoing surveillance (3).

Within Ontario, Canada, FDG-PET/CT became an insured indication for any new diagnosis of NSCLC within 2 years following a 2007 evidence review suggesting its utility (1). However, the downstream evidence to date that supports PET in terms of influencing staging, stage-appropriate therapy such as chemoradiation (CRT) and surgery for stage III disease, and potentially survival is limited. A subgroup analysis of the randomized PROCLAIM trial reported an improvement in both progression-free and overall survival in patients who had undergone a staging FDG-PET/CT scan, therefore underpinning the importance that FDG-PET/CT scans have as an early investigation for NSCLC (2). At a population-based level, the magnitude of effect of FDG-PET/CT on survival to our knowledge has not been previously investigated. Furthermore, information on the impact of FDG-PET/CT on therapy choice, and timing of FDG-PET/CT relative to curative treatment and outcomes has been limited.

Our objective was to conduct a population-based cohort study over the period of 2007-2017 in order to determine the magnitude of influence of FDG-PET/CT on both management and outcomes in patients with stage III NSCLC. Our hypothesis was that use of pre-treatment PET would result in improved survival as well as increased access to curative therapy such as CRT and surgery. Secondary objectives included the type and utilization of mediastinal biopsies, timing of FDG-PET/CT relative to initiation of any treatment in the entire cohort, as well as the years 2014-2017 using cut off dates of FDG-PET/CT scans ≤28 days prior to treatment start, 29-56 days prior to treatment, and >56 days prior to treatment.

Between 2007 and 2017, a total of 13 796 people were diagnosed with stage III NSCLC in Ontario, Canada ( Figure 1 ). A mean of 1254 diagnoses were made each year (range 1149-1391, Table 1 ). Males represented the majority of the cases at 54%. The age group with the largest proportion of diagnoses was 70-74 years, having 18% of cases. Figure 1 depicts the Cohort Identification diagram.

Pre-treatment FDG-PET/CT utilization steadily increased over the examined time period: 0% utilization was seen in 2007 prior to FDG-PET/CT-staging scans becoming standard of care. By 2017, utilization increased to 74% ( Figure 2 , Supplementary Table 4 ).

This report provides a retrospective look at the impact of pre-treatment FDG-PET/CT scans for staging of new cases of NSCLC in Ontario, Canada from 2007 to 2017. During that timeframe, utilization of pre-treatment FDG-PET/CT increased from 0% in 2007 to 74% in 2017. Mediastinal biopsy utilization increased from 41% to 53% in 2017, while mediastinoscopy usage decreased as more patients received EBUS biopsy. More stage III NSCLC patients who had FDG-PET/CT scans received at least one oncologic treatment, and received curative-intent treatment such as CRT +/- surgery. Overall survival for subjects undergoing an FDG-PET/CT scan was greater than those without. However, no observed relationship was seen between the timing of PET scan and OS.

Published data regarding population utilization of pre-treatment FDG-PET/CT in NSCLC are limited. A large, retrospective cohort study of 64 103 veterans treated within the United States Veteran Affairs (VA) health care system from 2000 to 2013 found that 40.1% (25 735) of the study population underwent an FDG-PET/CT in the year before diagnosis, while 64.3% (41 242) were scanned in the 5 years after diagnosis (8). Though the authors do not explicitly delineate pre-treatment FDG-PET/CT, nor show trends in utilization, they do show that approximately 75% of patients received FDG-PET/CT within 3 months of their diagnosis. This is quite similar to data presented herein which showed 74% FDG-PET/CT utilization for stage III NSCLC cases in Ontario for 2017. Another study from the United States using the Surveillance, Epidemiology, and End Results (SEER)-Medicare linked database reported that for new cases of NSCLC, 78% of non-Hispanic white people received FDG-PET/CT between 3 months before diagnosis and up to 30 days after initiation of therapy (9). The authors also found that 63% of black people and 70% of Hispanic people received FDG-PET/CT scans following the above criteria and argue that this disparity may explain differences in survival statistics between these ethnic groups. The cohort of patients reported in this project were not stratified by race; given the findings from the SEER database, this is an avenue for further research that should be pursued.

It is well established in the literature that pre-treatment FDG-PET/CT has a meaningful impact on treatment choice for new cases of NSCLC: work from 2001 found that the inclusion of FDG-PET/CT scans to staging investigations often demonstrated previously undetected metastases leading to appropriate staging and strongly influencing treatment strategy (10). Further work from 2001 on 105 cases of NSCLC where FDG-PET/CT was used for primary staging or re-staging showed that in 26% of the cases, results of the FDG-PET/CT scan led to a change in management plan from curative to palliative intent therapy (11). Indeed, as this was a prospective study, the authors determined that findings from the FDG-PET/CT scans changed or influenced management decisions in 67% of cases. The authors argue that these changes helped spare unnecessary treatment and ensure disease-appropriate management plans. More recent work from 2009 utilized questionnaire data from treating physicians to examine changes to treatment plans before and after FDG-PET/CT scans across all oncologic diagnoses, including lung cancer (12). Though not limited to NSCLC, the authors present data from 8240 cases and showed meaningful changes in planned treatment: 26.5% of cases where changes to type of therapy were made, 16.6% where changes were made to dose or duration of therapy, and 6.2% of subjects where the decision was made to change from therapy to observation or supportive care. Two more recent studies from 2020 point to the ability of FDG-PET/CT scans to aid in treatment choice of certain newly approved therapies including tyrosine kinase inhibitors (TKI) and immune checkpoint inhibitors (ICI): there was good performance in predicting epidermal growth factor receptor (EGFR) mutation status in NSCLC cases, potentially providing a faster alternative method for EGFR target therapy choice (13, 14).

The reasoning behind the observed increase in curative-intent therapies in patients who received a staging PET scan is challenging to explain and is likely multifactorial. One component may be related to provider comfort level in treating patients more aggressively that have been staged with PET-CT. Indeed, work examining radiation target volume delineation in patients with and without staging PET scans, showed reduced interobserver variability with fused PET-CT scans (15) and higher proportions of patients with PET scans receiving radical radiation doses (16). A second factor which may impact curative-intent treatment decision making is performance status: it is well established that pre-treatment performance status is a strong predictor of patient outcome with curative-intent NSCLC therapies (17), therefore patients with poorer performance status may be less likely to receive curative-intent therapy, and may also be less likely to receive staging PET-CT scans. Additionally, unequal patient access to PET scanners may represent a third contributing factor to the increase in curative-intent therapies in patients who received a staging PET scan. As northern parts of Ontario have poorer access to advanced scanning modalities such as PET-CT (18), this inequality may lead to fewer patients receiving curative-intent therapies, and ultimately having worse outcomes.

Looking at the trends for mediastinal staging, utilization of any mediastinal biopsy showed an increase over the timeframe examined, though not as dramatic as the increased PET utilization previously discussed. A major component of this trend likely represents clinicians seeking histopathological diagnosis for PET-avid lymph nodes. Targeting FDG-PET/CT positive lymph nodes with mediastinal biopsy has demonstrated reduced false-negative and false positive rates (19) and provides more accurate staging information than either modality alone (20). The observed trend of increasing EBUS usage when compared to mediastinoscopy for histopathological diagnosis is not a surprising finding given the less-invasive nature, lack of requirement for general anesthesia, and the lower complication rate of EBUS biopsies (21–23). The role that FDG-PET/CT may play in the increased adoption of EBUS is less clear. As previously discussed, EBUS biopsies guided by FDG-PET/CT data has been shown to have higher diagnostic accuracy than either technique alone (19). However, as PET-CT has now become standard-of-care for investigation and staging in NSCLC per ESMO guidelines, precise determination of the cause for the trend observed in the data presented here is unlikely.

The observed improved survival amongst patients undergoing FDG-PET/CT imaging is likely related to improved staging and the resulting pursuit of more aggressive therapies based on FDG-PET/CT results, rather than a therapeutic consequence of an FDG-PET/CT scan. While treatment decisions may play some role in improved survival, the impact related to appropriate staging must also be considered. Though FDG-PET/CT scans do not have a definitive role in diagnosing malignant pleural effusions currently (24), the remaining indications for M1 staging in NSCLC can be reliably detected and are key determining factors for appropriate staging. A large cohort study published in 2019 demonstrated that even though appropriate initial staging with FDG-PET/CT scans does account for stage-specific improvements in OS, the more impactful effect of FDG-PET/CT scans is through providing stage-appropriate therapies (8). In this report, the authors cite significant improvement in mortality across all NSCLC stages with the addition of FDG-PET/CT scans as their reason for the hypothesis that the association between accurate staging and decreased mortality is related to stage-appropriate therapy rather than related to appropriate staging pre-treatment. As the data presented herein is limited to stage III cases, the effect on OS by appropriate pre-treatment staging and or treatment choice cannot be determined, however it provides an exciting avenue for further investigation.

There was no significant difference seen in OS based on time comparing scans performed ≤28 days, 29-56 days, or >56 days prior to treatment. Though previous work from 2009 established that a shorter time to treatment is associated with improved OS in stage III NSCLC (25), literature examining the impact of delayed FDG-PET/CT scans is quite limited: a prospective trial from 2013 enrolled 82 people diagnosed with NSCLC who received two FDG-PET/CT scans between 8 and 176 days apart as part of their definitive chemoradiotherapy treatment (26). Among their findings, the authors found that a longer time gap between FDG-PET/CT and treatment led to larger radiation fields and increased tumour volume. An analysis of this kind would be beyond the scope of the project presented herein due to the number of subjects and inability to capture all radiation plans retrospectively. Diagnostic and treatment delays are expected to cause a substantial increase in the number of avoidable deaths due to cancer in the near future (27). Hopefully, the data presented in this project which demonstrated no significant relationship between delayed FDG-PET/CT scans and OS from 2007 to 2017 can be reassuring especially as delays from diagnosis to treatment are still present, and have been highlighted further during the COVID-19 pandemic.

The work presented here has limitations. First, during the timeframe studied (2007 to 2017), the AJCC Cancer Staging Manual went through three iterations (28–30): the 6^th edition (2002), the 7^th edition (2010), and the 8^th edition (2017). As each case was staged at the time of diagnosis, the updates between editions may have led to some variability between stage classifications. Secondly, the data analyzed was limited to stage III disease. This was an intentional decision by the authors to answer the research questions posed, however, this may limit the applicability of the results outside of stage III disease. Lastly, the geography and population distribution on Ontario can lead to challenges in providing equal access to cancer care across the province (18). As per Statistics Canada, Ontario has a land area of greater than 908 000 km² and a population over 14 million (2020) that is served by 13 FDG-PET/CT centres in 7 cities (Cancer Care Ontario, https://www.ccohealth.ca/en/). While the issue of unequal access to FDG-PET/CT scans across Ontario does not explicitly alter the data analyses presented herein, it speaks to a larger issue of diagnostic access in rural areas of Canada.

This project provided a look at the implementation of FDG-PET/CT scans as standard of care for NSCLC staging in Ontario, Canada over 10 years. Based on these data, the addition of pre-treatment FDG-PET/CT scans led to improved overall survival for patients with stage III NSCLC. This is likely due to improved stage-appropriate therapy and reduced stage misclassification, emphasizing the importance of pre-treatment FDG-PET/CT scans. Further and advocacy needs to be performed to reduce gaps in access, and to ensure that all patients who are eligible for curative-intent therapy receive an FDG-PET/CT scan as part of their initial work-up for suspected stage III NSCLC.

The original contributions presented in the study are included in the article/ Supplementary Material , further inquiries can be directed to the corresponding author.

Study ethics approval was obtained through the local institutional Hamilton Integrated Research Ethics Board and the Data Analytics Systems programs through IC/ES.

All co-authors contributed to conception and design of the study. GP organized the database. AS, CB & GP performed the statistical analysis. CB wrote the first draft of the manuscript. All authors contributed to manuscript revision, read, and approved the submitted version.