91桃色

Lung cancer accounts for nearly 27% of incident cancers in the United States and is the leading cause of cancer-related mortality [1]. The overall five-year survival in lung cancer patients is 17% but ranges from 55% for local tumors to 4% for distant tumors [1]. In 2011, the U.S. National Lung Screening Trial (NLST) found that three rounds of annual low-dose computed tomography (LDCT) as compared to chest x-ray reduced lung cancer mortality by 20% among persons 55 to 74 years old with ≥30 pack-years of smoking history and ≤ 15 years since quitting [2]. Based on these results, the United States Preventive Task Force (USPSTF) and Centers for Medicaid Services (CMS) recommend lung cancer screening (LCS) for high risk persons, however, USPSTF recommends screening current and former heavy smokers up to the age of 80 years annually and Medicare limits coverage to adults 55 to 77 years old and requires shared decision making before screening [3-7]. Core elements of this shared decisionmaking process involves discussion of benefits, harms and uncertainties of screening [8]. Because participants enrolled in NLST were younger, better educated, had lower number of comorbid conditions and were more likely to be former smokers as compared to the general population, the real-world evidence regarding the effectiveness of LCS remains unclear [9]. Moreover, uncertainty exists regarding the benefits and harms across diverse population groups, including stopping age for screening due to differences in demographic and clinical characteristics such as the burden of chronic co-existing illness and functional limitations.

What do professional guidelines recommend?

Professional guidelines reflect continued uncertainty regarding the stopping ages for LDCT screening. Whereas the American Society of Clinical Oncology, American College of Chest Physicians (ACCP), American Cancer Society (ACS) and the National Comprehensive Cancer Network (NCCN) guidelines are aligned with the NLST criteria of 77 years as the upper age limit [5,10-12], the United States.

Preventive Services Task Force and the American Association of Thoracic Surgery guidelines raises the cutoff to 80 years [13,14]. Overall, these guidelines offer limited guidance for individualizing lung cancer screening decisions as a function of life expectancy and coexisting illnesses. The American Association for Thoracic Surgery (AATS), ACCP, ACS and NCCN guidelines all incorporate life expectancy into some of their eligibility criteria for lung cancer screening; AATS and NCCN recommend screening among individuals with a 20 pack year smoking history and additional comorbidity that increases the risk of developing lung cancer whereas the ACS recommends that eligible individuals should be “in good health” [5,10,13]. The ACCP explicitly states that individuals with comorbidities that adversely influence the ability to tolerate screendetected findings or early- stage cancer treatment should not be screened [12,15]. Moreover, the American Academy of Family Physicians does not formally endorse LCS.7

Existing evidence on the impact of age, comorbidity and life expectancy on LCS outcomes

Evidence from NLST showed that the aggregate false positive rate in NLST was higher among older adults age 65 years compared to those younger than age 65 (65 and over: 27.7% vs. under 65: 22.0%); a higher proportion of invasive diagnostic procedures after false-positive screening was also observed by age (65 and over: 3.3% vs. under 65: 2.7%) [16]. Potential harms of LDCT screening include but are not limited to false positive results, over diagnosis as well as diagnostic and treatment complications due to older age and comorbidity [17-19].

Further, simulation modeling from the Cancer Intervention and Surveillance Modeling Network (CISNET) revealed that rate of over diagnosis increased with age, and hence stopping at earlier age might provide further benefits; however, this warrants further investigation [20]. This is an issue that was consistent across other trials and models and represents a major concern about developing LCS guidelines for community practices.

Crucially, nearly a third of the estimated 8.6 million LDCT lung cancer screening (LCS) eligible U.S. screening population present with significant chronic conditions (COPD, congestive heart failure, diabetes) [21,22]. In 2017, the American Thoracic Society convened a workshop to identify research gaps and future directions to optimize selection of candidates for lung cancer screening by accounting for coexisting chronic illness [9]. Experts from field of oncology, pulmonology, epidemiology and health services research concluded that competing causes of death, including smoking associated comorbid conditions like chronic-obstructive heart disease and cardiovascular disease, are highly prevalent among LCS eligible populations and may limit long term benefits of lung cancer screening due to their impact on overall health and life expectancy [9]. In the LCS context, Howard et al. also reported that the US population eligible for lung cancer screening may benefit less from early detection than NLST participants due to a higher risk of death from competing causes [22]. This emphasizes the need to obtain baseline information on comorbidities and functional status among screen eligible LCS populations to adequately quantify harms and benefits.

Previous reports from simulation models have identified significant impact of comorbidity and life expectancy in estimating screening harms and benefits and individualizing cancer screening decisions in the elderly [23,24]. While the majority of extant lung cancer risk prediction models rely primarily on age and smoking history, the PLCOM2012 model also includes several comorbid conditions, including COPD [25]. In another modeling analysis to predict life expectancy among high risk patients with COPD in a LCS setting, authors identified that lowering inclusion criteria for smoking packyears for these high risk patients may provide additional benefit in terms of life expectancy in LCS setting [26]. Furthermore, other studies have concluded that screening for smoking-associated comorbidities including COPD and emphysema improves lung cancer detection rates and improves the efficiency of lung cancer screening programs by targeting the highest risk individuals [27-30]. In France, researchers recommend that eligible LCS patients with peripheral arterial disease should be screened with a low dose scanner and confirmed with a blood test to identify non-solid tumor cells at an earlier stage [31].

In Italy, researchers recommend that lung cancer screenings should also evaluate cardiovascular abnormalities that might be relevant to a patient’s current health status [32]. Finally, Katki et al. recently compared stimulation modeling for participant selection using two risk based modeling strategies, and found that risk based modeling that incorporates age, smoking history, family history of lung cancer and the presence of comorbidities resulted in greater number of lung cancer deaths prevented over 5 years as compared to model based on USPSTF recommendation [33]. Finally, empirical evidence from various studies have led to the development of a COPD LCS score, which predicts lung cancer risk among patients with COPD eligible for screening [28,30,34]. On one hand, persons with COPD face a two-fold higher risk of lung cancer than smokers without COPD and may be more likely to benefit from LCS [35-39]. On the other hand, persons with advanced COPD are at a greater risk of complications during evaluation of pulmonary nodules [40] and are more likely to experience respiration-related surgical complications, have a higher 30-day mortality after resection of lung cancer (especially after thoracotomy) [25,41] and have a higher risk of non-lung cancer mortality [39,42]. Hence, it becomes crucial to measure both smoking and non-smoking related comorbidities among LCS populations and quantify benefits and harms associated with continued screening, including evaluation of downstream consequences such as rates of biopsies, surgeries, other treatment and rates of complications. Moreover, the continuing controversy over lung cancer screening indicates a need for quantitative data on benefits and harms in the general population, particularly within subgroupings of patients with varying degrees of life expectancy to guide informed choices [43,44].

What about frailty?

The World Health Organization in the last World Report on Ageing and Health, defined frailty as “an extreme vulnerability to endogenous and exogenous stressors that exposes an individual to a higher risk of negative health related outcomes” [45]. Evaluation of frailty is an important variable for the estimation of risk associated with clinical procedures among older adults [46]. Aging is associated with global decline in physiological reserves, leading to decrease tolerance to stress. Further, frailty is considered to be a marker of aging-associated decline in capacity to tolerate stress associated with chronic diseases leading to combined loss of mobility, sensory functions and cognition [47]. However, the underlining biology of frailty remains complex and multifactorial and includes aging-associated decline in functional status and loss of muscle mass. Frailty on the other hand, comprises of age related decrease in physiological reserves and increased vulnerability to stressors due to disturbances in processes associated with energy metabolism and neuromuscular changes [48].

The underlying mechanism of frailty mimics cachexia caused by disturbances in proportions of body fat, low- levels of leptin and high levels of pro-inflammatory cytokines. Domains that make up the frailty syndrome have been utilized in geriatric research over the years and includes combination of mobility, strength, balance, motor processing, cognition, nutrition, endurance and physical activity [49]. Hence a comprehensive assessment of both physical and cognitive domains helps cover global dimensions of frailty and its impact on health outcomes. Prior studies have highlighted frailty to be associated with poor screening outcomes, [50] response to chemotherapy and overall mortality and morbidity [51,52]. Additionally, frailty has shown to be associated with increased healthcare utilization and falls [53] as well as increased risk of multi-morbidity and polypharmacy and associated adverse events [54,55]. Moreover, as frailty impacts lifeexpectancy, understanding benefits and harms associated with LCS by levels of frailty remains unexplored and needs to be emphasized in future lung cancer screening studies.

Current evidence from community settings and future directions

Recently published studies provide conflicting evidence of success of lung cancer screening in community settings [56-63]. Although a couple studies indicated that screening was feasible and found results similar to those in the NLST [64,58,62], others reported hurdles regarding screening program implementation, including limited resources to manage abnormal findings [59], low patient volumes [60], or unanticipated extensive follow-up of incidental findings [64]. Moreover, limited evidence exists how individuals with limited life expectancy are ascertained in the context of LCS. Assessment of a person’s life expectancy can play a crucial role in identifying high risk individuals with favorable life expectancy that might benefit from early detection, as well as in limiting harms associated with over diagnosis and complications in individuals with limited life expectancy.

The uptake of lung cancer screening among the general population in the United States remains relatively low. Results from national studies show that only 1.9% of 7.6 million eligible current and former smokers underwent LDCT screening in 2016 [65]. Reasons for low uptake are multifactorial: limited knowledge of benefits and harms of LCS among primary care providers, including the presence of competing coexisting illnesses, financial barriers and system-level challenges such as preauthorization requirements, adequate access to specialty providers or low commitment at leadership levels [66,67]. Shared decision-making (SDM) is an important aspect of screening referral widely adopted after the USPSTF and CMS recommendations for the Medicare coverage of LCS [44]. Given that the goal of SDM is to provide patients with sufficient knowledge to make informed decisions about screening, updating current information about screening efficacy and contextualizing decisions in terms of comorbidity, frailty, functional status and life expectancy can help generate a more individualized approach towards improving benefits and reducing harms associated with LCS, especially among older adults with coexisting comorbidities [19].

The margin of benefit of LCS is highly uncertain among older adults with comorbidities and limited life expectancy. Further, LCS has not been empirically proven beneficial in individuals with diminished life expectancy due to comorbidity and those ages over 77 and these factors decrease the potential chance of screening benefit and likely increase the risk of harms. Given the heterogeneity in comorbidity and life expectancy among individuals undergoing LCS, it is important to optimize the selection of candidates for screening, especially among older adults beyond age 77 and with limited comorbidities. Research effort warrant evidence to inform life expectancy-based screening, particularly as screening programs disseminate into community practice.