Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Support Care Cancer (2015) 23:2341–2347 DOI 10.1007/s00520-015-2606-z ORIGINAL ARTICLE Validation and real-world assessment of the Functional Assessment of Anorexia-Cachexia Therapy (FAACT) scale in patients with advanced non-small cell lung cancer and the cancer anorexia-cachexia syndrome (CACS) Thomas W. LeBlanc & Greg P. Samsa & Steven P. Wolf & Susan C. Locke & David F. Cella & Amy P. Abernethy Received: 6 November 2014 / Accepted: 5 January 2015 / Published online: 14 January 2015 # Springer-Verlag Berlin Heidelberg 2015 Abstract Purpose Patients with cancer anorexia-cachexia syndrome (CACS) suffer a significant symptom burden, impaired quality of life (QoL), and shorter survival. Measurement of QoL impairments related to CACS is thereby important both in clinical practice and in research. We aimed to further validate the Functional Assessment of Anorexia-Cachexia Therapy (FAACT) scale in an advanced lung cancer population. Methods We tested the performance of the FAACT and its anorexia-cachexia subscale (ACS) within a dataset of patients This work was funded in part by a research grant from Helsinn to Duke University Medical Center. T. W. LeBlanc : A. P. Abernethy Duke Cancer Institute, Duke University School of Medicine, Durham, NC 27710, USA T. W. LeBlanc : G. P. Samsa : S. P. Wolf : S. C. Locke : A. P. Abernethy Center for Learning Health Care, Duke Clinical Research Institute, Durham, NC 27710, USA T. W. LeBlanc Division of Hematologic Malignancies and Cellular Therapy, Department of Medicine, Duke University School of Medicine, Durham, NC 27710, USA D. F. Cella Department of Medical Social Sciences, Northwestern University Feinberg School of Medicine, Chicago, IL 60611, USA A. P. Abernethy Division of Medical Oncology, Department of Medicine, Duke University School of Medicine, Durham, NC 27710, USA T. W. LeBlanc (*) 2400 Pratt Street, Suite 9000, Durham, NC 27705, USA e-mail: [email protected] with advanced non-small cell lung cancer (aNSCLC), using standard statistical methods. We then compared the performance of commonly used QoL measures stratified by CACS status and by patient self-report of appetite and weight loss. Results The FAACT and its ACS demonstrate internal validity consistent with acceptable published ranges for other QoL scales (Cronbach alpha=0.9 and 0.79, respectively). Correlation coefficients demonstrate moderate correlations in the expected directions between FAACT and ACS and scales that measure related constructs. Comparing patients with and without CACS, the ACS is more sensitive to change than other QoL instruments (mean score 33.1 vs. 37.2, p=0.011, ES= 0.58). Conclusion In patients with aNSCLC, the FAACT and its ACS performed well compared with other instruments, further supporting their validity and value in clinical research. FAACT and ACS scores covaried with symptoms and other QoL changes that are typical hallmarks of CACS, lending further support to their use as QoL endpoints in clinical trials among patients with CACS. Keywords Cancer anorexia cachexia syndrome . Quality of life instruments . Validity Introduction Patients with advanced lung cancer often experience anorexia (i.e., the subjective sense of poor appetite) and cachexia (i.e., the physiologic state of muscle catabolism and weight loss) [1]. These often occur together and constitute the “cancer anorexia-cachexia syndrome” or CACS [2, 3]. In an effort to 2342 promote standardization, a 2011 international consensus defined CACS, as “weight loss greater than 5 %, or weight loss greater than 2 % in individuals already showing depletion according to current bodyweight and height (body-mass index <20 kg/m2) or skeletal muscle mass (sarcopenia) [4].” We have previously shown that when using these criteria, patients with overt CACS suffer a significant burden of symptoms, marked impairments in quality of life (QoL) [5], and shorter overall survival [6]. The measurement of QoL impairments related to CACS is therefore important both in clinical practice and in research efforts [7] that seek to improve patients’ lived experience with advanced lung cancer. The Functional Assessment of Cancer Therapy (FACT) is a family of scales designed to measure health-related QoL and is among the best-validated set of instruments developed for use specifically among patients with cancer [8]. The FACT family is organized around the FACT-General (FACT-G) core instrument and its four subscales: physical well-being, emotional well-being, functional well-being, and social well-being. Often, the FACT-G is combined with additional subscales that include additional items of relevance to particular clinical contexts. For patients with lung cancer, for example, there is a “lung cancer subscale” (LCS); when scored together with the FACT-G, this yields the FACT-L [9]. Similarly, there is an anorexia and cachexia subscale (ACS) [10]. When combined with the FACT-G, this yields the Functional Assessment of AnorexiaCachexia Therapy (FAACT). While these scales are purported to capture important domains of health-related QoL among patients with cancer, some questions have been posed about the utility and validity of the FAACT scale [11]. The FACT-L has been extensively validated among patients with lung cancer. For example, a review article published in 2005 references 25 papers from 15 data sets including over 4000 patients [12]. Additional evidence has accumulated since then. The FAACT has also been widely studied, yet many validation studies do not report the proportion of patients with lung cancer and questions have emerged about whether the FAACT captures relevant domains regarding the psychosocial impact of cancer cachexia on patients’ experiences [11]. Furthermore, to our knowledge, none of these validation studies applied the recent consensus-based CACS definition to explore the performance of the FAACT scale among patients with overt cancer cachexia. Since the FAACT and its anorexia-cachexia subscale (ACS) are important markers of QoL in patients with advanced lung cancer and CACS, we sought to further explore their performance specifically in this population. This validation effort is unique for three reasons: (1) its exclusive focus on patients with advanced lung cancer, (2) the application of Support Care Cancer (2015) 23:2341–2347 the recent international consensus definition of CACS to stratify the sample, and (3) the comparative performance of FAACT with other instruments vis-à-vis responsiveness to symptoms and clinical anchors. Methods Design and participants From December 2007 through 2008, we screened patients in the Duke Oncology clinics who had a confirmed diagnosis of stage IIIB or stage IV non-small cell lung cancer. Eligible subjects were English-speaking, able to provide informed consent, and receiving cancer care at Duke University Medical Center. Ninety-nine patients were enrolled, and 97 patients completed at least the first study visit. The study was aimed at providing a broad understanding of the longitudinal experience of patients with advanced non-small cell lung cancer (aNSCLC) including their symptom burden, quality of life, physical function, and overall well-being [13]. Patients completed up to 4 study visits, in accordance with their scheduled routine outpatient cancer care. At each visit, they were asked to complete a battery of electronic patient-reported assessments including a detailed review of symptoms via the Patient Care Monitor version 2.0 and a quality of life assessment by several of the FACT questionnaires (including the FAACT and its anorexia-cachexia subscale). For this analysis, we utilized data only from the first visit, to maximize the available sample size, and to avoid bias due to nonrandom missingness of data at subsequent visits. We also grouped patients based on the weight-based international consensus CACS definition. Most patients had available weight data at 30, 60, 90, and 180 days prior to the first study visit. They were categorized as having CACS if adequate weight loss occurred within the 6 months prior to enrollment or as not having CACS if weight loss had not occurred during this period. Those with incomplete weight data who nevertheless clearly had at least 5 % weight loss were counted as meeting CACS criteria. Those with incomplete weight data who did not clearly lose over 5 % of their weight were considered as having an uncertain CACS classification and were excluded from analysis to minimize bias. Twenty-six patients were definitively identified as having CACS, 53 were definitively identified as not having CACS, and 18 had uncertain CACS status. These 18 patients could not be reliably categorized, as their weight data were not available for the 6 months prior to enrollment in the study. Therefore, data from 79 patients were included in the analysis. Support Care Cancer (2015) 23:2341–2347 Data elements The Patient Care Monitor Version 2.0 (PCM) is an instrument that primarily focuses on physical symptoms for patients with cancer [14]. Among its items (80 for males, 86 with females) are some that closely correspond to the symptom-based items used in the FAACT. Each item is scored on a 0–10 scale; higher scores are associated with more severe symptoms. The PCM can also generate 6 summated subscales (i.e., despair, distress, impaired ambulation, impaired performance, general physical symptoms, and treatment side effects), which we used as comparators in our analyses here. The Functional Assessment of Cancer Therapy-General (FACT-G) is a summated score of 27 items pertaining to physical well-being (7 items), emotional well-being (6 items), functional well-being (7 items), and social well-being (7 items). Each of the items uses a 5-point (0 to 4) scale. The FACT-G is typically used as a primary measure of general health status among patients with cancer, with its 4 subscales sometimes (albeit not always) used to provide additional detail. Higher scores are associated with better health-related quality of life. The lung cancer subscale (LCS) is a 9-item summated scale containing items specific to lung cancer (only 7 of these items are scored, however). Each item uses the same 5-point scale as above. Adding the 9 LCS items to the FACT-G produces the 36-item FACT-L. The Anorexia-Cachexia Subscale (ACS) is a 12-item summated scale containing items specific to patients’ perceptions of appetite and weight, also using the 5point scale. Adding the 12 ACS items to the FACT-G produces the 39-item Functional Assessment of AnorexiaCachexia Therapy (FAACT). Higher scores are associated with a higher QoL. Statistical analysis Patient characteristics were summarized using routine descriptive statistics. We then applied a multifaceted approach to instrument validation, combining traditional methods as follows. First, we assessed internal consistency with standardized Cronbach alpha coefficients. Next, we assessed convergent and divergent validity by calculating bivariate correlation coefficients between the FAACT (including the ACS), various versions of the FACT, and also the 6 subscales of the PCM. We divided patients into subgroups and compared their various summary scores to examine convergent and divergent validity, and used independent t tests to compare mean differences. For example, when comparing CACS and non-CACS patients, we expected a greater difference (in effect size comparisons) in the FAACT than in the other measures. Effect sizes were determined so that the group differences on the different scales could be easily compared using a common scale. Similarly, we repeated 2343 this comparison after dividing patients by their selfreport of weight loss, obtained from the PCM questionnaire. We repeated this comparison using patients’ selfreport of poor appetite, as well. Results Descriptive analysis Table 1 presents selected demographic and clinical characteristics of the patient population at baseline. Most patients were under the age of 65 (53 %), male (63 %), married (68 %), and Caucasian (78 %), with at least some college education (58 %). Of particular note is the distribution of functional status; 49 % had Karnofsky performance status (KPS) ratings of 80 and above. Patients with CACS, however, had lower (worse) KPS scores than those without CACS; this was expected given the known association between cancer cachexia, loss of lean body mass, and declining physical function. At enrollment, 35 % of CACS patients were receiving chemotherapy, compared to 45 % of control patients (P=0.47). Table 2 presents descriptive statistics for various scales derived from the FACT and the PCM. Of particular note is that the largest separation between the CACS and non-CACS group means occurs for the anorexia-cachexia specific (ACS) subscale of the FAACT (effect size=0.58). This suggests that among the instruments compared, the ACS-12 is uniquely suited to distinguish patients divided into those with or without CACS. There were no other statistically significant differences between the CACS and non-CACS groups for any of the other scales. Comparison of effect sizes also suggests a trend in differences between the PCM scales for despair (P= 0.12, ES=0.46) and distress (P=0.11, ES=0.38) among the CACS vs. non-CACS groups. Internal consistency Table 3 presents standardized Cronbach alpha coefficients, which quantify the internal consistency of the various scales within the FACT family. Coefficients were 0.79 and 0.90 for the ACS and the FAACT total, respectively. This is consistent with, even exceeding, values present in other published literature [10]. The other alpha coefficients are included to facilitate comparison to other published values. Correlations Table 4 presents correlations between the FAACT and various scales from the FACT and PCM. As anticipated, these correlations are of moderate magnitude (mostly ranging from about 0.3 to 0.6) and in the anticipated direction (note that high 2344 Support Care Cancer (2015) 23:2341–2347 scores are bad for the PCM and good for the FACT scales; thus, the correlations between the two instruments are expected to be negative). None of the correlations were adjusted for overlapping items. This explains the value of 0.96 between the FAACT and FACT-G, since the former contains all the items from the latter, plus the 12 additional ACS items. We also expected to see correlation among scales that measure related concepts, and there is indeed evidence of at least moderate correlation in several items that one would expect to be related. For example, the ACS correlates moderately with the FACT-functional subscale (0.35) and the FACT-physical subscale (0.54), as does the FAACT (at 0.85 and 0.62, respectively). Patients with CACS are known to have worse function and significant physical symptoms. Thus, if ACS and FAACT track with CACS status, they should also be at least moderately correlated with the FACT subscales that measure function and physical symptoms (such as fatigue). This is indeed what we found. Similarly, we would expect to see less correlation among items that are not directly related. For example, there is no evidence of appreciable correlation between the ACS and FACT-social (coefficient=0.06). The ACS focuses on specific questions about appetite, weight, food intake, Table 1 Patient characteristics at baseline Age under 65 years Female gender Marital status Married Widowed/divorced/separated Single Education Less than high school High school graduate Some college College graduate Ethnicity White/Caucasian Black/African-American Chinese Korean Spanish/Hispanic/Latino Ever smoked Currently smoke Karnofsky performance status 50 % 60 % 70 % 80 % 90 % Receiving chemotherapy physical appearance, interest in food, vomiting, early satiety, and pain. To the contrary, the FACT-social subscale focuses on relationships and closeness with friends and family, along with support derived from these relationships, so it makes sense that these would be essentially uncorrelated. Symptom-based grouping Tables 5 and 6 present additional analyses based on patientreported symptoms from the PCM that we expected to correlate with items measured in the FAACT and ACS scales. We grouped patients based on their responses to the two specific PCM questions about weight loss and appetite. The PCM items for weight loss (Table 5) and appetite loss (Table 6) were classified into “no problem” or “some problem,” based on whether the score for the item in question exceeded 0. This served to divide the sample into approximately equal halves. Grouping patients by the PCM item pertaining to weight loss (Table 5), the ACS was one of the two most sensitive instruments (the other being the PCM treatment side effects scale). Similarly, dividing the patients by the PCM item pertaining to appetite (Table 6), the ACS was again the most sensitive instrument. CACS (n=26) Non-CACS (n=53) Overall (n=79) P 46.2 % (12) 42.3 % (11) 56.6 % (30) 34.0 % (18) 53.2 % (42) 36.7 % (29) 0.38 0.47 0.65 73.1 % (19) 23.1 % (6) 3.8 % (1) 66.0 % (35) 24.5 % (13) 9.4 % (5) 68.4 % (54) 24.1 % (19) 7.6 % (6) 30.8 15.4 30.8 23.1 11.5 % (6) 28.8 % (15) 36.5 % (19) 23.1 % (12) 17.9 24.4 34.6 23.1 69.2 % (18) 23.1 % (6) 3.8 % (1) 0.0 % (0) 3.8 % (1) 88.5 % (23) 83.0 % (44) 15.1 % (8) 0.0 % (0) 1.9 % (1) 0.0 % (0) 84.9 % (45) 78.5 % (62) 17.7 % (14) 1.3 % (1) 1.3 % (1) 1.3 % (2) 86.1 % (68) 21.7 % (5) 13.6 % (6) 16.4 % (11) 3.8 % (1) 46.2 % (12) 23.1 % (6) 19.2 % (5) 7.7 % (2) 34.6 % (9) 0.0 % (0) 9.6 % (5) 30.8 % (16) 38.5 % (20) 21.2 % (11) 45.3 % (24) 1.3 % (1) 21.8 % (17) 28.2 % (22) 32.1 % (25) 16.7 % (13) 41.8 % (33) 0.17 % (8) % (4) % (8) % (6) % (14) % (19) % (27) % (18) 0.23 0.67 0.40 0.002 0.47 Support Care Cancer (2015) 23:2341–2347 Table 2 2345 Means (standard deviation) of the symptom and quality of life instruments by CACS category, at baseline Instrument CACS (n=26) Non-CACS (n=53) Total (n=79) P Effect sizea ACS FAACT total score FACT-social FACT-emotional FACT-physical FACT-functional FACT-G total score LCS FACT-L PCM despair PCM distress PCM impaired ambulation PCM impaired performance PCM general physical symptoms PCM treatment side effects 33.1 (7.7) 111.0 (17.4) 21.4 (6.1) 19.4 (4.1) 20.5 (5.3) 15.2 (6.3) 76.5 (14.3) 18.2 (4.7) 95.1 (17.0) 3.2 (5.0) 3.9 (6.5) 4.2 (6.2) 14.4 (12.7) 20.0 (13.2) 12.1 (8.4) 37.2 (6.5) 113.7 (20.9) 21.2 (5.5) 18.6 (5.0) 21.4 (4.4) 15.3 (6.7) 76.5 (17.0) 19.0 (3.9) 95.5 (20.1) 9.1 (15.0) 7.1 (9.1) 3.8 (6.1) 13.1 (10.0) 22.7 (14.2) 12.0 (9.2) 35.9 (7.1) 112.5 (19.8) 21.3 (5.7) 18.8 (4.7) 21.1 (4.7) 15.3 (6.5) 76.5 (16.0) 18.8 (4.2) 95.4 (19.0) 7.2 (12.9) 6.0 (8.4) 3.9 (6.1) 13.5 (10.8) 21.8 (13.8) 12.1 (8.9) 0.011 0.399 0.77 0.66 0.57 0.88 0.80 0.55 0.85 0.12 0.11 0.84 0.84 0.43 0.78 0.58 0.14 −0.04 −0.17 0.19 0.02 0.00 0.19 0.02 0.46 0.38 −0.07 −0.12 0.20 −0.01 a Effect size was calculated as (non-CACS mean–CACS mean)/SD of total sample Italicized numbers denote statistical significance at a threshold of p<0.05 Discussion Patients with advanced non-small cell lung cancer face a significant symptom burden and many such patients will develop overt CACS. While QoL assessment is an increasingly standard and important approach, assessment of QoL in the context of CACS is less established as a meaningful endpoint in clinical trials. Amidst recent positive interventional drug trials aimed at improving CACS [15], our efforts to further validate the FAACT and ACS in a lung cancer population, stratified by CACS status, are timely. The FACT family contains several well-validated instruments that are relevant to patients with advanced lung cancer; however, some questions have been posed about the validity of the FAACT scale and the ACS [11]. As such, we explored the cross-sectional validity of these scales. Our findings are Table 3 Cronbach’s alpha coefficients (standardized) Module Standardized alpha ACS (12-item anorexia and cachexia subscale) FAACT (ACS+FACT-G) FACT-G emotional well-being (6 item subscale) FACT-G functional well-being (7 item subscale) FACT-G physical well-being (7 item subscale) FACT-G family/social well-being (7 item subscale) FACT-G (27 item base FACT scale) LCS (9-item lung cancer subscale; only 7 items are scored) FACT-L (LCS+FACT-G) 0.79 0.90 0.84 0.87 0.78 0.87 0.90 0.49 0.91 encouraging—both scales demonstrate reasonable degrees of internal consistency, convergent validity, and divergent validity, consistent with those of the other FACT family of instruments. These results, when combined with the extensive validation of the FACT instruments outside the current context, provide reassurance about applying the FAACT and the ACS to patients with advanced lung cancer. Furthermore, by stratifying patients according to CACS status, we were able to test the ability of the FAACT scale and the ACS to detect meaningful differences known to exist among patients suffering with CACS (such as weight, appetite, and physical function, each of which is assessed through specific items in these scales). Our findings, highlighted in Table 2, suggest that the ACS tracks nicely with CACS status, being the only QoL measure in this analysis that was significantly different between the two groups. It is important to explicitly point out that we defined CACS by objective weight loss prior to study enrollment, whereas QoL was patient-reported. As such, these differences in ACS scores among CACS patients suggest that they were indeed aware of the weight, appetite, and functional considerations that define the syndrome. Our prior analyses of differences in survival, function, and symptoms among these patients lend further support to the notion that the two groups used in this analysis are quite meaningfully different [6] and are thus an appropriate validation set for these scales. Similarly, when we grouped patients by their responses to the PCM questions about appetite and weight loss (Tables 5 and 6), again there were significant differences in the ACS and FAACT scores by group. These differences are of sufficient magnitude to reach clinical significance as well. This suggests 2346 Support Care Cancer (2015) 23:2341–2347 Table 4 Correlation coefficients of the ACS and FAACT with the various FACT and PCM Scales at baseline ACS FAACT FACT-social FACT-emotional FACT-physical FACT-functional 0.06 0.33 0.54 0.35 0.56 0.74 0.62 0.85 FACT-G LCS FACT-L PCM despair PCM distress PCM impaired ambulation PCM impaired performance PCM general physical symptoms PCM treatment side effects 0.41 0.55 0.47 −0.34 −0.29 −0.41 −0.31 −0.34 −0.45 0.96a 0.76 0.97a −0.60 −0.52 −0.48 −0.45 −0.55 −0.29 Table 6 Means (standard deviation) of the symptom and quality of life instruments by response to the PCM decreased appetite question at baseline Instrument a Since the FAACT includes all items from the FACT-G, plus 12 additional questions from the ACS, and the FACT-L includes all items from the FACT-G, plus 9 additional questions from the LCS, a high correlation among these items is expected that the FAACT and ACS appropriately measure appetite, weight, or related constructs. Of further interest is the observation that a few additional QoL scales also differed based on responses to these PCM questions, including the FACTphysical subscale and the LCS (when grouped by the PCM weight responses), and the FACT-physical, FACT-functional, FACT-G, LCS, and FACT-L (when grouped by PCM appetite responses). These associations between physical and functional subscale changes along with changes in ACS and FAACT lend further support to the notion that the ACS and FAACT indeed track along with other QoL changes that we should Table 5 Means (standard deviation) of the symptom and quality of life instruments by response to the PCM weight loss question at baseline Italicized numbers denote statistical significance at a threshold of p<0.05 No decrease in Decreased appetite appetite (n=53) (n=26) P ACS 37.7 (7.2) 32.3 (5.5) 0.0001 FAACT FACT-social FACT-emotional FACT-physical FACT-functional FACT-G LCS FACT-L PCM despair PCM distress PCM impaired ambulation PCM impaired performance PCM general physical symptoms PCM treatment side effects 116.8 (20.6) 21.0 (6.0) 19.3 (4.6) 22.2 (4.5) 16.2 (6.9) 78.8 (16.9) 19.7 (4.2) 98.8 (20.0) 6.0 (12.2) 5.2 (8.0) 2.3 (4.1) 10.7 (9.5) 18.9 (11.3) 9.9 (7.0) 104.1 (14.9) 21.7 (5.1) 17.8 (4.8) 18.9 (4.5) 13.4 (5.2) 71.8 (13.1) 17.0 (3.6) 88.8 (15.1) 9.5 (14.2) 7.6 (9.3) 7.2 (7.9) 19.2 (11.1) 27.8 (16.7) 16.4 (10.7) 0.002 0.68 0.14 0.001 0.043 0.015 0.009 0.007 0.12 0.20 0.001 0.002 0.024 0.004 Italicized numbers denote statistical significance at a threshold of p<0.05 expect to find among patients with CACS. While these differences are statistically significant, however, it is worth noting that several of them do not reach clinical significance at published minimally interpretable difference levels for these scales, whereas the differences noted in ACS and FAACT are all clinically significant. There are a few limitations to our approach. While this study yielded broad and deep data about the experience of patients with aNSCLC, the relatively small sample size limited our Instrument No weight loss (n=52) Weight loss (n=27) P ACS FAACT FACT-social FACT-emotional FACT-physical FACT-functional FACT-G 38.3 (5.4) 115.8 (19.3) 20.5 (5.9) 19.3 (4.5) 21.8 (4.7) 15.7 (7.0) 77.3 (16.9) 31.4 (7.9) 106.4 (19.5) 22.7 (4.9) 18.0 (5.0) 19.7 (4.4) 14.5 (5.6) 75.0 (14.5) 0.0001 0.040 0.07 0.33 0.027 0.38 0.36 LCS FACT-L PCM despair PCM distress PCM impaired ambulation PCM impaired performance PCM general physical symptoms PCM treatment side effects 19.7 (4.0) 97.2 (19.9) 6.1 (11.4) 5.3 (7.3) 3.2 (5.4) 12.6 (10.5) 19.9 (12.6) 9.5 (7.0) 17.1 (4.0) 92.1 (17.0) 9.2 (15.5) 7.4 (10.4) 5.4 (7.1) 15.3 (11.3) 25.4 (15.5) 17.0 (10.2) 0.018 0.14 0.59 0.88 0.10 0.33 0.15 0.0002 Support Care Cancer (2015) 23:2341–2347 ability to detect significant differences between subsets of patients. As such, a larger study might yield additional insights about the performance of the FAACT and the ACS across the spectrum of those patients with or without CACS. This may explain why we did not see more striking QoL differences by CACS status. In addition, our ability to use all 97 patients was limited by the absence of weight data for some subjects prior to enrollment; this resulted in 18 patients who we could not classify by CACS status and were excluded from analysis. A larger sample size may have yielded additional findings. In conclusion, in a population of patients with aNSCLC, the ACS and FAACT demonstrate many hallmarks of instrument validity including good internal consistency, convergent validity, and divergent validity. Furthermore, when analyzed by patients’ CACS status, the ACS in particular appears to perform well in its ability to discriminate between groups. The performance of the ACS and FAACT in this analysis suggests that they track nicely with symptoms and other QoL changes that are typical hallmarks of CACS including weakness, poor appetite, weight loss, and declining physical function. These results lend further support to the notion that the ACS or FAACT are valid QoL endpoints for patients with CACS. Acknowledgments Dr. Abernethy reports personal fees (as ownership or employment) from Advoset, Orange Leaf Associates, Athena Health and Flatiron Health, Inc. Grants from Alliance for Clinical Trials in Oncology, American Cancer Society, Bristol-Myers Squibb, Celgene, DARA, Denderon, GlaxoSmithKline, Helsinn Healthcare, Helsinn Therapeutics, Kanglaite, Mayo Clinic, Medical College of Wisconsin, Memorial Sloan Kettering Cancer Center, Pfizer, University of North Carolina at Chapel HIll, University of South Florida and federal grants from NIH, National Cancer Institute, AHRA, National Institute for Nursing Research, National Institutes on Aging; personal fees from American Academy of Hospice and Palliative Medicine (as immediate past president), Pfizer, and ACORN Research. Dr. LeBlanc is a recipient of a Junior Career Development Award from the National Palliative Care Research Center (NPCRC), has received research support (paid to Duke University Medical Center) from Celgene and Helsinn Therapeutics, and honoraria from Helsinn (<$5000). The authors acknowledge the editorial assistance of Donald T. Kirkendall, ELS, a Duke-employed medical editor. Conflicts of interest The remaining co-authors have no disclosures or conflicts of interest to report. References 1. Evans WJ, Morley JE, Argiles J, Bales C, Baracos V, Guttridge D, Jatoi A, Kalantar-Zadeh K, Lochs H, Mantovani G, Marks D, Mitch WE, Muscaritoli M, Najand A, Ponikowski P, Rossi Fanelli F, Schambelan M, Schols A, Schuster M, Thomas D, Wolfe R, Anker SD (2008) Cachexia: a new definition. Clin Nutr 27(6):793–799. doi: 10.1016/j.clnu.2008.06.013 2347 2. Fearon KC, Voss AC, Hustead DS, Cancer Cachexia Study G (2006) Definition of cancer cachexia: effect of weight loss, reduced food intake, and systemic inflammation on functional status and prognosis. Am J Clin Nutr 83(6):1345–1350 3. George J, Cannon T, Lai V, Richey L, Zanation A, Hayes DN, Shores C, Guttridge D, Couch M (2007) Cancer cachexia syndrome in head and neck cancer patients: part II. Pathophysiology. Head Neck 29(5): 497–507. doi:10.1002/hed.20630 4. Fearon K, Strasser F, Anker SD, Bosaeus I, Bruera E, Fainsinger RL, Jatoi A, Loprinzi C, MacDonald N, Mantovani G, Davis M, Muscaritoli M, Ottery F, Radbruch L, Ravasco P, Walsh D, Wilcock A, Kaasa S, Baracos VE (2011) Definition and classification of cancer cachexia: an international consensus. Lancet Oncol 12(5): 489–495. doi:10.1016/S1470-2045(10)70218-7 5. Couch M, Lai V, Cannon T, Guttridge D, Zanation A, George J, Hayes DN, Zeisel S, Shores C (2007) Cancer cachexia syndrome in head and neck cancer patients: part I. Diagnosis, impact on quality of life and survival, and treatment. Head Neck 29(4):401–411. doi:10. 1002/hed.20447 6. LeBlanc TW, Nipp RD, Rushing CN, Samsa GP, Locke SC, Kamal AH, Cella D, Abernethy AP (2014) Correlation between the International Consensus Definition of the Cancer Anorexia Cachexia Syndrome (CACS) and Patient-Centered Outcomes in Advanced Non-small Cell Lung Cancer. J Pain Symptom Manag. doi:10.1016/j.jpainsymman.2014.09.008 7. Cella DF, Tulsky DS (1993) Quality of life in cancer: definition, purpose, and method of measurement. Cancer Invest 11(3):327–336 8. Cella DF, Tulsky DS, Gray G, Sarafian B, Linn E, Bonomi A, Silberman M, Yellen SB, Winicour P, Brannon J et al (1993) The functional assessment of cancer therapy scale: development and validation of the general measure. J Clin Oncol 11(3):570–579 9. Cella DF, Bonomi AE, Lloyd SR, Tulsky DS, Kaplan E, Bonomi P (1995) Reliability and validity of the Functional Assessment of Cancer Therapy-Lung (FACT-L) quality of life instrument. Lung Cancer 12(3):199–220 10. Ribaudo JM, Cella D, Hahn EA, Lloyd SR, Tchekmedyian NS, Von Roenn J, Leslie WT (2000) Re-validation and shortening of the Functional Assessment of Anorexia/Cachexia Therapy (FAACT) questionnaire. Qual Life Res 9(10):1137–1146 11. Wheelwright S, Darlington AS, Hopkinson JB, Fitzsimmons D, White A, Johnson CD (2013) A systematic review of health-related quality of life instruments in patients with cancer cachexia. Support Care Cancer 21(9):2625–2636. doi:10.1007/s00520-013-1881-9 12. Butt Z, Webster K, Eisenstein AR, Beaumont J, Eton D, Masters GA, Cella D (2005) Quality of life in lung cancer: the validity and crosscultural applicability of the Functional Assessment of Cancer Therapy-Lung scale. Hematol Oncol Clin North Am 19(2):389– 420. doi:10.1016/j.hoc.2005.02.009, viii 13. Suh SY, Leblanc TW, Shelby RA, Samsa GP, Abernethy AP (2011) Longitudinal patient-reported performance status assessment in the cancer clinic is feasible and prognostic. J Oncol Pract 7(6):374–381. doi:10.1200/JOP.2011.000434 14. Abernethy AP, Currow DC (2010) Need for mechanistically focused research of global systemic interventions in palliative care. J Pain Symptom Manag 40(3):e5–e8. doi:10.1016/j.jpainsymman. 2010.06.006 15. Temel J, Currow DC, Fearon K, Gleich L, Yan Y, Friend J, Abernethy AP (2014) Anamorelin for the treatment of cancer anorexia-cachexia in NSCLC: Results from the phase 3 studies ROMANA 1 and 2. Paper presented at the European Society of Medical Oncology, Madrid, 2014