Determinants of quality of life in patients with incurable

Low skeletal muscle attenuation was independently associated with poorer physical functioning OR, 1.67; 95% CI, 1.09-2.56, but muscle parameters were not independently associated with fa

Determinants of Quality of Life in Patients With Incurable

Cancer

Louise E Daly, PhD 1; Ross D Dolan, MD, MRCS, MSc, MA 2; Derek G Power, MD, MB, BCh, MRCPI3;

Éadaoin Ní Bhuachalla, RD, PhD1; Wei Sim, BSc2; Samantha J Cushen, RD, PhD1; Marie Fallon, MD, MBChB, MRCGP, FRCP4; Claribel Simmons, MD, MBChB, MRCP (UK), MRCGP4; Donald C McMillan, PhD 2; Barry J Laird, MD, MBChB4;

and Aoife M Ryan, RD, PhD 1

BACKGROUND: Optimizing quality of life (QoL) remains the central tenet of care in patients with incurable cancer; however,

determi-nants of QoL are not clear The objective of the current study was to examine which factors influence QoL in patients with incurable cancer METHODS: A multicenter study of adult patients with advanced cancer was conducted in Ireland and the United Kingdom

between 2011 and 2016 Data were collected from patients at study entry and included patient demographics, Eastern Cooperative Oncology Group performance status (ECOG-PS), nutritional parameters (the percentage weight loss [%WL]), muscle parameters assessed using computed tomography images (skeletal muscle index and skeletal muscle attenuation), inflammatory markers (modified Glasgow Prognostic score [mGPS]), and QoL data (the European Organization for Research and Treatment Quality-of-Life Questionnaire C-30) The relation between clinical, nutritional, and inflammatory parameters with QoL was assessed using the Spearman rank cor-relation coefficient and multivariate binary logistic regression Components of the European Organization for Research and Treatment Quality-of-Life Questionnaire C-30 (physical function, fatigue, and appetite loss) and summary QoL scores were mean-dichotomized for the logistic regression analyses RESULTS: Data were available for 1027 patients (51% men; median age, 66 years) Gastrointestinal

cancer was most prevalent (40%), followed by lung cancer (26%) and breast cancer (9%) Distant metastatic disease was present in 87% of patients The %WL, ECOG-PS, and mGPS were significantly correlated with deteriorating QoL functional and symptom scales (all

P < .001) On multivariate regression analysis, >10% WL (odds ratio [OR], 2.69; 95% CI, 1.63-4.42), an ECOG-PS of 3 or 4 (OR, 14.33; 95%

CI, 6.76-30.37), and an mGPS of 2 (OR, 1.58; 95% CI, 1.09-2.29) were independently associated with poorer summary QoL scores These

parameters were also independently associated with poorer physical function, fatigue, and appetite loss (all P < .05) Low skeletal muscle attenuation was independently associated with poorer physical functioning (OR, 1.67; 95% CI, 1.09-2.56), but muscle parameters were not independently associated with fatigue, appetite loss, or QoL summary scores CONCLUSIONS: The current findings indicate that QoL

is determined (at least in part) by WL, ECOG-PS, and the systemic inflammatory response in patients with advanced cancer Identifying early predictors of poor QoL may allow the identification of patients who may benefit from early referral to palliative and supportive care, which has been shown to improve QoL Cancer 2020;0:1-11 © 2020 American Cancer Society

KEYWORDS: incurable cancer, palliative care, performance score, quality of life, systemic inflammation, weight loss.

INTRODUCTION

The European Society of Medical Oncology1 advocates integrating supportive and palliative, patient-centered care into overall anticancer treatment at all stages of the disease The European Society of Medical Oncology acknowledges that oncology patients’ needs are not being adequately met and that oncology care should encompass patient-centered sup-portive and palliative care from initial diagnosis and throughout the entire trajectory of the disease Importantly, cancer care should not only aim to deliver the best quality anticancer treatment, but it should now also consider the effect of a cancer diagnosis and its treatment on each patient’s life.1

In patients who have an incurable cancer, the fundamental objective of treatment is to optimize quality of life (QoL) If this can be attained in unison with prolonged survival, then this is clearly desirable; however, if prolonged survival comes at the expense of impaired QoL, then this may not be in the best interests of patients Importantly, QoL

Corresponding Author: Louise E Daly, PhD, Office 135, School of Food & Nutritional Sciences, College of Science, Engineering, and Food Science, University College Cork,

Cork, Ireland (louisedaly@umail.ucc.ie).

1 School of Food and Nutritional Sciences, College of Science, Engineering and Food Science, University College Cork, Cork, Ireland; 2 Academic Unit of Surgery, University

of Glasgow, Glasgow, United Kingdom; 3 Department of Medical Oncology, Mercy and Cork University Hospital, Cork, Ireland; 4 Edinburgh Cancer Research Centre, Institute

of Genetics and Molecular Medicine, University of Edinburgh, Edinburgh, United Kingdom

The first two authors contributed equally to this work.

The last two authors contributed equally to this work as joint senior authors.

We acknowledge support from the Health Research Board Clinical Research Facility Cork and from Medical Research Scotland.

Additional supporting information may be found in the online version of this article

DOI: 10.1002/cncr.32824, Received: April 24, 2019; Revised: September 19, 2019; Accepted: January 24, 2020, Published online Month 00, 2020 in Wiley Online Library

(wileyonlinelibrary.com)

is increasingly being recognized as an important

prognos-tic indicator, and QoL has been associated with reduced

survival in various cancer sites, even after adjusting for

known prognostic clinical variables.2-5

The current almost routine adoption of patient-

reported outcome measures (PROMs) of QoL into

can-cer clinical trials has enhanced our understanding of

this area.6 The European Organization for the Research

and Treatment of Cancer (EORTC) has now developed

over 60 QoL modules, including the universal EORTC

Quality-of-Life Questionnaire C-30 (EORTC

QLQ-C30).7 By using this questionnaire, it has been shown that

both physical function (performance score) and measures

of the systemic inflammatory response (measured with

the modified Glasgow prognostic score [mGPS]) have a

differential association with QoL.8,9 In a large cohort of

2520 patients with advanced cancer, increasing mGPS

and deteriorating Eastern Cooperative Oncology Group

performance status (ECOG-PS) were associated with

deterioration in QoL parameters such as global health;

role, physical, and social functioning; fatigue; pain; and

appetite symptoms (P < .001) The association with

in-creasing systemic inflammation and poorer QoL

param-eters was independent of PS.8 It has also been reported

that other aspects, including weight loss (WL), body mass

index (BMI), and loss of muscle (sarcopenia) influence

QoL in patients with cancer.10-12

It has been argued that the host-tumor interaction

and the resulting systemic inflammatory response is key

in the genesis of how symptoms/QoL are influenced in

patients with cancer Indeed, work to date has supported

this hypothesis, demonstrating that the magnitude of the

systemic inflammatory response influences the

magni-tude of symptoms in patients with cancer.8 On the basis

of this work, markers of the systemic inflammatory

re-sponse are now advocated as key assessment criteria for

staging nutritional status13 and as stratification factors

in randomized clinical trials.14 In the same way that the

tumor is staged, it has been argued that the host should be

staged, as inflammatory status is likely to influence

treat-ment outcomes and magnitude of symptoms.15

However, a comparison of all factors known to

influ-ence QoL has yet to be done Elucidation of those factors

that adversely influence QoL may allow the

identifica-tion of patients who may benefit from early referral to

palliative and supportive care, which has been shown to

improve QoL.16,17 Therefore, the objective of the current

study was to examine the relation between clinical,

nu-tritional, and inflammatory factors and QoL in patients

with incurable cancer

MATERIALS AND METHODS

Study Sample

Data were collected across 18 sites in Ireland and Scotland (cancer centers, hospitals, and specialist palliative care units) over a period of 5 years (2011-2016) Patients were older than 18 years and had a diagnosis of incurable can-cer Incurable cancer was defined as metastatic disease

or locally advanced disease being treated with palliative intent Both inpatients and outpatients were recruited, and a convenience sampling approach was adopted Willing participants provided written informed consent Exclusion criteria included patients younger than 18 years and those who were unwilling or unable to participate because of cognitive impairment Ethical approval was given for the data collection at all sites and was conducted according to good clinical practice and applicable laws

Procedure and Assessment

Demographic data and clinical data were recorded and included primary tumor site, stage, and extent of metastatic disease (if present) The EORTC QLQ-C30 (version 3.0) was used to assess QoL.3 This 30-item, cancer-specific questionnaire includes 5 functional scales (physical, emotional, cognitive, social, and role),

3 symptom scales (fatigue, pain, and nausea/vomiting),

a global health/QoL scale, and 6 single items (dyspnea, insomnia, appetite loss, constipation, diarrhea, and financial impact of disease) The 28 items measuring functional and symptom scales use a numeric scale for scores of 1 (not at all), 2 (a little), 3 (quite a bit), and

4 (very much) The 2 items concerning global QoL use a scale from 1 (very poor) to 7 (excellent) The raw scores were linearly transformed to give standard scores

in the range of 0 to 100 for each of the scales and sin-gle items, as described by the EORTC.7 Higher scores for the functional or global QoL scales represent a high level of functioning or QoL, whereas higher scores on the symptom scales represent worse symptomatology The summary score of the EORTC QLQ-C30, which

is comprised from the mean of 13 of the 15 QLQ-C30 scales (the global QoL and financial impact scales are not included), was used to assess the overall summary QoL, with a maximum score of 100.18 The summary score was only calculated if all of the 13 required scale scores were available, and the scoring of the QLQ-C30 summary score was calculated as follows: QLQ-C30 summary score = (physical functioning + role functioning + so-cial functioning + emotional functioning + cognitive functioning + 100 − fatigue + 100 − pain + 100 − nausea_vomiting + 100 − dyspnea + 100 − sleeping

disturbances (insomnia) + 100 − appetite loss + 100

− constipation + 100 − diarrhea)/13.18

Nutritional parameters were also assessed Patient’s

weight, height, and BMI (weight [kg]/height [m2]) were

recorded Patients were categorized according to their

BMI as underweight (<20 kg/m2), normal weight

(20-24.9  kg/m2), overweight (≥25-29.9  kg/m2), or obese

(≥30  kg/m2) WL in the preceding 3  months was

re-ported by patients and, when possible, was verified from

patients’ medical records

C-reactive protein (CRP) (mg/L) and albumin

(g/L) were used as markers of the systemic inflammatory

response and were measured in a venous blood sample

drawn at the time of consent By using both CRP and

albumin, an mGPS was calculated accordingly.19 Patients

who had both elevated CRP (>10 mg/L) and

hypoalbu-minemia (<35 g/L) were assigned a score of 2 Patients

with only an elevated CRP (>10  mg/L) and without

hypoalbuminemia (albumin >35  g/L) were assigned a

score of 1 Patients with neither of these abnormalities

(ie, CRP <10  mg/L and albumin >35  g/L) were

as-signed a score of zero.20 The limit of detection of CRP

was <5 mg/L An increasing score is related to increasing

systemic inflammation.19

PS was assessed using the ECOG score.21 Scores

were assigned according to patient-reported daily physical

function as follows: 0, fully active with no restrictions; 1,

restricted in physically strenuous activity but ambulatory

and able to perform light work; 2, ambulatory and

capa-ble of all self-care but unacapa-ble to perform any work

activi-ties; 3, capable of only limited self-care; and 4, completely

disabled and totally confined to bed or chair

Body Composition Assessment

Abdominal computerized tomography (CT) images, taken

as part of routine patient care within 12  weeks of QoL

assessment, were used to assess body composition as

pre-viously described.22 The third lumbar vertebrae (L3) was

chosen as the standard landmark, and 2 consecutive

trans-verse CT images in which both transtrans-verse processes were

clearly visible were analyzed using OsiriX software version

4.1.1 (Pixmeo) for data collected in Ireland  and ImageJ

software (version 1.47; National Institutes of Health)  for

data collected in Scotland Both imaging software

pack-ages have been shown to provide excellent agreement for

body composition measures.23 L3 was used as a standard

landmark because it correlates best with whole body

meas-ures of muscle mass.24,25 Skeletal muscle area (SMA) (cm2)

was manually outlined, and segmentation of SMA was

based on Hounsfield unit (HU) thresholds (from −29 to

+150 HU).26 SMA was normalized for stature to compute the skeletal muscle index (SMI) (cm2/m2) Mean muscle at-tenuation (MA) in HU was assessed in all patients with a contrast-enhanced CT image and was reported for the en-tire SMA at L3 Sex-specific and BMI-specific cutoff points were used to define low SMI (sarcopenia) and low MA ac-cording to Martin et al.27 Measurements were performed by

2 individuals (R.D and L.E.D.), and inter-rater reliability was assessed in a sample of 20 patient images using interclass correlation coefficients (ICCCs) (SMA ICCC =  0.986; SMD ICCC  =  0.964) Investigators were blinded to patient’s demographic and clinicopathologic status

Statistical Analysis

Statistical analysis was conducted using SPSS (version 24.0; SPSS Inc) Data are expressed as the mean ± SD

or the median with interquartile range (IQR), where appropriate Comparisons between groups of patients were assessed using the chi-square test for categorical

variables and the unpaired t test and the Mann-Whitney

U test for differences in continuous variables Correlations

were investigated using the Spearman coefficient for non-parametric QoL data The correlation coefficient (ρ) was used to determine the strength of the correlations The Cohen guidelines were used when interpreting effect size and strength of correlations These suggest that ρ values from 0.1 to 0.29 indicate a small effect size or correlation,

ρ values from 0.3 to 0.49 indicate a medium effect size, and ρ values from 0.5 to 1.0 indicate a strong effect size or correlation Components of the EORTC-QLQ (physical function, fatigue, and appetite loss) and the summary QoL score were mean-dichotomized for the logistic regression analyses assessing clinical, nutritional, and inflammatory predictors of QoL Patients with a score below the mean for physical function and QoL summary scores and above the mean for fatigue and appetite loss scores were given a score of 1, whereas those with a score above the mean for physical function and QoL summary scores and below the mean for fatigue and appetite loss scores were given a score

of zero Thus, an odds ratio (OR) >1.0 indicate a greater likelihood of worse QoL Independent variables that had significance on univariate analysis were eligible for in-clusion in multivariate analysis All statistical tests were

2-sided, and P values <.05 were considered significant.

RESULTS

Patient Characteristics and Demographics

In total, 1027 patients with advanced cancer were re-cruited Baseline demographic, clinical, nutritional, and QoL characteristics are presented in Table 1 Patients

were a median of 4.6  months postdiagnosis when they entered the study (IQR, 3.0-13.0 months) In brief, 51%

of patients were men, and the median age was 66 years (IQR, 57-74 years) Gastrointestinal cancer was the most common type (40%), and metastatic disease was present

in 87% of patients In total, 830 patients (81%) were ac-tively receiving chemotherapy (chemotherapy in the pre-ceding 4 weeks)

Anthropometry and Body Composition

Patients exhibited a wide variation in BMI (range, 12.3-47.4 kg/m2) One-half (51%) of all patients were overweight or obese (BMI ≥25 kg/m2), whereas only 13% had a BMI <20.0 kg/m2 WL >5% in the pre-ceding 3 months occurred in 277 patients (29%), with 14% experiencing severe WL >10% In terms of body composition, CT scans within 12 weeks of QoL assess-ment were available in 428 patients (contrast-enhanced

CT images for MA assessment were available in 413 patients) Overall, 192 patients (45%) were considered

to have a low SMI (sarcopenia), and 223 (54%) had low MA

Relation Between Clinical, Nutritional, and Inflammatory Parameters With QoL

The relation between clinical, nutritional, and inflam-matory parameters to PROMs is displayed in Table 2 Within our cohort, female sex was significantly negatively correlated with poorer physical function (ρ  =  −0.112;

P  = .001), emotional function (ρ = −0.071; P = .024),

and summary QoL scores (ρ = −0.080; P = .012) and positively correlated with more nausea and vomiting (ρ = 0.123; P = .001) and pain (ρ = 0.068; P = .030) Overall, the strength of these correlations was small (ρ < 0.3) Increasing age was negatively correlated with poorer physical (ρ = −0.143; P = .001), and role func-tion (ρ = −0.063; P = .047) and was positively corre-lated with better emotional functioning (ρ  =  0.070;

P = .012) In terms of symptom scales, age was positively correlated with more fatigue (ρ = 0.70; P = .024), dysp-nea (ρ = 0.089; P = .005), and constipation (ρ = 0.073;

P = .020) The presence of distant metastatic disease (vs locoregional incurable disease) was not statistically signifi-cantly correlated with any EORTC functional or symp-tom scale

The %WL, ECOG-PS, and mGPS were negatively correlated with almost all EORTC functional scales

(P <  05) Importantly, medium-to-strong correlations (ρ  >  0.30) were observed between the ECOG-PS and mGPS with physical function (ρ = −0.557 [P < .001]

TABLE 1 Demographic and Clinical Characteristics

of the Patients Included in This Study

Emotional functioning 79.4  ± 22.7

Cognitive functioning 79.2  ± 24.7

Abbreviations: BMI, body mass index; ECOG PS, Eastern Cooperative

Oncology Group performance status; mGPS, modified Glasgow prognostic

score; QoL, quality of life.

a The other cancer group consisted of breast, gynecologic, genitourinary,

neu-rologic, and hematologic cancers, melanoma, unknown primary cancers, and

others.

b ECOG PS was available for 979 patients.

c mGPS was available for 821 patients.

d BMI was available for 949 patients.

e The percentage weight loss was available for 951 patients.

f Computed tomography scans were available for muscle mass (sarcopenia)

assessment in 428 patients.

g Contrast-enhanced computed tomography images were available for

mus-cle attenuation assessment in 413 patients.

*Presence or absense of distant metastatic disease was available for 994 patients.

No of Patients

20, 20-24.9, 25-29.9, or >30

a These

and ρ  =  −0.312 [P  <  001], respectively) and for the

ECOG-PS with role function (ρ = −0.494; P < .001),

social function (ρ  =  −0.334; P  <  001), global health

(ρ  =  −0.410; P  <  001), and, importantly, summary

QoL scores (ρ = −0.500; P < .001) The presence or

ab-sence of metastatic disease was not related to any of the

PROMs Interestingly, reduced EORTC-reported

phys-ical functioning was more strongly correlated with low

MA compared with low SMI (ρ = −0.244 vs ρ = −0.164,

respectively) Low SMI was not significantly associated

with any other PROMS, whereas low MA was

associ-ated with role function (ρ = −0.145; P = .003), global

health (ρ = −0.175; P < .001), and QoL summary score

(ρ = −0.135; P = .006)

Table 3 depicts the relation between the

symp-tom components of the EORTC-QLQ and clinical,

nutritional, and inflammatory parameters In line with

we what we observed in the PROMs functional scales,

%WL, ECOG-PS, and mGPS were associated with

increasing symptoms scores (P <  05) Medium

cor-relations (ρ > 0.30) were observed between ECOG-PS

and fatigue (ρ = 0.476; P < .001) and pain (ρ = 0.309;

P < .001) and, as expected, between %WL and anorexia

(ρ = 0.311; P < .001) Low MA was associated with more

fatigue (ρ = 0.150; P = .002) and dyspnea (ρ = 0.150;

P = .002)

In the multivariate logistic regression analyses, the

QoL summary score was dichotomized by the mean

score (73.8) ORs >1.00 were associated with poorer

overall QoL On multivariate regression analysis, %WL

(WL >5%: OR, 1.59; 95% CI, 1.01-2.51; P  =  048;

WL >10%: OR, 2.69; 95% CI, 1.63-4.42; P < .001),

ECOG-PS (ECOG-PS 2: OR, 3.32; 95% CI, 2.34-4.70;

P <  001; ECOG-PS 3-4: OR, 14.33; 95% CI,

6.76-30.37; P < .001), and mGPS (mGPS 1: OR, 2.05; 95%

CI, 1.26-3.32; P = .004; mGPS 2: OR, 1.58; 95% CI,

1.09-2.29; P = .0016) were independently predictive of

an overall QoL summary score below the mean (Table 4)

In terms of physical function (scores <68.4), WL

>10% (OR, 1.92; 95% CI, 1.16-3.19; P  =  039),

ECOG-PS (ECOG-PS 2: OR, 3.93; 95% CI, 2.77-5.58;

P <  001; ECOG-PS 3-4: OR, 18.07; 95% CI,

7.91-41.28; P < .001), an mGPS of 2 (OR, 2.01; 95% CI,

1.39-2.93; P < .001), and female sex (OR, 1.56; 95%

CI, 1.10-2.19; P =  011) were independent predictors

of poorer physical function on multivariate analysis (see

Supporting Table 1)

Examining predictors of fatigue (scores >42.3),

on multivariate analysis, WL >10% (OR, 2.53; 95%

CI, 1.53-4.19; P <  001), ECOG-PS (ECOG-PS 2:

OR, 2.89; 95% CI, 2.06-4.07; P <  001; ECOG-PS

3-4: OR, 18.67; 95% CI, 7.79-44.7; P < .001), and an mGPS of 2 (OR, 1.57; 95% CI, 1.09-2.25; P < .001)

were independent predictors of more fatigue (see Supporting Table 2)

On multivariate analysis, the factors associated with more appetite loss (scores >27.3) were WL (WL >5%:

OR, 2.38; 95% CI, 1.51-3.76; P <  001; WL >10%:

OR, 2.51; 95% CI, 1.58-3.99; P <  001), ECOG-PS

(ECOG-PS 2: OR, 1.86; 95% CI, 1.26-2.74; P = .002; ECOG-PS 3-4: OR, 2.59; 95% CI, 1.48-4.55; P = .001),

and mGPS (mGPS 1: OR, 1.72; 95% CI, 1.02-2.91;

P =  043; mGPS 2: OR, 1.64; 95% CI, 1.09-2.48;

P = .017) (see Supporting Table 3)

On assessment of the relation between muscle pa-rameters and QoL (n =  428), on univariate analysis, low SMI was associated with poorer physical

function-ing (OR, 1.72; 95% CI, 1.27-2.33; P <  0.001) but not fatigue, appetite loss, or summary QoL score (all

P >  05) However, on multivariate assessment (con-trolling for WL, ECOG-PS, mGPS, and low MA), low SMI was no longer associated with poorer physical

functioning (OR, 1.14; 95% CI, 0.74-1.73; P = .555)

On univariate analysis, low MA was associated with poorer physical function (OR, 2.31;[95% CI,

1.69-3.18; P < .001), fatigue (OR, 1.66; 95% CI, 1.22-2.25;

P = .001), appetite loss (OR, 1.94; 95% CI, 1.33-2.84;

P = .001), and poorer summary QoL scores (OR, 1.41;

95% CI, 1.03-1.92; P = .032) However, after adjust-ment for %WL, ECOG-PS, mGPS, and low SMI, low

MA was only independently associated with poorer physical functioning (OR, 1.67; 95% CI, 1.09-2.56;

P = .018)

DISCUSSION For the first time to our knowledge, the current study reports a comprehensive analysis of tumor and host fac-tors and their effect on QoL in a large cohort of patients with incurable disease Our findings indicate that QoL is determined (at least in part) by WL, performance, and the systemic inflammatory response in patients with ad-vanced cancer Muscle mass and attenuation were signifi-cantly associated with some QoL domains on univariate analysis; however, on multivariate analysis, there was no significant independent association with fatigue, appe-tite loss, or QoL summary score Our findings suggest that interventions to mitigate the systemic inflammatory response and WL in patients with incurable cancer might have a positive effect on patients’ QoL

No of Patients

Nausea and Vomiting

20, 20-24.9, 25-29.9, or

a These

As expected, better ECOG-PS (scores of 0-1)

cor-related with better physical, role, emotional, and

so-cial functioning; better global heath scores; and less

fatigue, pain, anorexia, and constipation (all P < .001)

Considering that ECOG-PS is designed to determine a

patient’s ability to perform activities of daily living and

general well being, it is no surprise that the ECOG-PS is

associated with items of the EORTC QLQ-C30, and this

relation has been reported previously.8,28,29

Our findings also demonstrate that the systemic

in-flammatory response, as evidenced by mGPS scores ≥1, is

correlated with almost all EORTC functional and

symp-tom scales Furthermore, the mGPS was independently

associated with physical functioning, fatigue, appetite

loss, and the QoL summary score Our findings echo

those previously reported in advanced cancer Laird et al

reported that CRP was significantly associated with all of

the functional components of the EORTC QLQ-C30

and several the symptoms, including appetite loss, pain,

and fatigue.30

In some instances, individual cytokines implicated in

the proinflammatory response have been associated with

clinical symptoms, eg, interleukin-6 (IL-6) and CRP with anorexia,31 IL-1ra with fatigue,31 and IL-6 with major depression.32,33 However, whether these cytokines exert their impact on symptoms in isolation or in combina-tion is unclear The reasons why systemic inflammacombina-tion worsens QoL in patients with cancer has recently been reviewed,34 and evidence from various preclinical and clinical studies suggest that the systemic inflammatory response has a direct role in the development of cancer- associated symptom clusters, including pain, fatigue, mood, anorexia, and physical function.34 Importantly, the effect of systemic inflammation on QoL was independent

of ECOG-PS, consistent with previous reports indicating that the systemic inflammatory response (mGPS) is asso-ciated with poorer QoL, even in those with a good perfor-mance score.8 Research is warranted to determine whether attenuating the systemic inflammatory response is capable

of producing clinically relevant improvements in symp-toms that may represent a new therapeutic approach to symptom management in patients with advanced cancer

We report herein that WL was associated with poorer QoL in almost all functional and symptom

TABLE 4 Clinical, Nutritional, and Inflammatory Parameters Related to Poor Quality-of-Life Summary Scores

(Below the Mean of <73.8) According to Multivariable Logistic Regression Analysis

Abbreviations: BMI, body mass index; ECOG PS, Eastern Cooperative Oncology Group performance status; mGPS, modified Glasgow prognostic score; OR, odds ratio.

aThese P values indicate statistical significance.

domains In particular, WL in excess of 10% in the

preceding 3 months was independently associated with

poorer physical function, fatigue, and appetite loss and

overall poorer QoL summary scores WL is a frequent

manifestation of malnutrition and is an important

cri-terion for the diagnosis of cancer cachexia, a

multifac-torial syndrome characterized by a negative protein

and energy balance driven by a variable combination

of reduced food intake and abnormal metabolism.35 In

patients with cancer, cancer cachexia is often defined

based on a single criterion: WL >5% over a period of

6 months The adverse impact of WL on QoL has long

been recognized in patients with cancer, and WL has

been associated with deterioration in patients’

perfor-mance status and psychosocial well being.36-38 In a

re-cent systematic review examining the impact of WL and

QoL, a negative relation between %WL and QoL was

reported in 23 of 27 studies included in the analysis.11

However, the mode by which WL exerts its influence

on QoL is not fully understood but may relate to

mus-cle atrophy associated with cachexia and WL leading to

fatigue or reduced functional capacity.39 Importantly,

interventions aimed at targeting nutritional status and

attenuating WL have proven successful in improving

aspects QoL in patients with cancer.40 In addition,

novel cachexia treatments, such as anamorelin, an

oral ghrelin-receptor agonist with appetite-enhancing

and anabolic activity, have shown a favorable clinical

response in alleviating anorexia-cachexia symptoms.41,42

When examining the effect of muscle parameters

and QoL outcomes, low SMI was associated with poorer

physical function and more insomnia, whereas low MA

was correlated with poorer physical function, role

func-tion, global health, and summary QoL scores and also

with more fatigue and dyspnea (all P < .05) Low MA

was independently associated with poorer

EORTC-reported physical functioning (hazard ratio, 1.67; 95%

CI, 1.09-2.56; P = .018), whereas low SMI was not This

is consistent with previous reports that low MA is

associ-ated with physical functional impairments, as evidenced

by improvements in timed-up-and-go, stair-climb, and

walking tasks.43 Inconsistent reports on this relation

be-tween muscle parameters and QoL have been published

in the literature.10,12,44,45 Parsons and colleagues reported

no significant associations between low SMI and

symp-tom burden or functional life domains assessed by the

MD Anderson Symptom Inventory in a cohort of 104

patients with advanced cancer.44 However, in a study of

734 patients with advanced lung cancer, low SMI was

nonlinearly associated with lower global QoL, physical function, and role function and was associated with more symptoms (fatigue and pain), whereas low MA was asso-ciated with poor physical function and more dyspnoea.10

An explanation for our findings may be that low SMI, at

a single time point, is not reflective of a dynamic measure

of loss and may be influenced by a patient’s intrinsic level

of muscularity Within our study, the composition of WL that influenced QoL was unknown, and perhaps losses of muscle over time may better reflect poor QoL A grow-ing body of evidence favors measures of muscle loss over time as prognostic of poor survival in patients with cancer compared with single-point measurements.46,47

The strengths of this study include the collec-tion of numerous variables measured with appropriate methods simultaneously in a relatively large sample of patients with incurable cancer In addition, using the QoL summary score to examine differences in QoL can avoid problems that may arise with multiple testing when otherwise making comparisons based on the 15 outcomes generated by the EORTC-QLQ question-naire.18 However, study limitations are also present The etiology of QoL is extremely complex given the web of determinants that influence it; and, although we accounted for several clinical and nutritional parame-ters, the list of variables examined was not exhaustive Given the convenient recruitment strategy, patients may have been at different time points in their disease trajectory when QoL was assessed (81% received had chemotherapy in the previous 4  weeks) In addition, patients may have received prior treatments, and this may have influenced QoL scores

Conclusion

In summary, the current findings provide evidence of the independent role of WL, ECOG-PS, and systemic inflam-mation (mGPS) in predicting poorer physical functioning, more fatigue and appetite loss, and poorer overall QoL sum-mary scores in patients with incurable cancer Our findings indicate potential targets for interventions aimed at safe-guarding the QoL of patients with advanced cancer Future work should focus on targeting the systemic inflammatory response, attenuating WL, and improving performance sta-tus in patients with incurable cancer as a means of improv-ing PROMs and reducimprov-ing symptom burden

FUNDING SUPPORT This work was funded in part with the financial support of Science Foundation Ireland (SFI) under grant SFI/12/RC/2273.

CONFLICT OF INTEREST DISCLOSURES

Claribel Simmons reports grants from Medical Research Scotland, during

the conduct of the study The remaining authors made no disclosures.

AUTHOR CONTRIBUTIONS

Louise E Daly: Conceptualization, data curation, formal analysis,

investigation, methodology, project administration, visualization,

writ-ing–original draft, and writing–review and editing Ross D Dolan:

Conceptualization, data curation, formal analysis, investigation,

meth-odology, project administration, visualization, writing–original draft,

and writing–review and editing Derek G Power: Data curation,

meth-odology, project administration, resources, supervision, and writing–

review and editing Éadaoin Ní Bhuachalla: Data curation,

investiga-tion, methodology, project administrainvestiga-tion, and writing–review and

editing Wei Sim: Data curation, investigation, methodology, project

administration, and writing–review and editing Samantha J Cushen:

Data curation, investigation, methodology, project administration, and

writing–review and editing Marie Fallon: Conceptualization, data

cura-tion, methodology, project administracura-tion, resources, supervision, and

writing–review and editing Claribel Simmons: Data curation,

inves-tigation, methodology, project administration, and writing–review and

editing Donald C McMillan: Conceptualization, data curation, formal

analysis, methodology, project administration, resources, supervision,

and writing–review and editing Barry J Laird: Conceptualization, data

curation, formal analysis, funding acquisition, methodology, project

administration, resources, supervision, visualization, writing–original

draft, and writing–review and editing Aoife M Ryan: Conceptualization,

data curation, formal analysis, funding acquisition, methodology, project

administration, resources, supervision, visualization, writing–original

draft, and writing–review and editing.

REFERENCES

1 Jordan K, Aapro M, Kaasa S, et al European Society for Medical

Oncology (ESMO) position paper on supportive and palliative care

Ann Oncol 2018;29:36-43.

2 Qi Y, Schild SE, Mandrekar SJ, et al Pretreatment quality of life is

an independent prognostic factor for overall survival in patients

with advanced stage non-small cell lung cancer J Thorac Oncol

2009;4:1075-1082.

3 Yang CJ, Roh JL, Kim MJ, et al Pretreatment quality of life as a

prog-nostic factor for early survival and functional outcomes in patients with

head and neck cancer Qual Life Res 2016;25:165-174.

4 Gotay CC, Kawamoto CT, Bottomley A, Efficace F The prognostic

significance of patient-reported outcomes in cancer clinical trials J Clin

Oncol 2008;26:1355-1363.

5 Kypriotakis G, Vidrine DJ, Francis LE, Rose JH The longitudinal

re-lationship between quality of life and survival in advanced stage cancer

Psychooncology 2016;25:225-231.

6 Cherny NI, Sullivan R, Dafni U, et al A standardised, generic,

vali-dated approach to stratify the magnitude of clinical benefit that can be

anticipated from anticancer therapies: the European Society for Medical

Oncology Magnitude of Clinical Benefit Scale (ESMO-MCBS) Ann

Oncol 2015;26:1547-1573.

7 Aaronson NK, Ahmedzai S, Bergman B, et al The European

Organization for Research and Treatment of Cancer QLQ-C30: a

qual-ity-of-life instrument for use in international clinical trials in oncology

J Natl Cancer Inst 1993;85:365-376.

8 Laird BJ, Fallon M, Hjermstad MJ, et al Quality of life in

pa-tients with advanced cancer: differential association with

perfor-mance status and systemic inflammatory response J Clin Oncol

2016;34:2769-2775.

9 Velikova G, Coens C, Efficace F, et al Health-related quality of life in

EORTC clinical trials—30 years of progress from methodological

de-velopments to making a real impact on oncology practice Eur J Cancer

(Suppl) 2012;10:141-149.

10 Bye A, Sjoblom B, Wentzel-Larsen T, et al Muscle mass and

associa-tion to quality of life in non-small cell lung cancer patients J Cachexia

Sarcopenia Muscle 2017;8:759-767.

11 Wheelwright S, Darlington AS, Hopkinson JB, Fitzsimmons D, White

A, Johnson CD A systematic review of health-related quality of life

instruments in patients with cancer cachexia Support Care Cancer

2013;21:2625-2636.

12 Nipp RD, Fuchs G, El-Jawahri A, et al Sarcopenia is associated with quality of life and depression in patients with advanced cancer

Oncologist 2018;23:97-104.

13 Arends J, Bachmann P, Baracos V, et al ESPEN guidelines on nutrition

in cancer patients Clin Nutr 2017;36:11-48.

14 Dolan R, Laird BJA, Horgan PG, McMillan DC The prognostic value

of the systemic inflammatory response in randomised clinical trials in

cancer: a systematic review Crit Rev Oncol Hematol 2018;132:130-137.

15 McMillan DC Cancer and systemic inflammation: stage the tumour

and stage the host Br J Cancer 2013;109:529.

16 Temel JS, Greer JA, El-Jawahri A, et al Effects of early integrated pal-liative care in patients with lung and GI cancer: a randomized clinical

trial J Clin Oncol 2017;35:834-841.

17 Bakitas M, Lyons KD, Hegel MT, et al Effects of a palliative care in-tervention on clinical outcomes in patients with advanced cancer: the

Project ENABLE II randomized controlled trial JAMA 2009;302:

741-749.

18 Giesinger JM, Kieffer JM, Fayers PM, et al Replication and valida-tion of higher order models demonstrated that a summary score for the

EORTC QLQ-C30 is robust J Clin Epidemiol 2016;69:79-88.

19 McMillan DC The systemic inflammation-based Glasgow prognostic

score: a decade of experience in patients with cancer Cancer Treat Rev

2013;39:534-540.

20 Proctor MJ, Morrison DS, Talwar D, et al An inflammation-based prognostic score (mGPS) predicts cancer survival independent of

tumour site: a Glasgow Inflammation Outcome Study Br J Cancer

2011;104:726-734.

21 Oken MM, Creech RH, Tormey DC, et al Toxicity and response

cri-teria of the Eastern Cooperative Oncology Group Am J Clin Oncol

1982;5:649-655.

22 Heymsfield SB, Wang Z, Baumgartner RN, Ross R Human body

composition: advances in models and methods Annu Rev Nutr

1997;17:527-558.

23 van Vugt JL, Levolger S, Gharbharan A, et al A comparative study of software programmes for cross-sectional skeletal muscle and adipose tis-sue measurements on abdominal computed tomography scans of rectal

cancer patients J Cachexia Sarcopenia Muscle 2017;8:285-297.

24 Mourtzakis M, Prado CM, Lieffers JR, Reiman T, McCargar LJ, Baracos

VE A practical and precise approach to quantification of body compo-sition in cancer patients using computed tomography images acquired

during routine care Appl Physiol Nutr Metab 2008;33:997-1006.

25 Shen W, Punyanitya M, Wang Z, et al Total body skeletal muscle and adipose tissue volumes: estimation from a single abdominal cross-

sectional image J Appl Physiol (1985) 2004;97:2333-2338.

26 Mitsiopoulos N, Baumgartner RN, Heymsfield SB, Lyons W, Gallagher

D, Ross R Cadaver validation of skeletal muscle measurement by

mag-netic resonance imaging and computerized tomography J Appl Physiol

(1985) 1998;85:115-122.

27 Martin L, Birdsell L, Macdonald N, et al Cancer cachexia in the age

of obesity: skeletal muscle depletion is a powerful prognostic factor,

independent of body mass index J Clin Oncol 2013;31:1539-1547.

28 Zimmermann C, Burman D, Swami N, et al Determinants of

qual-ity of life in patients with advanced cancer Support Care Cancer

2011;19:621-629.

29 Lam K, Chow E, Zhang L, et al Determinants of quality of life in advanced cancer patients with bone metastases undergoing palliative

radiation treatment Support Care Cancer 2013;21:3021-3030.

30 Laird BJ, McMillan DC, Fayers P, et al The systemic inflammatory response and its relationship to pain and other symptoms in advanced

cancer Oncologist 2013;18:1050-1055.

31 Paulsen O, Laird B, Aass N, et al The relationship between pro- inflammatory cytokines and pain, appetite and fatigue in patients with

advanced cancer PLoS One 2017;12:e0177620.

32 Breitbart W, Rosenfeld B, Tobias K, et al Depression, cytokines, and

pancreatic cancer Psychooncology 2014;23:339-345.

33 Du YJ, Zhang HY, Li B, et al Sputum interleukin-6, tumor necro-sis factor-α and salivary cortisol as new biomarkers of depression in