Burden of illness
Even though quality-of-life (QoL) assessments are highly valued by patients and constitute a necessary component of cancer care evaluation, they generally receive too little attention in both daily routine and the clinical trial setting . “QoL evaluation can provide information that is unique and not available otherwise,” emphasised Richard J. Gralla, MD, Department of Medicine, Albert Einstein College of Medicine, Bronx, New York, USA.
According to a survey conducted in 1998 among 154 oncologists, 87 % believed that QoL data and patient-reported outcomes (PROs) are important for patients with advanced cancer . However, fewer than 50 % of these physicians always or frequently formally obtained such data at baseline, and fewer than 45 % used them in following or monitoring patients. “There is absolutely no indication that we are doing better today,” Dr. Gralla stressed. Additionally, instead of assessments of the actual patient QoL or symptoms, surrogate markers such as laboratory findings that represent symptoms are often resorted to. Another possibility is derivative endpoints, such as time to deterioration.
In essence, the assessment of QoL and PROs aims at the quantification of the individual patient burden of illness. “We do not sufficiently assess the burden of illness and its impact on prognosis, or the effects of treatment on this burden,” Dr. Gralla pointed out. Imaging, clinical laboratory testing and molecular testing cannot replace the direct patient input. Likewise, ECOG performance status (PS) does not fully capture the burden of illness and is not a PRO. QoL measurements are frequently not included in the design of clinical trials. In larger studies, they often represent secondary endpoints, but physicians tend to neglect them during follow-up.
Questionnaires: available and practical instruments
Three validated lung-cancer-specific QoL assessment instruments are currently available. The Lung Cancer Symptoms Scale (LCSS) contains patient and observer forms and has been developed for clinical trials and patient management, while both EORTC QLQ-C30 and FACT-L contain general and lung cancer modules and have been developed for general use. As Dr. Gralla noted, these questionnaires are inexpensive, have extremely high patient acceptance, and are easy to apply. “Using electronic assistance, it takes as little as 2 minutes to complete them.” QoL and PROs should be assessed every 3 weeks in patients with lung cancer.
Important prognostic information can be obtained through these instruments. For instance, data demonstrated an association between baseline PRO factors according to the LCSS 3-item global index and survival (Table 1) . The 3-item global index includes symptom distress, interference with the activity level, and health-related quality of life. These clusters are a more reliable measure than symptoms whose patterns vary individually . Dr. Gralla pointed out that PS does not reflect this information to a sufficient degree, as several prognostic groups can be present within the same PS category.
This is illustrated by data on hospitalisation rates. Hospital admissions in patients with advanced lung cancer are due to either the disease and its complications (73 %) or toxicity (27 %) . Among 160 patients with advanced NSCLC included in a prospective assessment, one-third were hospitalised in the first 90 days. The LCSS 3-item global index baseline score was found to be highly predictive for cancer-related hospitalisations (p = 0.0001). The separate analysis of a patient cohort with ECOG PS 1 showed that the same relation exists within this group, which indicates that PS itself is not an accurate predictive measure. At 90 days, there was a 4:1 difference between the patient cohorts with the highest and the least risk according to the 3-item global index (48 % vs. 12 %; p = 0.025; Table 2)
Early determination of treatment benefits
Dr. Gralla concluded that QoL and PRO assessments have multiple roles in thoracic oncology. “Currently, they are secondary endpoints in large randomised trials and serve for informal assessment in daily practice.” Enhanced roles, however, might include identification of patients at increased risk, which allows for addressing their needs and reducing hospitalisation rates. More accurate baseline prognostic data than PS at the outset of clinical trials renders improvements in study design possible. Assuring high completion of practical QoL/PRO assessment of appropriate endpoints is mandatory.
Finally, evaluations might enable earlier determination of the benefit of a certain treatment in an individual patient. In this context, significant survival differences have been shown based on a 20 % decline in the 3-item prognostic index in NSCLC patients already after 2 cycles of chemotherapy (p = 0.01) . “If we evaluated PROs and QoL early on, we would get much more information that is very valuable both in clinical trials and in patient management,” Dr. Gralla concluded.
Lung cancer stigma affects all levels of care
The concept of health-related stigma is not new but highly relevant in the context of lung cancer. “We found that as many as 95 % of lung cancer patients feel stigmatised by other people,” reported Heidi A. Hamann, PhD, Department of Psychology, Department of Family and Community Medicine, University of Arizona Cancer Canter, Tucson, Arizona, USA . This stigma is often linked to the belief that behaviour, such as smoking, has caused the person’s illness, and is likely associated with several negative psychosocial and behavioural outcomes such as depression, poor treatment adherence, and underreporting of symptoms.
Dr. Hamann pointed out that stigma is relevant across the continuum of lung cancer care . “This includes prevention, detection, diagnosis, treatment, and survivorship.” Moreover, implications of stigma need to be dealt with on many levels of lung cancer care ranging from the individual patients to families, practice settings and even the national level, where policies and financial/political structures are provided .
The Lung Cancer Stigma Inventory (LCSI) has been developed in three phases that entailed interviews, generating and refining of items, and multi-site field testing among lung cancer survivors . Three factors related to patient-reported stigma were identified: internalised stigma, perceived stigma, and constrained disclosure (Table 3). The LCSI Measure is available for download on the NCI GEM database. “In terms of other psychometrics associated with this measure, a high test-retest correlation was found, as well as good convergent validity with the Cataldo Lung Cancer Stigma Scale, although the two tests were not completely overlapping,” explained Dr. Hamann. Ever smokers reported higher internalised stigma scores than never smokers, while these two groups did not differ with respect to perceived stigma. Also, there was a positive correlation between different aspects of stigma and the depression scale: higher levels of stigma correlated with higher levels of depression.
“Patient-reported stigma has multifaceted psychosocial impact and potentially affects multiple levels of lung cancer care,” Dr. Hamann summarised. “Addressing it is very important.” Further research is required regarding the care-related impact of stigma. “We need more data to truly understand how stigma affects treatment decisions, patient adherence, and clinical trial involvement.” Moreover, interventions, including those at the patient level, should be tested with a focus on internalised stigma and constrained disclosure.
- Gralla RJ, Quality of life: are we paying enough attention? WCLC 2018, MS21.05
- Morris J et al., The use of quality of life data in clinical practice. Quality of Life Research 1998; 7(1): 85-91
- Gralla RJ et al., Prediction of survival outcomes in NSCLC using a new PRO index from the LCSS (Lung Cancer Symptom Scale): Results of a 622-patient prospective trial. J Clin Oncol 2013; 31 (15_suppl): 8087-8087
- Gralla RJ et al., An evidence-based determination of issues affecting quality of life and patient-reported outcomes in lung cancer: results of a survey of 660 patients. J Thorac Oncol. 2014; 9(9): 1243-1248
- Gralla RJ et al., Predicting risk of hospitalization in patients with NSCLC receiving chemotherapy using the LCSS 3-item global index (3-IGI). WCLC 2017, P2.01-050
- Gralla RJ et al., Can benefit or futility in treating advances NSCLC be determined early using the LCSS 3-item global index (3-IGI) PRO? WCLC 2018, P2.01-39
- Hamann HA, Identifying consequences of stigma on lung cancer care delivery and patient outcomes. WCLC 2018, MS11.01
- Campo RA et al., Cancer prevention after cancer: changing the paradigm–a report from the American Society of Preventive Oncology. Cancer Epidemiol Biomarkers Prev 2011; 20(10): 2317-2324
- Taplin SH et al., A multilevel research perspective on cancer care delivery: the example of follow-up to an abnormal mammogram. Cancer Epidemiol Biomarkers Prev 2012: 21: 1709-1715
- Hamann HA et al., in press
Author: Judith Moser, MD
Lecture Board: Heidi A. Hamann, PhD
© 2018 Springer-Verlag GmbH, Impressum