Effect of early palliative care (PC) on quality of life (QOL), aggressive care at the end-of-life (EOL), and survival in stage IV NSCLC patients: Results of a phase III randomized trial

Reporter: J. Nicholas Lukens, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: June 7, 2010

Presenter: J. S. Temel, Massachusetts General Hospital, Boston, MA.


  • Patients with metastatic NSCLC often have a high symptom burden from the time of initial diagnosis, which greatly affects their quality of life.
  • These patients are often subjected to aggressive care at the end of life, which can contribute to patient distress, and the high cost of health care at the end of life.
  • Palliative care focuses on the physical and psychological well-being of the patient, with emphasis on educating patients about their illness, clarifying treatment goals, helping patients and their care-givers cope with the illness, and symptom management.
  • The traditional cancer care model focuses on symptom management and palliative care only once curative measures have been exhausted.
  • An early palliative care model is one in which palliative care is integrated with curative cancer care from the time of diagnosis.
  • The presenting authors had piloted an early palliative care model in a Phase II study (Temel, JCO, 2007), and found that it was feasible to provide simultaneous standard curative cancer care and palliative care.
  • The authors initiated this Phase III randomized trial to examine the impact of early palliative care on patient-reported and end-of-life care outcomes in newly diagnosed metastatic NSCLC patients.


  • Eligibility criteria: Patients with an ECOG Performance Status of 0-2, within 8 weeks of diagnosis of metastatic NSCLC.
  • Study design: 150 patients with newly diagnosed metastatic NSCLC were randomized in 1:1 manner to early palliative care (PC) with oncology care versus standard oncology care (SC).
  • Patients in the early PC arm met with the palliative care team within 3 weeks of enrollment, and at least monthly thereafter.
    • Visits with PC team were not scripted, but guidelines were provided according to the National Consensus Project that included a focus on illness understanding and education, symptom management, decision making, and coping with a life-threatening illness.
  • Patients in the SC arm were permitted to meet with the palliative care team upon request.
  • Study measures:
    • Quality of life: Measured using the FACT-Lung, which is a validated scale consisting of 5 sub-scales:
    • Trial Outcome Index (TOI) is a combination of lung cancer specific symptoms scale, and the functional and physical well-being scale.
    • Psychological distress: Measured using the HADS (Hospital anxiety and depression) scale, which measures symptoms of depression, and the Patient Health Questionnaire 9 (PHQ-9), which uses more strict DSM-IV criteria for depression.
  • Data collection:
    • Prior to randomization, patients completed baseline measures of QOL (FACT-Lung) and mood (HADS and PHQ-9).
    • QOL and mood assessments were performed at 12 weeks and patients were followed until death.
    • Data on care received at the EOL (hospice utilization, length of stay on hospice ? 3 days, chemotherapy in the last 14 days of life, and survival) were also collected.
  • The primary study objective was to measure the difference in QOL between the 2 study arms at 12 weeks, as measured by the FACT-Lung TOI.
  • Secondary objectives:
    • Psychological distress at 12 weeks,
    • Quality of end of life (EOL) care,
    • Resource utilization at EOL, and
    • Documentation of resuscitation preference in medical record.


  • 283 patients were assessed for eligibility between June 2006 and July 2009, and 119 patients did not participate in the study. Therefore, 151 patients (55% eligible patients) were enrolled.
  • Baseline characteristics:
    • The study group as a whole was relatively homogenous, with 95% of SC and 100% of PC patients being Caucasian.
    • There were no significant baseline differences between groups in terms of PS, brain metastases, initial anti-cancer therapy, and percentage of patients treated with radiation (35% in each arm).
    • Baseline QOL and psychological distress was equivalent between the 2 groups.
  • At 12 weeks, 27 patients had died (17 PC and 10 SC), and 107 (86%) patients completed assessments.
  • Primary outcome (QOL):
    • Patients assigned to early PC experienced better QOL, as measured by the FACT-Lung (98.0 v 91.5, p = 0.03) and TOI (59.0 v 53.0, p = 0.01) scales.
  • Secondary outcomes:
    • Patients randomized to early PC had lower rates of depression, as measured by both the HADS (15.8 v 38.3%, p = 0.01) and the PHQ-9 (3.5 v 17%, p = 0.02) compared to SC.
    • Change in QOL from baseline: Those randomized to SC had a decrease in their QOL, while those in the early PC arm had an improvement in their QOL from baseline (p=0.04 for TOI scale)
    • Fewer patients randomized to early PC received aggressive EOL care (no hospice, length of stay on hospice ? 3 days, chemotherapy in the last 14 days of life) compared with SC patients (33.3 v 53.6%, p = 0.05).
    • Despite receiving less aggressive care at the EOL, patients receiving early PC had longer median survival compared with SC (11.6 m v 8.0 m, p = 0.02).

Authors' Conclusions

  • The authors conclude that early and integrated palliative care in patients with newly diagnosed metastatic NSCLC led to:
    • Significant improvements in QOL,
    • Lower rates of depression,
    • Less aggressive care at end of life, and
    • Prolonged survival
  • They acknowledge certain study limitations:
    • Single, tertiary care center with specialized palliative care group
    • Study population lacked diversity
    • Some patients on SC arm were seen by palliative care (although this should only dilute study effect)
    • Survival was not a pre-specified endpoint, and there is a lack of data on mediators of medical outcomes.
  • The authors propose that the improvement in QOL seen with early palliative care is due to improved symptom management, and that decreased rates of depression may be due to better symptom management, and acceptance of illness.
  • They hypothesize that the improved survival in these patients may be due to earlier recognition of medical issues, improved QOL and mood, and more appropriate medical care at end of life (less chemotherapy).

Clinical Implications

  • This is an important study demonstrating that the early integration of palliative care with standard oncologic care in patients with metastatic NSCLC is not only feasible, but also leads to improvements in validated patient-reported measures of quality of life and mood, less aggressive end of life care, and potentially improved survival.
  • As such, it validates a paradigm shift from the traditional cancer care model, which focuses on symptom management only once curative measures have been exhausted, to a model of early integrated palliative care.
  • Although this study only examined patients with newly diagnosed metastatic NSCLC, the early palliative care model could appropriately be applied to a number of other disease sites characterized by a high burden of symptoms and short life expectancy.
  • The improvement in survival observed in patients receiving early palliative care is intriguing, but as the authors acknowledge, the study was not designed to test this hypothesis, and the factors contributing to improved medical outcomes were not analyzed.
  • In addition to the study limitations acknowledged by the authors (see above), another limitation is that only 55% of the eligible patients were actually enrolled on the study, and some of this was due to patient preference.
    • This suggests that the study patients may have been a self-selecting group who were more likely to derive a benefit from early palliative care, and that the study group is not representative of all patients with newly diagnosed metastatic NSCLC.
  • Nevertheless, this is a powerful and intriguing study that validates a paradigm shift to early and integrated palliative care, and poses interesting questions as to why this approach led to improved survival.


We Need to Know When to Stop
by Rodney Warner, JD
September 26, 2013

Why We Should Think About End-of-Life Wishes
by Marlys Johnson
September 28, 2016

Stay informed with the latest information from OncoLink!   Subscribe to OncoLink eNews
View our newsletter archives