Stopping Statins at the End of Life

Published

Kutner JS, et al. "Safety and benefit of discontinuing statin therapy in the setting of advanced, life-limiting illness: a randomized clinical trial". JAMA Internal Medicine. 2015. 175(5):691-700.
PubMed • Full text

Clinical Question

In patients with a life expectancy of <1 year, does discontinuing statin therapy for cardiovascular disease increase 60 day all-cause mortality when compared to continuation of statin therapy?

Bottom Line

Among patients with life expectancy <1 year who are not on a statin for active CVD, discontinuation of statin therapy does not increase 60 day all-cause mortality when compared to statin continuation. Discontinuation may be associated with improved quality of life (QOL) and reduced medication costs.

Major Points

Statins are widely used for primary and secondary prevention of ASCVD with modern guidelines recommending use of the medication class should individuals be at an elevated 10 year risk of ASCVD. For individuals with a limited life span, the medications are of unclear benefit.

Published in 2015, this study by Kutner and colleagues randomized 381 patients on long-term statin therapy with life expectancy <1 year to discontinuation or continuation of their statin therapy in an open label fashion. Individuals receiving the medication for active CVD were excluded. The trial was originally designed to test overall mortality but given the longer-than-expected survival in both arms, the primary outcome was changed to 60 day all-cause mortality. There was no difference in the primary outcome between groups though individuals in the discontinuation group scored higher on QOL measurements and had reduced costs of medications.

Guidelines

As of September 2015, no guidelines have been published that reflect the results of this trial.

Design

• Multicenter, open label, randomized controlled trial
• N=381
  • Discontinue statin therapy (n=189)
  • Continue statin therapy (n=192)
• Setting: 15 US palliative care centers
• Enrollment: 2011-2013
• Median follow-up: 18 weeks
• Analysis: Intention-to-treat
• Primary outcome: 60 day all-cause mortality

Population

Inclusion Criteria

• Age ≥18 years old
• English speaking
• Receiving a statin medication for ≥3 months for primary or secondary prevention of cardiovascular disease
• Documented diagnosis of advanced, life-limiting illness, defined by all of the following:
  • ≥1 physician who would not be surprised if the patient died in the following year
  • Life expectancy >1 month
  • Recent deterioration in functional status by the Australia-Modified Karnofsky Performance Scale (AKPS) to <80% in prior 3 months
• Cognitively intact (Short Portable Mental Status Questionnaire score ≤4) or has a legal surrogate who is willing and able to provide consent and study data related to the study participant

Exclusion Criteria

• Active CVD requiring statin therapy
• Myositis
• LFTs or creatinine >2.5x ULN

Baseline Characteristics

• Demographics: Age 74 years, male sex 55%, college graduate 34%
  • Race/ethnicity: White 83%, Black 14%, multiple 1%, other 3%, Hispanic 4%
• PMH: CVD 58%, cognitively impaired 22% (more randomized to statin discontinuation; 27% vs. 17%; P=0.02)
• Primary end-of-life diagnosis: Cancer 49%, other 51%
• Enrolled in hospice: 36%
• Duration of statin therapy: >5 years 69%, 1-5 years 26%, < 1 year 2%, unknown 3%
• # of non-statin medications: 11
• Generic statin: 75%
• McGill QOL score (higher QOL has higher score): 6.98 (total)
  • Overall: 6.12
  • Physical: 5.19
  • Psychological: 7.21
  • Well-being: 7.30
  • Support: 8.31

Interventions

• Randomized to discontinue statin therapy or continue to receive statin medication

Outcomes

Comparisons are discontinuation vs. continuation of statin therapy.

Primary Outcomes

60 day all-cause mortality

23.8% vs. 20.3% (90% CI -3.5 to 10.5; P=0.36)

Secondary Outcomes

Median time to death

229 vs. 190 days (P=0.60)

CV death: P=0.64 (proportions not given)

Proportion of patients experiencing cardiovascular-related event

13 vs. 11 patients (statistics not given)

McGill Quality of life (QOL) Questionnaire score

Higher scores indicate higher QOL. AUC is area under curve difference.

7.11 vs. 6.85 (95% CI 0.02-0.50; P=0.04)

Single-question QOL domains:

Overall: 6.53 vs. 6.35 (AUC 0.18; 95% CI -0.28 to 0.64; P=0.44)

Physical: 5.43 vs. 5.51 (AUC -0.08; 95% CI -0.43 to 0.26; P=0.64)

Psychological: 7.38 vs. 6.99 (AUC 0.39; 95% CI -0.02 to 0.80; P=0.06)

Well-being: 7.37 vs. 7.05 (AUC 0.32; 95% CI 0.00 to 0.64; P=0.05)

Support: 8.31 vs. 7.86 (AUC 0.53; 95% CI 0.16 to 0.90; P=0.005)

Patient satisfaction with current health care score

Out of 5, 5 indicating highest satisfaction of care.

4.63 vs. 4.55 (P=0.22)

Longitudinal AKPS score

Out of 100, 100 indicating highest performance status.

47.7 vs. 48.5 (P=NS)

Edmonton Symptom Assessment System scale

Out of 100, 100 indicating highest symptom burden.

32.4 vs. 34.8 (P=0.18)

Statin-specific: 7.0 vs. 7.2 (P=0.71)

Total number of nonstatin medications

10.1 vs. 10.8 (95% CI -1.3 to -0.1; P=0.03)

Adverse Events

No additional adverse events were determined to be study-related, other than those addressed in the primary and secondary outcomes.

Cost Savings Analysis

Average savings per patient from statin therapy discontinuation in 2012 dollars

$3.37 per day (95% CI 2.83-3.91)

If using only generic: $2.96 (no 95% given)

$716.46 per mean follow-up time of 212.6 days

If using only generic: $629.30

Projected annual US savings in 2014

$603 million

Criticisms

• The trial endpoint what changed from overall survival to 60 day mortality because of longer-than-expected median survival and the need to dramatically increase enrollment from ~1,200 to ~30,000 participants. It's unclear if this is a clinically-meaningful endpoint.
• The non-inferiority endpoint was inconclusive for the difference in mortality rates between the two study groups.
• Since the cost of patient prescription copays and third party prescription coverage were unknown, the cost analysis may inaccurately reflect actual patient cost savings.
• A greater number of patients in the discontinuation arm were cognitively impaired. Since these patients may not have been able to accurately express their QOL, the QOL results may not be reliable.
• Statistical data was not adequately provided for some important secondary outcomes, such as time until death, time until first cardiovascular event, and proportion of adverse cardiovascular events.
• Differences between the groups regarding QOL subscale results were statistically significant, yet small. This likely does not translate into a clinically significant difference. No significant difference was observed for physical symptoms or performance status, indicating that clinical benefits from discontinuing therapy were minimal.
• The study used a 90% CI to assess their primary outcome which allowed for greater margin of error.

Funding

The National Institute of Nursing Research