Let’s suppose you’re motivated to work on one of the several global challenges we face (God bless you). And suppose you’ve identified several opportunities to do that type of work but are uncertain about the best fit. 80,000 Hours suggests doing a series of “cheap tests” to gain information — talk with people about the job, spend some time on location, or do some type of trial work. When making important decisions, “cheap tests” requiring minimal time, effort, and resources can help to reduce uncertainty.
Some clinical trials use endpoints that directly measure outcomes of interest, while some others use surrogate endpoints. The FDA posits that surrogate endpoints are appropriate when:
- Clinical outcomes take a very long time to study
- The clinical benefit of improving the surrogate endpoint is well understood
- A clinical endpoint study would be unethical
- The surrogate measure is validated (i.e., can be relied upon to predict or correlate with a clinical benefit)
But what if the surrogate measure is cheaper yet inferior to another surrogate measure? Such is the case with the 6-minute walk test (6MWT), with implications for manufacturers of drugs in some classes that keep payers up at night.
A cheap test
The 6MWT is cheap. In fact, it doesn’t get much cheaper. Basically, all you need is a flat hallway with pre-measured marks, standardized instructions and encouragement, and a stopwatch. Ask a patient to walk as far as possible in 6 minutes and record the distance (an outcome known as 6MWD). It’s a cakewalk.
The 6MWT is a valid measure of functional exercise capacity. It is also reliable. That said, there is strong evidence that patients experience a learning effect when asked to perform the test two or more times; most patients walk further on the second 6MWT. For this reason, when the 6MWT is used to evaluate response to treatment or change over time, experts recommend two 6MWTs each time the test is performed and recording the longest distance.
There are several other problems afoot. The 6MWT is contraindicated for patients with a number of conditions and, of course, is not appropriate for those who cannot walk. There is a ceiling effect — that is, the 6MWT is less sensitive to measuring clinical change in patients with greater functional exercise capacity (e.g., a 6MWD greater than 450 meters, which is a little more than once around a 400-meter Olympic track). Age, gender, height, motivation, and effort can impact the 6MWD. And even the length of the hallway can also skew results — experts recommend (and evidence supports) a hallway or “course” that is at least 30 meters long. Additional turns on shorter courses take more time. Just imagine the dizzying number of turns while running a marathon on a balcony or in a small yard, as some ardent runners did during the COVID lockdown.
A better test
Cardiopulmonary exercise testing (CPET) is the gold standard for measuring functional exercise capacity. It provides a more accurate and sensitive assessment of functional exercise capacity than the 6MWT. Direct comparisons of the physiological demands of each test show that measures of peak exercise performance are similar between the 6MWT and CPET, but the 6MWT has substantially lower ventilatory requirements. Whereas CPET quantifies the factors limiting exercise, the 6MWT does not.
Both the 6MWD and CPET results correlate with cardiovascular endpoints, including heart failure severity, morbidity, and mortality and are, therefore, surrogate endpoints. A frequently cited strength of the 6MWD is that it correlates with CPET results. In that sense, the 6MWT is a surrogate of CPET. One can then view the 6MWT as a surrogate of a surrogate — two degrees of separation from the primary endpoints of greatest interest. That doesn’t sound like a strength!
Due to the time, effort, and resources needed, CPET costs significantly more to conduct than the 6MWT. CPET requires time to place the equipment, baseline measures at rest (2–3 minutes), a warm-up (2–3 minutes), exercise (8–12 minutes), a cool-down period, and then time to remove the equipment. CPET also requires more equipment (e.g., cycle ergometer or treadmill, a mask or mouthpiece, volume and gas sensors, ECG, and blood pressure and pulse monitors). And CPET requires specially trained staff. All said, it’s no walk in the park — and is why Medicare reimbursement for CPET is approximately five times greater than reimbursement for the 6MWT.
Costly consequences
Why do manufacturers use the 6MWT in clinical trials? Because they can. The FDA has given manufacturers the go-ahead to use this cheap test as an endpoint in clinical trials for an array of cardiovascular and pulmonary conditions. For example, it was a secondary endpoint in a recent trial studying how semaglutide, a GLP-1 agonist, could be used to treat patients with heart failure with preserved ejection fraction and obesity. Those results showed a statistically significant mean change in the 6MWD at 52 weeks — 21.5 meters with semaglutide vs. 1.2 meters with placebo (a difference of 20.3 meters or about 66 feet, 7 inches).
The semaglutide findings are encouraging. Patients with heart failure with preserved ejection fraction and obesity may have a new treatment option. Despite some uncertainty regarding outcomes, payers would be compelled to cover the GLP-1s with cardiovascular indications likely coming to market. As a result, GLP-1 utilization and costs would increase even faster than they are now, insurance premiums would likely rise, and manufacturer revenue and profits would surely go up.
Why should manufacturers go the extra meter and spend more on a better test that is more costly? First, they can afford it. Second, it can be differentiating. The GLP-1 market will likely be extremely competitive. If net costs were comparable, then it would be reasonable for payers to prefer GLP-1s with stronger evidence from trials that were larger, had longer durations, or used endpoints that directly measured the clinical outcomes of interest. Payers may also prefer drugs with evidence based on CPET or both CPET and the 6MWT over competitors with evidence based only on the 6MWT. So, if the FDA does not make the use of CPET mandatory (as it should but likely won’t), then forward-thinking and ambitious manufacturers may see an opportunity to gain a competitive advantage and do so voluntarily.
A cheap test that takes less time, effort, and resources is often desirable and sometimes necessary to reduce uncertainty. But when it comes to emerging competitive and costly markets, including GLP-1s with cardiovascular indications, a better test that takes more time, effort, and resources would increase certainty and should certainly be preferred by payers.
No Comments