Does the GeneSight® test change outcomes for all patients?
This is the number of patients who need to receive an intervention in order for one patient to respond due to the benefit of the intervention. Since placebo effect plays a significant role in medicine, it can be difficult to determine the direct benefit of an intervention over and above the placebo effect. The calculation to derive NNT eliminates the benefit of placebo effect, thus exposing the direct benefit of the intervention.
When evaluating how the GeneSight test affects NNT, it is first necessary to consider the NNT of the original intervention, then to evaluate the change in NNT when patient treatment is guided by pharmacogenomic testing.
The NNT for antidepressant medications is 10.
The Sequenced Treatment Alternatives to Relieve Depression (STAR*D) project has shown that the response rate for antidepressants for treatment-naïve patients is around 50%.1,2 This means that for every ten patients treated for depression, five will respond (Figure 1A). However, placebo effect plays a significant role in this response. Of the five patients who respond to treatment, four of them would have responded whether they were given an antidepressant or a placebo3 (Figure 1B). Therefore, only one patient in ten actually benefits from the direct effect of the medication (Figure 1C).
When treatment is guided by the GeneSight test, the NNT drops to 6.
All patients enrolled in clinical trials involving the GeneSight test have failed at least one medication trial. Therefore, they are not treatment-naïve and do not have a 50% chance of response. The STAR*D data has demonstrated that patients on a second medication trial have about a 29% chance of response, and those on a third trial have about a 17% chance of response.2 However, when patient treatment was guided by the GeneSight test, four out of ten patients responded, correlating with a 40% response rate4 (Figure 2A). For patients in the treatment as usual group, about two and a half out of ten responded4 (Figure 2B). This is a 25% response rate, which aligns with the response rate we would predict for patients on a second or third medication trial based on the STAR*D data. Therefore, since two and a half patients would have responded to treatment as usual, we can deduce that one and a half additional patients out of ten responded because of the effect of the GeneSight test (Figure 2C).
Precision prescribing using the GeneSight test resulted in a NNT of about six (Figure 3). Since the GeneSight test is not a medication, but rather a way to select more genetically appropriate medications for a patient, using the GeneSight test is equivalent to reducing the NNT with antidepressants from ten to six (Figure 3).
While having to treat six patients to get one to respond due to the direct effect of the medication may not sound encouraging, this is a significant improvement from where we have been in the past. It is worthwhile noting that the GeneSight test may not be able to help every patient achieve response, but that across a population it increases response by 71% compared to empiric prescribing.4
As the scientific community continues to learn more about how genetics and non-genetic factors can affect patient response to psychiatric medications, Assurex Health looks forward to continually updating and refining the impact of known gene-drug interactions on the GeneSight test, and to continue to advance the field of psychiatric pharmacogenomics by decreasing the NNT for patients with mental illness even further.
1. Warden, D., Rush, a J., Trivedi, M. H., Fava, M. & Wisniewski, S. R. The STAR*D Project results: a comprehensive review of findings. Curr. Psychiatry Rep. 9, 449–59 (2007).
2. Rush, A. J. et al. Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report. Am. J. Psychiatry 163, 1905–17 (2006).
3. Iovieno, N., Tedeschini, E., Ameral, V. E., Rigatelli, M. & Papakostas, G. I. Antidepressants for major depressive disorder in patients with a co-morbid axis-III disorder: a meta-analysis of patient characteristics and placebo response rates in randomized controlled trials. Int. Clin. Psychopharmacol. 26, 69–74 (2011).
4. Altar, C. A. et al. Clinical Utility of Combinatorial Pharmacogenomics-Guided Antidepressant Therapy: Evidence from Three Clinical Studies. Mol. Neuropsychiatry 1–11 (2015). doi:10.1159/000430915