Avapritinib for PDGFRA D842V-Mutant GIST

Discussion of the NAVIGATOR Trial

Obligatory—This is not medical advice

It’s rare that there are any real ‘game changers’ or home runs in medicine. Frequently, clinical trials with modest benefit result in FDA approval. Occasionally, however, something truly impressive appears. My bias is clear—if not already evidenced by the title of this substack—I treat patients with sarcoma. Therefore, breakthroughs in sarcoma are exciting to me. I consider avapritinib for patients with PDGFRA D842V-mutant gastrointestinal stromal tumors (GISTs) to be one of them. There are always areas to improve, certainly, but one must recognize that avapritinib provides effective treatment to patients who previously had limited options.

Background

As we’ve covered previously, the majority of GISTs have mutations in KIT or PDGFRA. Common mutations are shown in the figure below. Approximately 5% of GISTs harbor a mutation in PDGFRA D842V. This particular mutation happens in exon 18 and involves a substitution in the A-loop portion of an aspartic acid with a valine (hence D842V).¹ This simple change produces an innate resistance to imatinib and sunitinib, which have each been discussed in this substack before. Both of these medications had demonstrably conferred a survival benefit in pivotal clinical trials in broad populations of patients with GISTs. This benefit, however, was not true for patients who had PDGFRA D842V-mutant GISTs, whose overall survival was observed to remain a median of 12.7 months.²

Rizzo et al. Cancers. 2021. (Reference #1)

Studying Avapritinib

The NAVIGATOR trial was a two-part, open-label, non-randomized dose-escalation/expansion phase I study done at 17 sites in nine countries looking at the safety and efficacy of avapritinib in patients with advanced GIST.³ The first part looked for the maximum tolerated dose. Part 2 enrolled into three groups with a dose of 400mg daily (subsequently reduced).⁴

Groups:

1. Group 1: Disease progression following imatinib pluse one or more other kinase inhibitor and no D842V mutation

2. Group 2: PDGFRA D842V mutation regardless of previous therapy

3. Group 3: Patients who had received only imatinib and no known D842V mutation

Inclusion criteria otherwise were fairly simple with performance status and age ranges consistent with prior trials. Exclusion criteria likewise were standard, although patients with controlled HIV, seizure disorders, or TIA within the last year may be considered stringent. Wild type GIST was excluded.

Jones et al. Eur J Cancer. 2021. (Reference #4).

The primary endpoint was maximum tolerated dose for the dose escalation portion. The primary endpoint of the dose expansion part was the overall response rate by mRECIST 1.1.⁵ There were some minor modifications to mRECIST, which can be viewed in the protocol. My interpretation is that these do not affect the results.

What did it show?

In total, 250 patients were enrolled in parts 1 and 2. About 2/3 of these patients got the starting dose of 300mg daily. At data cutoff, 205 patients had discontinued. 51% of discontinuations were secondary to progression, while 15% were secondary to treatment-related adverse events (AEs). Of the 250 patients enrolled, 56 had PDGFRA D842V-mutant GIST.

Regarding the population of patients with PDGFRA D842V-mutant GIST:

1. 38 of 56 patients received a starting dose of 300 or 400 mg

   1. 8 of these patients stopped secondary to treatment-related AEs

2. All patients experienced at least one treatment-related adverse event (see table below)

3. Dose modifications occurred in 96% of cases (89% had interruptions, 73% had reductions).

4. One death on study was considered treatment related, and labeled as schizophrenia (or psychosis)

   1. 57% of patients had any-grade cognitive AEs in the PDGFRA D842V population

Jones et al. Eur J Cancer. 2021. (Reference #4)

Despite the toxicity, the rate of response was higher numerically than what has been shown in prior trials of other agents in more general populations of patients with GISTs. ⁶⁷ The overall response rate of the population with PDGFRA D842V mutations was 91%, with a response rate of 95% amongst those with doses of 300mg or 400mg daily. See the waterfall plot below. It’s striking. 3 of 4 patients who did not technically meet mRECIST criteria had reduced size of their tumors. Median overall survival was not reached.

Jones et al. Eur J Cancer. 2021. (Reference #4)

Jones et al. Eur J Cancer. 2021. (Reference #4)

These results established avapritinib as the preferred agent for first line treatment for patients with advanced PDGFRA D842V-mutant GIST.

Some questions remain:

1. What is the role of avapritinib in the adjuvant setting for patients with PDGFRA D842V mutated GIST?

2. What are the resistance mechanisms to this medication?

3. What are potential later line therapies after failure of avapritinib?

4. How do we best address the clear toxicity present with this medication that led to discontinuation in 19 patients (34%)?

5. What are means of minimizing the off-target effects for this or future medications? Dose reductions and interruptions work, but how can this be done safely?

These are just a few and I won’t further belabor it here. Avapritinib clearly works.

Conclusions

Avapritinib is the standard of care for patients with advanced PDGFRA D842V-mutant GIST. In most cases, a response is anticipated. This is a tremendous improvement for patients with PDGFRA D842V mutations which is hard to overstate. Many questions remain about how to best treat our patients in various contexts who have PDGFRA D842V mutant GIST, including in the adjuvant setting, or after failure of avapritinib. Additionally, there may be serious toxicity associated with avapritinib administration and dose alteration is necessary in nearly all patients. Providers should be very transparent about the potential risks and benefits of this agent. We are still in need of effective later line therapies for our patients with GISTs.

1. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7916155/

2. https://ascopubs.org/doi/10.1200/jco.2010.28.15_suppl.10051

3. https://www.thelancet.com/pdfs/journals/lanonc/PIIS1470-2045(20)30269-2.pdf

4. https://www.ejcancer.com/article/S0959-8049(20)31423-4/fulltext

5. https://radiologyassistant.nl/more/recist-1-1/recist-1-1

6. https://pubmed.ncbi.nlm.nih.gov/12181401/

7. https://pubmed.ncbi.nlm.nih.gov/17046465/