David S. Hong, MD

 key opinion leaders in the management solid tumor cancers consider the evolving treatment landscape for 

fusion-positive disease and share expert perspectives on the role of tyrosine receptor kinase (TRK) inhibitors for these patients. The 2 currently available TRK inhibitors, larotrectinib and entrectinib, have the potential to address important areas of unmet need, but investigators are continuing to assess how clinicians can best optimize therapeutic approaches with these agents. Ahead, David S. Hong, MD, highlights key efficacy and safety data on larotrectinib and examines how mechanisms of treatment resistance may inform future research into second-generation TRK inhibitors.

: What is the preferred method for detection of 

 fusion testing is NGS [next-generation sequencing], and preferably RNA-based [NGS]. There are a lot of commercial vendors out there that are DNA-based. I think FoundationOne is DNA-based. There are a number of them. An RNA-based NGS is preferred partly because sometimes the actual fusion partner genes can be much larger than what DNA-based NGS can pick up. Obviously, NGS is catching on at least here in the United States and more and more oncologists can order it. Usually, academic hospitals like The University of Texas MD Anderson Cancer Center have an NGS platform in-house, but there are lots of vendors out there specifically that do this. Foundation Medicine, Caris, and Tempus are all different types of vendors that can do that. In Europe there's usually an algorithm as to when you can order NGS because it’s usually socialized medicine, and they usually have to go through an algorithm where they can get FISH [fluorescence in situ hybridization] testing specifically for certain fusions. After that, if they're negative, then they can get NGS testing, but ideally RNA-based NGS testing.

: When is the best time to perform testing for NGS?

 This is a controversial question. I'm a phase 1 and early development doctor. I would argue that there is increasingly evidence that NGS testing or molecular testing in general should occur as soon as the patient is diagnosed with metastatic cancer or stage IV cancer. Now, not everybody would agree with that because some types of tumors don't really have targeted therapy. There are some diseases where I would argue that it is almost malpractice if you don't get NGS or some kind of wide-based molecular testing, such as lung cancer. We have multiple new therapies now so that if they have that alteration, they can get targeted therapy even in the frontline setting, as in lung cancer.

I think sometimes the challenge is the turnaround time of some of these NGS tests. After you get a biopsy and then send it off, it can take weeks for the turnaround time of it. Some folks would argue that it's better to get local testing, specifically targets that we know can be turned around fairly quickly in local hospitals, such as 

 testing. I would argue that, as soon as a patient is diagnosed with metastatic disease, it would be ideal to get the NGS testing.

: How does the long-term efficacy demonstrated with TRK inhibitors such as larotrectinib compare with other currently available agents?

 The latest data set that we have was from ESMO [European Society for Medical Oncology] 2021 this past year from Dr Alexander Drilon [Memorial Sloan Kettering Cancer Center, New York City, New York]. We were presented this data at ESMO looking at 

 fusion and larotrectinib. We had a total of 218 patients, and this was across a different number of prior lines of therapy. And again, this was all different tumor types across different ECOG [performance statuses], and the overall response rate remained around 75%. Our initial data cutoff showed around 80% response rates. The median duration of response was close to 49 months, median progression-free survival was 35 months, and median overall survival had not been yet reached. I would argue that there is no therapy out there, if you have metastatic cancer, that matches that.

In many ways, I would argue that, if you are a patient who is identified to have an 

 fusion, it's almost like a lottery ticket because these patients do so well on the study. The full data set did show that patients who had greater than 3 prior lines of therapy had a slightly lower median duration response or progression-free survival than somebody who had no prior therapy up-front. It's expected with the sheer burden of disease. With ECOG [performance status], there seems to be less likelihood of response, but again, that probably has to do with their overall condition and so forth. Overall, If you are able and identified to have an 

 fusion, I would argue that you need to try to get onto larotrectinib or entrectinib at some point.

: What implication might these results have for patient quality of life? How might that affect your treatment approach in some patients? What does this tell us regarding the importance of testing for 

 We've looked at this quality-of-life issue, we've looked at PROs [patient-reported outcomes] over periods of time, and it clearly shows that these patients do well on this study. From their previous state to actually going on larotrectinib, they have incredible, significant improvement in their overall quality of life. The overall adverse event profiles of these drugs are very benign. Probably the most common side effect tends to be this slight dizziness that patients develop, and over time, they actually get better. There are some grade 3 events such as an increase in LFTs [liver function tests], but usually these adverse events or side effects are manageable. You just hold the drug, you dose reduce them, and they still seem to respond to therapy, at least with larotrectinib. I don't know the data for entrectinib as well, but they're very similar drugs. I would argue that these are very well-tolerated, and their overall quality of life is very good over a period of time.

I'm an investigator, so I do mostly clinical trials, but I think that if patients are identified to have 

fusions, whether it's in the frontline setting and even the second-line setting, even though we don't necessarily have randomized data in every single tumor type, given this data at this point, I would argue that it makes sense to try these patients on TRK inhibitors compared with standard chemotherapy, which can be very toxic. In many ways, I see it as almost like a lottery ticket if you're a patient who has this 

fusion. It's worth giving larotrectinib or entrectinib a try.

You have to try to find these patients. I think we will find more and more of these patients as NGS testing becomes more common and almost reflexive, particularly in patients with metastatic cancer. I'm sure there are lots of patients out there who probably have 

 fusions who we’re not identifying. Ideally, we want to identify as many as we can.

: Regarding the on-target and off-target resistance that's been observed with the first-generation TRK inhibitors, how do these resistance mechanisms differ, and which type do you most commonly see in patients?

 There are really 4 different mechanisms of resistance. Again, the king of 

, Dr Alexander Drilon, published a great paper in 

 looking at these different types of mutations. One is something called solvent-front mutations and gatekeeper mutations. Both of these mutations, differing on where they are, and the third, called xDFG, really kind of affect that near-ATP [adenosine triphosphate] pocket. And then the last kind is the development of alternative pathways, whether it's BRAF or MET. These are alternative signaling pathways that can emerge from mechanisms of resistance.

The solvent-front mutation is probably the most common. Obviously, we don't have a huge number of patients to make that determination at this point, but several years ago we presented some of the data on selitrectinib, which was on this molecule called LOXO-195, which was initially developed by Loxo Oncology and then transitioned to Bayer. And probably the most common subset of patients who became resistant to larotrectinib and entrectinib, the first generation of NTRK inhibitors, was patients with the solvent-front mutation. The second generation tends to work in that group and tends to have efficacy, albeit not as robust as we'd hoped for compared with the first generation. Selitrectinib and also repotrectinib, which is, I think, produced by Turning Point Therapeutics, both target solvent-front and gatekeeper mutations. Selitrectinib also targets xDFG. Both of these drugs tend to have response rates around 40% to 50%, and that may again just be reflective of the disease burden in patients. The data from selitrectinib would also suggest that there was a little bit more toxicity associated with the second generation: more dizziness and ataxia associated with that based upon the April 2020 AACR [American Association for Cancer Research] presentation we gave several years ago. So we'll see what happens. There are other companies that are trying to develop second-generation inhibitors. There's a company called Pyramid Biosciences that is also developing a more brain-penetrant agent but also a molecule that exquisitely targets these solvent-front, gatekeeper, and xDFG mutations. All full disclosure, I'm involved in that trial.

There are hopes that there will be drugs for these patients who become eventually resistant to larotrectinib. The good news is that one of the challenges of these trials is these patients on larotrectinib and entrectinib are on for so long. They're benefiting, like I said; median duration of response was like 45 months, so these patients have been on for so long, and they benefit for so long from the first-generation drugs.

: How effective are the next-generation TRK inhibitors at overcoming resistance to first-generation TRK inhibitors?

 The most recent data sets were from repotrectinib, which is a ROS1/TRK inhibitor. They had a very small, limited data set. I think there were 17 patients in that ACR presentation. The overall response rate was around 45%, 43%, if you combine the phase 1 and 2 results. We will see. I think we'll see larger data sets coming out in the next year or so, and we'll see how response rates stand up. That's compared with the first-generation molecules for which you had response rates in the 80% range.

: What challenges have emerged regarding next-generation TRK inhibitors? What areas of research are further investigating overcoming TRK inhibitor-associated resistance?

 I think the main challenge of doing these trials is identifying new patients. It's been years now, and we've identified 218 patients on the first trial of larotrectinib, and I think that number will start ticking up as NGS becomes more common. But again, with the second generation, because the patients who get the first generation are rare, but also, they do so well. Finding enough patients for the second generation has been always a little bit of a challenge.

I do think that, as most things with targeted therapy, these patients, even in the second generation, may eventually become resistant. So how do we think about the next generation, or how do we improve upon the first generation even more? These are questions that we need to think through as we identify more and more of these patients.

Articles

In this companion article, David S. Hong, MD, considers the safety and efficacy of the first-generation TRK inhibitor larotrectinib and discusses how acquired resistance may affect research into second-generation agents. 

EP. 4B: Understanding Treatment Resistance in First-Generation TRK Inhibitors for the Treatment of NTRK Fusion-Positive Cancers 

The ideal preferred method of NGS testing is NGS or next-generation sequencing, and preferably RNA-based [NGS]. I would argue that there is increasing evidence that NGS testing or molecular testing in general should occur as soon as the patient is diagnosed with metastatic cancer or stage IV cancer.

The latest dataset that we have was from ESMO this past year. We presented this data at ESMO looking at TRK fusion and larotrectinib data We had a total of 218 patients, and this was across [a] different number of prior lines of therapy. And again, this [study included] all different tumor types and across different ECOGs [performance statuses], and the overall response rate remained around 75%. Our initial data cutoff showed around 80% response rates. The median duration of response was close to 49 months, median progression free survival was 35 months, and median overall survival had not been yet reached. I would argue that there is no therapy out there, if you have metastatic cancer, that matches that. If you are able [and] are identified to have [an] 

These patients do well on this study. Overall from their previous state to actually going on larotrectinib—they have significant improvement in their overall quality of life. The overall adverse event profile of these drugs [is] very benign. Probably the most common side effect tends to be this slight dizziness that patients develop, and over time, they actually get better. There are some grade 3 events such as increase in LFTs (liver function tests), but usually these adverse events or side effects are manageable. You just hold the drug, you dose reduce them, and they still seem to respond to therapy.

There are really 4 different mechanisms of resistance. One is something called solvent-front mutations and gatekeeper mutations. Both of these mutations, differing on where they are, and the third called xDFG, really kind of affect that near-ATP pocket. And then the last kind is [the] development of alternative pathways, whether it's BRAF or MET. These are alternative signaling pathways that can emerge from mechanisms of resistance.

The solvent-front mutation is probably the most common. Obviously, we don't have a huge number of patients to make that determination at this point, but several years ago, we presented some of th[ese] data on selitrectinib, which was this molecule called LOXO-195. Probably the most common subset of patients who became resistant to larotrectinib and entrectinib, the first generation of NTRK inhibitors, was [those with] the solvent-front mutation. The second generations tends to work in that [group and] tend to have efficacy, albeit not as robust as we'd hoped for, compared to the first generation. Selitrectinib and also repotrectinib both target solvent-front and gatekeeper mutations. Selitrectinib also targets xDFG. Both of these drugs tend to have response rates around 40% to 50%, and that may again just be reflective of just the disease burden in patients. The data from selitrectinib would also suggest that there was a little bit more toxicity associated with the second generation: more dizziness [and] ataxia associated with that based upon the AACR presentation we gave several years ago.

There are hopes that there will be drugs for these patients who eventually become resistant to larotrectinib. The good news is that one of the challenges of these trials is these patients on larotrectinib and entrectinib are on for so long and they benefit for so long from the first-generation drugs.

The main challenge of doing these trials is identifying new patients. It's been years now, and we've identified 218 patients on the first trial of larotrectinib, and I think that number will start ticking up as NGS becomes more common. But again, with the second generation, because the patients who get the first generation are rare, but also they do so well, finding enough patients for the second generation has been always a little bit of a challenge.

In the third video interview of the series, David S. Hong, MD, from the University of Texas MD Anderson Cancer Center discusses the safety and efficacy of first-generation TRK inhibitors and reflects on challenges affecting next-generation TRK inhibitors including research into mechanisms of treatment resistance.

EP. 4A: First -generation TRK Inhibitors and Treatment Resistance in NTRK Fusion-positive Cancers 

The next steps in understanding and NTRK are the mechanism resistances. We already have a clear idea of what the mechanism resistances are with larotrectinib and entrectinib: Patients develop what are called solvent front and gatekeeper and what are called 

 mutations that can confer resistance to the first generation of drugs. There’s an ongoing study that several colleagues around the world and I are running, which is also developed by Loxo Oncology, Inc in collaboration with Bayer AG, the trial with the molecule selitrectinib is ongoing. We are seeing robust responses in patients who have developed resistance to the first-generation molecules.

In many ways, this disease is analogous in the context of solid tumors to another disease, such as CML [chronic myelogenous leukemia], which was the first disease that we were able to target with a targeted-therapy imatinib. For patients with chronic CML who get imatinib and the next generation of molecules down the line, their life span essentially is normalized. They have the normal life span of a normal person. We’re not there with NTRK fusion–positive cancer patients, but as we become more cognizant, more patients are screened molecularly, and we identify these patients earlier, the disease course and eventually the survival course of these patients who develop solid tumor cancers and who have NTRK fusions may be similar to what we are seeing now with CML.

Videos

David S. Hong, MD, of The University of Texas MD Anderson Cancer Center, shares his enthusiasm for ongoing research efforts to optimize treatment for patients with advanced solid tumors who harbor NTRK gene fusions.

TRK Inhibitors for Advanced Solid Tumors: Next Steps in Research

Let me comment on the presentation of our most recent data set that we presented, concerning larotrectinib in the treatment of advanced-NTRK fusion thyroid cancer, presented by my colleague at [The University of Texas] MD Anderson [Cancer Center], Dr Kevin Nead. Stepping back, 1 of the most common tumor types actually is thyroid cancer. We’ve identified NTRK fusion in not only the more common papillary phenotype but also in follicular. Close to 25% of these patients with NTRK-fusion thyroid cancers are very aggressive anaplastic phenotypes. In this data set that Dr Nead has presented, there were close to 28 patients who were analyzed from our data set, 68% being papillary, a small percentage of 7% being follicular, and close to 25% being anaplastic. The majority of these patients interestingly, will either harbor NTRK1 or NTRK3. NTRK3 fusions can often be missed unless you have an RNA-based NGS [next-generation sequencing]. The majority of these patients were iodine refractory or systemic therapy refractory and had greater than 2 to 3 lines of therapy.

The overall response rate was close to 90%. In the anaplastic, the response rate was close to 30% or 29%. One could argue that that tells you the different histologies and the different biology of the tumors. Similar to the primary CNS cohort, these patients with anaplastic are very aggressive, and a response rate of 30% is fairly remarkable because they oftentimes do not respond to standard chemotherapy or other therapies. The duration of response in this cohort was again not reached. With a median follow-up of 10.2 months, the median progression-free survival again was not reached, the median follow-up of 12.8 months, and the overall median survival was 27.8 months in this subset.

Similar to the other patients in the larotrectinib safety profile, the vast majority of the patients who had received the drug experienced grade 1 or 2 toxicities, the most common being fatigue, constipation, and some dizziness. The vast majority of these patients tolerated therapy very well. These patients with NTRK fusion and thyroid tend to be, in this context, younger patients who do not have other alterations, such as BRAF or RET alterations. If you find a patient who is younger, who does not have all other alterations, in particular in the context of anaplastic, it is worth exploring and seeing if we can identify NTRK-fusion patients. Overall, these patients—and I’ve treated a majority of these patients in this data set with larotrectinib—have remarkable responses and will stay on study for years. Several of my patients have remained on study for 5, 6 years. On this trial, 1 of my anaplastic patients, which is an aggressive tumor, was able to stay on the study for close to a year with good quality of life. If these patients with ETV6, NTRK3, or any other NTRK alteration is identified, the benefit from this therapy can be life-changing for them.

Important takeaways from recent safety and efficacy data presented on the use of larotrectinib as treatment for advanced differentiated and anaplastic thyroid cancers with NTRK gene fusions. 

TRK Inhibitors for Advanced Thyroid Cancers

Increasing evidence in the last several years shows that lung cancer in particular is a molecular disease. As we’re increasingly being able to parse out the details of lung cancer, whether it’s ALK, ROS1, EGFR, or NTRK, we’re seeing that defining these tumors is important in our patients with NTRK1 fusion. In our data set, we’ve identified close to 14 patients. The overall response rate is very similar to the overall data set that we’ve seen in the larotrectinib cohort, and that is 71%. Interestingly, the majority of these patients, close to 50%, will present with CNS [central nervous system] metastasis. For those patients who have CNS metastasis, their response rate with a brain metastases is close to 60% or 57%. Overall, the duration we’ve seen is approximately 12.9 months; follow-up has not yet been reached. The median follow-up with 14.6 months with median progression-free survival has not been reached. The median overall survival also has not been reached for the medium fallout of about 12.6 months; 91% of these patients were alive at 12 months. These were patients who were refractory to standard platinum and immunotherapy, many of whom were in the data set that my colleague Dr [Alexander] Drilon presented with greater than 3 lines of therapy.

Dr David S. Hong reacts to updated survival data associated with the use of larotrectinib as treatment for TRK fusion–positive non–small cell lung cancer.

Treatment Considerations for TRK Fusion-Positive NSCLC 

The most recent data set we’ve presented was at our ESMO [European Society for Medical Oncology Congress] presentation by Ray McDermott, but we had also presented this similar data set in the 

 paper. In the most recent updated data set, we had close to 175 patients on that trial. Similar to the first data set that we published in the 

, the overall response rate was incredibly similar. The overall response rate was close to 80% or 78%. The complete response rate was 33%. The progression-free survival was close to 36.8 months, the duration response had not been reached, and the median overall survival had not been reached. That confirms the incredible efficacy of particularly the larotrectinib molecule. For entrectinib, the most recent data set is not much different from the approval data set, so we don’t have a clear idea how to compare these at this time. But larotrectinib is showing really incredible efficacy that has been sustained from the beginning when we published the phase 1 data to close to 175 patients.

We know that larotrectinib and also entrectinib can cross the blood-brain barrier. The most recent data set that was presented was presented by my colleague, Dr [Alexander] Drilon from MSKCC [Memorial Sloan Kettering Cancer Center] at ASCO [American Society of Clinical Oncology Annual Meeting], about 2 years ago. In that data set, we had 24 patients with intracranial disease; 18 had primary CNS [central nervous system] tumors, and 6 had nonprimary CNS. The overall response rate in this data set was excellent in 5 of the valuable patients. For the solid tumors with brain metastases, we had a response rate of approximately 60%. In those patients who had primary CNS metastasis, the response rate was approximately 36%. That may just reflect the biology of fairly aggressive disease in the brain. Overall, if you ask a neuro-oncologist, they’ll tell you 36% response rate in patients with primary-refractory CNS tumors is pretty remarkable.

The significance of recent data exploring larotrectinib, a TRK inhibitor, in a subset of patients with TRK fusion–positive tumors with CNS (central nervous system) metastasis.