Welcome international healthcare professionals

Welcome, this website is intended for healthcare professionals with an interest in the ECC Webcasts. By clicking the button below you are declaring and confirming that you are a healthcare professional.

You are here

When to Start and Sequencing MKIs to Maximize Patient Benefit

Speaker: Marcia Simpson Brose, MD, PhD (USA)

SlotAJ-3 Brosse, M.S.

Director of the Thyroid Cancer Therapeutics Program
Department of Otorhinolaryngology: Head and Neck Surgery and
Department of Medicine, Division of Hematology/Oncology
Abramson Cancer Center at the University of Pennsylvania School of Medicine
Philadelphia, Pennsylvania, USA


I want to thank everybody and especially Bayer for inviting me to talk today.  I going to be addressing a little bit of what is happening with the multikinase inhibitors, mostly in the arena of differentiated thyroid cancer.  Of course we are far behind our colleagues in renal cell and in hepatocellular carcinoma, but many of the same questions are coming up for us and we are of course learning very much from those who have gone ahead of us in this area. 

Full transcription

Full transcription

I want to thank everybody and especially Bayer for inviting me to talk today.  I going to be addressing a little bit of what is happening with the multikinase inhibitors, mostly in the arena of differentiated thyroid cancer.  Of course we are far behind our colleagues in renal cell and in hepatocellular carcinoma, but many of the same questions are coming up for us and we are of course learning very much from those who have gone ahead of us in this area.

The goal of cancer treatment

Of course for us the goal of treatment is to improve overall survival, but it is also to decrease morbidity, because many times, while this can be a chronically advancing disease and it can be indolent, there is considerable morbidity for patients when they have advancing refractory differentiated thyroid cancer.

History of Clinical Development of Sorafenib

The development of Sorafenib I think you have heard, as far as what is going on in the area of hepatocellular carcinoma, here is what was going on at renal cell carcinoma.  A lot of the Phase 1s and the Phase 2s that were predating the renal cell carcinoma approvals were going on at Penn at the time and so in 2004 we started our first Phase 2 study for radioactive iodine refractory differentiated thyroid cancer.

Eventually that led to a Phase 3 trial called DECISION, which was completed and led to FDA approval of Sorafenib in this indication in 2013 and subsequent other approvals following that.

How Do We Make the Right Decision?

When we look at our patients, similarly to what has been seen in hepatocellular, we of course always want to look at what evidence we have from clinical trials.  We are really in our infancy, both in the use of kinase inhibitors, but also even in the oncologic specialty of treating differentiated thyroid cancer.  In many countries this cancer is still treated primarily by nuclear medicine doctors as well as endocrinologists. Education has to happen on how to best manage adverse events etcetera, as many of these physicians are using these agents for the very first time.

If we want to make the right decision we first of course have to understand the scientific evidence from the clinical trials. That means understanding both the treatment efficacy and safety that we have in the clinical trials as well as, in the case of GIDEON done before us, understanding better the prognostic factors from observational studies.  We can also get information from that to inform us on the selection of treatment as well as the timing.

In addition when a doctor is going to use these agents they have to then assess the patient condition, their personal experience with the drug will affect whether or not they will be using it. More and more we are hearing in other countries around the world that availability of treatment and reimbursement is still having a major impact on the selection of treatment for our patients.

MKIs for the Treatment of RAI-Refractory DTC

In radioactive iodine refractory differentiated thyroid cancer, for those people who don’t treat thyroid cancer, radioactive iodine is the standard of care; it is analogous to the TACE of hepatocellular carcinoma.  It is the proven effective therapy and what was very clear initially, especially if we are going to look for additional therapies, is we wanted to make sure that this therapy was no longer effective.  This is radioactive refractory differentiated thyroid cancer; we will talk a little bit about defining when radioactive iodine should be stopped because, like TACE, there are still some questions about that in different treatment settings.

I have already mentioned that sorafenib was approved in 2013 and overlapping the Phase 3 for sorafenib was the SELECT study for lenvatinib, which then was announced in 2014 and ultimately approved this year.  Now we have two Phase 3 and approved drugs, sorafenib and lenvatinib for the treatment of radioactive iodine refractory differentiated thyroid cancer.

A third Phase 3 trial for vandetanib is currently underway. There was a randomised Phase 2 and now a Phase 3 that is attempting to prove efficacy.  We expect to hear the results of this perhaps at ASCO or at least in 2016.

Then additional Phase 2, cabozantinib in the first-line setting is also underway, a phase 2 at our institution.  There are also several other agents that have been shown to have activity – all of them VEGF receptor inhibitors for the most part; there is one exception. Including pazopanib, axitinib, sunitinib, so many of the kinase inhibitors that have shown efficacy in renal cell carcinoma have also shown activity.  This comes to be important when our patients do progress through the approved agent, sorafenib and lenvatinib, where can we go next? 

I also point out that BRAF V-600E mutations like melanoma play a very important role, because in papillary thyroid cancer, which is the majority of these differentiated thyroid cancers, BRAF V-600E mutations are present in up to 50-60% of patients. We have shown at our institution in a Phase 2 that vemurafenib, the V-600E inhibitor is also active, giving a good PFS in a Phase 2 study.

DECISION Trial: Sorafenib versus Placebo in Locally Advanced/Metastatic RAI-Refractory DTC

When we started we had to define who is an eligible patient. We spent about six months talking to endocrinologists and nuclear medicine physicians around the world, because we wanted to find criteria that best identified patients who really no longer got benefit from radioactive iodine and needed additional therapy.

There are two basic categories of eligibility we thought were important.  The first is of course that radioactive iodine was no longer effective. This actually can be easily shown if a radioactive iodine uptake scan or a therapeutic scan is performed and there is no uptake in the metastatic lesions.  Then, right away, we know that these patients are not candidate for therapy.

What is less understood is when people have uptake, when do we stop using radioactive iodine?  One of the other criteria that we developed was to say that patients who took up radioactive iodine following a treatment dose, who then went ahead and progressed in the next year fairly quickly following radioactive iodine, showed also that this therapy was no longer clinically effective.

Then the last was of course some people have had multiple treatments over the years, were starting to get higher levels and we know there are no complete responses for patients over 600 millicuries. There was data to suggest that, so that also could be used as a guideline for patients that we would take.

Then you will also talk to endocrinologists who will point out that some patients can have an indolent course and even following the lack of uptake of radioactive iodine, may not have progression for quite some time, sometimes months to years. For that reason a second criteria which has to do with progressive disease within 14 months and we use RECIST progression as our guideline for this, was also an eligibility criteria.

We wanted to make sure we didn’t end up with patients on our trial that weren’t going to progress for two years anyway and so that was important to identify patients who had clearly progressing disease and therefore would justify the use of a kinase inhibitor. 

This was the first trial so there were no other anti-cancer treatments allowed in this study. 

Randomisation was then one-to-one between sorafenib and placebo and then because of the wide availability of sorafenib, we felt that going for an overall survival endpoint was impractical because many patients would get sorafenib off study.  The primary endpoint chosen was progression-free survival.  Then at the time of progression we did allow for patients to be un-blinded and then, based on the investigator’s discretion, patients could be crossed over to sorafenib.  At this time over 71% of the initial and over 75% now of patients have crossed over and received sorafenib, even from the placebo arm which will definitely compound our ability to understand overall survival and other secondary endpoints. 

However, patients were going to be stratified by age; 60 to 65 is the middle of the age group of patients with RAI refractory cancer and stratification was prospective as well as by geographic region.

Sorafenib Treatment Was Associated WithSignificantly Longer PFS Versus Placebo

This was the primary endpoint for this study.  We see first of all in the placebo arm the patients progressed at 5.8 months, showing that we did identify effectively patients with advancing disease who we felt were definitely good candidates for systemic therapy.  While that was reassuring, the even better result of course was that sorafenib improved the progression-free survival from 5.8 to 10.8 months with a hazard ratio of 0.59 and a p-value of less than 0.0001.

Sorafenib Demonstrated a Consistent Trend for Greater FPS Across All Predefined Subgroups

We also have prospectively defined additional subgroups that we wanted to understand and, as this plot shows, all of the subgroups that we looked at had a benefit with a hazard ratio of less than one, regardless of the region, the age group, or the initial histology.  I will point out that this was done by central review.   Lung metastases, bone metastases, it didn’t matter whether they had this; everybody did better.  There are a few groups that crossed the mid-line cross – the confidence interval crosses one for the hazard ratio; in most of the cases that is when the subgroups were small in number, but it did not matter how many target lesions, whether they have lesions that were smaller or bigger than the median, it still seemed that all patients were benefitting.  There was no evidence from the original Phase 3 that any subgroup should be included or excluded from treatment with sorafenib; all of them seemed to benefit.

Sorafenib Subgroup Analysis in DTC

There was this consistent trend of the prolongation of the progression-free survival in the subgroups, but then we also wanted to understand a little bit more. Over the time of course of this study we also felt that some patients who their maximal lesion was only what was less than 1.5cm or 1.5cm and below, some of those patients sometimes could be observed a little bit longer prior to starting therapy.  We were interested in this outcome and unfortunately I apologise for not having this slide here today, but we did divide up in a post-hoc analysis the patients who had their largest lesion is 1.5 or less, versus greater and we did see that the patients whose disease was less than 1.5 did not seem to get a benefit from sorafenib, while the patients who had their largest lesion greater than 1.5 did. 

This is one of the most important results of the ad hoc analysis because this is a very good and useful guideline for us to use to translate over to the clinic.  However, I will point out there are exceptions to this because thyroid cancer can spread in a miliary pattern – not very often, but when it does you can have a patient with a very high tumour burden but no lesions that are greater than a centimetre. There will be exceptions to when this new guideline might be imposed and one is when there are multiple lesions or also when the lesions are in a particular dangerous spot, perhaps obstructing airways.  Last we noticed that patients who have pleural based disease tend to do worse as well.

Now we might start thinking about using the post-hoc analysis as a guideline for 1.5cm being what we would look for before we would start sorafenib.

Papillary histology: these patients did particularly well and lung-only metastases was also a group of patients that did particularly well on sorafenib.

Sorafenib Is Well-tolerated, With a Predictable and Manageable Safety Profile

When we look at the adverse events they were really what we expected, but we have to remember that the patients who were getting treated with sorafenib are on study for on average a year and many patients for many more than that – perhaps two or three years.

Comparing just the time of exposure from between sorafenib and placebo this group is exposed double what the placebo group is exposed. When we look at this what we see is that of course hand-food skin reaction is very common, as is diarrhoea, rash and hair loss.  There is also fatigue and weight loss in 50% of the patients, but reassuringly most of these are at fairly low rates when we look at the phase 3 for toxicity.

The one exception to this was the 20% of patients who had hand-foot skin reaction of grade 3 or 4. We felt that most of this is related to a learning curve because over 90% of the physicians had never given sorafenib before and therefore it was not uncommon for them to unfortunately, get to a grade 3 or 4 before they realised what was happening.

Most of us who are giving this agent now regularly would say that it is rare to get over 5% of grade 3 or 4 hand-foot skin reaction.

Diarrhoea is also fairly manageable, as well as the rash.  Fatigue and weight loss we also manage this. We encourage our patients to do weight lifting exercises as well as nutrition counselling.

There is a significant rate of grade 3 or 4 hypertension of almost 10% and for that reason patients are followed at the two week mark, to make sure that their blood pressure is well controlled.  Of course we want to get it into control before we start sorafenib, but because of this we do monitor these patients and make sure that we institute new therapy if needed.

Serious AEs and Deaths: Double-blind Period

When we look at the serious adverse events and deaths the most common serious adverse event is secondary malignancy with the known side effect of squamous carcinoma of the skin.  That basically makes up for the entire difference between the sorafenib and the placebo arm as far as second malignancies.  There were slightly more pleural infusions but again, if you take into account the timeframe I don’t think that this is a significant difference between the two arms. 

Grade 5 events – death are at 12 in the sorafenib arm and six in the placebo arm again in a longer time period.  Only one in both arms was thought to be attributed to the medication, in this case a patient who was on anticoagulation in the placebo arm fell and had a subdural haematoma and ultimately was put on hospice. While a patient on the sorafenib arm who had been on therapy for three years had a sudden death; no autopsy was performed, but he had had complaints of substernal chest pain for several days before, so it is presumed an MI death.  It is important to remember that and make sure we follow these patients and counsel them that chest pain is something to be pursued, of course, if they have it and not ignored just as if they had never been on any kind of kinase inhibitor therapy.

SELECT:  A Phase 3 Trial of Lenvatinib in RAI-Refractory DTC

The second trial to be reported, and this was reported a little over a year ago, was the SELECT trial which was a Phase 3 trial of lenvatinib in RAI-refractory differentiated thyroid cancer.  The definition of RAI-refractory disease is very similar to the DECISION trial, however, the one difference between the two studies is that the SELECT trial required a central review for progressive disease and for this reason added an extra rigour to the patients being evaluated for clear progression in the 13 months prior to treatment.

Because this came along second and many patients had been treated already, either on a Phase 2 for the other VEGF receptor inhibitors or on the sorafenib study, patients were allowed to have one prior VEGFR therapy.  Then they were randomised to lenvatinib two to one versus placebo and then they were followed by a primary endpoint, also progression-free survival and overall survival and safety also was secondary endpoints for this study.

Patients were then stratified by age and this time cut-off was 65 years as well as prior VEGF receptor therapy.

Progression-Free Survival

What we see first of all if we look at the placebo arm is that the patients here have more aggressive disease, progressing at 3.6 months which is a fairly short interval.  When we compare this to lenvatinib, lenvatinib prolonged the progression-free survival to 18.3 months with a hazard ratio of 0.21 in this population and a p-value of less than 0.0001.

PFS by Previous VEGF-Targeted Therapy

Interestingly the curves start to diverge shortly after the two month mark, so it is fairly early separation and even more interesting to us was the fact that regardless the top is first-line therapy with lenvatinib and the second is in the second-line therapy. In both cases the hazard ratio is 0.2 and 0.22 showing that there is a benefit in both settings for this agent, which was very reassuring, since now that lenvatinib is available many patients have already been exposed to sorafenib.  Of course, because of this curve it is also clearly very effective in the first-line setting as well.

Most Frequent AEs (>30%)

When we look at the adverse events we do have to remember that the adverse events in this study shows us that this agent is very different from sorafenib in its types of adverse events.  As you will notice, the top adverse events, hand-foot skin reaction isn’t even seen, however, hypertension is the most common with 68% and a little bit more concerning and we need to keep our eye on this is 42% of those were a grade 3 or above.  For this these patients do need to be very closely monitored for hypertension within the first week of starting therapy and this needs to be addressed very quickly.

There is, of course, diarrhoea and there is also a significant amount of fatigue and decreased appetite, nausea, vomiting and weight loss.

When we look at the grade 3s we see again that there is up to 10% fatigue and decreased weight, so these things do require active intervention again with nutritional support as well as exercise plans similar to what we were doing when we used sorafenib.

One other area that is different is that there is significant proteinuria in these patients and this also needs to be monitored when patients are on this agent.

Please get ready to vote

Now let’s get ready to vote.

Discussion (ARS)

The first question is what would be the best approach to hand-foot skin reaction management during treatment for those of you who are giving it?  Do you treat it when it is early/symptomatic, do you do a dose modification starting at grade 1, a dose reduction for grade 3 or both 1 and 3?  Go ahead and vote.


Many of you have clearly used sorafenib before; clearly number 1 and number 3 are approaches that we use regularly.  Most of the time for grade 1 we can usually treat that supportively with either creams; I use ibuprofen very effectively – it decreases the pain significantly in these patients.

Key Questions

I will also say that most of the time for grade 1 we do not need to decrease the dose, but we can usually treat it systemically.

A Post-Hoc Subgroup Analysis by Thyroid Carcinoma Symptoms at Baseline

When we go and look at more data that came from the DECISION trial, we also were faced with a lot of the people in the different countries saying “Given that it is an indolent disease we should wait until patients are symptomatic”. That is partly because, as I said, these patients have not been treated in an oncologic background where most of us who are oncologists would not necessarily wait for symptoms before we would treat patients with, say, renal cancer, lung cancer or other cancers.

This was still a very active question and still is a very active question in the thyroid community.

What we saw is when patients were symptomatic at baseline versus asymptomatic, at baseline both groups appeared to benefit from sorafenib showing that we really do not need to wait for symptomatic therapy to get a benefit in the overall progression-free survival for these patients.  I will also say anecdotally that it is not uncommon that if a patient comes into me already symptomatic, even with response, those symptoms may never completely resolve and for that reason we do not recommend waiting for symptoms before starting treatment with sorafenib.


When we look at RIFTOS, RIFTOS is a new trial that is going to be an observational study similar to GIDEON, to try to answer additional questions on what is the real-world use of sorafenib and now also lenvatinib that it is FDA approved?  When are patients getting started?  The primary question is, is there a benefit to starting immediately versus starting later, also similar to some of the observational studies used before?  Hopefully we are going to understand a little bit, how long does it take for symptoms to develop and whether or not sorafenib or lenvatinib starting early can change the time to overall symptomatic progression because that is, of course, felt to be a very important clinical endpoint.  We will also look at overall survival and some additional endpoints. 

This is currently enrolling in several countries and the cohort one will be the patients who are defined as asymptomatic and RAI refractory at the time of entry, but are started immediately on sorafenib versus cohort two, the patients who go under observation for a period first.

SWITCH: Phase 3 Open-Label Study – An Example of Sequencing Strategy Investigation

The other question we now are faced with, and of course this is a very dated slide, but just to remind you of the early days in renal cell carcinoma when sunitinib and sorafenib were both available and people wanted to understand which sequence was better.  At the end of the day it is not a sprint, it is a marathon and our patients will often get both agents that are available and in this case at this time it is sorafenib and lenvatinib, but we want to understand better which order is important.  When a randomised phase 3 was performed, when sorafenib was given versus sunitinib versus sunitinib and sorafenib the primary endpoint, total PFS, did not show any significant difference.  Now we don’t know if that is going to be true for the thyroid cancer world and sorafenib and lenvatinib are two very different drugs.  We are interested in this question, however, whether or not this is going to be a feasible study in the thyroid area, especially because RAI refractory differentiated thyroid cancer is so rare, it is yet to be seen?

Again this is still a question that we have with both drugs being approved in the first-line setting and we have some data showing that lenvatinib is active in the second-line setting; we don’t have yet data about whether sorafenib is also active in the second-line setting.

Please get ready to vote

Now we have another question. 


Which of the following parameters play a role in deciding treatment choice and sequence: tumour size and location? Symptoms?  Pre-existing comorbidities? Tumour growth rate? Tolerance and all of the above?


Yes.  Everybody here is a practical clinical oncologist; of course all of the above are going to weigh into your decisions about what actual medication to use and what the treatment sequencing will be.  I have pointed out some differences as we said between the way the adverse event profiles are, there are definitely patients that you might avoid one agent and pick the other. I think all of the information that we know from the trials, as well as tolerability will go into of course how things are sequenced in patients.           

Everolimus Plus Sorafenib for Patients With DTC Who Progress on Sorafenib Alone

I also want to bring up a new approach to sequencing, which is a phase 2 that we have performed at our institution and reported at ASCO this year, which is that we did have some tissue data that suggested that the nodules that were progressing on sorafenib showed increased activity of the PI3 kinase pathway.  We also noted that if we switched an earlier patient to an everolimus alone, which was an mTOR inhibitor, that while the new lesions that were progressing stopped growing, the old lesions that had been wonderfully under control under sorafenib started growing we felt it was a rational for trying to continue sorafenib at the time of progression and add in everolimus as an approach to see what happened.

We did an inter-patient dose escalation which allowed us to get to the maximum tolerated dose on a person-by-person basis and also looked at our primary endpoints.

Efficacy and Safety: Everolimus Plus Sorafenib for Patients With DTC Who Progress on Sorafenib Alone

This is the result of the trial which showed that the progression-free survival of patients when they went on this study was an additional 13.9 months from adding in everolimus at the time of progression.  What we see is that while there are still quite a large percentage of grade 1, not unlike what we see in the DECISION trial, there are markedly few grade 3/4 adverse events, showing that this approach can be quite tolerable and does allow for perhaps up to an extra year to 14 months of progression-free survival in this patient population.

This approach is of great interest to us and we are currently in the process of doing a companion study of adding everolimus into patients who progress on lenvatinib at the time of progression. This is to see whether the same approach would also allow us to extend the efficacy of that drug as well.

Starting and Sequencing MKIs to Maximise Patient Benefit

When we talk about starting and sequencing MKIs to maximise patient benefit, of course we want to prolong survival while maintaining quality of life and we are going to use data from clinical trials as well as experience in managing AEs. It is very important that doctors are aware of all of the different approaches and the best ways to minimise adverse events for our patients, because if we start decreasing doses for adverse events just because the doctor doesn’t know how to manage them well, we will not get as good results as we did in the studies.

It is important to monitor patients; clear evidence of progressing disease.  It might be time to initiate treatment, but also there may be additional effects if patients have symptoms at baseline, but more importantly we also eventually understand what is the optimal treatment sequence – at this point we don’t have it; we definitely have some options.

The one other treatment I didn’t mention was the fact that the Phase 2 of the BRAF inhibitor for the BRAF V-600E containing lesions also resulted in over three months’ progression-free survival.  Again, another agent that can be active in this setting for patients with that molecular characteristic. 

We need to continue to do work.  We need to understand better on whether or not there is some sequencing choices, but if we can’t do that we need to at least continue to develop the other agents like renal cell.  We are now in the throes of so many new therapies, even announced at this meeting this week; it is a very exciting time for some of these tumours that previously did not have good therapies.

With that I will thank you for your attention.