Phase III – ADAURA

Trial Data - Top-Line Results

Change to LOA: 0%

Abstract LBA5: Osimertinib as adjuvant therapy in patients (pts) with stage IB–IIIA EGFR mutation positive (EGFRm) NSCLC after complete tumor resection: ADAURA

Context

In April 2020, an Independent Data Monitoring Committee recommended for the Phase III ADAURA trial to be unblinded two years early based on its determination of overwhelming efficacy. At the time of data cut-off, overall survival (OS) data favoured Tagrisso, but were not mature. The trial will continue to assess OS as a secondary endpoint.

Tagrisso is approved for the 1st-line treatment of patients with locally advanced or metastatic EGFRm NSCLC in the US, Japan, China, the EU and many other countries around the world.

Design

Per the abstract, eligible pts: ≥18 years (Japan/Taiwan: ≥20), WHO PS 0/1, primary non-squamous stage

       IB/II/IIIA NSCLC, confirmed EGFRm (ex19del/L858R), complete resection of primary NSCLC with full      

recovery from surgery; postoperative chemotherapy was allowed. Pts were randomized 1:1 to osimertinib 80 mg once daily orally or PBO to receive treatment for up to 3 years and stratified by stage (IB/II/IIIA), mutation type (ex19del/L858R), and race (Asian/non-Asian).

Endpoints

The primary endpoint is disease-free survival (DFS) in patients with Stage II and IIIA disease. Secondary endpoints include OS and DFS results in the overall trial population.

Results

Detailed results from the Phase III ADAURA trial showed AstraZeneca’s Tagrisso (osimertinib) demonstrated a statistically significant and clinically meaningful improvement in DFS in the adjuvant treatment of patients with early-stage (IB, II and IIIA) EGFRm NSCLC after complete tumour resection with curative intent.

In the primary endpoint of DFS in patients with Stage II and IIIA disease, adjuvant treatment (after surgery) with Tagrisso reduced the risk of disease recurrence or death by 83% (based on a hazard ratio [HR] of 0.17; 95% confidence interval [CI] 0.12, 0.23; p<0.0001). DFS results in the overall trial population, Stage IB through IIIA, a key secondary endpoint, demonstrated a reduction in the risk of disease recurrence or death of 79% (based on a HR of 0.21; 95% CI 0.16, 0.28; p<0.0001).

At two years, 89% of patients in the trial treated with Tagrisso remained alive and disease free versus 53% on placebo. Consistent DFS results were seen across all subgroups, including patients who were treated with surgery followed by chemotherapy and those who received surgery only, as well as in Asian and non-Asian patients.

Most Common Adverse Events

The safety and tolerability of Tagrisso in this trial was consistent with previous trials in the metastatic setting. Adverse events at Grade 3 or higher from all causes occurred in 10% of patients in the Tagrisso arm versus 3% in the placebo arm as assessed by the investigator.

Conclusion

Per the abstract, adjuvant osimertinib is the 1st targeted agent in a global trial to show a statistically significant and clinically meaningful improvement in DFS in pts with stage IB/II/IIIA EGFRm NSCLC after complete tumor resection and adjuvant chemotherapy, when indicated.

Comment

These first numerical results from the Phase III ADAURA study are very promising. Tagrisso demonstrated a 83% reduction in the risk of disease recurrence or death compared with placebo in patients with Stage II and IIIA EGFR-mutated NSCLC. Additionally, with a two-year disease-free survival (DFS) rate of 90% and

a three-year DFS rate of 80%, Tagrisso showed excellent preliminary durability of response.

These results will likely be practice changing, but the OS data is still immature and it remains to be seen if the DFS data alone will be sufficient for regulatory approval. Tagrisso is already a best-selling EGFR inhibitor and, assuming these data are sufficient for a label expansion into the lucrative adjuvant setting, this success will further solidify Tagrisso’s position in the market.

Of note is that the DFS results were consistent across all subgroups, including patients who were treated with surgery alone and patients who were treated with surgery followed by adjuvant chemotherapy.

However, based on the data presented for the secondary endpoint of DFS in patients with Stage IB-IIIA disease, it appears that patients with Stage Ib disease may not have benefited from treatment with Tagrisso as much as the later-stage patients did. It will be interesting to see if AstraZeneca includes a subgroup analysis in the ASCO presentation, particularly for patients with Stage Ib disease. We will provide an update after the full presentation during the plenary session on Sunday, May 31st.

Comment (Updated 05/31/2020)

Detailed results, including a breakdown of DFS by stage, were presented at ASCO during the plenary session on Sunday. As expected, Tagrisso was most effective at preventing disease recurrence or death in patients with Stage II and Stage III disease, although a significant benefit was still seen in patients with Stage IB disease (HR=0.50). Additionally, a consistent improvement in DFS was seen across all stages of disease regardless of whether a patient received prior adjuvant chemotherapy.

Although the early OS results looked intriguing, the data are only 5% mature. Given the early unblinding of the study, the discussant expressed concerns about the future overall survival data showing the true OS advantage of Tagrisso. Dr. Spigel also mentioned that further research is needed to determine if Tagrisso eliminates residual disease or if it merely delays the growth of disease.

Although these impressive data support the use of Tagrisso as an adjuvant treatment, a number of questions remain, particularly in regard to a regulatory path forward. AstraZeneca is discussing these data with regulatory authorities in order to determine if the DFS data are sufficient to support filings.

A false dawn in adjuvant EGFR mutant lung cancer? Dr Ross Camidge provides a critical review of ADAURA

Not in Chicago: A hallmark of the annual meeting of the American Society of Clinical Oncology (ASCO) is “practice changing”

clinical trial data often featured in the plenary session.

This year one of the noteworthy phase 3 trials presented at the meeting (link to ASCO20 Abstract LBA5), was the AstraZeneca sponsored “ADAURA” trial for osimertinib as adjuvant therapy in patients with stage 1B-IIIA EGFR mutation-positive NSCLC after complete tumor resection.

We’ve been following the clinical development of osimertinib since the initial presentation of the phase 1 data in 2013 (link).

At first glance it’s hard not to be wowed by the separation of the disease-free survival (DFS) curves in ADAURA, which show a benefit for patients who received the EGFR inhibitor osimertinib compared to those who received placebo. A 0.17 hazard ratio is certainly not something we see every day.

Indeed, if you were in the media and listened to Dr Herbst on the #ASCO20 press briefing last week – to use a “Britishism” – you would have thought this trial was “the best thing since sliced bread.” The data monitoring committee recommended unblinding the study early.

Anyone leaving the story there and doing a superficial report about this data is, however, doing a disservice to their readers. The US academic lung cancer community are not all singing Handel’s Hallelujah chorus for the ADAURA trial and in this post, we take a critical look at why this might be the case.

For good measure, we interviewed a global thought leader who was prepared to offer some candid expert commentary.

Dr Ross Camidge is Professor of Medicine/Oncology and holds the Joyce Zeff Chair in Lung Cancer Research at the University of Colorado school of medicine. He kindly spoke to BSB and shared his perspective on adjuvant therapy in EGFR mutant lung cancer.

Dr Camidge characterized the disease-free survival in the ADAURA trial as a potential “false dawn” and told BSB:

“I do not believe the data should be practice changing or at least not yet. I think when you show there is an overall survival benefit then it will be practice changing…

So far there is no reason to suggest that disease free survival is going to translate into an overall survival advantage as it has not in any other comparable targeted therapy trial in EGFR mutant lung cancer. If this trial is the exception though, it will certainly not be of the same magnitude as the DFS benefit. However, the real unanswered questions are who needs this drug in this setting and if they need it, who can stop it safely and when.”

To learn more about the ADAURA trial presented at ASCO20 and read our latest expert interview, subscribers can log-in or you can click to gain access to BSB Premium Content.

“The dictum that you cannot have your cake and eat it, too, has always struck me as sheer defeatism.”

~ Victor Mollo

Introduction

Dr Camidge would have made a great discussant of the ADAURA data at ASCO20, not only does he possess considerable translational research experience with targeted therapies and EGFR inhibitors, but his analysis and commentary are fair and objective, as well as being based on the underlying science and clinical data.

Indeed, it was his discussion of the first phase 1 data for osimertinib (then known as AZD9291) at the European Cancer Congress in Amsterdam back in 2013 (See post: ECCO 2013: AZD9291 shows early promise in T790M NSCLC) that inspired me to follow the development of osimertinib over several years.

I still vividly recall him showing of a picture of the foot print of an astronaut on the moon while concluding:

“By addressing acquired resistance at the molecular level potentially creating one small step in the EGFR treatment paradigm, third-generation inhibitors like 9291 are likely to represent one giant leap forward in the treatment of EGFR mutant disease.”

His discussion inspired a whole series of BSB posts on the AstraZeneca vs. Clovis race to bring a third-generation EGFR inhibitor to market.

As a result, we’ve written extensively about osimertinib since 2013, and I encourage you to check out some of the prior posts (too numerous to mention) if you’re interested in revisiting this story.

Back to ASCO20 and a masterclass from Dr Camidge in how to review an area where there is a lot of nuance and detail, which are all too easy to overlook. We’ll use his interview as a roadmap for this post, so grab a cup of coffee and follow along…

Background – Prior studies with adjuvant TKIs in EGFR mutant lung cancer

Dr Camidge started off by highlighting several studies where adjuvant use of an EGFR inhibitor has led to a disease-free survival (DFS) advantage, in other words ADAURA while it may be a global study, is not breaking new ground, nor as we’ll see are the DFS results seen are far off what has been seen with other EGFR inhibitors in the adjuvant setting.

Dr Camidge: “So there have been in the past at least three other adjuvant EGFR TKI studies conducted in populations that include EGFR mutant non-small cell lung cancer.

One of them was single arm, that was called SELECT and that was 2 years of erlotinib. (link to paper in JCO 2017). You got erlotinib for 2 years after completion of adjuvant standard chemotherapy if you had a stage 1A to 3A non small cell lung cancer. It was 100 patients, and the key thing here is at the end of 2 year, when you’re still on the drug (it’s a single arm study), the disease-free survival was 88%. Almost exactly the same as in the ADAURA trial.

If you looked at the [DFS] curve, when you stop the drug, it then starts to deteriorate and you then end up with a certain number of people who are cured, and it takes on average 24 months for the disease to recur once you stop the TKI

That single arm study was consistent with what we’ve seen over the past 10 years, which is that these tyrosine kinase inhibitors can depopulate the cancer but they do not kill 100% of the cells that are technically sensitive to the drug.

And we’ve know that for a very long time with a very elegant study from Memorial Sloan Kettering that Greg Riely was the first author of back in 2007 (link), where you had patients who were progressing on erlotinib or gefitinib, they had a CT scan when they were progressing, they stopped the drug and they had another CT scan 2 weeks later where it was still progressing and then they went back on the same drug at the same dose, had another CT scan 2 weeks later, and what you saw was you had partial shrinkage.

Now how do you interpret that?

What it means is that when you were initially progressing on the drug you have some acquired resistance clones growing but some clones are still alive, but still sensitive and just suppressed and so when you stop the drug, now you have two sets of clones growing, but when you go back on the drug you re-suppress the ones that were still sensitive to the drug.

And the reason that matters is the impression then was that this was not a curative modality, it was a suppressive modality, and the SELECT study in the adjuvant setting gave the impression again that whilst you were on the drug, you could suppress people who had not been cured by the surgery and chemotherapy, but once you stop that TKI if the disease was still there, it would come back.

Consistent with that we then moved to randomized studies, so there was a subgroup analysis of RADIANT.

RADIANT was the study of two years of adjuvant erlotinib in a non-mutation selected population, but they were able to do a subgroup analysis (link).

In the EGFR mutant group, although, we’re down to relatively small numbers, the 2 year disease free survival for those who got erlotinib with a proven EGFR mutation was 75%. Certainly not that different from ADAURA again. And again, following the cessation of the erlotinib, the DFS curves come right back together again.

Then we get to the CTONG1104/ADJUVANT study, which was a Chinese study in resected lung cancer, prospectively selected for EGFR mutations, with at least stage II to IIIa disease, with patients randomized to 2 years of gefitinib or adjuvant chemotherapy. The experimental arm did not get adjuvant chemotherapy, and in that group the 2 year DFS if you look at the figure was in the high 60% range.”

The results for CTONG11O4 “CTONG” were published by Prof Wen-Zhao Zhong Guangdong Lung Cancer Institute, Guangdong Lung Cancer Institute and colleagues in a Lancet Oncology 2018 paper (link).

There is also an accompanying editorial by Drs Ng and Camidge (link), which is well worth reading. They pointed out the DFS benefits don’t appear to continue once you cease tyrosine kinase inhibitor (TKI) therapy:

“After 24 months, the Kaplan-Meier curves for disease- free survival began to converge, meeting by 36 months, with no apparent tail of non-recurrent patients in either treatment group by 48 months.”

ADAURA 2-year DFS is comparable to other studies

We’ve just heard that many other adjuvant studies of TKI’s in EGFR mutant lung cancer had a high DFS rate, so how does the

ADAURA study results reported at ASCO20 stack up compare to them? It turns out they are comparable.

Back to Dr Camidge:

“The 2-year disease free survival in ADAURA was 89%, and bear in mind in SELECT it was 88%, in RADIANT 75%, and

CTONG was in the high 60s.

The concept of being on a targeted therapy when you have an EGFR mutation and having a disease-free survival rate of 89% is not unprecedented.

The issue is all of those studies based on longer follow up after TKI cessation seem to imply that when you stop the TKI the disease-free survival curves start to fall, and in randomized studies versus placebo, come back together.”

Disease Free Survival is not a Surrogate for Overall Survival

The gold standard in oncology, especially in the potentially curative setting of adjuvant therapy, remains overall survival (OS); the top-of- mind question every cancer patient will ask of their oncologist is “will I live longer if I take this drug?”

In their 2018 Lancet Oncology editorial, Ng and Camidge cogently make the point:

“Positive data for overall survival rather than just disease-free survival would strengthen the case for adjuvant tyrosine kinase inhibitor treatment, but overall survival data are not mature for the ADJUVANT study and are unreported for RADIANT. Notably, overall survival is the primary endpoint of the ongoing ALCHEMIST study (NCT02194738) of adjuvant EGFR and ALK tyrosine kinase inhibitors in EGFR-mutant and ALK-positive disease.”

In case you missed it, at ASCO20 this weekend (link to abstract 9005) there was updated data for CTONG1104/ADJUVANT that showed the final overall survival analysis for this randomized phase 3 trial.

In their abstract conclusion, the researchers rather stridently offered a warning to future trial readouts:

“  “The DFS survival advantage did not translate to OS difference in ADJUVANT trial.”

This is a really important statement and raises the question of whether it will also be the case with other adjuvant trials and if we don’t see OS, then how can we justify these expensive treatments?

In many European countries and those with public health systems, obtaining a survival benefit is required for reimbursement. Let’s not forget even in 2020 there are some countries around the world where erlotinib, gefitinib, nor osimertinib are approved for EGFR+ metastatic disease, never mind asking them to pay for several years in early stage disease.

We must inject the key caveat here that CTONG1104 had a relatively small patient sample size, but it is a randomized phase 3 trial and does raise the question for consideration, even if you can’t translate the findings from this to other trials with different EGFR inhibitors.

We also saw at this year’s ASCO that – as was reported in 2018 – the 3 year and 5-year DFS curves continued to converge

over time, as Dr Camidge noted earlier.

Given we’ve seen this happen with several TKI inhibitors, it raises the question of whether this will happen over time with the ADAURA trial too.

I encourage you to review the CTONG1104 presentation at this year’s ASCO if you have access, there’s also a post-hoc analysis included, which looks at subsequent treatment on disease progression. Dr Azzoli (Brown) provided some informative commentary on this trial as the ASCO20 discussant.

Dr Camidge takes our story forward:

“The hypothesis is disease free survival equals overall survival and at this year’s ASCO, the CTONG study was able to show more prolonged follow-up and there was zero difference in overall survival, with a hazard ratio of 0.98. So it’s not like it was a non-statistically trend towards overall survival, there was no trend towards overall survival.

So the two year disease free survival rate in ADAURA is really not that different in what we’ve seen in other studies.

And then you say well the hazard ratio looks a little better and really we’re comparing it to the hazard ratio to the subgroup analysis in RADIANT, which was something like 0.6, the hazard ratio in CTONG was about 0.6, yet this is 0.17, why is that so much better?

Well, sometimes you have to say who exactly went in and how well staged they were? I note there was a suggestion that these patients had to have a complete resection but I don’t know if they drilled down into saying how much the lymph nodes had to be actually sampled, was there a lymph node dissection mandated, I couldn’t see that.

The other thing is you are getting 3 years of TKI, rather than 2 years [in CTONG] and osimertinib we know also has greater activity in the brain than some of these other first-generation drugs, and we know from the analysis in CTONG that when people were relapsing, they were often relapsing in the brain.

ADAURA has not come out of the blue, ADAURA is built on the back of these studies and these other studies have suggested that a disease free survival does not automatically translate into an overall survival advantage.

The real question here is if you are just kicking the can down the road with adjuvant osimertinib, and then you stop it at 3 years and everything starts over again. The only way that’s going to translate into an overall survival advantage is if the group who are in the placebo arm would be expected to die during the period of time that the other group is on osimertinib and they are not,  and given that that’s a 3 year period, and we saw from SELECT there was about 24 months after you stopped the TKI when things are beginning to recur, the question is if you are delaying things by about 5 years, is that going to change overall   survival?

One of the issues is EGFR mutant patients tend to live a long period of time, they have plenty of time to play catch-up with going on these treatments. Maybe 5 years is a long enough time that it might have an effect on overall survival, but there is no way

that the extent would be anywhere near the extent from the disease free survival rate.

The fact this [trial design] was approved by the agency [FDA] with disease free survival as a primary endpoint almost certainly means the drug will get licensed, however I think the true value of ADAURA will not become manifest until we have seen the overall survival data because the CTONG update suggests that we should be practicing a buyer beware approach with these data.”

The big question that everyone will want to know from ADAURA is will it lead to an overall survival benefit? We may, however, never know. Back to Dr Camidge:

“The analogy here is let’s say you had a group of stage IV with EGFR mutant disease and you put them on osimertinib or placebo, well not surprising your progression free survival will look dramatically different in the first couple of years. We know the PFS of osimertinib in the first-line [metastatic] setting is about 19 months, so you would expect it to look dramatically different, essentially PFS of nothing in the placebo arm and 19 months in the osi arm, and that’s fine but then the whole idea that you might say

….and great your disease has been controlled, now we’re going to stop the drug

…and you go, what, that’s illogical,

… no we’ve cured you, you have a great response, we’re going to stop the drug!

The trouble is I think that’s the danger we might have here in focusing on DFS in ADAURA.”

Other ongoing phase 3 studies such as ALCHEMIST may offer additional insights in the future on whether there is an overall survival benefit or not:

Dr Camidge: “Even though the FDA approved [the study design for ADAURA] with disease free survival as a primary endpoint, there’s an NCI study called ALCHEMIST which is still listed as a trial in progress, where the NCI had recognized that the unanswered question was does adjuvant TKI therapy improve overall survival, so you have two government agencies, FDA and NCI giving somewhat conflicting opinions in the endpoint that’s most relevant, yes we’re interested in disease free survival as an endpoint, no we’re interested in overall survival as the most impotant endpoint.”

ADAURA Study Design

A good starting point when analyzing any clinical trial is to look at the study design:

The first question which jumped to my mind when I looked at this, is why was this phase 3 study conducted with a placebo control group when the standard of care in the United States at least, is a cisplatin-based chemotherapy doublet (often with navelbine)?

Dr Camidge again:

“If it was my Mom or your Mum I would urge her to have the adjuvant chemo, which we know has an overall survival advantage. That’s been proven over years.”

If CTONG, a Chinese-led phase 3 trial, can run such an adjuvant trial with chemo as the control arm then why didn’t AstraZeneca? After all, the lucrative market for more expensive therapies such as osimertinib is likely not countries such as Poland, Romania, Russia, and the Ukraine, where there were trial sites.

However, looking at the baseline characteristics for ADAURA, we learn a few startling things.

Maybe a challenge was that two thirds of the patients enrolled were of Asian ethnicity (Asian n=434, non-Asian n=248), and granted, there is some reluctance towards chemo in Asia.

One solution to the common problem of regional differences in standards of care is to conduct trials in each region based on the prevailing therapy choices. This way, AstraZeneca could have chosen to go the route of conducting say, ADAURA-EAST with Asia and Eastern Europe evaluating osimertinib against placebo plus a separate ADAURA-WEST study in the US with a head-to-head trial of osimertinib versus investigators choice of platinum-based chemotherapy – with the comparator problem solved in a trice.

Who knows, maybe the thinking was that if you could get away with a placebo control arm in many countries then why not take advantage of it, and by doing so, generate the best shot on goal and chance of a positive trial, or there was a desire to accrue the trial as fast as possible?

We have written on BSB about the potential for chemotherapy to influence the immune system – see interview with Prof Kees Melief

(link), “Chemotherapy is Immunotherapy” and this point was also taken up by Camidge and Ng in their Lancet 2018 editorial:

“Chemotherapy is not a cure in advanced NSCLC, but in the adjuvant setting it can either depopulate the microscopic cancer or potentially render it immunogenic enough for the immune system to sometimes take over lifetime control, or both of these. Therefore, another challenge for adjuvant tyrosine kinase inhibitor treatment could be that cancers susceptible to tyrosine kinase inhibitors, the cell death induced by these drugs, or both, might be less immunogenic than those cancers benefiting from classic adjuvant chemotherapy.”

In other words, we should not dismiss the influence chemotherapy, which is why it is surprising the enrollment criteria for the ADAURA study did not exclude prior adjuvant chemotherapy!

If you look closely at the DFS subgroup data presented by Dr Herbst, more than half the patients appear to have received adjuvant chemo at some point (n=378 for prior adjuvant chemo, n=304 for no prior chemo). To be fair, the two arms were balanced for this:

Osi plus prior chemo (55%) Osi w/out prior chemo (45%)

Placebo plus prior chemo (56%) Placebo w/out prior chemo (44%)

The analogy that comes to mind is a marathon where some are wearing the latest running shoes [osimertinib] while others have no shoes at all [placebo without prior chemo]!

Meanwhile US doctors will need to factor into their analysis of benefit, the lack of an active control arm as comparator. Are they

concerned about the risk of over-treatment of stage 1B patients, the financial toxicity from taking a targeted therapy for three years, some other factor, or is it not knowing if a patient will live longer as a result?

Let’s move and briefly consider some of these issues.

Financial Toxicity Can’t be Ignored

One challenge to anyone declaring ADAURA to be the new standard of care is the potential financial toxicity associated with prolonged therapy, especially for a treatment that so far has not shown anyone would be living any longer, as a result.

In his discussion of the CTONG1104 OS data at ASCO, Dr Azzoli (Brown) showed a slide of the cumulative cost of continuous adjuvant therapy, and while that is great from a company commercial perspective, I challenge anyone to justify why someone should mortgage their house or raid their pension in the absence of an overall survival benefit showing they will live longer on osimertinib.

It is worth remembering that many people in Asia have to pay for drugs themselves.

We heard Prof Tony Mok (Hong Kong) at ESMO 2018 on episode 26 of the Novel Targets Podcast draw attention to this when talking about lung cancer immunotherapy, but the same sentiments could perhaps apply to osimertinib in the adjuvant setting too:

“Patients, especially in a non-reimbursed system, may have to spend a lot of their life saving on a drug that may not work at all, and you have the responsibility to explain to your patient, why it may work, why it may not work, what is the likelihood of working, such that they will invest appropriately.”

The suggestion from ASCO20 is that AstraZeneca will most likely seek the broadest approval possible for osimertinib in the adjuvant setting in all three early stages (Ib-IIIa).

Back to Dr Camidge:

“The real question in adjuvant is do we have a way of telling who is not cured from the get go and therefore is there a benefit of going on the TKI, not as adjuvant but as early metastatic control, and not stopping it versus identifying a group of people who are truly cured and don’t need this at all? If you look at the placebo arm in the main analysis population in ADAURA as many as 40% of the patients were already cured and did not need anything more.

And that comes down to the idea of some kind of baseline analysis, a highly sensitive cfDNA has been floated for that. Not just a

low sensitivity one. The need to look for a subgroup that is deriving additional benefit in studies such as ADAURA and also whether in those people the drug can ever reasonably be stopped is the academic burning question.”

Is ADAURA practice changing?

Dr Herbst in his ASCO20 presentation clearly thinks so, concluding, “adjuvant osimertinib provides a highly effective, practice changing treatment for patients with stage 1b/II/IIIa EGFRm NSCLC after tumor resection.”

The ASCO20 discussant, Dr David Spigel (Sarah Cannon) also went there concluding, “adjuvant osimertinib substantially improves DFS in early stage EGFRm NSCLC and should be a new standard of care.”

It will be interesting to see how many US doctors share Dr Camidge’s considered perspective on the use of osimertinib in the adjuvant setting:

“I think it will undoubtedly get an FDA license. I do not believe the data should be practice changing or at least not yet. I think when you show there is an overall survival benefit then it will be practice changing…

So far there is no reason to suggest that disease free survival is going to translate into an overall survival advantage as it has not in any other comparable targeted therapy trial in EGFR mutant lung cancer. If this trial is the exception though, it will certainly not be of the same magnitude as the DFS benefit. However, the real unanswered questions are who needs this drug in this setting and if they need it, who can stop it safely and when.”

It doesn’t get any more unequivocal than that from someone well respected in the EGFR field!

Additional Commentary by @MaverickNY

One of the challenges repeatedly seen with targeted therapies for the adjuvant treatment of early stage disease has been the crowning achievement everyone craves, namely overall survival. This endeavour has turned into quite the Holy Grail of searches for many a company and researcher. After all, it’s tempting to believe the oft repeated hypothesis — the earlier the treatment, the better the outcomes – but are they?

If want to put early EGFRm positive lung cancer into context we can quickly look at a couple of examples as a guide, such as trastuzumab and pertuzumab in HER2+ breast cancer, as well as imatinib in GIST.

In both cases, they produced impressive results in metastatic disease which truly changed the course of the disease and became the standard bearers for others to follow in their wake. This is precision medicine at its best – the target and biomarker in each case are the same and select patients most likely to respond, leading to improved overall survival.

When they moved into the adjuvant setting, however, the case overall is far less compelling.

At SABCS last year we saw how with six years of follow-up in the APHINITY trial, which investigated the addition of pertuzumab to trastuzumab plus chemotherapy) there were fewer deaths observed in patients treated with pertuzumab, although the survival benefit was not statistically significant at that time (link to abstract).

The initial results were published in the NEJM in 2017, illustrating the improved rates of estimated 3-year invasive disease–free survival (DFS) compared with those in the placebo arm (94.1% vs 93.2%, respectively) in the ITT analysis. There was a clearer significant benefit in node-positive disease, where the 3-year rate of invasive DFS was 92.0% in the pertuzumab group compared with 90.2% in  the placebo group (HR 0.77; P=0.02).

Now in this particular case it was always going to be a be a huge hurdle to demonstrate superiority over trastuzumab because Piccart  et al (2005) previously demonstrated a significant benefit in DFS over chemo alone with just one year of trastuzumab treatment. It was another six years later before Slamon et al., (2011) were able to demonstrate a significant benefit in overall survival in favour of trastuzumab over chemo alone, but this came at the expense of increased cardiotoxicity:

“The rates of congestive heart failure and cardiac dysfunction were significantly higher in the group receiving AC-T plus trastuzumab than in the TCH group (P<0.001).”

It is also worth noting it took over 3,200 women to be randomly assigned and followed over 5.5 years in order to have sufficient power to garner the OS benefit for trastuzumab. Pertuzumab may still have a chance of showing a benefit over trastuzumab with longer follow-up, although we should not hold our breath, as this will be very tough to do.

What I do like about both pertuzumab combo and the original trastuzumab trials is they were added to chemotherapy and compared to chemotherapy plus placebo. Part of me feels strongly that AstraZeneca could have at least had one trial in the US comparing osimertinib to SOC chemo rather than placebo. Clearly others have successfully incorporated it in the trial design as the control arm and… succeeded.

Moving on, we can look briefly at the case of imatinib for the treatment of gastrointestinal stromal tumours (GIST) and see what happened there.

Several phase 3 trials have been conducted exploring the role of imatinib as adjuvant therapy after surgical resection in GIST, including the following:

NCI ACOSOG Z9001 – 1 year of adjuvant imatinib therapy versus observation EORTC trial – 2 years of adjuvant imatinib therapy versus observation

Italian study (ImadGist) – Maintenance or interruption after 3 years of adjuvant treatment Scandinavian Sarcoma Group/AIO trial – 1 vs. 3 years of adjuvant imatinib therapy

It soon became obvious that patients with KIT exon 11 deletions assigned to 1 year of adjuvant imatinib had a longer RFS than those with exon 9 deletions, but the trials included both subsets before this became known, thereby diluting the enriched population.

Of those four GIST trials, however, while each in turn demonstrated a clear RFS benefit in favour of imatinib treatment over observation (chemo is not effective in GIST), only one (the Scandivanian study deftly overseen by Dr Heikki Jonsuu) produced a significant overall

survival result and even then, it barely squeaked in with a P-value of 0.05 in favour of 3 years of therapy over 1 year. Note that in each case, the trials were not unblinded and long-term follow-up was conducted to determine the OS benefit, and the same is also true for both the pertuzumab and Herceptin studies.

In addition, there is another ongoing phase 3 trial exploring longer duration of therapy, although this one is currently in the enrollment phase of the study so it will be a long while before we know if OS is improved:

Scandinavian Sarcoma Group – 3 vs 5 years of imatinib

We can now clearly see how incredibly hard it is to achieve the nirvana of overall survival in early stage disease. Of the three examples shown here (trastuzumab, pertuzumab, and imatinib), only trastuzumab was truly successful thus far in showing an unparalleled (and uncontroversial) benefit in favour of the targeted therapy.

For osimertinib, we may never know if there is improved overall survival since the trial was stopped early on successful attainment of DFS, which is a shame if we really want to answer the question of do patients live longer when taking such a therapy early in the course of the disease.

While there is clearer benefit in favour of stage III rather than stage I disease, the ITT analysis across all stages demonstrated a significant benefit in favour of the TKI. The company are highly likely to seek FDA approval for all three stages on the basis of the ADAURA trial, as Dr Camidge noted in hi

While there is clearer benefit in favour of stage III rather than stage I disease, the ITT analysis across all stages demonstrated a significant benefit in favour of the TKI. The company are highly likely to seek FDA approval for all three stages on the basis of the ADAURA trial, as Dr Camidge noted in his interview.

Whether we will ever see an overall survival benefit from the ADAURA trial is the key commercial question, as without it, the company are unlikely to see widespread usage of osimertinib in Europe in the adjuvant setting for EGFR mutant lung cancer. The US is a different kettle of fish entirely.

EGFR 突变Ⅰ B~ Ⅲ A 期NSCLC

奥希替尼辅助将复发或死亡风险降低83%

 美国耶鲁癌症中心、Smilow 癌症中心Herbst 等报告ADAURA 试验结果显示， 对EGFR 突变 的Ⅰ B~ Ⅲ A 期非小细胞肺癌 （NSCLC）患者，奥希替尼辅助 治疗可将疾病复发或死亡的风险降低83%。其中几乎不吸烟或没有 吸烟史的肺癌患者，奥希替尼或 展现新希望。（摘要号LBA5） 约30% 的NSCLC 患者在诊断 时手术，但手术联合基于顺铂的 辅助化疗后5 年复发率仍很高， 范围从Ⅰ B 期患者的45% 至Ⅱ期 的62%。Ⅲ期NSCLC 患者的比例 为76%。Osimertinib 相较于厄洛替 尼或吉非替尼，可改善EGFR 突 变晚期NSCLC 患者以及中枢神经 系统转移患者的无进展生存期和 总生存期。

该项在美国、欧洲、亚洲和澳 大利亚进行的Ⅲ期随机双盲试验 共入组682 例完全切除的Ⅰ B 期、 Ⅱ 期或Ⅲ A 期NSCLC 患者， 可 有或没有按计划开展的辅助化疗， 并按疾病分期、EGFR 突变和种 族（亚洲人与非亚洲人）分层。 患者等比随机接受奥希替尼80 mg qd 或安慰剂，计划最长治疗3 年。 独立数据监察委员会于2020 年4 月举行了一次会议，认为结果明 显支持奥希替尼；因此建议揭盲。 揭盲时，研究已完成招募，所有 患者均已随访至少1 年。

结果显示，在Ⅱ ~ Ⅲ A 期疾 病患者中，奥希替尼组和安慰剂组 的主要终点中位无病生存期（DFS） 分别为尚未达到和20.4 个月 （HR=0.17，P<0.0001）。在包括 Ⅰ B 期患者在内的总人群中，次 要终点DFS 的数字相似，分别为 尚未达到和28.1 个月（HR=0.21， P<0.0001）。在所有人群中，根据 性别、年龄、吸烟状况、种族、 疾病分期、EGFR 突变和辅助化疗 （是或否）划分亚组，奥希替尼 在DFS 方面均具有明显优势。

奥希替尼耐受性良好，药物安 全性与已知的一致，未见死亡事 件，且任何种类的3~4 级不良事 件的发生率都很低。奥希替尼组 中共有10 例（3%）患有间质性肺 疾病。奥希替尼组22 例（7%）和 安慰剂组4 例（1%）患者发生QT 间隔延长。

有研究者表示：奥希替尼已是 EGFR 突变晚期NSCLC 患者的一 线治疗标准，该研究结果支持将 其用于早期疾病。但该研究结果 应被认为是初步结果，目前尚不 清楚这种疗法是否真的能提高患 者的治愈率，仍需更多的无病生 存数据来支撑论断。