The background to the DETERMINATION study was really built on the success of the triplet combinations incorporating an immunomodulator, proteasome inhibitor and a steroid. In this case this was lenalidomide, bortezomib and dexamethasone - we call that RVd. Given that that regimen had demonstrated a favourable tolerability profile and remarkable efficacy with response rates up to 100% and high-quality responses in about half of patients, with about 75% of patients achieving what we call very good partial response that appeared to be sustained, both in terms of progression-free and overall survival. The question then arose, in newly diagnosed transplant eligible patients, where does transplant best belong? Can it be used early? Should it be used always early? Can it be kept in reserve in some selected patients? And can it be used in a delayed fashion in others who may wish to choose to keep the transplant option as a back pocket option, as opposed to a front pocket option, if you excuse the metaphor. But I think the important question was that, obviously high dose melphalan stem cell transplant has been the standard of care, and continues to be; I think we just need to better understand, or we hope to better understand, with this trial, how its evolving position in the treatment of newly diagnosed myeloma patients would transpire to land. 

So how did you look at this?

We looked at this by conducting a 56 centre multi-centre trial, conducted across the United States. We were supported by the Clinical Trials Network and sponsored by them, and also endorsed by the alliance. The study itself was led by Dana Farber, so truly an investigator-sponsored trial. We had tremendous support from pharma partners and also grants from the NIH and NCI, so this was truly a study of public-private partnership if you will, for want of a better term. Importantly, it was a very comprehensive but really carefully conducted trial. We had the benefit of a very high-quality contract research organisation; there were over 28,000 monitored visits. This was a study which was really conducted to almost industry standards, but at the same time was investigator sponsored. So a challenging question, but one done in a real-world sense

What we did was to assign patients to either early transplant after successful RVd induction with RVd consolidation and then lenalidomide maintenance, versus RVd induction remission therapy for a total of eight cycles, harvesting of stem cells, and then lenalidomide maintenance with transplant used later as needed. Very importantly, another aspect of this trial was the use of lenalidomide maintenance continuously. In the myeloma space there’s been some controversy about how long to give lenalidomide and in the US our data best supported continuous treatment. In Europe, for regulatory reasons, there was more of a push for a fixed duration of exposure. 

Our study design was actually built in close partnership with our dear friends and partners in the IFM, and the IFM ran a parallel study, identical to ours, but critically their study only used lenalidomide for one year, and also – and we’ll come to this in a minute – because of the nature of the way the IFM is structured, they were also able in their delayed transplant arm, to have a much higher use of transplant. These comparative data, we feel, will potentially be very helpful and very informative.

 Indeed that’s kind of what transpired; what we were able to show was, to date, the best progression-free survival for either arm, either the early transplant arm in which we saw a dramatic 67 month median progression free survival, so in excess of five years, and very importantly, we also showed that for the delayed transport arm, their progression-free survival was around 46 months. So 21 months less overall, but still the longest seen to date with that approach. So both approaches did very well, but this big difference in favour of transplant really caught our attention. 

This suggests that melphalan matters, and autologous stem cell support may reset the immunological milieu in a way that’s favourable. The important thing was, however, with long follow-up – we’re now approaching almost seven years of follow-up median – we saw no overall survival difference. It was identical at 80% in both arms. When we drilled down on that and looked at our patient salvage regimens, we were struck by the fact that in the delayed transplant arm only 28% of the patients had a transplant to date. This is fascinating because it suggests that there’s no overall survival difference, yet there’s this big progression free survival difference. Now, clearly that’s a win for the early transplant patients, but at the same time it does mean, given the absence of a clear survival benefit, that if a patient chose to keep transplant in reserve, that would be a very legitimate next step. 

You may say, well, why would you do that? Well, I think the real fact here is that transplant, as successful as it is, is nonetheless associated with significant toxicities, and we showed that. We showed a higher rate of grade 3 and grade 4 events clearly in the transplant arm versus the non-transplant arm. We also saw a small difference in transplant related versus non-transplant mortality. There was 1.6% deaths in the transplant early arm, versus 0.3% in the delayed arm. This was not meaningful, and very rare, but nonetheless important as one is thinking. When we drilled deeper and looked at quality of life, we’re able to show clearly there that quality of life took a clinically meaningful and significant drop during transplant and then recovered. The good news is that it recovered, and it recovered well, so that was great. 

When we looked then at second primary malignancies, both arms had the same rate, which was fascinating because there was no big difference overall, and fortunately relatively rare. Interestingly, though, we showed a significant difference in the instance of AML MDS against the transplant arm. There were ten cases in the early transplant arm versus zero in the RVd alone arm, and that had a significant p-value. You may say well, that’s rare, isn’t it? It is, the challenge is that the MDS AML risk doesn’t stop – it increases slowly over time, so rare, but important. So we’ve got to better understand what’s going on there. 

With all of that in mind, if you’re thinking of big PFS gain but OS is the same, it really informs a conversation with a patient, and we can start to say well look, there are these factors in your favour, there are these factors that may mean keeping transplant in reserve makes sense. But, conversely, there may be a patient who’s got high-risk disease, it’s been proving difficult to control with the triplet, that may be just the patient in whom autologous transplant really matters and you need to use it early. 

In terms of what we saw for useful correlative science, a particularly important aspect of the trial was the assessment of minimal residual disease. What we were able to show using next generation sequencing to a sensitivity of 10-5 was that, if you achieved MRD negativity, regardless of which arm you were on, you did very well. So our preliminary analysis of about 100 patients in each arm has shown that you get MRD negativity in about 52% of the patients transplanted early, versus around 40% for those in whom transplant is delayed. So that difference is real, but it’s not enormous. It shows that both arms are performing well, but clearly transplant adds more. The good news is that regardless of which arm you were assigned to, that was associated with progression-free survival benefit, so that’s a very important observation. More to follow on that but this continues, in my mind, to exemplify why MRD testing is clearly a very important research tool, and is gaining some traction now in practice because it can help inform treatment choices in settings such as this. 

The other thing with subgroup analyses in our study – we looked at various different patient populations, and some of our a priori pre-planned analyses suggested that high risk patients in particular might benefit from early transplant, but not all high risk is the same, and that 17p deleted patients didn’t do very well regardless of which arm they were assigned to. So we need to do more for 17p deletion. 

The second thing was that with the new therapies that are coming, obviously that’s going to change all of this. Fascinatingly, in certain other groups, for example African American patients – we had the highest representation of any phase III study to date as I mentioned previously, 20%, – in that patient population the progression free survival benefit for transplant was less marked, with a hazard ratio that was only just above 1.0. So you can see how in that patient population, absolute equivalent access to treatments, so that was not the issue, for pathobiological reasons we presume, transplant isn’t adding as much as it is in other populations. We need to better understand that but I’m sure you can see how this is helping us think things through, and we can then counsel patients accordingly as to what may be a best direction to take in accordance and tailored to their needs. 

So it’s added to your options?

Without a doubt, and then in terms of future directions, now we’ve got the quadruplets, we’ve got CAR-T therapy, we’ve got bio specifics, all these new agents. That’s changing the playing field yet again, and making the opportunity for patients to have multiple choices, and to tailor their needs over not just a tactical view of immediate therapy, to a strategic view of the whole natural history of their disease over, God willing, the next ten, fifteen, twenty years, this really helps make sense of that.