Loxo Oncology is without question the furthest ahead in the field of NTRK-driven tumours. What was not known, until the weekend, was the threat that Roche presented to this lead. New data on the companies' respective assets – larotrectinib and entrectinib – at Esmo on Sunday suggest that the US biotech does not have too much to worry about.
Not that the data on entrectinib, the first to be presented in patients with an NTRK gene fusion, were poor, but its overall response rate of 57% is notably lower than the 81% achieved by larotrectinib. Important distinctions between these datasets offer some explanation for the different response rates, but these relative performances seem to confirm that hitting a target much more specifically makes a difference.
Larotrectinib is a kinase inhibitor that selectively blocks the tropomyosin receptor encoded by the NTRK gene, while entrectinib targets the same receptor as well as Ros1 proteins. Ros1 is predominantly seen in lung cancer, while NTRK gene fusions are tumour agnostic; in the NTRK setting the studies of both agents recruited patients from around 10 different cancer types.
The Loxo asset, which has been partnered with Bayer and is awaiting news on US approval next month, has a more extensive dataset. At Esmo Loxo revealed results from a further 67 patients on top of the 55 treated initially, and which formed the basis of the marketing application. The patients were recruited into two studies, one adult and one paediatric.
|Larotrectinib (Navigate and Scout trials)|
|Data in marketing application||Data presented at Esmo 2018|
|Primary data||Primary data||Supplementary data||Integrated data|
|Central assessment||Investigator assessment||Investigator assessment||Investigator assessment||Investigator assessment|
|DoR||Not reached, 8.3mth follow-up||Not reached, 17.6mth follow-up||Not reached, 7.4mth follow-up||Not given|
|mPFS||Not reached, 9.9mth follow-up||28.3mth*||Not given||Not given|
|*Judged not statistically stable owing to low number of progressions. ORR = objective response rate, PR = partial response, CR = complete response, DoR = duration of response, mPFS = median progression-free survival.|
|Source: Esmo 2018.|
Reassuringly, ORR did not drop with more patients added, and most impressively median duration of response has not yet been reached.
This is particularly notable, suggested Dr Allan Jordan of Cancer Research UK’s Manchester Institute, who discussed both datasets at the conference. Designing drugs for specificity can mean compromising on durability of response, he said, but this does not seem to have been the case with larotrectinib.
Comparing these data against the first look at entrectinib’s efficacy in NTRK patients is made tricky by the slightly different measures used, on top of all the usual caveats of cross-trial comparison. Importantly, the Roche trials enrolled a sicker patient population, almost half of whom had received at least two prior lines of treatment, and included more patients with metastases in the brain.
This could have contributed to a weaker headline response rate and shorter duration of response, which had matured at a median of 13.1 months' follow-up.
It is unclear whether the fact that these data were in adult patients makes a difference, although this was emphasised by the presenter of the entrectinib data at Esmo. Larotrectinib generated stronger results in children, with a response rate of 92% compared to 75% for adults.
This could be because several adults treated with larotrectinib ended up not having an NTRK fusion, according to Dr Ulrik Lassen of Rigshopitalet in Copenhagen, who presented the larotrectinib data Esmo.
“I think everyone would respond if they have that fusion; I would expect adults and children to have exactly the same response rate,” he told Vantage.
The selectivity of larotrectinib could make it a more potent agent, he said, adding that response rates are so high because the NTRK fusion is the only activating signal involved in this cancer type.
“This fusion is very oncogenic, and we don’t any other mutations at the same time, which is why it is so important.”
|Entrectinib (ALKA and Starktrk 1 & 2 trials)|
|Blinded independent central review|
|IC ORR (n=11)||54.5%|
|DoR (n=31)||10.4 mos|
|PFS (n=54)||11.2 mos|
|OS (n=54)||20.9 mos|
|ORR = objective response rate, SD = stable disease, PD = progressive disease, IC ORR = intracranial objective response rate, DoR = duration of response, PFS = progression free survival, OS = overall survival. Source: Esmo 2018.|
Entrectinib’s intracranial response rate was highlighted as notable by Dr Jordan, although the numbers here are small. A comparable number is not available for larotrectinib, but impressive cases of CNS activity have been reported; in a case profiled at Esmo over the weekend brain metastases in a lung cancer patient shrunk by 95% after treatment.
Dr Lassen said intracranial data was being prepared on larotrectinib, and it will be interesting to see how this stacks up against the Roche product.
Another important point is that larotrectinib’s response rate fell to 75% when assessed by independent reviewers, as can be seen in the data submitted to regulators, above. Even so, it is clear that this is a highly impressive agent. Leerink analysts commented that larotrectinib seemed to have turned NTRK-fusion cancer into a state of chronic disease.
The sellside sees the product generating sales of $776m by 2024, according to consensus data from EvaluatePharma, a figure that could well now travel upwards.
2024 consensus for entrectinib sales stands at $375m, and there are much data still to come. After reporting a very impressive response rate of 77% in Ros1-positive lung cancer patients at World Lung last month the project is far from a disappointment.
However, Roche paid the princely sum of $1.7bn to get its hands on entrectinib. Whether this was too rich is also a question yet to be answered.