Tesaro Enjoys A 'Ta-Dah' Moment

| About: Tesaro (TSRO)

The presentation at Esmo of Tesaro’s (NASDAQ:TSRO) ENGOT/Nova phase III trial of niraparib as maintenance therapy in ovarian cancer was one of the most eagerly anticipated at this year’s conference and, in the event, it did not disappoint.

Although the study, in platinum-sensitive patients, was already known to have been positive, detailed results presented on Saturday show a highly statistically and clinically significant benefit in all of the patient subgroups examined, including one thought unlikely to respond to a PARP inhibitor (see table below).

Tesaro reported outline results in June showing a very significant benefit for niraparib in germline BRCA mutant patients. Although this had been expected based on the drug’s mechanism, the magnitude came as something of a surprise.


Niraparib showed a near threefold increase in median progression-free survival (PFS) of 21 months versus 5.5 months for placebo. The hazard ratio was 0.27, which is equivalent to a 73% improvement in relative risk of progression, a figure described as unprecedented in ovarian cancer.

More good news came in the fact that the drug’s benefit extended to two other groups of patients who were not germline BRCA mutation carriers: those who were positive for homologous recombination deficiency (HRD) and those who were HRD negative. The statistical analysis plan allowed sequential testing of these two subgroups in turn, if the prior one had been positive.

The new data, which were included in the Esmo presentation and the study’s simultaneous publication in the NEJM show results in new subgroups as part of an exploratory analysis: HRD-positive patients with the rarer somatic BRCA mutations, BRCA mutants who are HRD positive and, perhaps most importantly, BRCA mutant negatives who are also HRD negative.

This last group saw, in relative terms, the least benefit, with a hazard ratio of 0.58. But it represents the largest patient demographic, accounting for some 50% of ovarian cancers. And the fact that a benefit is clearly there calls into question the utility of the HRD diagnostic test, much to the chagrin of its developer, Myriad Genetics (NASDAQ:MYGN).

To test or not to test

The original hypothesis underlying the development of this test and its use in the Nova and other ovarian studies was that it would be able to identify a wider pool of patients than just BRCA mutants, who might best respond to the drug.

Tesaro intends to file niraparib in the fourth quarter and will seek “the widest possible indication”, something that presumably will cover both HRD-positive and -negative patients, thus rendering the HRD test unnecessary. This could mean that regulators have to consider whether to approve a drug without the companion diagnostic test that was used as a part of its development, but turned out unexpectedly not be needed.

Myriad has expressed strong views against this, but the Nova study’s principal investigator, Dr Mansoor Rava Mirza, of the Nordic Society of Gynecologic Oncology, takes the alternative position. He believes that HRD testing would be unnecessary, since all platinum-sensitive ovarian cancer patients should be offered the drug irrespective of HRD status. Dr Mirza stresses that testing for germline BRCA mutation remains justified, since close female relatives of the patient could be offered genetic counseling.

Given its strong efficacy data, the utility of the HRD test likely to be the most controversial aspect of the registration dossier.

The Nova study subgroups

PFS months

PFS hazard ratio, p value




Germline BRCA mutant

21.0 (12.9, NR)

5.5 (3.8, 7.2)

0.27 (0.173-0.410), p<0.0001

Non-gBRCA mutant

9.3 (7.2, 11.2)

3.9 (3.7, 5.5)

0.45 (0.33, 0.6-07), p<0.0001

Non-gBRCA mutant, HRD positive

12.9 (8.1,15.9)

3.8 (3.5, 5.7)

0.38 (0.243, 0.586), p<0.0001

Somatic BRCA mutant, HRD-positive

20.9 (9.7, NR)

11.3 (2.2, NR)

0.27 (0.087, 0.903), p=0.0248

Non-BRCA mutant, HRD-positive

9.3 (5.8, 15.4)

3.7 (3.3, 5.0)

0.38 (0.23, 0.828), p=0.0001

Non-BRCA mutant, HRD-negative

6.9 (5.6, 9.6)

3.8 (0.7, 5.6)

0.58 (0.361, 0.922), p=0.0226

95% confidence intervals shown in parentheses.

Click to enlarge
Click to enlarge