All adverse events (irrespective of relationship to study treatment)
Table 8 on page 57 shows all-causality AEs reported in ≥ 10% of patients in either arm. Most prominent41 AEs for the palbociclib arm (relative to placebo) were:
neutropenia (79.5%, versus 6.3% for the placebo arm), in the context of more:
laboratory abnormalities of ‘neutrophils decreased’ (95% versus 20%, including 67% versus 2% grade 3-4 findings42, as per CSR Table 53)
infection (59.7% versus 42.3%), including grade 3-5 infection (6.5% versus 4.5%); and
febrile neutropenia (2.5% versus 0%; grade 3-4 in 1.8% versus 0%)4344
leucopenia (39% versus 2.3%);
alopecia (32.9% versus 15.8%);
stomatitis (30.4% versus 13.5%);
anaemia (24.1% versus 9%), alongside:
laboratory abnormalities of ‘anaemia’45 in 78% versus 42% (including grade 3-4 findings in 6% versus 2%, as per CSR Table 53)
thrombocytopaenia (15.5% versus 1.4%), noting also an increase in:
laboratory abnormalities of ‘platelets decreased’46 (63% versus 14%, including grade 3-4 findings in 2% versus 0%)
treatment-related epistaxis (6.3% versus 2.7% in the CSR, 9.2% versus 6.3% in the Clinical Overview for Dossier version ; the sponsor states that no grade 3-4 thrombocytopenia AEs were associated with bleeding events, but in the sponsor’s s31 response to CER-ES safety Q24, multiple low-grade AEs of epistaxis were temporally associated with thrombocytopenia)
13. In relation to alopecia, please characterise this further. For affected patients in the palbociclib-containing arm, was there typically ‘complete’ hair loss as might be seen with some chemotherapies or ‘thinning’ as might be seen with, for example, letrozole? Was there scalp tenderness? Did hair re-grow normally on treatment discontinuation?
14. It was noted from the CSR that there were several cases of intracranial haemorrhage in the palbociclib-containing arm (patient IDs 13331001 and 10131001). Were there other cases of significant bleeding? Was any case associated with thrombocytopenia of any grade?
Also of note, regarding rarer AEs:
cataracts were reported in 3.4% versus 0.5%; one case in the palbociclib arm was grade 3 and required bilateral cataract surgery (overall on treatment, 9 patients in the palbociclib-containing arm and 1 in the placebo-containing arm had cataract surgery). While these were not necessarily mediated by hyperglycaemia, there was a distinct imbalance across arms (beyond what can easily be explained by imbalance in treatment duration) suggesting a treatment effect.
Question 15 for sponsor
15. Was sufficient information gathered to understand whether the imbalance in cataract surgery extended more than 28 days after the last dose?
interstitial lung disease or pneumonitis was reported in 0.7% (3/444) versus 0.5% (1/222), although all three AEs in the palbociclib arm were considered related and the one AE in the placebo arm was not considered related
The following graph indicates AEs with wider risk differences:
Figure 17: AE with wider risk differences
There was a lower reporting frequency of headache (21.4% versus 26.1%; including grade 3-4 headache, 0.2% versus 1.8%) and hot flushes (20.9% versus 30.6%) in the palbociclib-containing arm. There was also a lower reporting frequency of grade 3-4 hypertension (3.4% versus 5.9%).
Comment: Headache, hot flushes and hypertension are commonly reported with letrozole. The lower frequency of these AEs in the palbociclib + letrozole arm versus the placebo + letrozole arm, despite no PK interaction, raises the suspicion that there may be a PD interaction between the two drugs, which might lessen some activities of letrozole.