Table 4 Food and Drug Administration reviewer comments on deficiencies related to patient-reported outcome measures

Faslodex (Fulvestrant)

FACT-B TOI

2002

• Insufficient data collection

• Difficult to reach a meaningful conclusion

Tykerb (Lapatinib)

FACT-B, EQ-5D

2007

• No comments

Ixempra (Ixabepilone)

FBSI

2007

• Difficult to reach a clinically meaningful conclusion

• Poor compliance after the baseline evaluation

Perjeta (Pertuzumab)

FACT-B TOI

2012

• Questionable content validity

• Questionable sensitivity to detect differences in symptoms between treatment arms due to long interval between questionnaire administration.

Kadcyla (Ado-trastuzumab emtansine)

FACT-B TOI

2013

• No comments

Ibrance (Palbociclib)

mBPI-sf

2015

• PRO measure is not comprehensive or sensitive

• Open label study design complicates PRO interpretation.

Kisqali (Ribociclib)

EORTC QLQ-C30, EORTC QLQ-BR23, EQ-5D-5 L

2017

• PROs were not alpha allocated and considered exploratory.

Nerlynx (Neratinib)

EQ-5D, FACT-B

2017

• FACT-B’s combined assessments of disease symptoms and global health status complicates PRO interpretation.

• Lack of clear guidance on different levels of severity

• PRO data were no longer collected after Amendment 9.

Verzenio (Abemaciclib)

EORTC QLQ-C30, EORTC QLQ-BR23, mBPI-sf, EQ-5D-5 L

2017

• Lack of prespecified objectives

• Treatment related symptom data do not add significantly to the existing adverse event data.

Talzenna (Talazoparib)

EORTC QLQ-C30, EORTC QLQ-BR23

2018

• PRO analysis was subjective and not validated in the patient population.

• PRO outcomes are likely to be biased due to the open label study design (i.e., EMBRACA).

• Missing type-1 error adjustment.

Piqray (Alpelisib)

BPI-sf, EQ-5D-5 L, EORTC QLQ-C30

2019

• PRO analysis was not controlled for multiple comparisons and considered exploratory.

• Outcomes in PRO analysis was not clearly defined

• Not adequately powered to support a non-inferiority conclusion.