Reply to “Methodological considerations in assessing the impact of CGRP monoclonal antibodies on blood pressure”

We appreciate the thoughtful comments from Mascarella et al. ¹ regarding our systematic review on the impact of calcitonin gene-related peptide (CGRP) monoclonal antibodies (mAbs) on blood pressure (BP). We acknowledge the importance of real-world data (RWD), particularly for evaluating long-term safety outcomes in clinical populations. However, we would like to provide methodological justifications for excluding observational studies without control groups from our review.

Lack of control group and causality assessment

The primary limitation of observational studies without a control group is their inability to establish causality. Without a comparator, it is challenging to determine whether observed BP changes are attributable to CGRP-mAbs or other confounding factors such as migraine, as an a cardio/cerebrovascular risk factor itself, age, comorbidities or concomitant medications. Temporal associations alone do not imply causation, limiting the interpretability of such studies.

Risk of bias and internal validity:

The absence of a control group increases susceptibility to selection bias because patients receiving CGRP-mAbs may differ systematically from those who do not. Additionally, variations in BP measurement protocols and the potential for measurement bias further compromise the internal validity of these studies. By including only studies with a control group in our review, differences in BP measurement protocols are minimalized. Ensuring internal validity is essential when assessing potential cardio- and cerebrovascular effects because unreliable data can lead to inaccurate conclusions, as might be the case in observational studies without a control group. ² Furthermore, studies often include patients requiring adjustments to their antihypertensive therapy during follow-up, introducing a major source of bias in the causal pathway that leads to changes in systolic and diastolic blood pressure. Patients who initiated or modified antihypertensive medication should be excluded from further analysis beyond that point, as was correctly performed in only one study. ³

Confounding and baseline comparisons

Even with statistical adjustments, uncontrolled observational studies cannot fully account for confounding variables. Without a control group, there is no baseline reference to evaluate whether BP changes are due to treatment or other underlying factors. This limitation is particularly critical when assessing safety outcomes, where precise attribution is necessary.

Alignment with systematic review methodology

Although PRISMA guidelines do not explicitly mandate the exclusion of uncontrolled observational studies, they emphasize minimizing bias to ensure the reliability of evidence synthesis. Including studies without control groups may compromise the overall quality of the evidence and reduce the robustness of conclusions regarding the safety profile of CGRP-mAbs.

In conclusion, our decision to exclude observational studies without control groups was based on the need to prioritize studies with stronger methodology to ensure reliable evidence synthesis. We recognize the complementary role of RWD in post-marketing surveillance but maintain that studies with appropriate comparison groups provide more reliable insights into the potential cardiovascular effects of CGRP-mAbs. We appreciate the opportunity to clarify our rationale and thank the authors for their engagement in this important discussion.