Subscription is FREE for qualified healthcare professionals in the US.

At Stake: The Possible Long-Term Side Effects of CGRP Antagonists

As AHS holds its annual meeting, very specific questions are arising about the use of CGRPs in migraine prevention. The author poses key considerations and sample cases for prescribing this new class of monoclonal antibodies.

A commentary for clinicians involved in the treatment of chronic headache and migraine

The novel class of monoclonal antibodies (mAbs) targeting calcitonin gene-related peptide (CGRP) are a valuable addition to our preventives for migraine. However, there are significant conceivable long-term adverse effects that need to be considered as these new products gain approval and enter the market. We will have a better feel for the true risk in 10 years. For each patient, we have to decide whether the benefits outweigh the possible risks. With luck, it may come to pass that the biologics targeting CGRP carry very few long-term risks. Certainly, these have been safe compounds for the short-term. Three-year safety data has recently been presented (see also “A New Frontier in Migraine Management: Inside CGRP Inhibitors & Migraine Prevention”). Herein, I  discuss some of the possible long-term issues with these long-awaiting medications.

The following are presented as a series of questions related to core systems (eg, cardiovascular, central nervous, gastrointestinal, reproductive, and more) that need to be addressed before CGRP antagonists are used widely. Most of these questions do not have answers at this time. Much of the CGRP research to date has been conducted in animal models, which, as we know, does not always correlate with effects in humans. Hopefully, over time, the community will be able to determine which of our patients may be at increased risk for long-term adverse effects. (You can also see a Pro/Con debate on CGRP inhibitors which took place among Drs. Alan M. Rapoport and Robert B. Cowan at the 2019 AAPM meeting.)


Cardiovascular and Pulmonary System

1. CGRP plays an important role in resisting the onset of hypertension (HTN); how relevant is this when prescribing to young patients, particularly those at higher risk for HTN? How much does vascular dilation redundancy matter (with other vasodilator mediators, such as PGs and NO, compensating for the loss of CGRP)?

2. With the onset of HTN, there is a compensatory release of CGRP: how relevant is this, and what effects do the antagonists have? In the face of HTN, CGRP release may become attenuated over time. There have been conflicting studies as to the amount of plasma CGRP present in those with HTN. To date, the antagonists have not appeared to affect blood pressure. Will CGRP antagonists be studied in those with HTN? Will these be evaluated in the face of poorly controlled HTN? Deletion of RAMP 1, for example, has been associated with cytokine production (proinflammatory) and HTN. Is this clinically relevant?

3. CGRP may delay or protect against the development of cardiovascular disease. For which patients is this relevant? CGRP is the most potent of all the vasodilators, so how might this influence prescribing for higher risk patients?

4. The effect of CGRP on the expression of endothelial nitric oxide synthase (eNOS): depleting CGRP may lead to enhanced loss of eNOS; what is the clinical relevance?

5. CGRP depletion may produce oxidative stress in the aorta; how clinically relevant is this?

6. If CGRP is knocked out, and the vasodilator effects are diminished, do other compounds (eg, nitrous oxide, substance P, prostaglandins) help to compensate (primarily at the resistance vessel level)? One study indicated that both substance P and CGRP had to be blocked for there to be a loss of vasodilatation.

7. There is polymorphism with the CALC 1 gene (this gene encodes CGRP and calcitonin, linked to essential HTN); is this clinically relevant in light of mAb use?

8. Evidence from KO mice indicates that reduction of CGRP on the cardiovasvular system may become pathologically relevant primarily in conjunction with compromised vasculature. Infusions of CGRP improve the circulation in the presence of heart disease. We need angiographic (and other) studies in patients with cardiovascular disease (CAD), ideally prior to and after treatment with the antagonist. Are further studies planned?

9. Could smaller cardiac or cerebral infarcts become more dangerous resulting from the protective effects of CGRP being blocked? CGRP protects against ischemia, cell death, and vascular inflammation in various organs (heart, brain, GI, kidney). When CGRP is seriously depleted, there is an increased susceptibility to injury via ischemia. How clinically relevant is blocking CGRP? This will need more research.

10. CGRP plays a role in heart failure. Infusion of CGRP improves circulation in the face of heart disease. Regarding microvascular growth, CGRP is an angiogenic facilitator. Should patients at high risk for failure, or with actual heart failure, not be prescribed these medications?

11. There is evidence that CGRP helps to protect the heart, and this effect is lessened in the presence of diabetes. For patients with both diabetes and CAD, should CGRP inhibitors be withheld?

12. CGRP levels decline with age (although there may be a bimodal effect) and CGRP helps to protect the myocardium; should CGRP inhibitors be withheld in older patients, particularly for those with heart disease?

13. Do the Amylin 1 receptors (or other calcitonin-group receptors) help to “cover” for the loss of beneficial effects, particularly vasodilatory, after the blocking of CGRP?

14. With regards to the cardiovascular system, is there a difference between antagonizing the ligand of CGRP, and blocking the receptor?

15. Regarding advanced CAD, how important is CGRP as a vasodilator? CGRP levels are increased during myocardial infarction. Could antagonizing CGRP lead to a more severe infarct? There was an erenumab-aooe (Aimovig, Amgen/Novartis - the first FDA approved CGRP mAb for migraine prevention) study of 90 patients with stable angina, who were given 140 mg IV as a one-time dose. There were no problems found in the 3 months post-infusion. Are further studies planned?

16. How clinically relevant is CGRP in the cerebral vasculature? CGRP is an inhibitor of platelet aggregation, through cAMP activity. Inhibiting CGRP could lead to embolic cardiac or cerebral events.

17. Is CGRP a vasodilator in both smaller and larger cerebral arteries?

18. CGRP and pulmonary HTN: CGRP is abundant in the lung; for high- risk individuals, would blocking CGRP increase the chance of developing pulmonary HTN? There is evidence that CGRP is beneficial in those with pulmonary HTN. Should patients with pulmonary HTN be excluded from receiving CGRP antagonists?

Central Nervous System (within the Blood-Brain Barrier)

1. There is slight penetration of these large-molecule mAbs into the CNS, from 0.1% to 1%; is this clinically relevant as to the mechanism of action of the mAbs? Probably not, but certainly it is possible.

2. Botox undergoes transcytosis (tracking along the axon from the trigeminal ganglion, into the brainstem): does this also occur with mAbs? Most likely it does not, but it may happen to a small degree. Radioisotope studies to identify elements of the mAb in the brainstem would be helpful.

3. How effectively do the peripheral (trigeminal ganglion) effects of the mAb dampen down central sensitization, and/or cortical spreading depression?

Central Nervous System (outside the Blood-Brain Barrier)

1. The anterior pituitary contains CGRP. Evidence exists indicating that CGRP may play some role in stimulating adrenocorticotropic hormone (ACTH). What are the possible effects on CGRP antagonism for the various hormones (GH, TSH, FSH, LH, ACTH, MSTH, and prolactin)? Should we be hesitant to prescribe for adolescents (off-label), due to possible effects on growth hormone? Should we measure hormone levels in those adolescents prescribed the mAbs? I am not aware of the various levels being drawn during the Phase 2 and 3 studies, but are there plans to evaluate these post-approval?

2. We do not know about the effect of this new mediation class on individuals with thyroid disease: should the mAbs be used with caution in diagnosed patients until more is known about the mAbs' effect on thyroid-stimulating hormone, or TSH?

3. Should studies be done evaluating FSH, LH, and ACTH levels before and after these antagonists?

4. What is the effect on prolactin? Should those with pituitary microadenomas be restricted from use?

5. Might there be an effect on melatonin levels?

6. The choroid plexus: could CGRP knockout affect cerebrospinal fluid (CSF) production? Would the CSF inflammatory homeostasis, partially controlled by the choroid plexus, be affected? Can this be evaluated?

7. The median eminence: could CGRP knockout affect hypothalamic hormone release (of CRF, TRH, DA, GHRH, and GnRH)? Should these be tested?

8. Area postrema (part of the circumventricular organs): would regulation of nausea/vomiting be affected? Other circumventricular organs: would homeostasis of cardiovascular or immune functions, or fluid regulation, or thirst/feeding, be affected? How can this be evaluated?

Gastrointestinal System

1. (Beta) CGRP is primarily present in the GI system (versus alpha CGRP), and CGRP is important for mucosal protection. What is the effect of antagonizing CGRP on the GI mucosa?

2. CGRP is involved in the healing of GI ulcers. Should antagonists be restricted for those with ulcers? This most likely depends upon how recently the ulcer was present, and if the patient is at high risk for recurrence. Should these not be given to those with a recent past history of ulcers, or those at high risk?

3. For those with, or at high risk for Inflammatory Bowel Disease (IBD), should these antagonists be restricted? Caution is prudent in considering the mAbs for those with IBD, or at high risk.

4. CGRP acts in a biphasic manner on GI motility. Should the antagonists be used with caution for those with moderate or severe IBS? Particularly with IBS-C, the mAbs may exacerbate constipation. Less often, diarrhea may be worsened (in theory).

Skin, Wound Healing, Burns, and Bone Health

1. CGRP contributes to flushing and thermoregulation; what are the effects of blocking CGRP on these functions? In blushing syndromes (such as hot flushes), CGRP release is involved. Would blocking CGRP affect these syndromes? Is there an effect on Raynaud’s symptoms?

2. What clinical effect results from dampening the CGRP effects on local skin edema and itch? CGRP can inhibit allergic conditions, such as certain types of dermatitis (irritant dermatitis). What effects on dermatitis might be seen by inhibiting CGRP? Loss of alpha CGRP-containing nerves may be associated with cold hypersensitivity. CGRP may have a role in temperature regulation. Could inhibiting CGRP be clinically relevant with these issues?

3. CGRP facilitates tissue repair and wound healing. These effects are mediated via vasodilation, upregulating VEGF expression, and by limiting inflammatory processes. CGRP also promotes revascularization. What effect does blocking CGRP have on wound healing? Which receptor does CGRP engage with to facilitate wound healing? Should at-risk patients for wound healing be prescribed these antagonists with caution?

4. CGRP plays some role in regeneration of the skin, via promoting proliferation of keratinocytes. Will skin be able to regenerate as well after CGRP is diminished?

5. For those with burns, CGRP and SP facilitate acute edema formation. What is the clinical relevance of knocking out CGRP for those with more severe burns?

6. CGRP may regulate bone metabolism through stimulation of osteoblastic differentiation, as well as an effect on osteoclastic formation. Amylin and calcitonin are also vital for bone health. Might the CGRP antagonists inhibit normal bone growth and metabolism? Does diminishing CGRP play a role in the healing of bone?

Renal, Sepsis, Diabetes/Obesity

1. Renal effects: during dialysis, CGRP levels are raised, possibly as a defense mechanism. How does antagonizing CGRP affect the person undergoing dialysis? CGRP may protect against renal damage in certain pathological conditions. In light of kidney disease, should the CGRP antagonists be used sparingly?

2. CGRP levels are raised during sepsis. CGRP acts on macrophages and other cells partially via upregulating IL-10, and by decreasing inflammation. CGRP also may mediate hypotension during sepsis. What effect does blocking CGRP have on these effects? If a patient on an antagonist becomes septic, would the therapy change?

3. CGRP is active within the pancreas, and is involved with the regulation of insulin release; the effect may be to reduce insulin levels, which (in theory) may result in hyperglycemia. Would antagonizing CGRP theoretically help with diabetes? In the presence of diabetes, CGRP is lessened (through nerve growth factor, NGF) in sensory neurons; what is the relevance for peripheral neuropathy? In diabetics with cardiovascular disease, should extra precautions be taken regarding antagonizing CGRP?

4. CGRP knockout may affect metabolism, energy use, and body weight. Could the CGRP mAbs affect body weight? Could this be included in long-term post-approval studies?

Pregnancy, Arthritis and Other Pains, and Aging

1. Early in pregnancy, CGRP levels are minimal in the fetus: what are the risks if CGRP antagonists are given prior to pregnancy? Later in pregnancy, CGRP may play a role in mediating the adrenal glucocorticoid response to acute stress in the more mature fetus. Circulating CGRP levels (in the mother) are increased during pregnancy, peaking in the last trimester. CGRP levels are lower with pre-eclampsia. There was a prenatal and postnatal study in monkeys with Aimovig. The animals received 50 mg/kg of Aimovig every 2 weeks. No effects were apparent on the fetus or infant, with regards to growth and development. The follow-up was through 6 months after delivery. Towards the end of pregnancy, CGRP plays a role in cervical ripening, and is present in the placenta and fetus: how would lessening CGRP affect the latter stages of pregnancy? Could a CGRP mAb render it more difficult to become pregnant?  Do the mAbs affect sperm in any fashion? There is a CGRP pregnancy registry that is being organized: is it coordinated among the various companies, and how does one access it?

2. Arthritis: could CGRP antagonism possibly help with rheumatoid or osteoarthritis? How about other pain syndromes, such as fibromyalgia, or peripheral neuropathy? What is the state of studies for these conditions? CGRP does show suppression of TNF alpha, through upregulation of other pathways. Is this clinically relevant? Arthritis patients have increased levels of CGRP in plasma and synovial fluid, and CGRP causes cytokine production in both rheumatoid arthritis and osteoarthritis. Could inhibiting CGRP help alleviate arthritis, or help in various pain syndromes?

3. CGRP levels may decline as one ages, although circulating levels may be increased in certain individuals. There may be a bimodal effect. Would the mAbs have more (or less) risk at age 70? At age 85 or 90? This population is at an increased risk for cardiovascular disease. Does this influence prescribing in the elderly? With declining stores of CGRP as one ages, the CGRP protective effect also (presumably) declines. Does this affect our prescribing after a certain age? Could eliminating some of the effects of CGRP actually help aging (there is some experimental evidence for this).

Amylin, Adrenomedullin, and CGRP Receptors

1. The AMY 1 receptor (and to a lesser degree the AMY 2 receptor), along with the ADM 1 and ADM 2 receptors, also have affinity for the CGRP ligand (although with lower specificity). Amylin, while mostly involved with glucose regulation, may be important in other functions (it is also located in the trigeminal ganglion and brainstem).

2. CGRP also functions at the amylin receptors; what is the result of blocking CGRP on the functioning of the amylin receptor. If one blocks the CGRP receptor, versus the ligand, is there a clinically relevant difference? By blocking the CGRP receptor, versus the ligand, CGRP may still attach to the amylin receptors. 

3. Adrenomedullin (ADM) competes with CGRP at the receptor site, and under certain conditions, ADM may actually compete with and displace CGRP from the receptor. If the CGRP antagonists affect the actions of ADM, what clinical effects might we see, over the long-term?

4. The ADM 1 and 2 receptors also have affinity for the CGRP ligand. What effects may occur from lowering CGRP, with regards to these other receptors?

5. Intermedin (IMD) is a peptide with affinity for this family of receptors. What effect may occur from blocking some of the IMD effects?

6. What are the effects, after blocking CGRP, on these other ligands and receptors with regard to the vasodilator effects?

7. The CGRP receptors are complex. CGRP 1 is primarily CLR and RAMP1. The pharmacology is complex, as the other peptides in the calcitonin family may attach to the CGRP receptor. CGRP may also activate the amylin receptor. With this overlapping pharmacology, what should we know about the effects of 1) knocking out the CGRP receptor, and 2) knocking out CGRP ligand?

Theoretical “CGRP Risk Scale”

Serious adverse events from CGRP antagonists have not yet occurred. When patients who have been prescribed these antagonists do suffer from a GI ulcer, a myocardial infarction, hypertension, or any number of conditions, the “cause and effect” may be difficult to determine. Clinicians should consider developing a “CGRP Risk Scale” as a basis for assessing risk going forward. A numerical scale of risk could easily be developed.

However, we want to determine risk first, including those patients who might be identified as low, medium, or high risk for the antagonist. Some of this work is beginning and as our knowledge increases over time, clinical risk assessment will be more accurate.

Miscellaneous Issues

1. The receptor occupancy of Aimovig is approximately 89%. The blocking of the CGRP ligand (by the other three mAbs under regulatory review: eptinezumab, ALD403, Alder Pharmaceuticals; fremanezumab,TEV-48125, Teva Pharmaceuticals; and galcanezumab, LY2951742, Eli Lilly) is approximately 85%. Do these occupancy levels steadily decline over the weeks/months, or is there a precipitous fall off at some point? Is this clinically relevant?

2. NGF influences CGRP. What clinical relevance, if any, does NGF have regarding the mAbs? Also, TRPV1 agonists may help to regulate CGRP; what is the importance of this?

3. Differences between the ligand antagonists (the three compounds in development noted above) and the receptor antagonist (Aimovig, on the market): receptors (that CGRP may attach to) other than the CGRP receptor may compensate for loss of the CGRP receptor; on the other hand, antibodies directed at the ligand of CGRP would also block the effects at the other (particularly AMY 1) receptors.  What is the clinical relevance of these differences between the ligand and the receptor antagonist?

4. With other meds (example: methysergide), we had patients take a “drug holiday” every 6 months. Does that make sense with these CGRP antagonists, at least until we are sure of long-term safety? Would doing this produce more antibodies, after re-introduction?  There would also be the risk that, after re-introduction, the mAb would not be as effective.

5. Informed consent: should we obtain this from patients (ideally, yes), and if so, what should be included in the informed consent?

6. Monitoring of adverse events: we should encourage reporting to the company or to the FDA. As with a pooled pregnancy registry, are there any plans for a pooled “mAb” adverse event registry? This would be a good idea. In addition, evaluation of other beneficial effects should be encouraged (such as the effect on other pain syndromes).

7. There have been a number of patients who have experienced moderate or severe fatigue/asthenia after the Aimovig injection. The fatigue and asthenia usually is short-lived. In addition, some patients have had muscle and/or joint pains. The hypothalamic-pituitary-adrenal(HPA) axis may be the culprit. The HPA axis is not protected from the CGRP mAbs by the blood-brain barrier. In addition, the hypothalamic-pituitary-thyroid axis may be involved as well. With low cortisol levels, muscle or joint pains may also occur. It will be helpful to have studies investigating various hypothalamic and pituitary hormones, particularly cortisol, in these patients.  Some patients have also experienced stroke-like symptoms as well. It would be helpful to investigate the etiology of these symptoms.


Sample Case Scenarios

In several years, we will have more information, regarding long-term safety and physiological effects of the CGRP antagonists. Short-term, these have been well tolerated. Until we know more, clinicians will have to decide which patients should not be given a CGRP antagonist using a combination of available evidence, clinical judgment, patient preference, and risk versus benefit. It is not fair to compare these new antagonists to “nothing:” patients are not on “nothing.” Other medications consumed by patients may have significant short and long-term adverse effects. Quality of life issues also play a role. Many migraineurs have significantly diminished quality of life due to poorly controlled headaches.

The following are sample scenarios where clinicians may or may not choose to prescribe the CGRP antagonists. The selections posed include the author's opinion alone.

CASE #1: Heather is 18 years old, with mild chronic migraine (CM) for 1 year. Heather recently sustained a fracture to her arm. Topiramate and amitriptyline have not been helpful, but the triptans do work abortively. Heather has not tried Botox. In this situation, I would suggest trying several other preventives, particularly Botox, and ARBs or beta-blockers. Petadolex may be worth prescribing. Pregnancy is always a concern in younger women, but the other medications also carry risks with pregnancy. mAbs would certainly be a consideration, but not until several others have been utilized. Her recent fracture of a bone plays a role in our decision, as CGRP is involved in bone healing. 

CASE #2: Eric is a 32-year-old man with severe chronic migraine, and a history of a gastric ulcer 4 years ago. He has IBS and a strong family history of inflammatory bowel disease (IBD). He is refractory to many preventives, including Botox. This is another difficult decision. The family history of IBD, and his (remote) history of an ulcer may increase (in theory) risk for the mAbs. While it may be better to first try other refractory approaches, this is in that “gray zone.”

CASE #3: Sally is a 63-year-old insulin-dependent diabetic with a history of angina. She has high cholesterol. Sally has failed 4 preventives, including Botox. This is a tough call; with DM and angina, the lowering of the CGRP vasodilatation (among other effects) may increase (in theory) the risk for mAbs. If her life is devastated by migraine attacks, and she is informed of possible risks, it may be reasonable to prescribe the antagonist. However, it would be better if we can avoid the mAbs, until more is known about cardiovascular effects in those with DM and coronary disease.

CASE #4: John is a 52-year-old with chronic migraine, and a history of mild DM Type 2. He has a mild peripheral neuropathy, and a cut on his big toe is very slow to heal. John has been on 3 preventives, which did not help. He has not tried Botox. There are some (theoretical) potential problems from the use of mAbs: possible increased risk for cardiac disease, and diminished wound healing. Utilizing a mAb would be reasonable, but if possible Botox may be a better choice in this situation.

CASE #5: Caitlin is a 39-year-old with hypothyroidism and an increased prolactin due to a small pituitary microadenoma. She has been refractory to many preventives, including Botox. Triptans work well for Caitlin. In view of the pituitary dysfunction, mAbs should be used with some caution (until we know more about the possible effects of diminishing CGRP on the pituitary hormones). If we do use a CGRP antagonist, I would suggest closely monitoring the hormonal levels. It is possible that we should evaluate hormone levels in most (or all) patients.


The CGRP antagonists for migraine prevention and certain chronic headache indications are potentially terrific options for patients with these conditions. Patients and the healthcare community at large desperately need improved treatments in this area. However, CGRP is involved in a multitude of physiological processes and we are only dealing with theoretical side effects at this time. It is the author's opinion that it is prudent to screen patients, using the limited knowledge available, before prescribing this new class of medications. Over the next 5 to 10 years, we will be in a better place to determine who is at risk for these antagonists and who may see life-changing benefit.

Last updated on: June 6, 2019
Continue Reading:
Special Report: CGRP Monoclonal Antibodies for Chronic Migraine
close X