Cardiac cephalalgia: one case with cortical hypoperfusion in headaches and literature review

We described the clinical features as well as the neuroimaging data of CC patients, and searched PubMed database using the terms “cardiac cephalalgia”, “cardiac cephalgia”, “headache and angina”, “headache and acute coronary syndrome”, and “headache and myocardial infarction”. The following limitations were exercised: full text, English language only, and published after 1997.

A 40-year-old male presented a 4-year history of episodic bitemporal headaches before he was seen for neurological consultation in outpatient. The headaches were rated as 7–10 in severity on the visual analog scale, pulsatile, tight in quality, and occasionally radiating to upper limbs. The headaches were sometimes also associated with chest discomfort, palpitations, cold sweating, and facial pallor. However, the patient denied nausea, vomiting, photophobia, and phonophobia.

The symptoms attack occurred 2–3 times per month, elicited by exertion, cold stimuli, and sexual activities, lasting 5–10 min, and relieved after treatment with nitrates. Coughing, sneezing, or having a bowel movement did not trigger the pain.

The patient self-administered aspirins and statins post-diagnosis of acute non-ST-elevation myocardial infarction (NSTEMI) in 2009 and nifedipine to control hypertension since 2001. Additionally, he presented 20 years of smoking history with 30 cigarettes/day but has ceased smoking for 5 years before the start of headache.

Physical examination revealed normal blood pressure, heart rate and rhythm, systemic and neurological examination results. The cardiac enzymes were in normal range at the time of headache attack. The estimation of catecholamines and their metabolites were normal. The ECG showed inverted T wave (Fig. 1).

The patient underwent brain MR examinations with routine clinical sequences including axial T1W, T2 FLAIR, diffusion-weighted imaging (DWI), and MRA (Magnetic Resonance Angiography) on a 3.0 T MR system (Discovery MR750, GE Healthcare, Milwaukee, WI, USA) equipped with an 8-channel head coil to receive signals. Perfusion-weighted images (PWI) were obtained using a 3D pCASL technique.

T1, T2, FLAIR and DWI weighted images of brain MRI were negative (Fig. 2a). However, the PWI revealed cerebral hypoperfusion during the headache attacks (Fig. 2b, c).

Owing to the headaches provoked by exertion, cold stimuli, and sexual activities and relieved after administration of nitrates, CC diagnosis should be suspected according to the international classification of headache disorder (ICHD-3β) in 2013 [2]. Thus, a coronary angiography was performed, which demonstrated complete occlusion at the middle segment of left anterior descending (LAD) and proximal right coronary arteries (RCA), 80% stenosis at the middle segment of left circumflex (LCX), and 80% stenosis at the bifurcation of the first diagonal (Fig. 3). Percutaneous transluminal coronary angiography (PTCA), stenting of LAD, and bifurcation of the first diagonal and the RCA were carried out successfully. Six months following the operation, the patient did not report any recurrence of a headache.

The demographic data and the clinical manifestations of the 35 cases were indicated in Tables 1 and 2. A male predominance was observed (male:female ratio 1.5:1). The mean age was 62-years-old.

The clinical manifestations of the headache were displayed in Table 1. In more than half of the patients, the headache could be trigged in conditions with high myocardial oxygen consumption, such as exertion, sexual activity, and emotional fluctuation. However, 6 cases have been reported; wherein the headache appeared during rest. The symptom was not localized in a specific area but involved frontal, temporal, parietal, and occipital regions. Moreover, the headache may be unilateral or bilateral. The quality of a headache was varied including shooting, bursting, dull, and squeezing. A maximum number of patients had a severe intensity of headache, and only two patients presented mild or moderate. In some cases, the headache had no accompanying symptoms while in others, it was accompanied by photophobia, phonophobia, osmophobia, and nausea or vomiting. Notably, all the reviewed cases of CC described the resolution of a headache after reinstating the flow in cardiac vessels by medical or surgical interventions.

The clinical manifestations of cardiac ischemia were illustrated in Table 2. About 1/2 of the CC patients (18 cases) were without typical angina symptoms, which included chest pain or tightness, palpitations, and dyspnea. Most of the patients showed pathological alterations of the baseline ECG trace, such as ST-segment elevations or depressions and T-wave inversions, as well as, elevated cardiac enzymes. However, three separate cases presented normal ECG at rest [3, 4], and one case presented a negative ECG even under stress [5]. Thirty-one out of 35 cases underwent coronary artery stenosis or occlusion. However, normal coronary angiography results could not exclude the diagnosis of CC as two cases continued to present normal coronary angiography [4, 6]. After intra-arterial injection of acetylcholine in 1 of the cases, coronary artery’s contraction was observed. These two cases were suggested as potential variants of angina.

In “typical” cases that were triggered by exertion or sexual activity or emotion fluctuation and accompanied by the symptoms of angina, a diagnosis of CC could be established according to the medical history demonstrating the exact onset of headache concurrently with acute myocardial ischemia, abnormality of ECGs performed at rest or under stress, elevating cardiac markers (CPK-MB, myoglobin, and troponin) and coronary angiography illustrating coronary arteries occlusion or stenosis. However, the majority of cases were “atypical”. Headaches may occur as the sole symptom without the symptoms of angina, absence of triggers, absence of ECG abnormalities (Table 2) when acute myocardial ischemia onset. In “atypical” cases, the doctors might be prone to omit the cardiac examinations (cardiac marker and coronary angiography) and make an incorrect diagnosis of benign headaches related to exertion, cough, migraine, and even orgasm. Especially, when some cases resembled migraine without aura, we should be able to make a diagnosis and prescribe triptans, which are vasoconstrictors aggravating the cardiac ischemic. However, there were more clinical clues implying the diagnosis of CC. We observed that most of the patients were above 50 years of age, and the majority of them presented cardiac risk factors such as hypertension, hyperlipidemia, diabetes, and smoking. Therefore, it is suggested that the patients suffering a new headache, who are over 50-years-old or display cardiac risk factors, should be suspected of CC. We also found that the headaches of all the patients reacted to vasodilators. Also, we demonstrated that the experimental treatment by vasodilators might be an efficient method in the event of difficulty in differentiating CC from other headaches. Moreover, even the patient showed a normal ECG; thus, the diagnosis of CC should not be excluded, because CC patients could have a normal ECG even under the stress test [3, 4]. Therefore, we suggested that coronary angiography should be recommended to all patients with suspected CC for accurate diagnosis.

Since the diagnosis had been introduced in 1997, several theories have been proposed about the pathogenesis of CC. Based on the current case, we proposed a new hypothesis about the mechanisms of CC vessel constriction hypothesis.

In the current case, the PWI of brain MRI confirmed the cerebral hypoperfusion; however, MRA was normal during the headache attack, which had not been reported. Thus, this phenomenon might lead to headache by two possible physiological mechanisms. Firstly, the hypoperfusion may suggest reversible microvessels constriction during headache. The reversible cerebral vasoconstriction syndrome (RCVS) has been classified with moderate to a severe headache, which was also accompanied by reversible vasoconstriction of the cerebral vasculature [2]. Since the reversible vessels constriction could be restored both in CC and RCVS, we speculated that they might share some common mechanisms. Several previous studies revealed that vasoconstriction of RCVS was associated with sympathetic over-activity based on clinical observations or hypotheses including central vascular tone changes, aberrant sympathetic response, pheochromocytoma, and autonomic dysreflexia [7‐10]. The mechanism of CC might also be linked to sympathetic hyperfunction due to the following reason: when myocardial ischemia occurred, the cardiac sympathetic afferents nerve could be stimulated. Hence, the activation of sympathetic afferents nerve could increase the sympathetic outflow through cardiac sympathetic nerve reflexes, which has been confirmed by several physiological tests [11‐13].

On the other hand, sympathetic hyperfunction and parasympathetic hypofunction were found in migraine patients [14, 15]. And the autonomic nervous system imbalance might be derived from the abnormal functional connections between the hypothalamus and other brain structures involved in autonomic function including the brain stem [16]. Thus, we speculated that the sympathetic hyperfunction in CC patients might also be associated with abnormal hypothalamic functional connectivity. Although the distinguishing characteristics of RCVS were attributed to constriction of medium- and large-sized arteries, the extent of headache and vasoconstriction in RCVS was asynchronous [17‐19]. This implicated that the headache might not be derived from medium- or large-sized arteries constriction. Ducros et al. demonstrated that headache could onset 1 week before vasoconstriction of large- and medium-sized arteries appeared, and in the same period the complications of cortical subarachnoid hemorrhage (cSAH), intracerebral hemorrhage (ICH), and reversible posterior leukoencephalopathy syndrome (RPLS) were observed. The ischemic events, including TIAs and cerebral infarction, often occurred after approximately 7 days. Based on the temporal pattern, their study illustrated that the underlying disturbance in the control of cerebral arterial tone first involved small distal arteries responsible for hemorrhages and RPLS and then progressed towards medium- and large-sized arteries responsible for ischemic events. Furthermore, the study also speculated that the headache of RCVS might primarily be due to the involvement of small distal arteries, with sudden changes in caliber (constriction or dilatation) that could stimulate perivascular pain-sensitive fibers [18, 20]. Thus, we inferred that myocardial ischemia might activate the sympathetic system, causing small intracranial arteries constriction and leading to a headache attack. Secondly, the cortical hypoperfusion might induce the occurrence of cortical spreading depression (CSD) in primary headaches, especially in a migraine [21]. Several experiments showed that CSD could contribute to the headache via activating meningeal nociceptors [22, 23] and activating or disinhibiting the second-order neurons in the trigeminocervical complex (TCC) [22, 24, 25]. Hence, according to the phenomenon of cerebral hypoperfusion, we proposed another possible mechanism: intracranial arteries constriction derived from myocardial ischemia lead to a CSD, which in turn caused a headache.

The field also proposed several other hypotheses to illustrate the mechanisms of CC. The first hypothesis was the convergence projection mechanism based on the fact that visceral afferent nerves from the heart and somatic afferent nerves from the head converged in the same spinal segments (C1 and C2 segments) belonging to TCC [26, 27]; when cardiac ischemic occurred, TCC was activated, leading to headache, similar to the reference pain of angina. The second hypothesis is hyper intracranial pressure mechanism: the sudden reduction of cardiac output associated with cardiac ischemia increased pressure in the left ventricle and in the right atrium, which might cause a decrease in venous return from the brain and subsequently an elevation of intracranial pressure, causing the headache [1, 4, 28, 29]. The third hypothesis is neurochemical mediator mechanism: during cardiac ischemia, several chemical mediators such as bradykinin, histamine, and substance P are released into the blood. These vasodilators could give rise to headache by dilation of the cerebrovasculature [1, 30]. However, these hypotheses were contrary to the fact that nitrates, also vasodilators, could relieve CC. Moreover, the PWI in the current case suggested microvessels constriction rather than vessel dilation during headache, which was also converse to this hypothesis.

We presented a CC patient with cortical hypoperfusion during headache attack, which had never been reported. Based on the phenomenon, we proposed vessel constriction hypothesis including two possible physiological mechanisms to explicate the pathophysiology of CC. Firstly, when myocardial ischemia occurred, the sympathetic system was activated causing small intracranial arteries constriction and leading to a headache, similar to the mechanisms of RCVS. Secondly, hypoperfusion might induce CSD that might contribute to the headache confirmed by several experiments.