Chapter 48 Raynaud’s Phenomenon
In its simplest form, local syncope is a condition perfectly compatible with health. Persons who are attacked with it are ordinarily females. Under the least stimulus, sometimes without appreciable cause, one or many fingers become pale and cold all at once; in many cases, it is the same finger that is always first attacked; the others become dead successively and in the same order. It is the phenomenon known as “dead finger.” The attack is indolent, the duration varies from a few minutes to many hours. The determining cause is often the impression of cold; but that which is only commonly produced under the influence of the most severe cold, appears in the subjects of whom I speak on the occasion of the least lowering of temperatures; sometimes even a simple mental emotion is enough … the skin of the affected parts assumes a dead white or sometimes a yellow colour; it appears completely exsanguine. The cutaneous sensibility becomes blunted, then annihilated; the fingers become like foreign bodies to the subject … the slight importance of this local abolition of the circulation is probably due to the fact that it is so transient … the attack is followed by a period of reaction, which is often very painful, and which gives place to a sensation quite analogous to that of being numbed by cold … and in the more pronounced cases, which the patients compared to tingling from cold, or to the stinging of nettles…Finally, a patch of deep red is formed on the extremities of the fingers. This patch gives place to the normal pink colour, and then the skin is found to have entirely returned to the primitive condition.
Episodic vasospastic ischemia of the digits was first described by Maurice Raynaud in the quotation above1 (Fig. 48-1). Raynaud’s phenomenon comprises sequential development of digital blanching, cyanosis, and rubor following cold exposure and subsequent rewarming2 (Fig. 48-2). Emotional stress also precipitates Raynaud’s phenomenon. The color changes are usually well demarcated and primarily confined to fingers or toes. Blanching, or pallor, occurs during the ischemic phase of the phenomenon and is secondary to digital vasospasm. During ischemia, arterioles, capillaries, and venules dilate. Cyanosis results from the deoxygenated blood in these vessels. Cold, numbness, or paresthesias of the digits often accompany the phases of pallor and cyanosis. With rewarming, digital vasospasm resolves, and blood flow dramatically increases into the dilated arterioles and capillaries. This “reactive hyperemia” imparts a bright red color to the digits. In addition to rubor and warmth, patients often experience a throbbing sensation during the hyperemic phase. Thereafter, the color of the digits gradually returns to normal. Although the triphasic color response is typical of Raynaud’s phenomenon, some patients may develop only pallor and cyanosis. Others may experience only cyanosis.
Figure 48-1 A patient with Raynaud’s phenomenon.
(From Raynaud M: Local asphyxia and symmetrical gangrene of the extremities, London, 1862, New Sydenham Society. Courtesy Boston Medical Library in the Francis A. Countway Library of Medicine.)
(From Creager MA: Raynaud’s phenomenon. Med Illus 2:84, 1983.)
The classification of Raynaud’s phenomenon is broadly separated into two categories: (1) the idiopathic variety, termed primary Raynaud’s phenomenon, and (2) the secondary variety, associated with other disease states or known causes of vasospasm (Box 48-1). Secondary causes of Raynaud’s phenomenon include collagen vascular diseases, arterial occlusive disease, thoracic outlet syndrome, several neurological disorders, blood dyscrasias, trauma, and several drugs.
Primary Raynaud’s phenomenon, or idiopathic episodic digital vasospasm, is the most common diagnosis of patients who present with Raynaud’s phenomenon.2 The diagnosis is based on criteria originally established by Allen and Brown,3 including (1) intermittent attacks of ischemic discoloration of the extremities, (2) absence of organic arterial occlusions, (3) bilateral distribution, (4) trophic changes—when present, limited to the skin and never consisting of gross gangrene, (5) absence of any symptoms or signs of systemic disease that might account for the occurrence of Raynaud’s phenomenon, and (6) symptom duration for 2 years or longer. If a normal erythrocyte sedimentation rate (ESR), normal nailfold capillary examination, and negative test for antinuclear antibodies (ANAs) are added to these criteria, the diagnosis is more secure.
Women are affected approximately five times more frequently than men. In one large study, the average age of onset of Raynaud’s phenomenon was 31 years; 78% of the patients were younger than 40 when symptoms began.4 Onset of symptoms in women may occur between menarche and menopause. Raynaud’s phenomenon is also known to occur in young children.5,6 Prevalence of primary Raynaud’s phenomenon varies with climate, with 4.6% of the population affected in warm climates, compared with 17% in cooler climates.6 There is a significant familial aggregation of primary Raynaud’s phenomenon. Approximately 26% of patients may know of one or more relatives who have the phenomenon, suggesting a genetic predisposition.7
In the vast majority of patients, the fingers are the initial sites of involvement.2 At first, blanching or cyanosis may involve only one or two fingers (Fig. 48-3). Later, color changes may develop in additional fingers, and symptoms occur bilaterally. In about 40% of patients, Raynaud’s phenomenon involves the toes as well as the fingers. Isolated Raynaud’s phenomenon of the toes occurs in only 1% to 2% of patients. Rarely, the ear lobes, tip of the nose, or tongue are affected.
Episodes of Raynaud’s phenomenon are usually precipitated by exposure to a cool environment or by direct exposure of the extremities to low temperatures. Some patients may experience Raynaud’s phenomenon during either cold exposure or emotional stress; infrequently, emotional stress may be the only precipitating factor. Duration, frequency, and severity of Raynaud’s phenomenon increase during cold months.
Several studies have correlated Raynaud’s phenomenon with migraine headaches and variant angina, suggesting a common mechanism for vasospasm.8–10 An association with vasospasm in the kidney,11 retina,12 and pulmonary13 vessels has also been described. Further evidence is the report of a family with three generations of systemic arterial vasospastic disease involving Raynaud’s phenomenon, variant angina, and migraine headaches.14 Differences in the responses of pharmacological intervention make the hypothesis of a common mechanism less appealing.15 Propranolol has been successfully used to prevent migraine headaches.16 In contrast, β-adrenoceptor blockers are not beneficial in variant angina and may cause Raynaud’s phenomenon.17,18 Similarly, nitrates are used for variant angina but are not beneficial in Raynaud’s phenomenon and often cause headaches. Ergot alkaloids are effective for treating migraine headaches but can cause coronary and digital vasospasm.19,20
Physical examination of patients with primary Raynaud’s phenomenon is often entirely normal. Sometimes the fingers and toes are cool and may perspire excessively. The pulse examination is normal; radial, ulnar, and pedal pulses should be easily palpable. Trophic changes such as sclerodactyly (thickening and tightening of the digital subcutaneous tissue) have been reported in up to 10% of patients, but these studies preceded nailfold capillaroscopy and ANA tests. The physical examination is most important to exclude secondary causes of Raynaud’s phenomenon.
Of all the forms of Raynaud’s phenomenon, primary Raynaud’s phenomenon has the most benign prognosis. In the group of patients identified by Gifford and Hines4 followed for a period of 1 to 32 years (average 12 years), 16% reported worsening of their symptoms, and 38%, 36%, and 10%, respectively, reported no change, improvement, or disappearance of symptoms. Sclerodactyly or trophic changes of the digits occurred in approximately 3% of patients during follow-up, and less than 1% of patients lost part of a digit. In some patients, scleroderma may develop after Raynaud’s phenomenon has been present as the only symptom for more than 20 years. Wollersheim et al.21 reported that measuring ANAs by immunofluorescence and immunoblotting in patients with Raynaud’s phenomenon had a positive predictive value of 65% and 71%, and a negative predictive value of 93% and 83%, respectively, for development of a connective tissue disease.
The precise cause of Raynaud’s phenomenon has not been clearly identified. It is quite likely that a variety of physiological and pathological conditions may contribute to or cause digital vasospasm2 (Box 48-2 and Fig. 48-4).
Figure 48-4 Pathophysiology of digital vasospasm.
Digital vasospasm may be due to vasoconstrictive stimuli, decreased intravascular pressure, or both. Mechanisms that contribute to exercise vasoconstriction include local vascular hypersensitivity to vasoactive stimuli (e.g., increased α-adrenoceptor sensitivity), sympathetic efferent activity, and local or circulating vasoactive hormones such as angiotensin II (Ang II), endothelin-1 (ET-1), serotonin, or thromboxane A2 (TxA 2). Low blood pressure, even in a healthy young person, may predispose to Raynaud’s phenomenon when the person encounters vasoconstrictive stimuli. Pathological conditions that may decrease intravascular pressure include arterial occlusion in proximal arteries (e.g., atherosclerosis), digital vascular occlusion (e.g., scleroderma), or hyperviscosity. TAO, thromboangiitis obliterans.
Normally, regulation of peripheral blood flow depends on several factors that include intrinsic vascular tone, sympathetic nervous system activity, hemorrheological properties such as blood viscosity, and various circulating hormonal substances. In contrast to other regional circulations that are supplied by both vasoconstrictor and vasodilator sympathetic fibers, the cutaneous vessels of the hands and feet are innervated only by sympathetic adrenergic vasoconstrictor fibers. In these vascular beds, neurogenic vasodilation occurs by withdrawal of a sympathetic stimulus. Cooling evokes reflex sympathetic-mediated vasoconstriction in the hands and feet via neurons originating in cutaneous receptors. Environmental cooling or cooling of specific body parts, such as the head, neck, or trunk, normally causes a reduction in digital blood flow. Local digital cooling also induces vasoconstriction, but digital vasoconstriction caused by local cooling is not mediated by the sympathetic nervous system. Thus digital vasoconstriction may be a physiological response to local cooling or to reflex activation of the sympathetic nervous system by environmental cold exposure or emotional stress.
Raynaud’s phenomenon is not a normal physiological response but rather an episode of digital artery vasospasm causing cessation of blood flow to the digits. The term vasospasm must be distinguished from vasoconstriction. Vasoconstriction may be defined as the expected reduction in vessel lumen size as a result of endogenous neural, hormonal, or metabolic factors that cause smooth muscle contraction. Vasospasm implies an excessive vasoconstrictor response to stimuli that would normally cause modest smooth muscle contraction, but that instead has resulted in obliteration of the vascular lumen. Patency of the digital artery depends on a favorable balance between the contractile forces of the muscular wall of the digital artery and its intraluminal pressure. Thus a situation in which there is excessive vasoconstrictive force or decreased intravascular pressure upsets this balance and results in vasospasm. It is with these rather simple concepts that several theories have been proposed to explain the episodic digital vasospasm that defines Raynaud’s phenomenon.
Several theories implicate excessive vasoconstrictive stimuli as a cause of Raynaud’s phenomenon. Postulated causes include local vascular hyperreactivity, increased sympathetic nervous system activity, elevated levels of vasoconstrictor hormones (e.g., angiotensin II (Ang II), serotonin, thromboxane A2(TxA2)), and exogenously administered agents such as ergot alkaloids and sympathomimetic drugs.
The observation that episodic digital vasospasm occurs during cold exposure has led several investigators to consider the possibility that Raynaud’s phenomenon occurs as a result of a local vascular hyperreactivity. In 1929, Sir Thomas Lewis observed that following exposure of the finger to cold, vasospasm could be produced even after nerve blockade or sympathectomy.22 These experiments were repeated and confirmed 60 years later.23 Therefore, the vasospastic response of the Raynaud’s phenomenon may occur in the absence of efferent digital nerves. The possibility of local vascular hyperreactivity was examined by Jamieson et al.24 They compared the magnitude of reflex vasoconstriction in each hand following application of ice to the neck while one hand was kept at 26 °C and the other at 36 °C.9 At 36 °C, the reflex vasoconstrictor response was comparable in normal subjects and patients with primary Raynaud’s phenomenon. In the hand cooled to 26 °C, however, reflex vasoconstriction was exaggerated in patients with Raynaud’s phenomenon. This response led these investigators to hypothesize that digital α1 adrenoceptors were sensitized by cold exposure.
A series of studies by Vanhoutte et al.25 have supported the hypothesis that cooling potentiates the vascular response to sympathetic nerve activation. Vasoconstriction, in response to exogenous norepinephrine, also is increased by cooling. Augmentation of adrenergic-mediated vasoconstriction by cooling occurs despite generalized depression of contractile machinery and diminished release of norepinephrine from sympathetic nerve endings in the vessel wall. The most likely hypothesis is that cold causes changes at the level of the adrenoceptor, such as an increase in the affinity for norepinephrine or greater efficacy of the agonist/receptor complex. Vanhoutte et al.25 have reported that α2 adrenoceptors are more sensitive than α1 adrenoceptors to temperature change. Whereas cooling slightly depresses α1 adrenergic–mediated vasoconstriction, it markedly augments α2 adrenergic–mediated responses. Conversely, warming augments α1-adrenergic vasoconstriction and depresses α2-adrenergic vasoconstriction.26
These experimental observations may have important implications regarding the pathophysiology of Raynaud’s phenomenon. Flavahan et al.27 examined the distribution of α1 and α2 adrenoceptors in arterial tissue from amputated limbs of patients who did not have vascular disease. They reported that α2 adrenoceptors were more prominent in digital arteries. Chotani et al.28 found that human dermal arterioles selectively expressed α2C adrenoceptors. Jeyaraj et al.29 observed that cooling redistributed α2C adrenoceptors from the Golgi to the plasma membrane in human embryonic kidney cells. It is therefore an intriguing observation by Keenan and Porter that the density of α2 adrenoreceptors is increased in platelets from patients with Raynaud’s disease.30
In support of these findings, Coffman and Cohen reported that α2 adrenoceptors were more important than α1 adrenoceptors in mediating sympathetic nerve–induced vasoconstriction in the fingers.31 They administered the α1-antagonist prazosin and the α2-antagonist yohimbine to patients with Raynaud’s phenomenon during reflex sympathetic vasoconstriction caused by body cooling. Whereas prazosin caused no significant change in finger blood flow or finger vascular resistance, yohimbine significantly increased finger blood flow and decreased finger vascular resistance. This study confirmed that postjunctional α2 adrenoceptors are present in human digits and strongly suggested that these receptors contribute to digital vasoconstriction during environmental cooling in patients with Raynaud’s phenomenon.
Thereafter, Coffman and Cohen demonstrated that compared to normal subjects, patients with Raynaud’s phenomenon were hypersensitive to the vasoconstrictor effects of clonidine, an α2-adrenoceptor agonist, but not to phenylephrine, an α1-adrenoceptor agonist.31 Cooke et al.32 found that both α1– and α2-adrenoceptor antagonists induced digital vasodilation in patients with acute Raynaud’s phenomenon, yet did not inhibit digital vasoconstriction caused by local digital cooling. Although still speculative, these studies suggest that episodic digital vasospasm may be secondary to a predominance of postjunctional α2 adrenoceptors in digits of patients with primary Raynaud’s phenomenon.
Although appealing as a potential mechanism for digital vasospasm, the concept of exaggerated reflex sympathetic vasoconstrictor responses to cold environment has not been convincingly demonstrated. Increased concentrations of epinephrine and norepinephrine in peripheral venous blood at the wrist were found to be higher in patients with primary Raynaud’s phenomenon than in normal subjects by one investigator,33 but others found normal local levels of norepinephrine in brachial arterial and venous blood samples.34 The latter group of investigators reported that the reflex vasoconstrictor response of the hand to a cold stimulus in affected patients is similar to that in a control group, and there were comparable vasoconstrictor responses to the intraarterial infusion of tyramine, a drug that causes vasoconstriction by releasing norepinephrine from sympathetic nerve terminals. Central thermoregulatory control of skin temperature has also been reported to be comparable in normal individuals and patients with primary Raynaud’s phenomenon.35 Finally, microelectrode recordings of skin sympathetic nerve activity do not demonstrate an abnormality in patients with primary Raynaud’s phenomenon.36 There was no hypersensitivity of the vessels to strong sympathetic stimuli or abnormal increase in sympathetic outflow.
Raynaud’s phenomenon is observed frequently in individuals treated with β-adrenoceptor antagonists.37–39 It may be inferred from this observation that β-adrenergic vasodilation normally attenuates digital vasoconstrictor tone. Cohen and Coffman40 examined the effect of isoproterenol and propranolol on fingertip blood flow after vasoconstriction had been induced by a brachial artery infusion of norepinephrine or angiotensin, or reflexly by environmental cooling. Intraarterial isoproterenol administration increased fingertip blood flow during infusions of norepinephrine and angiotensin, but not during reflex sympathetic vasoconstriction. Conversely, propranolol served to potentiate vasoconstriction caused by intraarterial norepinephrine, but not that caused by reflex sympathetic vasoconstriction. These investigators concluded that a β-adrenergic vasodilator mechanism may be active in human digits, but does not modulate sympathetic vasoconstriction. There is no evidence to support the contention that decreased sensitivity or number of β adrenoceptors contributes to the pathophysiology of Raynaud’s phenomenon in the absence of pharmacological blockade of β adrenoceptors.
Various neurotransmitters, hormones, and platelet release byproducts are capable of constricting vascular smooth muscle and causing digital vasoconstriction. These include Ang II, serotonin, TxA2, and endothelin-1 (ET-1). It would be difficult to attribute all causes of Raynaud’s phenomenon to excessive levels of these vasoconstrictor agents, but in some secondary causes of Raynaud’s phenomenon, any one of them might contribute to vasoconstriction.
Serotonin (5-hydroxytryptamine [5-HT]) is a neurotransmitter that is synthesized and released by selective neurons and enterochromaffin cells. Serotonin can cause vasoconstriction by directly activating serotoninergic receptors on the smooth muscle cells (SMCs). Vasoconstriction may also be caused by direct activation of α adrenoceptors on SMCs or indirectly by facilitating release of norepinephrine from adrenergic nerve terminals. Although some evidence implicates a role for serotonin in the pathophysiology of Raynaud’s phenomenon, the contribution of serotonin to digital vasospasm remains speculative.
The possibility that vasoconstrictors released during platelet aggregation may be pertinent to the pathophysiology of Raynaud’s phenomenon has been further evaluated by studies that have either measured levels of TxA2 or administered a thromboxane synthetase inhibitor.41,42 Coffman and Rasmussen compared the thromboxane synthetase inhibitor dazoxiben to placebo in patients with either primary or secondary Raynaud’s phenomenon.41 Dazoxiben did not affect total fingertip blood flow or fingertip capillary blood flow, whether measured in a warm (28.3 °C) or cool (20 °C) environment. With chronic treatment, there was a small decrease in frequency of vasospastic episodes in patients with primary Raynaud’s phenomenon. To date, however, there is insufficient evidence to support a role for TxA2 in digital vasospasm.
Plasma concentration of the potent vasoconstrictor Ang II is rarely elevated in patients with Raynaud’s phenomenon. This hormone is therefore unlikely to contribute to the pathophysiology of digital vasospasm in most patients.
Endothelin-1 is an endothelium-derived, powerful, and prolonged-acting vasoconstrictor agent suggested to play a part in the pathogenesis of Raynaud’s phenomenon. It rises in response to a cold pressor test and constricts cutaneous blood vessels.43 Studies measuring ET-1 in primary or secondary Raynaud’s phenomenon have been conflicting.44 Controlled clinical trials of ET-1 receptor antagonism in the treatment of Raynaud’s disease have achieved little success.45,46 It is therefore doubtful that it plays a role in Raynaud’s phenomenon.
Patency of a blood vessel requires balance between arterial wall tension (favoring closure of the vessel) and intravascular distending pressure. Landis measured intravascular pressure in patients with Raynaud’s phenomenon by introducing a micropipette into a large digital capillary.47 During cyanosis, capillary pressure fell to approximately 5 mmHg, and flow ceased. These findings suggested that the site of closure was proximal to the capillaries at the arterial level. Interestingly, Thulesius reported that brachial artery blood pressure in patients with primary Raynaud’s phenomenon was significantly lower than that in a normal control population.48 Cohen and Coffman also found that blood pressure was lower in patients with primary Raynaud’s phenomenon compared with normal subjects.49 In addition to lower brachial blood pressure, systolic blood pressure (SBP) measured at the proximal and distal digital arteries averaged 18 mmHg less than that in normal digits.
A low digital artery pressure may occur in various disorders associated with Raynaud’s phenomenon, such as large-vessel arterial occlusive disease secondary to atherosclerosis, embolism, or thoracic outlet syndrome. When extrinsic vasoconstrictor force is applied, these vessels may collapse and cause digital ischemia. Distal vascular occlusions secondary to thromboangiitis obliterans (TAO), vasculitis, or vibration injury may also reduce digital arterial pressure distal to the diseased vascular segment.
Hyperviscosity may reduce blood flow velocity in digital vessels, leading to a decrease in intravascular pressure. Indeed, Raynaud’s phenomenon occurs in patients with hyperviscosity due to polycythemia vera or Waldenström macroglobulinemia.50,51 In patients with Raynaud’s phenomenon secondary to disorders such as cryoglobulinemia and cold agglutinin disease, hyperviscosity caused by cooling may contribute to digital vasospasm.52–54 Indeed, cooling has been shown to abolish hand blood flow in patients with cold agglutinins, possibly because the vessels become occluded by agglutinated red cells.54 Data invoking hyperviscosity as a cause of Raynaud’s phenomenon in patients who do not have an established blood dyscrasia, however, are less compelling.
The secondary causes of Raynaud’s phenomenon include collagen vascular diseases, arterial occlusive disorders, thoracic outlet syndrome, several neurological disorders, blood dyscrasias, trauma, and several drugs (see Box 48-1).
Raynaud’s phenomenon occurs in 80% to 90% of patients with systemic sclerosis; it may be the presenting symptom in approximately 33% of patients. In some patients, scleroderma may develop after Raynaud’s phenomenon has been present as the only symptom for many years. Frequency and severity of Raynaud’s phenomenon in patients with systemic sclerosis is often worse than that observed in patients with primary Raynaud’s phenomenon. The incidence of digital ulceration and gangrene is increased, possibly leading to amputation. Diagnosis of systemic sclerosis is suggested by the appearance of typical sclerotic skin changes. These include tightness, thickening, and nonpitting induration involving the extremities, face, neck, or trunk. When present in the digits, these abnormalities produce changes in the contour of the fingers and toes, referred to as sclerodactyly. Other manifestations of systemic sclerosis include pitting scars of the tips of the digits, normal skin pigmentation, and telangiectasia. Visceral manifestations include pulmonary fibrosis, esophageal dysmotility, and colonic sacculation. The kidney and heart may also be involved. As the disease progresses, skin and subcutaneous tissue of the fingers become stiffer, joints become immobile, and contractures develop. A variant of systemic sclerosis is the CREST syndrome, a form of limited scleroderma that includes calcinosis, Raynaud’s phenomenon, esophageal dysmotility, sclerodactyly, and telangiectasia in the absence of internal organ involvement.