To assess the early signs, risk factors, and rate of transition from primary Raynaud's phenomenon (primary RP) to secondary RP.
To assess the early signs, risk factors, and rate of transition from primary Raynaud's phenomenon (primary RP) to secondary RP.
A clinical sample of 307 consecutive patients with RP was included in a prospective followup study. After an initial screening, 244 patients were classified as having primary RP, of whom 236 were followed up for a mean ± SD of 11.2 ± 3.9 years. Patients classified according to the screening as having suspected secondary RP underwent an extended screening program annually until transition to secondary RP occurred.
The initial prevalence of secondary RP was 11%. The annual incidence of transition to suspected secondary RP was 2%, and the annual incidence of transition to secondary RP was 1%. Overall, 46 patients were classified as having suspected secondary RP, and 23 of these later were classified as having secondary RP. Older age at onset of RP (hazard ratio 2.59, 95% confidence interval [95% CI] 1.40–4.80), shorter duration of RP at enrollment (hazard ratio 0.87, 95% CI 0.81–0.94), and abnormal findings on thoracic outlet test (hazard ratio 2.69, 95% CI 1.12–6.48) were associated with an increased risk for transition to secondary RP. Compared with patients with suspected secondary RP, those diagnosed as having secondary RP had a higher number and earlier occurrence of pathologic findings. Furthermore, antinuclear antibodies at a titer of ≥1:320 and positive findings in specific serologic subsets were associated with a significantly increased risk for developing a connective tissue disease.
Patients diagnosed initially as having primary RP may actually comprise 1 of 3 groups: those with idiopathic RP, those with a rather benign disease course, and those with a more severe course of the disease.
Raynaud's phenomenon (RP) is defined as bouts of reversible vasospastic ischemia of the fingers and toes; it was first described by Maurice Raynaud in 1862 (1). It is typically manifested by an initial white discoloration (pallor) of the digits as a reaction to cold, which leads to cyanosis, pain, and numbness, followed by postischemic red flush upon rewarming. The etiology as well as the pathophysiology of the disease have not been completely established (2–4). Differentiation between primary RP and secondary RP does not reflect a diagnosis in the strict sense, but rather a description of the current findings in an ongoing screening process. Classifying a patient as having primary RP means that, up to that point in time, no underlying disease has been identified. If findings suggest an underlying disease, the patient's condition would be classified as suspected secondary RP, while the condition would be classified as secondary RP if an underlying disease can be definitively established.
Primary RP is a functional vascular disorder. When and why it transforms into a structural disease is largely unpredictable. The diagnosis of secondary RP includes association with a wide spectrum of different diseases of vascular and nonvascular origin, ranging from peripheral arterial diseases and connective tissue diseases to paraneoplastic disorders (2).
Reports on the rate of transition from primary RP to secondary RP (identified by diagnosis of an associated disease) vary widely (5). The incidence as well as the prevalence of secondary RP also seem to depend on the origin of the patient sample (5–15). Despite the inconsistencies, the rate of this transition is high enough to make followup studies of clinical interest.
Several risk factors for transition to secondary RP have been implicated (5, 16), but there is still a lack of adequate prognostic evidence regarding the course of the disease. Early prognosis is important since secondary RP may be due to a severe underlying disease that may even be associated with reduced life expectancy. More than 90% of patients with scleroderma present with RP as an early symptom. For patients with other types of connective tissue diseases, this proportion ranges between 40% for systemic lupus erythematosus (SLE) and 85% for mixed connective tissue disease (MCTD) (17, 18). RP is not uncommon in the general population, with only more severe cases reaching clinical attention (9).
This study was designed to address the following questions: What is the incidence of transition from primary RP to suspected secondary RP and to definite secondary RP in a clinic-based sample without preselection except for a verified diagnosis of RP and an initial lack of identification of an underlying disease? Are there risk factors for transition to secondary RP that are already present at the initial screening? Do patients with RP have features that can be identified initially or during followup that will determine the further progression of the disease?
Sample size was determined by setting the minimal number of critical events (transitions to secondary RP) at 50, of which 20 were assumed to occur during followup and 30 in patients already initially classified as having suspected secondary RP. Based on past experience about accrual of patients with RP (∼60 patients per year) and an estimated overall incidence density of transition to secondary RP of 32 per 1,000 patient-years (5), a sampling duration of 5 years and a followup period of 10 years were established.
In the time between 1990 and 1994, a total of 307 patients were diagnosed as having RP at the outpatient angiology unit of Hanusch Hospital, Vienna. Diagnosis of RP was established based on the presence of typical clinical symptoms, including discoloration of the skin of fingers and/or toes induced by cold, wet, or stressful conditions (with or without subsequent cyanosis and/or hyperemia), substantiated by a hot–cold provocation test. This test consisted of oscillography and temperature measurements of all digits spontaneously and after 5 minutes of warming in water of 37°C and subsequent exposure to cold water maintained at 12°C for 5 minutes. A prolonged (>20 minutes) time to reach the initial temperature established the diagnosis.
After an extensive patient history was obtained, all patients with RP underwent initial diagnosis for differentiation between primary and secondary RP. This procedure had already been applied in an earlier study (19) and was continued in this investigation in only slightly modified form as detailed below. The procedure consists of the segments described below (with criteria for establishment of abnormality shown in parentheses).
Clinical examination consisted of pulse palpation in the upper limb (pathologic findings in typical location); the Allen test (delayed or reduced intensity of red coloration of hands after release of blood flow); and inspection of hands and feet for signs of connective tissue disease (thickening, tightening, induration, sclerosis, sclerodactyly, digital pitting scars, loss of digital pad tissue, and necrosis of fingertips).
Angiologic studies included Doppler ultrasonography with blood pressure measurements of brachial, radial, and ulnar arteries (pathologic Doppler shift curve, differences of >30 mm Hg between the 2 sides, Doppler index <1); electronically amplified acral oscillography of all fingers on both hands in resting position (pathologic resting oscillography and/or vasospasms persisting for >5 minutes after dosing with 0.8 mg nitroglycerin lingual spray [G. Pohl-Boskamp, Hohenlochstedt, Germany]); acral oscillography during 1) the Adson test, in which the patient's head is kept in dorsal flexion and simultaneously rotated toward the affected side, 2) the hyperabduction or “costoclavicular” test, in which the patient is asked to take a full breath and to take back the shoulders with simultaneous hyperabduction of the arm flexed at right angle, or 3) the adduction or “pectoralis” test, with forced adduction of the arms raised upright over the head against an enclosed resistance (classified as an abnormal result on a thoracic outlet test if there is a significant reduction of the oscillographic amplitude in at least 1 of the preceding 3 maneuvers); and vital microscopy of nailfold capillaries (classification of pathologic findings according to the semiquantitative grading proposed by Maricq ).
Radiographic evaluation consisted of chest radiography in posteroanterior and lateral projection (pleural effusion, pericardial effusion, cardiac dilatation, speckled/striped infiltration, pulmonary fibrosis, rheumatoid nodules, atelectasis) and radiography of both hands in dorsovolar and zither player projection (diffuse soft tissue atrophy, swan-neck deformity, changes in soft tissue of fingertips, subcutaneous calcifications, bony erosion, etc.).
Laboratory evaluation included measurement of the erythrocyte sedimentation rate (ESR) according to the Westergren method (reproducible pathologic findings in 2 blood samples obtained within a period of 3 weeks), and red and white blood cell counts with differential cell counts. Renal function was assessed by measurement of blood urea nitrogen, creatinine, and urinary protein. Protein electrophoresis was performed using the quantitative immunoglobulin assay. Serologic tests for rheumatic diseases included testing for levels of cryoglobulins, antistreptolysin O, C-reactive protein, rheumatoid factor (by Rose-Waaler test), antinuclear antibodies (ANAs), and anticentromere antibodies with subset specification upon positive findings specified below.
If no abnormalities were detected in this screening, the patient was classified as having primary RP, i.e., RP with no currently detectable underlying disease. If at least 1 finding was positive, the patient was diagnosed as having suspected secondary RP and an extended screening program was initiated. This extended program comprised the following components: 24-hour urine sampling and analysis of proteinuria for 3 consecutive days, renal ultrasonography, esophageal radiography, scintigraphy of the parotid gland, transthoracic echocardiography, ophthalmoscopy, and Schirmer's test. Depending on the findings of the screening and the extended screening program, additional tests were carried out to establish a diagnosis (e.g., for suspected Hashimoto thyroiditis). Several criteria have been used to differentiate between suspected secondary RP and secondary RP (9, 11, 13, 15). Here we applied the same method that proved useful in a previous study (19).
Figure 1 shows a flow chart of the procedure. Patients diagnosed as having primary RP underwent screening at annual intervals. If, at any time, a patient was classified as having suspected secondary RP, the patient underwent extended screening and if no underlying disease could definitely be established the patient remained classified with this status, with annual screening plus extended screening until transition to secondary RP or disappearance of the abnormalities that led to a diagnosis of suspected secondary RP. In the latter case, the patient was reclassified as having primary RP.
The same serologic methods were applied initially and during followup. ANAs were measured using a test kit (Immunoconcepts, Sacramento, CA) based on indirect immunofluorescence and serial dilution. Results were provided in titers ranging from 1:20 to 1:2,560. ANA differentiation was assessed with a line immunoassay (Innogenetics, Ghent, Belgium).
Time of transition to secondary RP and suspected secondary RP was analyzed by Kaplan-Meier estimates of the cumulative proportion of patients who remained classified as having primary RP. Prognostic factors (initial parameters as well as time-contingent covariates) were analyzed by the Cox proportional hazards model. Results were expressed as hazard ratios, relating the hazard of transition to secondary RP during the next time interval to the baseline hazard. Applicability of the proportional hazards model was tested by including an interaction term between the predictor and the diagnosis time. In addition, comparisons between diagnostic groups were performed by nonparametric methods: Mann-Whitney U test, Kruskal-Wallis test, or chi-square test, as appropriate. P values less than 0.05 were considered significant.
Of 307 consecutive patients with RP, 244 were classified as having primary RP at the initial screening, while 63 patients initially presented with pathologies sufficient for diagnosis of suspected secondary RP. Of these 63 patients, 34 were classified as having secondary RP following the extended screening program (initial prevalence 11%). Diagnoses in these cases were thromboangiitis (7 patients), rheumatoid arthritis (RA) (6 patients), peripheral arterial occlusive disease (5 patients), scleroderma (5 patients), undifferentiated connective tissue disease (UCTD) (4 patients), MCTD (3 patients), SLE (2 patients), and Sjögren's syndrome (2 patients). During the followup period, another 12 of the remaining 29 patients were diagnosed as having secondary RP, with diagnoses of scleroderma (4 patients), MCTD (3 patients), RA (2 patients), UCTD (1 patient), SLE (1 patient), and Sjögren's syndrome (1 patient). All 63 patients with initial suspected secondary RP were excluded from the primary RP followup study.
Of the 244 patients initially classified as having primary RP, 2 were misclassified (no RP) and 6 did not undergo the full screening program and were excluded, leaving a sample of 236 patients with primary RP for the followup. The patients' mean ± SD age was 40.2 ± 13.9 years. One hundred eighty of the patients were female (76%) and 56 were male (24%). The mean ± SD duration of followup was 11.2 ± 3.9 years.
Table 1 shows the patients' characteristics at enrollment, with stratification according to diagnosis during followup. Patients who developed suspected secondary RP were older, and fewer had a history of smoking. In 8 patients, RP disappeared and did not reappear after 2 cold seasons. In these patients, results of the thoracic outlet test were less frequently abnormal at enrollment and the incidence of spontaneous vasospasms in the initial oscillography was lower. Patients diagnosed as having secondary RP during followup had symptoms of RP for a significantly shorter duration (an average of ∼5 years) at the initial screening, and the majority of them had abnormal thoracic test results.
|Primary RP (n = 182)||Suspected secondary RP (n = 23)||Secondary RP (n = 23)||Total (n = 236)†|
|Age at enrollment, mean ± SD years||38.0 ± 13.8||50.3 ± 14.8‡||44.1 ± 11.1||40.2 ± 13.9|
|Age at onset of RP, mean ± SD years||25.6 ± 10.5‡||38.6 ± 14.7||39.4 ± 10.9||28.5 ± 12.3|
|Duration of RP, mean ± SD years||12.4 ± 9.3||11.7 ± 11.5||5.3 ± 6.9‡||11.6 ± 9.5|
|Duration of RP at diagnosis, mean ± SD years||–||19.7 ± 12.1||11.9 ± 6.8§||16.1 ± 10.5|
|BMI, mean ± SD kg/m2||23.6 ± 3.4||23.3 ± 3.1||24.3 ± 3.2||23.6 ± 3.4|
|Systolic BP, mean ± SEM mm Hg||121 ± 19||130 ± 17||133 ± 22||123 ± 19|
|Diastolic BP, mean ± SEM mm Hg||72 ± 11||74 ± 9||75 ± 10||72 ± 11|
|Symptoms triggered by cold, %||88.5||91.3||91.3||89.0|
|Symptoms triggered by wetness, %||41.8||34.8||65.2||42.8|
|Symptoms triggered by stress, %||29.7||26.1||52.2||31.4|
|Abnormal thoracic outlet test result, %||40.7||47.8||60.9§||42.4|
|Spontaneous spasms, %||75.1||65.2||73.9||72.5|
Table 2 shows the results of the Cox proportional hazards model. Older age at the onset of RP, shorter duration of RP at enrollment, and abnormal thoracic outlet test results at the initial screening were associated with an increased hazard for transition to secondary RP.
|Covariable||Hazard ratio (95% confidence interval)||P|
|At initiation of followup|
|Age, decades||1.10 (0.82–1.47)||0.533|
|Age at onset of RP, decades||2.59 (1.40–4.80)||0.003|
|Duration of RP||0.87 (0.81–0.94)||<0.001|
|BP, systolic||1.02 (0.99–1.05)||0.071|
|BP, diastolic||0.99 (0.94–1.04)||0.722|
|Alcohol consumption||0.40 (0.05–3.03)||0.375|
|Symptoms triggered by cold||0.97 (0.20–4.60)||0.968|
|Other triggers||2.18 (0.84–5.72)||0.109|
|Spontaneous spasms||0.68 (0.25–1.87)||0.449|
|Abnormal thoracic outlet test||2.69 (1.12–6.48)||0.026|
|Capillary microscopy||11.54 (4.17–31.95)||<0.001|
|Chest radiography||44.21 (11.29–173.16)||<0.001|
|Hand radiography||14.99 (6.02–37.32)||<0.001|
|Differential cell count||16.58 (6.91–39.75)||<0.001|
|Renal function||12.81 (5.22–31.44)||<0.001|
|Protein electrophoresis||26.26 (11.06–62.31)||<0.001|
|Rheumatoid factor||15.76 (6.56–37.86)||<0.001|
|ANA titer >1:160||68.22 (16.64–279.70)||<0.001|
Figure 2 shows Kaplan-Meier curve estimates of the cumulative proportion of patients who still had primary RP, as well as those whose disease transitioned from suspected secondary RP to definite secondary RP during followup. The annual incidence of transition to suspected secondary RP was 2%, and the annual incidence of transition to secondary RP was 1%. Overall, 46 patients were classified as having suspected secondary RP, and 23 of these later were classified as having definite secondary RP. The median duration until establishment of a specific concomitant diagnosis after transition to suspected secondary RP was 5 years.
Twenty of these 23 patients with suspected secondary RP were women. The mean ± SD age at onset of symptoms in the 23 patients was 38.6 ± 14.7 years. At the time of enrollment, they had already had RP for a mean ± SD of 11.7 ± 11.5 years. Table 3 shows the number of patients with specific positive findings during the followup screenings in the 23 patients with suspected secondary RP. Twenty-three of the patients had an elevated ESR (100%), 14 had ANAs (61%), 11 had an abnormal differential cell count (48%), and 7 had abnormal renal function (30%). All other findings were seen in <5 patients. None of the patients developed cryoglobulinemia, and none had abnormal findings on chest or hand radiographs. ANA titers were mostly <1:80. A titer of 1:160 was found in only 2 patients, and only 1 patient was positive for specific serologic subsets.
|Variable||Suspected secondary RP (n = 23)||Secondary RP (n = 23)||Suspected secondary RP + secondary RP (n = 46)|
|Capillary microscopy||4 (2)||4 (2)||8 (4)|
|Hand radiography||0||8 (1)†||8 (1)|
|ESR||23 (17)||21 (14)||44 (31)|
|Differential cell count||11 (2)||11 (1)||22 (3)|
|ANA||14 (2)||18 (4)||32 (6)|
|No. of findings, mean ± SEM||2.9 ± 1.0||4.6 ± 2.1†||3.8 ± 1.6|
|No. of findings per year, mean ± SEM||0.34 ± 0.13||0.75 ± 0.52†||0.54 ± 0.38|
|Years until earliest findings, mean ± SEM||7.5 ± 1.9||6.2 ± 2.6‡||6.9 ± 2.3|
During followup, 23 patients were diagnosed as having definite secondary RP. Underlying diseases detected during the extended screening program, diagnosed using previously described criteria when applicable (21–25), were MCTD (5 patients), scleroderma (5 patients), RA (4 patients), Sjögren's syndrome (2 patients), SLE (2 patients), UCTD (1 patient), thyroiditis (1 patient), paraproteinemia (1 patient), autoimmune anemia (1 patient), and embolism (1 patient). Twenty-one of the patients were women. The mean ± SD age at onset of symptoms in the 23 patients was 39.4 ± 10.9 years, and they had had RP symptoms for 5.3 ± 6.9 years at the time of enrollment. Table 3 shows the number of patients with specific positive findings during the followup screenings. Twenty-one patients had an elevated ESR (91%), 18 had ANAs (78%), 12 had an abnormal protein electrophoresis result (52%), 11 had an abnormal differential cell count (48%), 10 were rheumatoid factor positive (43%), 8 had immunoglobulins (35%), 8 had abnormal renal function (35%), and 8 had abnormalities detected on radiographs of the hands (35%). All other findings were seen in <5 cases. ANA titers were ≥1:160, with a median titer of 1:450. Fifteen of 18 patients were also positive for specific serologic subsets.
Results of the Cox proportional hazards model, summarized in Table 2, indicate that all parameters tested in the screening program were associated with a significantly increased risk of developing secondary RP during the following years. ANA titers >1:160 were associated with the highest risk of subsequent diagnosis of secondary RP.
The present investigation in a clinical sample of outpatients with RP demonstrates the suitability and usefulness of systematic diagnostic screening and thorough followup. With such an approach, ∼20% of patients initially diagnosed as having primary RP were found to have transitioned to either suspected secondary RP or secondary RP during the followup period of >10 years.
It should, however, be stressed that the observed incidence of transition to secondary RP does not directly generalize to all patients with primary RP. Patients seen in a clinical setting are probably a specific group that cannot, based on the evidence available to date, be precisely characterized within the total primary RP population in a community. To estimate the prevalence and incidence of RP in the general population, the Framingham offspring study (26) may be used. Interestingly, only approximately one-third of those with RP had persistent symptoms and only a small fraction, approximately one-tenth of those with persistent symptoms, showed increasing severity. Hence the annual incidence of severe persisting RP is ∼90 cases among a population of 100,000.
Little is known about whether these patients seek consultation with a physician; however, due to the severity of symptoms, it can be assumed that most do. In a population-based study in England (27), 2% of patients with RP had consulted a doctor. If this proportion is similar in our area, it would mean that the RP patients referred to our clinic comprise almost all RP patients who have consulted a physician. Sources of referral were as follows: 63% from a general practitioner, 12% from specialists of internal medicine, dermatology, or orthopedics, and 10% from other departments of our hospital; 15% presented without a referral. Since our sample is composed of patients with initially moderate to severe symptoms, with the majority referred by their general practitioner, we assume that the incidence of transition to suspected secondary RP and secondary RP is ∼10 times higher than would be the case in a sample of patients with primary RP with persistent symptoms from the general population.
In ∼3% of patients enrolled in the study, the symptoms of RP resolved. Although this is consistent with clinical experience, there are no recent data on the frequency of remission in clinically based samples. In a retrospective study in 1957, including adolescent patients, spontaneous remission occurred in 10% of the patients with RP (28). The somewhat higher proportion as compared with our study may have been due to the higher average age of our patients. The higher frequency of remission in community-based investigations indicates a higher overall severity of the disease in patients seeking medical advice at a hospital-based outpatient unit (26). Remission was seen in patients with less severe symptoms, as demonstrated by a much lower prevalence of spontaneous vasospasms and a lower frequency of triggers other than cold. Moreover, abnormal thoracic outlet test results were less frequent in these patients, which is significant because irritation of the sympathetic nerve is a known sensitizing factor for RP attacks (29).
Patients maintaining the status of suspected secondary RP showed some similarities with, but also relevant differences from, patients with an established diagnosis of secondary RP. Similarities included the comparatively higher age at onset of RP (∼39 years, as compared with 26 years in patients with primary RP). A higher age at manifestation of RP has been implicated as one of the factors associated with increased risk for secondary RP (30, 31). The prevalence of smoking was significantly lower in patients with suspected secondary RP. There is evidence for an association between cigarette smoking and SLE (32), RA (33), and other underlying diseases. However, in SLE, no association with smoking preceding diagnosis was found in a recent meta-analysis (32). A number of abnormal findings in the followup screening program were seen with similar frequency in patients with suspected and those with established secondary RP: abnormal capillary microscopy results, elevated ESR, abnormal differential cell count, and elevated creatinine levels (abnormal renal function).
Patients with an established diagnosis of an underlying disease differed from those remaining in the subset with suspected secondary RP during followup by a significantly higher number of abnormal findings during the followup screening. On an annual average, more than twice as many of the parameters showed abnormal results in patients with secondary RP. As already demonstrated in an earlier study (19), accumulation of positive findings in >3 parameters is almost inevitably associated with a transition to secondary RP. Moreover, the first abnormal findings in patients with definite secondary RP occurred an average of 1 year earlier in the followup, compared with the first abnormal finding in patients with suspected secondary RP.
Another significant difference was observed with respect to ANA titers. While positive ANA titers were seen with virtually identical frequency in patients with suspected secondary RP and those with definite secondary RP, none of the patients with suspected secondary RP had a titer >1:320, while in 14 of 18 patients with secondary RP who were positive for ANA, titers exceeded this level. In addition, almost all of these patients were also positive for specific serologic subsets. The prognostic value of positive ANA has been questioned because of the changes (improvements) over time in the sensitivity of laboratory methods (2). In our study, however, these methods remained unchanged throughout the followup period, allowing for the definition of an appropriate threshold. In our study, a titer of 1:320 and a positive finding for specific serologic ANA subsets are clearly associated with an increased risk for developing a connective tissue disease. Whether this should be taken as an indication for early initiation of immunosuppressive therapy is beyond the scope of the present study.
The significantly shorter duration of symptoms at study enrollment in patients with secondary RP suggests a more rapid development and severe course of the disease. While patients still classified as having suspected secondary RP had the symptoms for ∼12 years, patients with secondary RP presented at our outpatient unit ∼5 years after the onset of RP. These figures are consistent with those reported in a large meta-analysis (5).
The earliest findings in the screening that led to a classification into the suspected secondary RP group were elevated ESR, positive ANAs, and abnormal patterns on capillary microscopy. These results are consistent with those of previous studies (7, 9, 16).
In all patients with secondary RP except for 1 with an embolism, which was suspected to have developed concurrently with RP, abnormal findings were seen earlier than were abnormalities in patients with suspected secondary RP. Suspected secondary RP might be just an intermediate condition that ultimately will lead to a definite diagnosis. However, not only did 3 patients have to be reclassified as having primary RP because the pathologic findings resolved, but, considering the differences outlined above, it may also be hypothesized that there are actually 2 groups of patients with suspected secondary RP.
Our findings in this long-term study indicate that patients with RP may be categorized into 3 groups. We observed patients remaining in the study with no signs of an underlying disease for >10 years but who still had RP (idiopathic RP). Patients who show signs of an underlying disease during followup can be subdivided into 2 groups. One is characterized by earlier presentation to a specialist, initial presence of an abnormal result on a thoracic outlet test, earlier occurrence of abnormal findings during followup (in particular, high ANA titers with positivity for specific serologic ANA subsets), and an overall higher number of abnormal findings occurring in shorter sequence (more severe course of the disease). The other group can be characterized by a much later onset of abnormal findings that remain stable for long periods of time and do not allow for a definite diagnosis. In many cases, these abnormal findings may be related to aging and have no direct implication regarding an underlying disease (a rather benign disease course).