Diagnosis and long-term clinical outcome in patients diagnosed with hand ischemia

Transcription

1 Diagnosis and long-term clinical outcome in patients diagnosed with hand ischemia Robert B. McLafferty, MD, James M. Edwards, MD, Lloyd M. Taylor, Jr., MD, and John M. Porter, MD, Portland, Ore. Purpose: The long-term clinical outcome of patients diagnosed with digital artery obstruction and symptomatic hand ischemia is largely unknown. Our long-term experience with the diagnosis of symptomatic digital artery obstruction and the long-term natural history of this condition forms the basis for this report. Methods: From 1971 to 1985, 44 patients with symptomatic hand ischemia and palmar or digital arterial obstruction underwent arteriography and digital photoplethysmography (PPG). Patients were grouped according to severity of hand ischemia symptoms, including ulceration and digital amputation and the presence of a connective tissue disease (CTD). Arteriography was compared with PPG by creating an objective severity scale (digital obstruction index [DOI]). Results: Average follow-up was 15.2 years (range 10 to 22 years). Initially 21 patients (48%) had moderate symptoms, and 23 patients (52%) had severe symptoms of hand ischemia. Follow-up symptoms in 28 patients improved (13 of 26 with CTD, 15 of 18 without CTD), in 15 patients (12 of 26 with CTD, 3 of 18 without CTD) remained unchanged, and in only 1 patient (1 of 26 with CTD) worsened. Seventeen (65%) patients with CTD (n = 26) had development of one or more ulcers, and six (24%) underwent one or more digital amputations. Four (22%) patients without CTD (n = 18) had finger ulceration (p < compared with patients with CTD), and one patient (6%) underwent subsequent digital amputation (p = NS). The arteriography-doi and PPG- DOI were equally accurate in determining severity of finger ischemia as manifested by severity of symptoms or ulcer development.,conclusions: The favorable long-term prognosis of symptomatic finger artery occlusion described herein mandates the avoidance of premature finger amputation. Patients with CTD fare worse, but even in this group tissue loss is modest. Finger PPG is as accurate as arteriography for determining severity of hand ischemia. (J VAsc SURG 1995;22: :361-9.) Hand ischemia may be caused by either large- or small-artery disease. Large-artery diseases (defined as occlusive arterial disease proximal to the wrist) have been studied extensively, and a large body of literature describes the diagnosis and treatment of these conditions. 1 For the purpose of this study, smallartery disease is defined as arterial occlusion distal to From the Division of Vascular Surgery, Department of Surgery, Oregon :Health Sciences University, Portland. Supported in part by grants RR00334 and HL05828, General Clinical Research Centers Branch, Division of Research Resources~ National Institute of Health, Bethesda, Md. Presented at the Tenth Annual Meeting of the Western Vascular Society, Phoenix, Ariz., Jan , Reprint requests: James M. Edwards, MD, Division of Vascular Surgery, 3181 S.W. Sam Jackson Park Rd., Portland, OR Copyright 1995 by The Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter /95/$ /6/66214 the wrist with normal proximal arteries. By far the most prevalent patient group with symptomatic small-artery disease in our population have normal arteries at rest and are symptomatic only with exposure to cold or emotional stress induced by palmar and digital artery vasospasm. 2 These patients have vasospastic Raynaud's syndrome, which for a large majority represents only a nuisance condition and is not further considered in this report. Patients with permanent small-artery obstruction of the palmar and digital arteries often have persistent symptoms of hand ischemia, including rest pain and ulceration in addition to marked sensitivity to cold and Raynaud's attacks, specifically in this case termed obstructive Raynaud's syndrome. 3 Symptoms in these patients represent much more than a nuisance condition and on occasion lead to significant tissue loss. 4 Most literature on occlusive disease of the palmar 361

2 362 McLafferty et al. October 1995 Table I. Hand ischemia categories Mild symptoms Moderate symptoms Severe symptoms Raynaud's attacks associated with no effect on fifestyle and minimal pain Moderate pain with Raynaud's attacks requiring fifestyle modifications to avoid cold exposure Ischemic finger ulceration associated with severe pain; unable to use hands if exposed to cold and digital arteries is significantly focused on the epidemiology of associated diseases, s8 Patients with small-artery occlusive disease in the hands are well known to have a high prevalence, as high as 50% in some series, of associated connective tissue disease(s)(ctds). The few publications addressing long-term follow-up in these patients have appeared in the internal medicine literature and have been concerned almost exclusively with the prevalence and incidence of CTDs in patients initially diagnosed with hand ischemia. 6,9 To date, we are aware of no published information on the long-term clinical outcome in these patients, including symptomatic severity, development of ischemic ulcers, and the need for amputation, or the ability of available invasive (arteriography) and noninvasive (digital photoplethysmography [PPG]) tests to predict long-term outcome. The hand ischemia clinic at the Oregon Health Sciences University has been devoted to research in the diagnosis, treatment, pathophysiology, and natural history of upper extremity small-artery disease since I,11 We have monitored a significant number of patients for greater than 10 years, allowing us to observe the long-term history of this disease. The long-term symptomatic outcome of patients with small-artery obstruction and hand ischemia, as well as a comparison of objective arteriographic and photoplethysmographic criteria and the correlation of these test results with long-term follow-up, form the basis for this report. METHODS Patients who had arteriography for symptomatic hand ischemia before 1985 in whom the arteriogram demonstrated palmar or digital artery obstruction with normal proximal arteries were selected for entry into this study. The patients were monitored at least annually in the clinic. The medical records of these patients were reviewed, and a standardized questionnaire regarding the patient's hand ischemia was administered by a physician within the last 6 months. The data collected included the patient's history at initial presentation, with particular attention paid to the severity of symptoms, the presence of one or more ischemic finger ulcers, tobacco use, and whether the diagnosis of a connective tissue disorder had been established. The diagnosis of a connective tissue disorder was made by use of standard American Rheumatologic Association criteria An ulcer was defined as skin breakdown greater than 1.0 mm in diameter on the finger with or without surrounding cellulitis. Short-term traumatic lesions were excluded. The clinical course of the patient's hand ischcmia since initial presentation was obtained, specifically addressing the development of ischemic finger ulcers during any time in the follow-up period, smoking status, performance of digital amputation, diagnosis of connective tissue disorder during followup, symptomatic severity of hand ischemic symptoms, and use of medications. Hand ischemia symptoms were divided into three categories, mild, moderate, and severe (Table I). The techniques of magnification hand arteriography and digital PPG have been previously described for this group of patients. 18,19 Vasodilating medications, primarily tolazoline hydrochloride (Priscoline), are used as necessary. All arteriograms were reviewed specifically for this study by a vascular surgery staff physician and resident surgeon. Both were blinded to the results of digital PPG performed just before arteriography, as well as to patient history and questionnaire. A positive history of ulcer was recorded if the patient initially was diagnosed with an ischemic digital ulcer or had development of an ulcer at any time in the follow-up period. No patient was initially diagnosed with a history of prior digital amputation, therefore need for digital amputation during follow-up constituted a positive history. Amputations were performed in such a fashion that finger length was preserved. Strategies used for this purpose included debridement of just the necrotic tissue without trying to obtain skin closure (allowing healing by secondary intention) and delaying amputation to allow possible regrowth in a similar fashion to frostbite care. An objective classification scheme for scoring the severity of digital artery obstruction by arteriography was devised. This consisted of assigning scores for patency of each digital artery of each phalanx in each finger, followed by calculation of a total that was divided by a normal score to generate an index. Patent proximal segments were assigned higher

3 Volume 22, Number 4 _McLafferty et al. 363 Individual digital arteries graded Normal arteriogram = 59 points DOI = arteries patent (points) 59 Fig. 1. Technique for calculating A-DOI. Each individual artery is given points depending on how far it extends without interruption into digit. For example, if both digital arteries extend past proximal phalanx, 8 points are given, whereas if only one artery extends past first phalanx, 4 points are given. scores than patent distal segments. The method of calculation is illustrated in Fig. 1. A normal score of 59 required one complete digital artery in the thumb and two complete digital arteries in the four remaining digits. All digital arteries had to extend to the distal phalangeal segment and be continuous for a normal score. The patency of the deep or superficial arch was not directly scored because of frequent anatomic variations. Arch patency was indirectly scored by requiring that there be a continuously patent vessel from the arm into the proximal digit for the proximal digital artery segment to be scored as patent. In digits two through five, if one digital artery were obstructed and the other digital artery were continued into the next segment, a half score was assigned to each of the distal segments. Requirement for a normal score in any phalangeal segment required both arteries to enter the segment and be continuous. Phalangeal segmental scores of each hand were added and divided by a normal score of 59 to obtain the arteriogram-digital obstruction index (A-DOI). The plethysmographic digital obstruction index (P-DOI) was constructed in a similar fashion to the A-DOI and was obtained by adding the assigned scores of each digit of the hand and dividing by a perfect score of 40 (Fig. 2). A normal digital waveform received eight points if the upstroke time was less than 0.2 seconds and a dicrotic notch was present on the downstroke. A score of four was given if the upstroke time was greater than 0.2 seconds with loss of the dicrotic notch. A score of two was assigned if minimal pulsatile waveform activity was detected, and no points were given in the absence of any pulsatile waveform. Hands were divided into groups for analysis, including presence and absence of a connective tissue disorder, and by symptomatic severity. Patients with Buerger's disease were included in the group of patients with CTD not because we believe that Buerger's disease represents a CTD, rather because the underlying arteritis is similar in both Buerger's disease and the CTDs. Additionally, there were too few patients with Buerger's disease to allow separate analysis. For each of these groups, the Mann- Whitney rank sum test was used to determine significant differences in the distributions of the A-DOI and P-DOI separately in patients with or without a positive history of ulcer or amputation. Chi-square analysis was used to compare patients with and without a connective tissue disorder who had development of ulcers or required amputation. RESULTS Forty-four patients fulfilled the study entry criteria, including the presence of symptomatic hand ischemia, minimum clinical follow-up of 10 years, and digital artery obstruction confirmed by hand arteriography. There were 23 males and 21 females with an average age at the time of arteriography of 44 years (range 15 to 75). Mean follow-up of patients was 15 years (range 10 to 22). At initial presentation, no patients had mild symptoms (symptomatic severity defined in Table I). Twenty-one patients (48%, 10 of 26 with CTD, 11 of 18 without CTD) had moderate symptoms, and

4 364 A/lcLafferty et al. October 1995 i = ~ ~22 ~22 r~ 2222 ~2 ~ 2 2 ~ 2 2 ~ 222 ".o _~:::::: x: ~ 2~ 2' x~ xi x22:~ Fig. 2. Waveform criteria for photoplethysmographic digital obstruction index (P-DOI). A, Eight points, normal waveform, upstroke less than 0.2 seconds, dicrotic notch present. B, Four points, upstroke greater than 0.2 seconds, loss of dicrotic notch. C, Two points, markedly flattened waveform. 23 patients (52%, 16 of 26 with CTD, 7 of I8 without CTD) had severe symptoms of hand ischemia. Twenty-eight patients' symptoms improved (13 of 26 with CTD, 15 of 18 without CTD), 15 patient's symptoms remained unchanged (12 of 26 with CTD, 3 of 18 without CTD), and only one patient's symptoms worsened over the subsequent 10 or more years (1 of 26 with CTD). At their last follow-up visit, 20 patients had mild symptoms (45%, 7 of 26 with CTD, 13 of 18 without CTD), 16 patients had moderate symptoms (36%, 12 of 26 with CTD, 4 of 18 without CTD), and 8 patients had severe symptoms (18%, 7 of 26 with CTD, 1 of 18 without CTD). Nine patients (20%) received medication (most frequently nifedipine) for their hand ischemia, and 16 patients (36%) continued to see a physician specifically for treatment of hand ischemia. Of the 21 patients with moderate symptoms at initial presentation, 12 (57%, 3 of 10 with CTD, 9 of 11 without CTD) had improvement to mild symptoms at most recent follow-up, 8 (38%, 6 of 10 with CTD, 2 of 11 without CTD) continued to have moderate symptoms, and i (5%, 1 of 10 with CTD) had progression to severe symptoms. Of the 23 patients with severe symptoms at initial presentation, 8 (35%, 4 of 16 with CTD, 4 of 7 without CTD) had improvement to mild symptoms, 8 (35%, 6 of 16 with CTD, 2 of 7 without CTD) had improvement to moderate symptoms, and 7 (30%, 6 of 16 with CTD, i of 7 without CTD) continued to have severe symptoms. Of the conditions of 26 patients with a CTD, 13 (50%) had improvement of symptoms, 12 (46%) remained unchanged, and 1 (4%) worsened. Of the 18 patients without a connective tissue disorder, 15 (83%) had improvement of symptoms, and 3 (17%) had no change in symptoms. At the time of arteriography, 17 patients had been diagnosed with connective tissue disorder. Nine patients were subsequently diagnosed with a connective tissue disorder during follow-up. Of the 26 patients with a connective tissue disorder, l i had scleroderma, 4 had systemic lupus erythematosus, 4 had rheumatoid arthritis, 3 had mixed CTD, 2 had undifferentiated CTD, and 2 had Buerger's disease. Similarly, the patients without CTD had a variety of associated conditions, including three patients with vibration disease, two each with atherosclerosis and hypercoagulable states, and one with posttraumatic, one with embolic, and nine with no identifiable cause or source of digital artery obstruction. Twenty-one patients had one or more ischemic digital ulcers. Nineteen patients (15 with CTD, four without CTD) had an ulcer present at study entry (time of arteriography), and two patients (both with CTD) with no ulcer at study entry had development of an ischemic digital ulcer during follow-up. Eleven patients (10 with CTD, one without CTD) had recurrent ulcers during follow-up. Four patients required partial finger amputation within months of study entry, and three patients had partial amputation during the follow-up period (six with CTD, one without CTD). Five patients required multiple amputations. Seventeen of the 26 patients with connective tissue disorder (65%) had ischemic digital ulcerations. Only four of 18 (22%) patients without a connective tissue disorder had ulcers in follow-up (p < compared with patients with CTD). Similarly, six patients (24%) with connective tissue

5 Volume 22, Number 4 McLaJfeH;y et ai. 365 Table II. Arteriographic A-DOI and P-DOI by symptom severity at initial presentation and final follow-up Initial Final Moderate Severe Mild~Moderate Severe A-DOI _ p < ± p = NS P-DOI 0.75 _ _ p < p = NS disorder required amputation as opposed to one patient (6%) without (p = NS). At study entry 62% of patients smoked. In follow-up, 35% of patients continued to smoke, representing a significant decline (p = 0.02). The A-DOI and P-DOI both demonstrated significant differences in the distributions of patients with moderate and severe symptoms and with or without an ulcer history (Tables II, III, and IV). Patients "with severe symptoms initially had lower A-DOI and P-DOI than patients with moderate symptoms (2o = 0.005, Table II). Patients with ulcers had a significantly lower A-DOI and P-DOI. This was true for patients with a connective tissue disorder (p = for A-DOI, p = for P-DOI), patients without a connective tissue disorder (p < for A-DOI, p = 0.02 for P-DOI), and the entire group (2o < for A-DOI, p = for P-DOI) (Table III). Only the A-DOI was significantly lower in patients requiring amputation after arteriography or in follow-up. This difference was not present in P-DOI scores. This was true for patients with a connective tissue disorder (p = for A-DOI, p = for P-DOI) and the entire group (p = for A-DOI,p = for P-DOI) (Table IV). Statistical calculation could not be performed for amputation in patients without a connective tissue disorder because of small numbers (n = 1). Use of an A-DOI less than 0.65 in patients with a CTD to predict nicer development yielded a sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of 76%, 100%, 100%, and 73%. Similarly, a P-DOI less than 0.65 in patients vdth a connective tissue disorder predicted digital ulceration with a sensitivity, specificity, PPV, and NPV of 77%, 100%, 100%, and 73%, respectively. An A-DOI less than 0.50 in patients with a connective tissue disorder predicted amputation with a sensitivity, specificity, PPV, and NPV of 71%, 72%, 29%, and 94%, respectively (Table IV). A-DOI and P-DOI scores were not useful in predict- ing clinical symptom severity during follow-up, although there was a trend for patients with lower A-DOI scores to have severe symptoms (p = 0.10). DISCUSSION Symptomatic hand ischemia occurs infrequently and accounts for less than 5% of patients with extremity ischemia being examined by a physician. As noted, most patients with hand ischemia have no fixed arterial obstruction and experience only coldinduced vasospastic symptoms, typically only a nuisance condition. Even in the small subset of patients with fixed arterial obstruction, ischemic ulceration, rest pain, and amputation are infrequently encountered. Because of the small number of patients with moderate or severe symptomatic hand ischemia, there is a dearth of objective data concerning the natural history of this condition on which to base decisions for evaluation and treatment. The available literature on the natural history of Raynaud's syndrome has focused on the prevalence of CTD in these patients and the risk for development of a CTD with time. s-s We are aware of no prior publications specifically addressing the clinical outcome of patients with symptomatic digital artery occlusion. During the past 22 years we have evaluated and monitored more than 1100 patients in our hand ischemia clinic at Oregon Health Sciences University. The data presented herein describe the long-term clinical outcome of a small and highly selected subset of patients with palmar and digital arterial obstruction as documented by arteriography. Holmgren et al.i9 compared digital PPG with hand arteriography in patients with digital artery obstruction. In this study PPG was sufficiently accurate to detect digital artery obstruction when the wavcform was flattened with an upstroke greater than 0.2 seconds or loss of the dicrotic notch. The classification criteria for both plethysmography and arteriography was restricted and analyzed by digit instead of the whole hand as in this study. Other authors have suggested different PPG criteria for

6 366 2PicLafferty et al. October 1995 Table III. Mean, SD, and SE values of the A-DOI and P-DOI in patients with and without an ulcer either at initial presentation or in follow-up Group Ulcer(+~-) Mean SD SE p Value All patients A-DOI p < A-DOI P-DOI p = P-DOI Patients with CTD A-DOI p = A-DOI P-DOI p = P-DOI Patients without CTD A-DOI p < A-DOI P-DOI p = P-DOI ~Mann-Whitney rank sum test. Table IV. Mean, SD, and SE values of the A-DOI and P-DOI in patients undergoing amputation in follow-up ~ Group Amputation(+~-) Mean SD SE p Value~ All patients A-DOI p = A-DOI O P-DOI p = P-DOI - O O.04 Patients with CTD A-DOI p = A-DOI P-DOI p = P-DOI *Patients without CTD are not included because only one patient underwent amputation. tmann-whitney rank sum test. Raynaud's syndrome, but these criteria have not been used to predict outcome. 21,22 Our results demonstrate that the outcome of moderate and severe hand ischemia caused by digital artery obstruction is generally quite favorable. Approximately half of the patients after 10 years will have only mild symptoms, and less than one fifth will have severe symptoms,.\, In this study symptoms progressed in only on~,patient (2%) with the diagnosis of scleroderma in'whom ulcers developed during follow-up. Two patients without ulceration initially had development of ulcers, and recurrent ulceration was seen in 11 of 19 (58%) patients admitted with ulceration. Only seven patients (15%) required amputation. The test of choice to evaluate patients with hand ischemia appears to be digital PPG. The data obtained from this simple, inexpensive, noninvasive test compare favorably with arteriography. The determination of the P-DOI appears to permit accurate prediction of the clinical course of patients with hand ischemia, but before acceptance this must be validated in a prospective trial. In this study a P-DOI of less than 0.65 predicted future ischemic digital ulceration with a high specificity and PPV. We continue to use arteriography in two groups of patients with symptomatic hand ischemia. The first is in patients with unilateral hand ischemia in whom there may be a source of proximal emboli, and the second is in patients who do not manifest a systemic illness to explain the hand ischemia. In both of these groups arteriography is performed as a diagnostic test. All patients undergo extensive screening for autoimmune diseases and hypercoagulable states in

7 Volume 22, Number 4 McLafferty et al. 367 an effort to determine the cause of hand ischemia before arteriography. A variety of noninvasive tests continue to be obtained as described previously. In conclusion, over a follow-up period exceeding 10 years, the clinical outcome of symptomatic hand ischemia is remarkably benign for patients without CTD, with only 18% of patients continuing to have significant symptoms, 58% of those patients admitted with ischemic ulceration having recurrent ulceration, and 15% requiting amputation. Patients with connective tissue disorder fared worse, with only 50% of patients having an improvement in symptoms compared with 83% improved in the group without CTD. Both arteriography and PPG can be used to objectively measure hand ischemia in patients with palmar and digital artery obstruction. The A-DOI and the P-DOI appear to predict the future development of digital ulceration. Only the A-DOI appears sufficient to accurately predict the need for amputation. The benign nature of digital artery obstruction in long-term follow-up supports the philosophy of minimal digital artery amputation in these patients. The clinical follow-up information presented herein should be valuable both from the standpoint of formulation on therapeutic plans and, of equal importance, as a basis for patient counseling. REFERENCES 1. Machleder HI. Arterial diseases. In: Machleder HI ed. Vascular disorders of the upper extremity. Mt Kisco NY: Futura Publishing, 1989: Porter JM, Edwards JM, Taylor LM Jr. Upper e ttemity vasospastic disease. In: Stanley JC, Ernst CB, eds. Current therapy in vascular surgery. Philadelphia: BC Decker, 1987: Edwards JM, Porter IM. Update on Raynaud's syndrome. Semin Vase Surg 1990;3: Mills IL, Friedman EI, Taylor LM Jr, Porter IM. Upper extremity ischemia caused by small artery disease. Ann Surg 1987;206: Kallenberg C, Wouda A, Hanw The T. Systemic involvement and immunologic findings in patients presenting with Raynaud's phenomenon. Am J Med 1980;69: Harper FE, Maricq HR, Turner RE, Lidman RW, Leroy EC. A prospective study of Raynaud phenomenon and early connective tissue disease: a five-year report. Am J Med 1982;72: Priollet P, Vayssairat M, Housset E. How to classify Raynand's phenomenon: long-term follow-up study of 73 cases. Am J Med 1987;83: Edwards JM, Porter JM. Associated diseases with Raynaud's syndrome. Vase Med Rev 1990;1: Vayssairat M, Fiessinger J, Housset E. Phenomene de Raynand: Etude prospective de 100 cas. Paris: La Nouvelle Presse Medicale 1979;26: Porter }'M, Bardana El, Baur GM, Wesche DJ, Andrasch RI-I, Rtsch I. The clinical significance of Raynaud's syndrome. Surgery I976;80: ll. Porter JM, Rivers 8P, Anderson CJ, Baur GM. Evaluation and management of patients with Raynaud's syndrome. Am J Surg 1981;142: Ropes MW, Bennett EA, Cobb S, et al revision of diagnostic criteria for rheumatoid arthritis. Bull Rheum Dis 1958; Tan EM, Cohen AS, Fries JF, et al. The 1982 revised criteria for the classification of systemic lupus erythematosis. Arthritis Rheum 1982;25: I Mikerji B, Hardin JG. Undifferentiated, overlapping, and mixed connective tissue diseases. Am J Med Sci 1993;305: i5. Fox RI, Saito I. Criteria for diagnosis of Sjogren's syndrome. Rheum Dis Clin North Am 1994;20: Rodnan GP, Iablonska S, Medsger TA Jr. Classification and nomenclature of progressive systemic sclerosis (scleroderma). Clin Rheum Dis 1979;5: Arnett FC, Edworthy SM, Bloch DA, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arth Rheum 1988;31: Rosch J, Porter JiM, Gralino BI. Cryodynamic hand angiography in the diagnosis and management of Raynaud's syndrome. Circulation 1977;55: Holmgren K, Baur GM, Porter JM. The role of digital photoplethysmography in the evaluation of Raynand's syndrome. Bruit 1981;5: Sumner DS, Sttandness DE. An abnormal finger pulse associated with cold sensitivity. Ann Surg 1972;175: Alexander S, Cummings C, Figg-Hoblyn L, Taylor LM Jr, Porter JM. Usefulness of digital peaked pulse for diagnosis of Raynaud's syndrome. J Vase Tech 1988;12:71-5. Submitted Feb. 10, 1995; accepted May 5, DISCUSSION Dr. Herbert I. Machleder (Los Angeles, Calif.). All of the patients in this study underwent hand arteriography and digital plethysmography. The authors constructed a numeric index to describe the analogue findings. Even though all of the patients with bilateral hand ischemia underwent arteriography, you affirm, "we presently reserve arteriography for patients with unilateral disease suspected of having a surgically correctable proximal arterial lesion."

8 368 21/icLafferty et al. October 1995 However, I recall your statement in the American Journal ofsurgery,i1 "Arteriography is seldom necessary to establish the diagnosis, especially since the advent of digital plethysmography. We currently utilize arteriography only to evaluate a suspected surgically correctable proximal arterial lesion." What was the continued value in obtaining angiograms in these patients with bilateral hand ischemia? What really is the proper approach (in view of this apparent inconsistency)? You indicate that the DOI may be useful in predicting future digital amputation or ulceration. Have you tried to use this index prospectively to validate its predictive value? One conclusion emphasized by the authors is the apparent benign nature of hand ischemia in the long term. However, I wonder if the group presented is representative. No mention is made of patients lost to follow-up or,those who may have died. One would have expected a certain attrition of the more seriously afflicted patients. How many patients were lost to follow-up, and would you suspect that they were the patients with more severe disease? Could this selection skew your interpretation of the data? You derive support for your conservative approach to these problems on the basis of your interpretation of benign consequences. However, 48% had moderate symptoms and 52% had severe symptoms of hand ischemia. Forty-eight percent had is chemic ulcers, 25 % had recurrent ulcers, 16% had finger amputation, and 71% of these had more than one amputation. Even more striking is that, after an average follow-up of 15 years, 20% of your patients with CTD continued to receive vasodilator drugs for hand ischemia and that 36% "continued to see a physician, specifically for treatment of hand ischemia." Can this be considered a remarkably benign condition? Is it reasonable to commingle the patients without CTD with the patients with CTD when you're looking at long-term clinical outcome? These patients represent such a heterogenous population of hypercoagulability, thrombocytosis, gammopathy, malignancy, and so forth that extrapolating from them as a group seems problematic. What diseases were represented in your patients in this group and would you have expected them to have recurrent hand ischemia after the diagnosis was established and appropriate therapy instituted? You indicate that the purported benign nature of the condition supports a philosophy of minimal digital artery amputation in these patients. I don't quite follow that argument. In our own population we have found that autoamputation results in far better tissue preservation than formal amputation, and we prefer this course unless forced to amputate by uncontrolled ischemic pain. So for us the issue of whether to be conservative is immaterial. How do you view these two different approaches to amputation? Dr. James M. Edwards. With regard to the role of arteriography, at the time that most of these patients were enrolled in the study, we had an NIH-supported grant in which all patients underwent bilateral arm arteriography. Because of that study we reached the conclusion that arteriography was not necessary in all patients for diagnosis. We reserve arteriography for those patients with unilateral hand ischemia where there may be a proximal embolic source that is surgically correctable. Patients diagnosed with bilateral hand ischemia who have a CTD or who have laboratory abnormalities consistent with CTD do not undergo arteriography, We have not yet tried to use the plethysmographic index, although we are starting to collect those data. Now that we know that there does seem to be a correlation, we've started using it on a prospective basis over the last 6 months. You asked whether a 50% amputation rate and recurrence of ulcers is really benign. In some ways it's not benign in that we don't regard amputation as a benign consequence, but the patients who have ulcers are often convinced that they will invariably lose all their fingers. To tell them that they only have approximately a 50% chance of having recurrent ulcers is benign. Several of the patients with scleroderma did die during follow-up, but you asked about patients lost to follow-up. No patients who had greater than 10 years follow-up by the study criteria were lost to follow-up. You asked about mingling CTD and non-ctd. If we looked at any individual non-ctd we would have had a maximum of five patients. So we chose to mingle them, although we emphasize the differences between the two groups. With regard to autoamputation versus formal amputation, we follow the same philosophy as you do in that we will postpone amputation as long as possible because a lot of these will heal or will heal with a little stub of bone that you can cut back with a rongeur without taking any tissue. But because these patients almost always have marked relief of pain with fingertip amputation, we perform that when we believe it's necessary. Dr. Albert D. Hall (Greenbrae, Calif.). What is your experience with corkscrew ulnar artery? Do you believe the digital arteries are at continued risk for microembolization when a corkscrew ulnar artery caused by a trauma is shown by arteriography? What is your indication for resection and microvascular reconstruction in such cases? Dr. Edwards. There were no patients like that in this group, although we see somebody like that an average of twice a year and have previously published a series of approximately a dozen patients. If we have a patient who has had significant hand or finger embolus and a patent ulnar artery that is obviously deformed as seen by arteriography, we consider that an indication for surgery, and we do a saphenous vein bypass across the wrist rather than a microvascular repair within the palm. If it's occluded and the patient has ongoing ischemia, we will bypass it. Dr. Calvin B. Ernst (Detroit, Mich.). I know of no group, other than the Portland group, that has a better database as it relates to upper extremity surgical problems and digital ischemia, and I think this report bears close

9 Volume 22, Number 4 31cLafferty et al. 369 scrutiny. Is it your policy to obtain bilateral arteriograms even on individuals with unilateral symptoms; if so, why and what are you looking for? In the classic surgical and orthopedic literature, there is still ongoing controversy regarding sympathectomy. Some favor a digital sympathectomy in individuals who have mild digital ischemia with the idea that perhaps digital sympathectomy will augment healing and prevent the ischemic events you depicted; namely, the terminal gangrene. What are your thoughts regarding, not dorsal sympathectomy, but digital sympathectomy under these circumstances? Dr. Edwards. With regard to arteriography, we always order bilateral arteriography, and specifically we ask for nondigital magnification films of the hand because the digital films don't give us a lot of detail. The reason for that is even in patients with unilateral hand symptoms, we often find bilateral disease, and that makes us much more confident that we're not missing something that is surgically correctable. With regard to the role of microvascular digital periarterial digital sympathectomy, we have referred two patients over the last 4 to 5 years for this and have seen probably half a dozen more who have had it done before they have come to see us. The results in this group seem to be the same as in everyone else. So we have not been impressed, and we're not aware of any prospective randomized study that proves any benefit. Dr. John M. Porter (Portland, Ore.). Early death of patients excluded them from the study. The patients had to undergo a 10-year follow-up to get into this particular study. If they died at 5 years, they were not included.