Case 14-2005 — A 38-Year-Old Man with Fever and Blurred Vision
http://www.100md.com
《新英格兰医药杂志》
Presentation of Case
A 38-year-old man was admitted to this hospital because of blurred vision and fever.
The patient had ulcerative colitis but had been well while receiving treatment with mercaptopurine and balsalazide until two weeks before admission, when a chronically draining pilonidal cyst was excised. Amoxicillin–clavulanate was prescribed after the operation. One week later, the patient noted the gradual onset of severe right anterior thigh pain; he had difficulty walking and a temperature as high as 38.3°C. He was admitted to another hospital.
On examination, the temperature was 38.8°C, and the abdomen was moderately distended but not tender. The right thigh was swollen, tender, and erythematous, and the patient could not flex his leg. An ultrasonographic study of the right thigh showed a superficial saphenous-vein thrombosis extending into the pelvis. Computed tomographic (CT) scanning of the abdomen and pelvis revealed a fluid collection and stranding of the soft tissues adjacent to the cecum. Inflammatory changes extended within the retroperitoneum along the inferior vena cava and the right external iliac vein, anterior and lateral to the psoas muscle. Cultures of multiple specimens of blood and urine were negative for pathogens. The white-cell count was 18,900 per cubic millimeter, with 82 percent neutrophils and 6 percent band forms. The serum creatinine level was 1.7 mg per deciliter (150.3 μmol per liter). Ampicillin, metronidazole, and heparin were administered. Klebsiella species and candida grew from cultures obtained from the pilonidal cyst. Ampicillin was discontinued, and piperacillin and caspofungin were started.
Daily spiking fevers continued. A magnetic resonance imaging (MRI) study of the right thigh showed two ring-enhancing lesions within the muscle tissue, consistent with the presence of inflammation or abscess. There was no evidence of osteomyelitis or deep-vein thrombosis. Heparin was discontinued. Fever persisted; vancomycin was started, and mercaptopurine was discontinued.
On the sixth hospital day, the patient became short of breath and could not complete full sentences. The next day, he experienced double vision, first in his right eye and then in his left. The conjunctivae were injected, and a topical antibiotic was administered. He remained febrile and short of breath and was transferred to this hospital.
The patient worked in an office and did not use alcohol or tobacco. A diagnosis of ulcerative colitis had been made 18 years earlier. It was minimally active until eight months before this admission, when the patient had a flare-up of the disease. The symptoms responded to a tapering dose of prednisone, which was discontinued three months before admission.
On examination, he was in mild distress. The temperature was 37.5°C, the blood pressure 115/52 mm Hg, the heart rate 109 beats per minute, the respiratory rate 24 breaths per minute, and the oxygen saturation 95 percent while the patient was breathing room air. The oral cavity was clear; the sclerae were injected, and the conjunctivae were pink. The visual acuity was 20/800 in the right eye and 20/200 in the left eye. There were 4+ cells in the anterior chambers of both eyes. Funduscopic examination showed 1+ to 2+ vitritis bilaterally, with diffuse retinal hemorrhages and perivascular exudates. Examination of the chest and abdomen showed no abnormalities. There was serosanguinous drainage from the pilonidal-cyst wound. There was 2+ edema of both legs; the peripheral pulses were easily palpable, and there was no thigh tenderness. Raised purpuric lesions were present on the right leg. The right third proximal interphalangeal joint, left elbow, and right knee were slightly warm to the touch, without effusion. The neurologic examination revealed no abnormalities. Levels of serum electrolytes, calcium, phosphorus, magnesium, urea nitrogen, and creatinine were normal; the results of other laboratory tests are shown in Table 1 and Table 2.
Table 1. Blood Chemical Values.
Table 2. Hematologic and Serologic Laboratory Values.
A transthoracic ultrasonographic study of the heart and both MRI and magnetic resonance angiography of the brain were performed and showed no abnormalities. High-dose methylprednisolone, acyclovir, and heparin were given intravenously. Cultures of specimens of skin, the cyst, and blood for viral, fungal, and bacterial pathogens yielded no growths.
Over the next few days, abdominal pain and pain in the left testis developed; the patient passed bloody diarrheal stools. A colonoscopic examination revealed multiple deep ulcers in the ascending colon and cecum, with shallow ulcers in the transverse colon, distal sigmoid, and rectum.
Diagnostic procedures were performed.
Differential Diagnosis
Dr. Sekar Kathiresan: This patient's clinical course suggests a bimodal illness that began with superficial thrombophlebitis after the resection of a pilonidal cyst, followed by a systemic illness with fever, rash, arthralgias, and visual loss. To identify a unifying cause of these disparate symptoms, the general medical team, of which I was a member, focused on the patient's three active medical issues on transfer to this hospital: fever, rash, and eye disease. The primary problem was whether his symptoms represented an infectious or an immunologically mediated inflammatory process.
Schemes that have been designed to classify the causes of fever and rash generally start with the nature of the rash.1 In this patient, the eruption was purpuric, indicating extravasation of red cells. Purpura can be caused by disorders of platelets, the coagulation cascade, or vascular integrity, such as vascular infection and noninfectious vasculitis. Purpura may be divided into two major categories: nonpalpable and palpable (Table 3).2 Most causes lead to flat or nonpalpable purpura, whereas far fewer causes lead to palpable lesions. The palpability is due to localized swelling associated with the purpuric lesions; this usually occurs when an inflammatory infiltrate of the vessel wall leads to the bleeding. In this patient, the purpuric lesions involving the right leg were clearly palpable (Figure 1).
Table 3. Causes of Purpura.
Figure 1. Typical Lesions of Palpable Purpura (from the Arm of Another Patient).
Purple, raised skin lesions are present. The lesions are palpable because of localized swelling that usually occurs when an inflammatory infiltrate of the vessel wall leads to the bleeding, as in infection or vasculitis. (Photograph courtesy of Dr. Marc Sabatine, Division of Cardiology, Brigham and Women's Hospital.)
Palpable purpura is caused by a limited number of infections that directly involve the vascular wall and by noninfectious immune-mediated processes. The infections include acute meningococcemia (purpura fulminans), disseminated gonococcal infection, Rocky Mountain spotted fever, and subacute bacterial endocarditis. Of the noninfectious vasculitides, small-vessel vasculitis (involving vessels smaller than arteries, such as arterioles, venules, or capillaries) characteristically leads to palpable purpura.3
The disease in the eye could also have had either an infectious or an inflammatory cause. Viral infections, including infection with a member of the herpes simplex virus family, such as herpes simplex virus or varicella–zoster virus, can cause vitritis and retinal hemorrhage. Septic thrombophlebitis, which the patient may have had, could have been complicated by bacterial endocarditis with embolic involvement of the eye, and this possibility was considered.4 An inflammatory cause of eye disease, including a primary systemic vasculitis with eye involvement, was also considered.
Overall, the constellation of fever, rash, and severe eye disease led us to consider either bacterial endocarditis or a noninfectious vasculitis as the most probable unifying diagnosis. Prominent involvement of the eye is rare in meningococcemia or gonococcemia. The patient did not report headache, a history of a tick bite, or recent travel, and thus Rocky Mountain spotted fever seemed to be unlikely. Herpesvirus infections may cause severe eye disease, but we found that tying in the other systemic manifestations with a herpesvirus infection was difficult. This patient had several risk factors for vasculitis, including a recent infection, antibiotic therapy, and inflammatory bowel disease.
We ordered blood cultures to evaluate the possibility of bacterial endocarditis, meningococcemia, and gonococcemia. Measurement of the patient's complement levels, which are low in systemic vasculitis, was requested. Measurement of antineutrophil cytoplasmic antibodies (ANCA) and cryoglobulins, serologic assays for hepatitis B and C, and tests for antinuclear antibody and rheumatoid factor were obtained to look for specific types of small-vessel vasculitis. A transthoracic echocardiogram was obtained to detect any vegetations on cardiac valves. We asked for consultation from our colleagues in ophthalmology, rheumatology, and gastroenterology.
Dr. C. Stephen Foster: A diagnostic vitrectomy was performed on the right eye because of the concern about infectious processes; there was no clinical evidence of features, such as the acute retinal necrosis syndrome, that are seen in patients with varicella–zoster virus or herpes simplex virus infection. In the fundus of the right eye (Figure 2A), there were multifocal hemorrhage, perivascular exudates indicative of vasculitis, optic neuropathy, and a macular infarct. All quadrants were involved, and the process predominantly involved arterioles rather than veins. There was pallor of the left optic nerve and foci of frank infarction of the retina. Fluorescein angiography was impossible in the right eye, but in the left eye (Figure 2B) staining of the vasculature in the late phase was apparent, in a pattern indicative of retinal vasculitis, specifically retinal arteritis.
Figure 2. Ophthalmologic Examination on Admission.
In the fundus of the right eye (Panel A), there are multifocal hemorrhage (arrows), perivascular exudates indicative of vasculitis (arrowheads), optic neuropathy, and a macular infarct. All quadrants are involved, and the process predominantly involves arterioles rather than veins. Fluorescein angiography of the left eye (Panel B) shows staining of the vasculature in the late phase (arrows) in a pattern indicative of retinal vasculitis.
The causes of retinal vasculitides fall into four categories: infection, trauma, conditions associated with cancer, and autoimmune diseases. Vasculitis that is secondary to trauma or cancer is rare and improbable in this patient. Infectious and autoimmune diseases cause most cases of retinal vasculitis. In distinguishing among these, the predominant involvement of arteries or veins can be useful. If the arteries are predominantly involved, systemic lupus erythematosus, polyarteritis, the Churg–Strauss syndrome, and so-called frosted-branch angiitis, as well as infection with varicella–zoster virus, herpes simplex virus, and syphilis, are the primary entities to consider. If the veins are predominantly involved, sarcoidosis is by far the most common cause. Other considerations include Eales disease, a localized phlebitis that is often associated with tuberculosis, a localized autoimmune ocular disease called birdshot retinochoroidopathy, and infections with pathogens such as parvovirus, toxoplasma, and the human immunodeficiency virus. Three entities involve arteries and veins relatively equally: multiple sclerosis, Beh?et's disease, and Wegener's granulomatosis.
In this patient with predominantly arterial involvement and a systemic illness, without evidence of infection, my leading diagnosis based on the ocular findings would be polyarteritis nodosa.
Dr. Allen C. Steere: In this patient, a complex array of symptoms developed two weeks after the excision of a pilonidal cyst and the start of antibiotic treatment. These symptoms included devastating visual problems, pain in the right thigh, subtle oligoarticular arthritis, palpable purpura on the right leg, and during his hospitalization here, bloody diarrhea, abdominal pain, elevated pancreatic enzymes, and testicular pain. The ophthalmogist has diagnosed a retinal vasculitis, and the skin lesions of palpable purpura on his right leg were suggestive of a small-vessel vasculitis. As the rhematology consultant, I had to consider what category of vasculitis this disorder best fit, and how to treat the patient.
The classification of vasculitides that was adopted by the 1994 Chapel Hill Consensus Conference is delineated in Table 4.5 The vasculitides are categorized according to the size of the blood vessels affected and the histologic appearance of the lesions. Broadly, the different types are divided into those that affect the large, medium-sized, and small blood vessels, and the latter group is further divided into pauci-immune and immune-complex–mediated types. It should be acknowledged that no classification scheme for the vasculitides has been perfect. Some types of vasculitis affect more than one size of vessel, and some diseases are not included in the current listing.
Table 4. Classification of Vasculitis.
Retinal vasculitis is a well-described manifestation of the most common medium-sized–vessel vasculitis, polyarteritis nodosa.6 Moreover, in this patient there was evidence of pancreatic involvement, muscle inflammation in the right leg, testicular pain, and subtle arthritis, all of which occur in polyarteritis nodosa.3 Abdominal pain and bloody diarrhea may also be manifestations of polyarteritis, although because of the patient's history of ulcerative colitis, the cause of his current bowel symptoms was unclear. However, he had palpable purpura, which suggested involvement of microscopic blood vessels.7 Before the 1994 classification was available, the term "polyarteritis nodosa" was used to describe a broader array of systemic necrotizing vasculitides, and a patient with symptoms such as retinal vasculitis and palpable purpura would probably have been classified as having either polyarteritis nodosa or the polyangiitis overlap syndrome.8 However, polyarteritis nodosa is now defined more strictly as a vasculitis affecting only medium-sized vessels, and the term "polyangiitis overlap syndrome" has been eliminated from the classification.
A second consideration, then, was whether the patient had one of the three forms of pauci-immune vasculitis that are often associated with involvement of both small and medium-sized vessels and a positive test for ANCA.3 Wegener's granulomatosis often causes destructive lesions in the upper respiratory tract and lungs, along with glomerulonephritis. The Churg–Strauss syndrome may be associated with purpura, arthritis, or abdominal pain, but the characteristic clinical feature is hypereosinophilic asthma, which was missing in this case. Finally, microscopic polyangiitis is associated with renal disease in most patients and often associated with pulmonary involvement. In contrast, this patient had a negative test for ANCA, and he did not have the common clinical features of these diseases.
Beh?et's syndrome is associated with inflammation in veins and arteries of all sizes.9 This entity might explain the patient's eye findings, subcutaneous thrombophlebitis, bloody diarrhea, and subtle arthritis. However, he did not have oral aphthous ulcers, which are characteristic of Beh?et's syndrome. Is there a link between ulcerative colitis and vasculitis? Ulcerative colitis has been associated with both small-vessel vasculitis (leukocytoclastic vasculitis) and large-vessel vasculitis (Takayasu's arteritis).10 However, the type of widespread vasculitis seen in this patient has not been reported.
Could this patient have a particularly aggressive form of immune-complex–mediated small-vessel vasculitis?3 As the name implies, the small-vessel vasculitides usually involve only small blood vessels, but they may also affect medium-sized arteries. Henoch–Sch?nlein purpura, the most common vasculitis of childhood, is associated with purpuric rash on the legs, arthritis in large joints, abdominal cramping with bloody diarrhea, and IgA-containing immune complexes in the affected tissues. Although this entity may occur in adults, the patient's age made this possibility unlikely. Mixed cryoglobulinemia may cause palpable purpura and organ dysfunction, but the patient had no clinical signs of cryoglobulins or of hepatitis C infection, which is the usual cause of this entity.
Could he have had immune-complex–mediated vasculitis resulting from either the klebsiella infection of his pilonidal cyst or from the amoxicillin–clavulanate therapy used to treat it? Drug-induced vasculitis usually develops within 7 to 21 days after treatment begins,3 which fits the timing in this case. Both of these vasculitides usually affect only the skin, but they may be associated with systemic disease, including gastrointestinal hemorrhage.3 However, this patient's widespread involvement, including retinal vasculitis, as far as I know, goes beyond previous reports of clinical manifestations of drug-induced vasculitis.
Thus, this patient exemplifies the challenge often faced by clinicians in the diagnosis of vasculitis, which is that some patients have conditions that do not fit well into defined categories. Regardless, it was clear that the vasculitis in this case was very aggressive, which necessitated an aggressive approach to diagnosis and treatment. For this reason, we performed biopsies of the skin and colon, followed by abdominal angiography.
Dr. Peter B. Kelsey: I was asked to see this patient to address the question of whether his new onset of diarrhea and lower gastrointestinal bleeding was due to a flare-up of his ulcerative colitis or to another process. His history was not typical for a patient with chronic ulcerative colitis. He had been asymptomatic while receiving mild immunosuppressive therapy since his initial diagnosis 18 years earlier. Eight months before admission, he had what was interpreted as a flare-up, which was rapidly controlled with prednisone. In the weeks before this admission, he had had no recurrence of symptoms of ulcerative colitis. Thus, before the colonoscopy, I was concerned that there might be another cause of his gastrointestinal symptoms.
At colonoscopy, I saw a background of normal colonic mucosa punctuated by discrete areas of ulceration. There was no evidence of chronic inflammation, such as scarring or thickening of the wall, to suggest past episodes of healed colitis. Several of the ulcers had visible blood vessels at their bases, one of which had a large adherent clot. Because the bowel appeared very fragile and because there were multiple potential sites of bleeding, I did not attempt to ligate any of the exposed blood vessels. I obtained several biopsy specimens, which were submitted for microscopical examination.
Dr. Kathiresan: While we were awaiting the results of the evaluation of the biopsy specimens, we requested an angiogram.
Dr. Marcio S. Curvelo: Visceral digital-subtraction arteriography was performed (Figure 3). There were alternating areas of stenosis and ectasia in the distal branches of the right hepatic artery and the superior rectal artery, with microaneurysms in the distal branches of the superior mesenteric artery and both renal arteries. The angiographic finding of areas of alternating stenosis and ectasia and microaneurysms involving medium-sized vessels in several vascular beds is consistent with a diagnosis of a vasculitis. Microaneurysms of the renal-artery branches are considered fairly specific for polyarteritis nodosa; however, the differential diagnosis includes vasculitides that can involve both small and medium-sized vessels.
Figure 3. Selective Renal Arteriogram.
There are areas of vessel irregularity and microaneurysms consistent with a diagnosis of a vasculitis. This finding is fairly specific but not pathognomonic of polyarteritis nodosa.
Dr. Nancy Lee Harris: Could we have the medical students' diagnosis?
A Medical Student: The constellation of symptoms and laboratory findings indicates an acute inflammatory process and suggests a systemic vasculitis. We thought that in this patient with long-standing ulcerative colitis, the emergence of a vasculitis, such as polyarteritis nodosa, could have been masked by his immunosuppressive therapy. On the withdrawal of the prednisone before surgery, the vasculitis became clinically manifest.
Clinical Diagnosis
Systemic vasculitis, not definitely classifiable: polyarteritis nodosa versus immune-complex–mediated vasculitis (infection-induced or drug-induced).
Pathological Discussion
Dr. James R. Stone: The diagnostic procedures were biopsies of the skin and colon. Histologic examination of the skin-biopsy specimen (Figure 4A and Figure 4B) revealed a necrotizing leukocytoclastic vasculitis involving small vessels (arterioles, capillaries, and venules) as well as small arteries (Figure 4C), which are classified as medium-sized vessels. Examination of the colonic biopsy specimen (Figure 5) revealed fragments of fibrin and inflammatory cells, indicating the presence of an ulcer. There was a leukocytoclastic vasculitis involving small vessels of the colonic submucosa. Neither biopsy showed eosinophil infiltration, granuloma formation, or evidence of prior vasculitic episodes (vascular scarring).
Figure 4. Skin-Biopsy Specimens (Hematoxylin and Eosin).
A low-power view (Panel A) shows the location of the superficial dermal vasculature (arrow) in relation to the deep dermal vasculature (arrowhead). A leukocytoclastic vasculitis involving arterioles, capillaries, and venules is revealed in the superficial vascular plexus (Panel B). The deep vascular plexus (Panel C) contains a small artery involved by vasculitis.
Figure 5. Colon-Biopsy Specimens (Hematoxylin and Eosin).
The mucosa is intact (Panel A), but blood vessels in the submucosa show involvement by leukocytoclastic vasculitis (Panel B).
The current approach to the subclassification of systemic vasculitides starts with consideration of the sizes of the vessels involved (Table 4).3,5 Large-vessel vasculitides principally involve the aorta and major arterial branches to the head and extremities, but they may also involve medium-sized vessels. Medium-sized–vessel vasculitides involve small and medium-sized arteries without involvement of small vessels. The small-vessel vasculitides are the only types now recognized to involve small vessels.3 An important distinction is that the small-vessel vasculitides frequently also involve medium-sized vessels — small-to-medium-sized arteries — and can thus, on occasion, result in the formation of microaneurysms (or micro-pseudoaneurysms) visible on angiography.11,12,13 Because evaluation of the biopsy specimens of this patient's skin and colon showed involvement of small vessels, the vasculitis in this case was classified as a small-vessel vasculitis.
The small-vessel vasculitides include pauci-immune and immune-complex vasculitides.3 Two of the pauci-immune vasculitides, Wegener's granulomatosis and the Churg–Strauss syndrome, are histologically distinguished by the presence of granulomatous inflammation, and in the case of the Churg–Strauss syndrome, by prominent eosinophil infiltration. Neither was seen in this case.
Small-vessel immune-complex vasculitides are characterized histologically by the presence of immune-complex deposition in vascular walls, as can be observed by immunofluorescence. These entities are further characterized on the basis of the source of the immune complexes (Table 4). Although drug-induced immune-complex vasculitis is often limited to the skin, so-called cutaneous leukocytoclastic angiitis, systemic symptoms are present in 10 to 20 percent of cases, and occasionally severe, life-threatening systemic complications occur.14,15,16,17 The -lactam antibiotics are one of the classes of medication most frequently implicated in drug-induced immune-complex vasculitis.
Given the lack of clinical or histologic evidence of prior episodes of vasculitis, it is quite possible that the vasculitis in this case is an immune-complex vasculitis caused by either the recent infection or the antibiotic treatment.17,18 The presence of immune complexes in the tissue could not be established, since frozen tissue for immunofluorescence analysis was not available. Since 10 percent of patients with microscopic polyangiitis are negative for ANCA, this disorder cannot be definitively excluded, nor can the possibility that the vasculitis was secondary to ulcerative colitis.19
Discussion of Management
Dr. Harris: Drs. Johnsen, Steere, and Foster, can you tell us how you treated this patient?
Dr. Alyssa K. Johnsen (Internal Medicine): The therapeutic interventions that were initiated on the patient's admission included a regimen of broad-spectrum antibiotics and acyclovir. In addition, because of the severity of the eye findings and the potential for bilateral vision loss, immunosuppressive therapy with high-dose, intravenously administered methylprednisolone was initiated. However, the tests for viral, bacterial, and fungal pathogens in the biopsy specimen of the vitreous of the patient's right eye were all negative, and therefore, antiviral and antibiotic agents were discontinued.
Dr. Steere: Because of the severe, widespread vasculitic process and the strong clinical suspicion of polyarteritis nodosa, according to current recommendations for treatment, the patient received cyclophosphamide and methylprednisolone, both administered intravenously.20 In addition, plasmapheresis was performed three times. His systemic symptoms and the vision in his left eye improved rapidly. Methylprednisolone administered intravenously was changed to prednisone administered orally and tapered gradually. On hospital day 22, he was discharged and was instructed to take orally administered prednisone (50 mg, four times a day), and two weeks later, he was to begin taking cyclophosphamide orally (2 mg per kilogram of body weight per day). At the time of his discharge, his visual acuity was 20/60 in his left eye, and he was virtually blind in the right eye.
Dr. Foster: After the patient was discharged, the prednisone was quickly tapered to a dose of 60 mg daily, and this dose was subsequently tapered by 10 mg each month over the next six months. He continued to take cyclophosphamide orally (in various doses) for six months, after which azathioprine was substituted and continued for six months. In addition, prophylactic trimethoprim–sulfamethoxazole was prescribed to prevent the development of pneumocystis infection. While taking the prednisone, the patient required insulin coverage for diabetes. Within 10 months, the visual acuity in his left eye improved to 20/25. He did not recover vision in his right eye.
Dr. Steere: Although recurrence is common in patients with Wegener's granulomatosis or microscopic polyangiitis, recurrence occurs in only about 10 percent of cases of polyarteritis nodosa. The intriguing question remains whether this patient's vasculitis was triggered by therapy with amoxicillin–clavulanate. In most cases, drug-induced vasculitis is treated simply by stopping the offending medication. In severe cases with systemic involvement, such as this patient had, corticosteroids may be given until the symptoms resolve. Regardless of the specific diagnosis in this patient, who had remission within a few months, it would seem reasonable to stop immunosuppressive therapy after one year, and it would seem prudent that he not receive treatment with amoxicillin–clavulanate again.
Anatomical Diagnosis
Systemic small-vessel vasculitis involving small vessels and small arteries; probably immune-complex vasculitis.
Dr. Kathiresan reports receiving consulting fees from Pfizer Health Solutions and Novartis and grant support from GlaxoSmithKline Cardiovascular Research Foundation; he was the recipient of an Adult Cardiology Research Fellowship supported by the American College of Cardiology and Merck. Dr. Kelsey reports receiving lecture fees from AstraZeneca and research support from Pentax.
Source Information
From the Cardiology Division (S.K.), the Gastroenterology Division (P.B.K.), and the Rheumatology Division (A.C.S.), Department of Medicine, and the Departments of Radiology (M.S.C.) and Pathology (J.R.S.), Massachusetts General Hospital; the Department of Ophthalmology, Massachusetts Eye and Ear Infirmary (C.S.F.); and the Departments of Medicine (S.K., P.B.K., A.C.S.), Ophthalmology (C.S.F.), Radiology (M.S.C.), and Pathology (J.R.S.), Harvard Medical School — all in Boston.
References
Levin S, Goodman LJ. An approach to acute fever and rash (AFR) in the adult. Curr Clin Top Infect Dis 1995;15:19-75.
Bologna JL, Braverman IM. Skin manifestations of internal disease. In: Braunwald E, Fauci AS, Kasper DL, Hauser SL, Longo DL, Jameson JL, eds. Harrison's principles of internal medicine. 15th ed. New York: McGraw-Hill, 2001:329.
Jennette JC, Falk RJ. Small-vessel vasculitis. N Engl J Med 1997;337:1512-1523.
Case Records of the Massachusetts General Hospital (Case 7-2003). N Engl J Med 2003;348:834-843.
Jennette JC, Falk RJ, Andrassy K, et al. Nomenclature of systemic vasculitides: proposal of an international consensus conference. Arthritis Rheum 1994;37:187-192.
Morgan CM, Foster CS, D'Amico DJ, Gragoudas ES. Retinal vasculitis in polyarteritis nodosa. Retina 1986;6:205-209.
Lotti TM, Comacchi C, Ghersetich I. Cutaneous necrotizing vasculitis: relation to systemic disease. Adv Exp Med Biol 1999;455:115-125.
Lightfoot RW Jr, Michel BA, Bloch DA, et al. The American College of Rheumatology criteria for the classification of polyarteritis nodosa. Arthritis Rheum 1990;33:1088-1093.
Sakane T, Takeno M, Suzuki N, Inaba G. Beh?et's disease. N Engl J Med 1999;341:1284-1291.
Bansal R, Aggarwal P, Handa R, Biswas A, Bandhu S, Wali JP. Ulcerative colitis associated with Takayasu arteritis. Int J Cardiol 2003;88:91-93.
Moutsopoulos HM, Avgerinos PC, Tsampoulas CG, Katsiotis PA. Selective renal angiography in Wegener's granulomatosis. Ann Rheum Dis 1983;42:192-195.
Costedoat-Chalumeau N, Cacoub P, Maisonobe T, et al. Renal microaneurysms in three cases of hepatitis C virus-related vasculitis. Rheumatology (Oxford) 2002;41:708-710.
Inoue M, Akikusa B, Masuda Y, Kondo Y. Demonstration of microaneurysms at the interlobular arteries of the kidneys in microscopic polyangiitis: a three-dimensional study. Hum Pathol 1998;29:223-227.
ten Holder SM, Joy MS, Falk RJ. Cutaneous and systemic manifestations of drug-induced vasculitis. Ann Pharmacother 2002;36:130-147.
Somer T, Finegold SM. Vasculitides associated with infections, immunization, and antimicrobial drugs. Clin Infect Dis 1995;20:1010-1036.
Blanco R, Martinez-Taboada VM, Rodriguez-Valverde V, Garcia-Fuentes M. Cutaneous vasculitis in children and adults: associated diseases and etiologic factors in 303 patients. Medicine (Baltimore) 1998;77:403-418.
Hannedouche T, Fillastre JP. Penicillin-induced hypersensitivity vasculitides. J Antimicrob Chemother 1987;20:3-5.
Lum PNL, Woo PCY, Wong SSY, Yuen K. Leukocytoclastic vasculitis complicating Klebsiella pneumonia bacteremia. Diagn Microbiol Infect Dis 2000;37:275-277.
Iannone F, Scioscia C, Musio A, Piscitelli D, Lapadula G. Leucocytoclastic vasculitis as onset symptom of ulcerative colitis. Ann Rheum Dis 2003;62:785-786.
Fauci AS, Haynes BF, Katz P. The spectrum of vasculitis: clinical, pathologic, immunologic and therapeutic considerations. Ann Intern Med 1978;89:660-676.(Sekar Kathiresan, M.D., P)
A 38-year-old man was admitted to this hospital because of blurred vision and fever.
The patient had ulcerative colitis but had been well while receiving treatment with mercaptopurine and balsalazide until two weeks before admission, when a chronically draining pilonidal cyst was excised. Amoxicillin–clavulanate was prescribed after the operation. One week later, the patient noted the gradual onset of severe right anterior thigh pain; he had difficulty walking and a temperature as high as 38.3°C. He was admitted to another hospital.
On examination, the temperature was 38.8°C, and the abdomen was moderately distended but not tender. The right thigh was swollen, tender, and erythematous, and the patient could not flex his leg. An ultrasonographic study of the right thigh showed a superficial saphenous-vein thrombosis extending into the pelvis. Computed tomographic (CT) scanning of the abdomen and pelvis revealed a fluid collection and stranding of the soft tissues adjacent to the cecum. Inflammatory changes extended within the retroperitoneum along the inferior vena cava and the right external iliac vein, anterior and lateral to the psoas muscle. Cultures of multiple specimens of blood and urine were negative for pathogens. The white-cell count was 18,900 per cubic millimeter, with 82 percent neutrophils and 6 percent band forms. The serum creatinine level was 1.7 mg per deciliter (150.3 μmol per liter). Ampicillin, metronidazole, and heparin were administered. Klebsiella species and candida grew from cultures obtained from the pilonidal cyst. Ampicillin was discontinued, and piperacillin and caspofungin were started.
Daily spiking fevers continued. A magnetic resonance imaging (MRI) study of the right thigh showed two ring-enhancing lesions within the muscle tissue, consistent with the presence of inflammation or abscess. There was no evidence of osteomyelitis or deep-vein thrombosis. Heparin was discontinued. Fever persisted; vancomycin was started, and mercaptopurine was discontinued.
On the sixth hospital day, the patient became short of breath and could not complete full sentences. The next day, he experienced double vision, first in his right eye and then in his left. The conjunctivae were injected, and a topical antibiotic was administered. He remained febrile and short of breath and was transferred to this hospital.
The patient worked in an office and did not use alcohol or tobacco. A diagnosis of ulcerative colitis had been made 18 years earlier. It was minimally active until eight months before this admission, when the patient had a flare-up of the disease. The symptoms responded to a tapering dose of prednisone, which was discontinued three months before admission.
On examination, he was in mild distress. The temperature was 37.5°C, the blood pressure 115/52 mm Hg, the heart rate 109 beats per minute, the respiratory rate 24 breaths per minute, and the oxygen saturation 95 percent while the patient was breathing room air. The oral cavity was clear; the sclerae were injected, and the conjunctivae were pink. The visual acuity was 20/800 in the right eye and 20/200 in the left eye. There were 4+ cells in the anterior chambers of both eyes. Funduscopic examination showed 1+ to 2+ vitritis bilaterally, with diffuse retinal hemorrhages and perivascular exudates. Examination of the chest and abdomen showed no abnormalities. There was serosanguinous drainage from the pilonidal-cyst wound. There was 2+ edema of both legs; the peripheral pulses were easily palpable, and there was no thigh tenderness. Raised purpuric lesions were present on the right leg. The right third proximal interphalangeal joint, left elbow, and right knee were slightly warm to the touch, without effusion. The neurologic examination revealed no abnormalities. Levels of serum electrolytes, calcium, phosphorus, magnesium, urea nitrogen, and creatinine were normal; the results of other laboratory tests are shown in Table 1 and Table 2.
Table 1. Blood Chemical Values.
Table 2. Hematologic and Serologic Laboratory Values.
A transthoracic ultrasonographic study of the heart and both MRI and magnetic resonance angiography of the brain were performed and showed no abnormalities. High-dose methylprednisolone, acyclovir, and heparin were given intravenously. Cultures of specimens of skin, the cyst, and blood for viral, fungal, and bacterial pathogens yielded no growths.
Over the next few days, abdominal pain and pain in the left testis developed; the patient passed bloody diarrheal stools. A colonoscopic examination revealed multiple deep ulcers in the ascending colon and cecum, with shallow ulcers in the transverse colon, distal sigmoid, and rectum.
Diagnostic procedures were performed.
Differential Diagnosis
Dr. Sekar Kathiresan: This patient's clinical course suggests a bimodal illness that began with superficial thrombophlebitis after the resection of a pilonidal cyst, followed by a systemic illness with fever, rash, arthralgias, and visual loss. To identify a unifying cause of these disparate symptoms, the general medical team, of which I was a member, focused on the patient's three active medical issues on transfer to this hospital: fever, rash, and eye disease. The primary problem was whether his symptoms represented an infectious or an immunologically mediated inflammatory process.
Schemes that have been designed to classify the causes of fever and rash generally start with the nature of the rash.1 In this patient, the eruption was purpuric, indicating extravasation of red cells. Purpura can be caused by disorders of platelets, the coagulation cascade, or vascular integrity, such as vascular infection and noninfectious vasculitis. Purpura may be divided into two major categories: nonpalpable and palpable (Table 3).2 Most causes lead to flat or nonpalpable purpura, whereas far fewer causes lead to palpable lesions. The palpability is due to localized swelling associated with the purpuric lesions; this usually occurs when an inflammatory infiltrate of the vessel wall leads to the bleeding. In this patient, the purpuric lesions involving the right leg were clearly palpable (Figure 1).
Table 3. Causes of Purpura.
Figure 1. Typical Lesions of Palpable Purpura (from the Arm of Another Patient).
Purple, raised skin lesions are present. The lesions are palpable because of localized swelling that usually occurs when an inflammatory infiltrate of the vessel wall leads to the bleeding, as in infection or vasculitis. (Photograph courtesy of Dr. Marc Sabatine, Division of Cardiology, Brigham and Women's Hospital.)
Palpable purpura is caused by a limited number of infections that directly involve the vascular wall and by noninfectious immune-mediated processes. The infections include acute meningococcemia (purpura fulminans), disseminated gonococcal infection, Rocky Mountain spotted fever, and subacute bacterial endocarditis. Of the noninfectious vasculitides, small-vessel vasculitis (involving vessels smaller than arteries, such as arterioles, venules, or capillaries) characteristically leads to palpable purpura.3
The disease in the eye could also have had either an infectious or an inflammatory cause. Viral infections, including infection with a member of the herpes simplex virus family, such as herpes simplex virus or varicella–zoster virus, can cause vitritis and retinal hemorrhage. Septic thrombophlebitis, which the patient may have had, could have been complicated by bacterial endocarditis with embolic involvement of the eye, and this possibility was considered.4 An inflammatory cause of eye disease, including a primary systemic vasculitis with eye involvement, was also considered.
Overall, the constellation of fever, rash, and severe eye disease led us to consider either bacterial endocarditis or a noninfectious vasculitis as the most probable unifying diagnosis. Prominent involvement of the eye is rare in meningococcemia or gonococcemia. The patient did not report headache, a history of a tick bite, or recent travel, and thus Rocky Mountain spotted fever seemed to be unlikely. Herpesvirus infections may cause severe eye disease, but we found that tying in the other systemic manifestations with a herpesvirus infection was difficult. This patient had several risk factors for vasculitis, including a recent infection, antibiotic therapy, and inflammatory bowel disease.
We ordered blood cultures to evaluate the possibility of bacterial endocarditis, meningococcemia, and gonococcemia. Measurement of the patient's complement levels, which are low in systemic vasculitis, was requested. Measurement of antineutrophil cytoplasmic antibodies (ANCA) and cryoglobulins, serologic assays for hepatitis B and C, and tests for antinuclear antibody and rheumatoid factor were obtained to look for specific types of small-vessel vasculitis. A transthoracic echocardiogram was obtained to detect any vegetations on cardiac valves. We asked for consultation from our colleagues in ophthalmology, rheumatology, and gastroenterology.
Dr. C. Stephen Foster: A diagnostic vitrectomy was performed on the right eye because of the concern about infectious processes; there was no clinical evidence of features, such as the acute retinal necrosis syndrome, that are seen in patients with varicella–zoster virus or herpes simplex virus infection. In the fundus of the right eye (Figure 2A), there were multifocal hemorrhage, perivascular exudates indicative of vasculitis, optic neuropathy, and a macular infarct. All quadrants were involved, and the process predominantly involved arterioles rather than veins. There was pallor of the left optic nerve and foci of frank infarction of the retina. Fluorescein angiography was impossible in the right eye, but in the left eye (Figure 2B) staining of the vasculature in the late phase was apparent, in a pattern indicative of retinal vasculitis, specifically retinal arteritis.
Figure 2. Ophthalmologic Examination on Admission.
In the fundus of the right eye (Panel A), there are multifocal hemorrhage (arrows), perivascular exudates indicative of vasculitis (arrowheads), optic neuropathy, and a macular infarct. All quadrants are involved, and the process predominantly involves arterioles rather than veins. Fluorescein angiography of the left eye (Panel B) shows staining of the vasculature in the late phase (arrows) in a pattern indicative of retinal vasculitis.
The causes of retinal vasculitides fall into four categories: infection, trauma, conditions associated with cancer, and autoimmune diseases. Vasculitis that is secondary to trauma or cancer is rare and improbable in this patient. Infectious and autoimmune diseases cause most cases of retinal vasculitis. In distinguishing among these, the predominant involvement of arteries or veins can be useful. If the arteries are predominantly involved, systemic lupus erythematosus, polyarteritis, the Churg–Strauss syndrome, and so-called frosted-branch angiitis, as well as infection with varicella–zoster virus, herpes simplex virus, and syphilis, are the primary entities to consider. If the veins are predominantly involved, sarcoidosis is by far the most common cause. Other considerations include Eales disease, a localized phlebitis that is often associated with tuberculosis, a localized autoimmune ocular disease called birdshot retinochoroidopathy, and infections with pathogens such as parvovirus, toxoplasma, and the human immunodeficiency virus. Three entities involve arteries and veins relatively equally: multiple sclerosis, Beh?et's disease, and Wegener's granulomatosis.
In this patient with predominantly arterial involvement and a systemic illness, without evidence of infection, my leading diagnosis based on the ocular findings would be polyarteritis nodosa.
Dr. Allen C. Steere: In this patient, a complex array of symptoms developed two weeks after the excision of a pilonidal cyst and the start of antibiotic treatment. These symptoms included devastating visual problems, pain in the right thigh, subtle oligoarticular arthritis, palpable purpura on the right leg, and during his hospitalization here, bloody diarrhea, abdominal pain, elevated pancreatic enzymes, and testicular pain. The ophthalmogist has diagnosed a retinal vasculitis, and the skin lesions of palpable purpura on his right leg were suggestive of a small-vessel vasculitis. As the rhematology consultant, I had to consider what category of vasculitis this disorder best fit, and how to treat the patient.
The classification of vasculitides that was adopted by the 1994 Chapel Hill Consensus Conference is delineated in Table 4.5 The vasculitides are categorized according to the size of the blood vessels affected and the histologic appearance of the lesions. Broadly, the different types are divided into those that affect the large, medium-sized, and small blood vessels, and the latter group is further divided into pauci-immune and immune-complex–mediated types. It should be acknowledged that no classification scheme for the vasculitides has been perfect. Some types of vasculitis affect more than one size of vessel, and some diseases are not included in the current listing.
Table 4. Classification of Vasculitis.
Retinal vasculitis is a well-described manifestation of the most common medium-sized–vessel vasculitis, polyarteritis nodosa.6 Moreover, in this patient there was evidence of pancreatic involvement, muscle inflammation in the right leg, testicular pain, and subtle arthritis, all of which occur in polyarteritis nodosa.3 Abdominal pain and bloody diarrhea may also be manifestations of polyarteritis, although because of the patient's history of ulcerative colitis, the cause of his current bowel symptoms was unclear. However, he had palpable purpura, which suggested involvement of microscopic blood vessels.7 Before the 1994 classification was available, the term "polyarteritis nodosa" was used to describe a broader array of systemic necrotizing vasculitides, and a patient with symptoms such as retinal vasculitis and palpable purpura would probably have been classified as having either polyarteritis nodosa or the polyangiitis overlap syndrome.8 However, polyarteritis nodosa is now defined more strictly as a vasculitis affecting only medium-sized vessels, and the term "polyangiitis overlap syndrome" has been eliminated from the classification.
A second consideration, then, was whether the patient had one of the three forms of pauci-immune vasculitis that are often associated with involvement of both small and medium-sized vessels and a positive test for ANCA.3 Wegener's granulomatosis often causes destructive lesions in the upper respiratory tract and lungs, along with glomerulonephritis. The Churg–Strauss syndrome may be associated with purpura, arthritis, or abdominal pain, but the characteristic clinical feature is hypereosinophilic asthma, which was missing in this case. Finally, microscopic polyangiitis is associated with renal disease in most patients and often associated with pulmonary involvement. In contrast, this patient had a negative test for ANCA, and he did not have the common clinical features of these diseases.
Beh?et's syndrome is associated with inflammation in veins and arteries of all sizes.9 This entity might explain the patient's eye findings, subcutaneous thrombophlebitis, bloody diarrhea, and subtle arthritis. However, he did not have oral aphthous ulcers, which are characteristic of Beh?et's syndrome. Is there a link between ulcerative colitis and vasculitis? Ulcerative colitis has been associated with both small-vessel vasculitis (leukocytoclastic vasculitis) and large-vessel vasculitis (Takayasu's arteritis).10 However, the type of widespread vasculitis seen in this patient has not been reported.
Could this patient have a particularly aggressive form of immune-complex–mediated small-vessel vasculitis?3 As the name implies, the small-vessel vasculitides usually involve only small blood vessels, but they may also affect medium-sized arteries. Henoch–Sch?nlein purpura, the most common vasculitis of childhood, is associated with purpuric rash on the legs, arthritis in large joints, abdominal cramping with bloody diarrhea, and IgA-containing immune complexes in the affected tissues. Although this entity may occur in adults, the patient's age made this possibility unlikely. Mixed cryoglobulinemia may cause palpable purpura and organ dysfunction, but the patient had no clinical signs of cryoglobulins or of hepatitis C infection, which is the usual cause of this entity.
Could he have had immune-complex–mediated vasculitis resulting from either the klebsiella infection of his pilonidal cyst or from the amoxicillin–clavulanate therapy used to treat it? Drug-induced vasculitis usually develops within 7 to 21 days after treatment begins,3 which fits the timing in this case. Both of these vasculitides usually affect only the skin, but they may be associated with systemic disease, including gastrointestinal hemorrhage.3 However, this patient's widespread involvement, including retinal vasculitis, as far as I know, goes beyond previous reports of clinical manifestations of drug-induced vasculitis.
Thus, this patient exemplifies the challenge often faced by clinicians in the diagnosis of vasculitis, which is that some patients have conditions that do not fit well into defined categories. Regardless, it was clear that the vasculitis in this case was very aggressive, which necessitated an aggressive approach to diagnosis and treatment. For this reason, we performed biopsies of the skin and colon, followed by abdominal angiography.
Dr. Peter B. Kelsey: I was asked to see this patient to address the question of whether his new onset of diarrhea and lower gastrointestinal bleeding was due to a flare-up of his ulcerative colitis or to another process. His history was not typical for a patient with chronic ulcerative colitis. He had been asymptomatic while receiving mild immunosuppressive therapy since his initial diagnosis 18 years earlier. Eight months before admission, he had what was interpreted as a flare-up, which was rapidly controlled with prednisone. In the weeks before this admission, he had had no recurrence of symptoms of ulcerative colitis. Thus, before the colonoscopy, I was concerned that there might be another cause of his gastrointestinal symptoms.
At colonoscopy, I saw a background of normal colonic mucosa punctuated by discrete areas of ulceration. There was no evidence of chronic inflammation, such as scarring or thickening of the wall, to suggest past episodes of healed colitis. Several of the ulcers had visible blood vessels at their bases, one of which had a large adherent clot. Because the bowel appeared very fragile and because there were multiple potential sites of bleeding, I did not attempt to ligate any of the exposed blood vessels. I obtained several biopsy specimens, which were submitted for microscopical examination.
Dr. Kathiresan: While we were awaiting the results of the evaluation of the biopsy specimens, we requested an angiogram.
Dr. Marcio S. Curvelo: Visceral digital-subtraction arteriography was performed (Figure 3). There were alternating areas of stenosis and ectasia in the distal branches of the right hepatic artery and the superior rectal artery, with microaneurysms in the distal branches of the superior mesenteric artery and both renal arteries. The angiographic finding of areas of alternating stenosis and ectasia and microaneurysms involving medium-sized vessels in several vascular beds is consistent with a diagnosis of a vasculitis. Microaneurysms of the renal-artery branches are considered fairly specific for polyarteritis nodosa; however, the differential diagnosis includes vasculitides that can involve both small and medium-sized vessels.
Figure 3. Selective Renal Arteriogram.
There are areas of vessel irregularity and microaneurysms consistent with a diagnosis of a vasculitis. This finding is fairly specific but not pathognomonic of polyarteritis nodosa.
Dr. Nancy Lee Harris: Could we have the medical students' diagnosis?
A Medical Student: The constellation of symptoms and laboratory findings indicates an acute inflammatory process and suggests a systemic vasculitis. We thought that in this patient with long-standing ulcerative colitis, the emergence of a vasculitis, such as polyarteritis nodosa, could have been masked by his immunosuppressive therapy. On the withdrawal of the prednisone before surgery, the vasculitis became clinically manifest.
Clinical Diagnosis
Systemic vasculitis, not definitely classifiable: polyarteritis nodosa versus immune-complex–mediated vasculitis (infection-induced or drug-induced).
Pathological Discussion
Dr. James R. Stone: The diagnostic procedures were biopsies of the skin and colon. Histologic examination of the skin-biopsy specimen (Figure 4A and Figure 4B) revealed a necrotizing leukocytoclastic vasculitis involving small vessels (arterioles, capillaries, and venules) as well as small arteries (Figure 4C), which are classified as medium-sized vessels. Examination of the colonic biopsy specimen (Figure 5) revealed fragments of fibrin and inflammatory cells, indicating the presence of an ulcer. There was a leukocytoclastic vasculitis involving small vessels of the colonic submucosa. Neither biopsy showed eosinophil infiltration, granuloma formation, or evidence of prior vasculitic episodes (vascular scarring).
Figure 4. Skin-Biopsy Specimens (Hematoxylin and Eosin).
A low-power view (Panel A) shows the location of the superficial dermal vasculature (arrow) in relation to the deep dermal vasculature (arrowhead). A leukocytoclastic vasculitis involving arterioles, capillaries, and venules is revealed in the superficial vascular plexus (Panel B). The deep vascular plexus (Panel C) contains a small artery involved by vasculitis.
Figure 5. Colon-Biopsy Specimens (Hematoxylin and Eosin).
The mucosa is intact (Panel A), but blood vessels in the submucosa show involvement by leukocytoclastic vasculitis (Panel B).
The current approach to the subclassification of systemic vasculitides starts with consideration of the sizes of the vessels involved (Table 4).3,5 Large-vessel vasculitides principally involve the aorta and major arterial branches to the head and extremities, but they may also involve medium-sized vessels. Medium-sized–vessel vasculitides involve small and medium-sized arteries without involvement of small vessels. The small-vessel vasculitides are the only types now recognized to involve small vessels.3 An important distinction is that the small-vessel vasculitides frequently also involve medium-sized vessels — small-to-medium-sized arteries — and can thus, on occasion, result in the formation of microaneurysms (or micro-pseudoaneurysms) visible on angiography.11,12,13 Because evaluation of the biopsy specimens of this patient's skin and colon showed involvement of small vessels, the vasculitis in this case was classified as a small-vessel vasculitis.
The small-vessel vasculitides include pauci-immune and immune-complex vasculitides.3 Two of the pauci-immune vasculitides, Wegener's granulomatosis and the Churg–Strauss syndrome, are histologically distinguished by the presence of granulomatous inflammation, and in the case of the Churg–Strauss syndrome, by prominent eosinophil infiltration. Neither was seen in this case.
Small-vessel immune-complex vasculitides are characterized histologically by the presence of immune-complex deposition in vascular walls, as can be observed by immunofluorescence. These entities are further characterized on the basis of the source of the immune complexes (Table 4). Although drug-induced immune-complex vasculitis is often limited to the skin, so-called cutaneous leukocytoclastic angiitis, systemic symptoms are present in 10 to 20 percent of cases, and occasionally severe, life-threatening systemic complications occur.14,15,16,17 The -lactam antibiotics are one of the classes of medication most frequently implicated in drug-induced immune-complex vasculitis.
Given the lack of clinical or histologic evidence of prior episodes of vasculitis, it is quite possible that the vasculitis in this case is an immune-complex vasculitis caused by either the recent infection or the antibiotic treatment.17,18 The presence of immune complexes in the tissue could not be established, since frozen tissue for immunofluorescence analysis was not available. Since 10 percent of patients with microscopic polyangiitis are negative for ANCA, this disorder cannot be definitively excluded, nor can the possibility that the vasculitis was secondary to ulcerative colitis.19
Discussion of Management
Dr. Harris: Drs. Johnsen, Steere, and Foster, can you tell us how you treated this patient?
Dr. Alyssa K. Johnsen (Internal Medicine): The therapeutic interventions that were initiated on the patient's admission included a regimen of broad-spectrum antibiotics and acyclovir. In addition, because of the severity of the eye findings and the potential for bilateral vision loss, immunosuppressive therapy with high-dose, intravenously administered methylprednisolone was initiated. However, the tests for viral, bacterial, and fungal pathogens in the biopsy specimen of the vitreous of the patient's right eye were all negative, and therefore, antiviral and antibiotic agents were discontinued.
Dr. Steere: Because of the severe, widespread vasculitic process and the strong clinical suspicion of polyarteritis nodosa, according to current recommendations for treatment, the patient received cyclophosphamide and methylprednisolone, both administered intravenously.20 In addition, plasmapheresis was performed three times. His systemic symptoms and the vision in his left eye improved rapidly. Methylprednisolone administered intravenously was changed to prednisone administered orally and tapered gradually. On hospital day 22, he was discharged and was instructed to take orally administered prednisone (50 mg, four times a day), and two weeks later, he was to begin taking cyclophosphamide orally (2 mg per kilogram of body weight per day). At the time of his discharge, his visual acuity was 20/60 in his left eye, and he was virtually blind in the right eye.
Dr. Foster: After the patient was discharged, the prednisone was quickly tapered to a dose of 60 mg daily, and this dose was subsequently tapered by 10 mg each month over the next six months. He continued to take cyclophosphamide orally (in various doses) for six months, after which azathioprine was substituted and continued for six months. In addition, prophylactic trimethoprim–sulfamethoxazole was prescribed to prevent the development of pneumocystis infection. While taking the prednisone, the patient required insulin coverage for diabetes. Within 10 months, the visual acuity in his left eye improved to 20/25. He did not recover vision in his right eye.
Dr. Steere: Although recurrence is common in patients with Wegener's granulomatosis or microscopic polyangiitis, recurrence occurs in only about 10 percent of cases of polyarteritis nodosa. The intriguing question remains whether this patient's vasculitis was triggered by therapy with amoxicillin–clavulanate. In most cases, drug-induced vasculitis is treated simply by stopping the offending medication. In severe cases with systemic involvement, such as this patient had, corticosteroids may be given until the symptoms resolve. Regardless of the specific diagnosis in this patient, who had remission within a few months, it would seem reasonable to stop immunosuppressive therapy after one year, and it would seem prudent that he not receive treatment with amoxicillin–clavulanate again.
Anatomical Diagnosis
Systemic small-vessel vasculitis involving small vessels and small arteries; probably immune-complex vasculitis.
Dr. Kathiresan reports receiving consulting fees from Pfizer Health Solutions and Novartis and grant support from GlaxoSmithKline Cardiovascular Research Foundation; he was the recipient of an Adult Cardiology Research Fellowship supported by the American College of Cardiology and Merck. Dr. Kelsey reports receiving lecture fees from AstraZeneca and research support from Pentax.
Source Information
From the Cardiology Division (S.K.), the Gastroenterology Division (P.B.K.), and the Rheumatology Division (A.C.S.), Department of Medicine, and the Departments of Radiology (M.S.C.) and Pathology (J.R.S.), Massachusetts General Hospital; the Department of Ophthalmology, Massachusetts Eye and Ear Infirmary (C.S.F.); and the Departments of Medicine (S.K., P.B.K., A.C.S.), Ophthalmology (C.S.F.), Radiology (M.S.C.), and Pathology (J.R.S.), Harvard Medical School — all in Boston.
References
Levin S, Goodman LJ. An approach to acute fever and rash (AFR) in the adult. Curr Clin Top Infect Dis 1995;15:19-75.
Bologna JL, Braverman IM. Skin manifestations of internal disease. In: Braunwald E, Fauci AS, Kasper DL, Hauser SL, Longo DL, Jameson JL, eds. Harrison's principles of internal medicine. 15th ed. New York: McGraw-Hill, 2001:329.
Jennette JC, Falk RJ. Small-vessel vasculitis. N Engl J Med 1997;337:1512-1523.
Case Records of the Massachusetts General Hospital (Case 7-2003). N Engl J Med 2003;348:834-843.
Jennette JC, Falk RJ, Andrassy K, et al. Nomenclature of systemic vasculitides: proposal of an international consensus conference. Arthritis Rheum 1994;37:187-192.
Morgan CM, Foster CS, D'Amico DJ, Gragoudas ES. Retinal vasculitis in polyarteritis nodosa. Retina 1986;6:205-209.
Lotti TM, Comacchi C, Ghersetich I. Cutaneous necrotizing vasculitis: relation to systemic disease. Adv Exp Med Biol 1999;455:115-125.
Lightfoot RW Jr, Michel BA, Bloch DA, et al. The American College of Rheumatology criteria for the classification of polyarteritis nodosa. Arthritis Rheum 1990;33:1088-1093.
Sakane T, Takeno M, Suzuki N, Inaba G. Beh?et's disease. N Engl J Med 1999;341:1284-1291.
Bansal R, Aggarwal P, Handa R, Biswas A, Bandhu S, Wali JP. Ulcerative colitis associated with Takayasu arteritis. Int J Cardiol 2003;88:91-93.
Moutsopoulos HM, Avgerinos PC, Tsampoulas CG, Katsiotis PA. Selective renal angiography in Wegener's granulomatosis. Ann Rheum Dis 1983;42:192-195.
Costedoat-Chalumeau N, Cacoub P, Maisonobe T, et al. Renal microaneurysms in three cases of hepatitis C virus-related vasculitis. Rheumatology (Oxford) 2002;41:708-710.
Inoue M, Akikusa B, Masuda Y, Kondo Y. Demonstration of microaneurysms at the interlobular arteries of the kidneys in microscopic polyangiitis: a three-dimensional study. Hum Pathol 1998;29:223-227.
ten Holder SM, Joy MS, Falk RJ. Cutaneous and systemic manifestations of drug-induced vasculitis. Ann Pharmacother 2002;36:130-147.
Somer T, Finegold SM. Vasculitides associated with infections, immunization, and antimicrobial drugs. Clin Infect Dis 1995;20:1010-1036.
Blanco R, Martinez-Taboada VM, Rodriguez-Valverde V, Garcia-Fuentes M. Cutaneous vasculitis in children and adults: associated diseases and etiologic factors in 303 patients. Medicine (Baltimore) 1998;77:403-418.
Hannedouche T, Fillastre JP. Penicillin-induced hypersensitivity vasculitides. J Antimicrob Chemother 1987;20:3-5.
Lum PNL, Woo PCY, Wong SSY, Yuen K. Leukocytoclastic vasculitis complicating Klebsiella pneumonia bacteremia. Diagn Microbiol Infect Dis 2000;37:275-277.
Iannone F, Scioscia C, Musio A, Piscitelli D, Lapadula G. Leucocytoclastic vasculitis as onset symptom of ulcerative colitis. Ann Rheum Dis 2003;62:785-786.
Fauci AS, Haynes BF, Katz P. The spectrum of vasculitis: clinical, pathologic, immunologic and therapeutic considerations. Ann Intern Med 1978;89:660-676.(Sekar Kathiresan, M.D., P)