A 33-year-old man with type 1 diabetes mellitus (DM) since the age of 14, dyslipidemia and hypertension, presented with a 2-week history of progressive pain and swelling of his left lower extremity. He denied fever, chills, insect bites, trauma, or insulin administration in that limb. He was sexually inactive without history of recent travel or a preceding diarrheal or upper respiratory tract illness. His DM was poorly controlled, complicated with retinopathy for which he received prior laser treatment in both eyes, and nephropathy with laboratory evidence of significant proteinuria. There was no clinical evidence of diabetic neuropathy. He had been receiving insulin through a pump since the age of 25 years. His other home medications included aspirin 81 mg and lisinopril 10 mg daily. He was not on statin therapy for dyslipidemia because of elevation of hepatic transaminases.
On admission, he had a blood pressure of 126/84 mmHg, a regular pulse of 104 /min, and a temperature of 36.1°C. Examination of the left lower extremity revealed a tenderness to touch, redness, warmth, and swelling starting from mid-thigh down with no areas of fluctuance, in addition to a left knee effusion with marked limitation of range of motion. The distal arterial pulse and perfusion of both lower extremities were intact. Laboratory workup revealed a white blood cell count of 15 700/mcl, erythrocyte sedimentation rate 125 mm/h, C-reactive protein 12 mg/dl, and urine microalbumin/creatinine ratio of 1942 mg/g Cr. Hemoglobin A1C was 11.5% on admission and 12.9% 6 months earlier. There was a slight elevation of the aspartate transaminase 66 (N=10–49) and creatine phosphokinase (CPK) was normal. During his hospital course, his random blood sugar was satisfactory controlled with fasting levels between 80 and 164 mg/dl and pre-meal levels ranging between 59 and 211 mg/dl. Table 1 reveals the pertinent laboratory investigations on admission, 10 days and 2 months after discharge.
Blood cultures were negative and the knee effusion was noninflammatory with negative Gram stain and no crystals visualized on polarized light microscopy. Duplex examination of the left lower extremity revealed a normal arterial tree and ruled out deep venous thrombosis. Vancomycin and piperacillin/tazobactam were started for an initial working diagnosis of cellulitis. Owing to the lack of response to antibiotics, MRI was performed which revealed increased T2 signal throughout the quadriceps muscle with enhancement centered on the vastus intermedius and the surrounding soft tissue (Fig. 1) suggestive of diabetic myonecrosis. The patient was managed conservatively during his hospitalization and discharged on non-steroidal anti-inflammatory drugs with an emphasis about the importance of strict DM control. Two months later, he was seen by his primary care physician who noted that he was ambulating normally.
Epidemiology and pathogenesis
Diabetic myonecrosis is an uncommon and likely under-recognized complication of DM, first described by Angervall and Stener in 1965 1. Since then, around 100 cases have been reported. Middle-aged individuals appear to be more susceptible as evident in most case series 2 with a slight predilection for female patients and an approximate female to male ratio of 1.7 : 1 3. Interestingly, in a few described cases, it was the first presentation of DM. Angervall and Stener 1 hypothesized that ischemia due to arteriosclerosis and microangiopathy was the underlying pathophysiology, later, Banker and colleagues supported their hypothesis 4,5. Subsequent studies also suggested that muscle ischemia is a predisposing factor as evident by advanced peripheral arterial occlusive disease in the affected muscle; the resulting anoxia and subsequent mild compartmental syndrome further worsened the muscle ischemia 5,6. Nevertheless, Bjornskov et al.7 demonstrated a normal vasculature in histological examination of the affected muscle. Similarly, Kattapuram et al.2 excluded any evidence of vasculitis, embolic occlusion, or microvessel angiopathy on histopathological examination. Bjornskov et al. and Kattapuram et al. agreed that an interaction of endothelial damage from microangiopathy, activated coagulation cascade and impaired fibrinolysis is the likely underlying mechanism. The later hypothesis was also proposed by others 8,9. However, the exact underlying pathophysiologic mechanism for diabetic myonecrosis is still debated.
Diabetic myonecrosis presents with sudden onset of severe pain in the affected muscles. The pain is usually present at rest, worsens with passive or active movement and radiates to the surrounding structures as the knee or the calf. The disease usually has a unilateral distribution, however, in up to one-third of the cases, bilateral involvement can occur 10. There is no prior history of trauma and no correlation with local injection of insulin 11. The disease has characteristic predilection to the thigh muscles, yet, occasionally, the calf muscles can also be affected 3, but the upper limb is rarely involved. The muscles mostly affected include the quadriceps (62%), hip adductors (13%), hamstrings (8%), and hip flexors (2%) 12. Patients are usually afebrile with no signs of systemic infections and nontoxic looking. Symptoms usually persist for weeks followed by spontaneous resolution. Physical examination reveals swelling and tenderness of the affected muscles and occasionally, a palpable mass may be detected 13. The range of motion may be limited because of pain. Distal pulsations are intact and neurological exam including the motor power and sensation is normal.
Establishing the diagnosis
Laboratory workup in diabetic myonecrosis is nonspecific with limited value in confirming or ruling out the diagnosis. Leukocytosis, elevated erythrocyte sedimentation rate and elevated CPK enzyme were found in 35, 75, and 45% of cases, respectively, in a review by Kapur et al.10. Khoury et al.14 and Barohn and Kissel 15 confirmed the expected elevation of CPK enzyme when assessed early in the course of the disease reflecting active muscle breakdown. In his series, Kattapurum et al.2 noted elevation of CPK in only two of eight patients. The lack of the expected CPK elevation might be attributed to the delayed presentation of most patients. The majority of patients have uncontrolled DM evident with an elevated hemoglobin A1C level. Aspartate transaminase, alanine transaminase, and lactate dehydrogenase levels are usually normal.
Imaging studies such as plain X-ray, ultrasound, and computed tomography scan of the limbs are beneficial mainly in ruling out other pathologies. However, Nagdev et al.16 reported a case of diabetic myonecrosis diagnosed by ultrasound in the emergency room which revealed abnormal tissue architecture of the affected muscle. Delaney-Sathy et al.17 also demonstrated the utility of ultrasonography which revealed a well-defined, predominately hypoechoic lesion with mixed echogenicity. The lack of either a predominantly anechoic area or internal motion under transducer pressure was a discriminating factor between diabetic myonecrosis, abscess, or necrotic neoplasm.
The cornerstone for establishing the diagnosis is MRI. On T1-weighted images, there is an isointense swelling of the affected muscles with the surrounding facial planes mildly displaced but well maintained differentiating it from tumors, which tend to infiltrate and disrupt fascial planes 2. On T2-weighted images, there is increased signal intensity of the involved muscles with enhancement after administration of intravenous gadolinium; nonenhancing areas represent necrotic muscle fibers. MRI also reveals extensive edema within the muscle, subcutaneous edema, and interfacial edema 18,19.
Histological examination of the affected muscle reveals areas of necrosis and regenerative change with hyalinization and partial or total occlusion of the walls of small vessels and atherosclerosis of medium-sized vessels 9,11,20. Although biopsy is the gold standard for the diagnosis, it carries a risk for delayed healing, infection, and hematoma formation 13,21. As opposed to needle biopsy, excisional biopsy is associated with higher rates of recurrent hemorrhage and reoperation 18.
Treatment and prognosis
Treatment of diabetic muscle necrosis is mainly conservative. Analgesics, bed rest, and nonweight bearing on the affected limb are mandatory to relieve symptoms and avoid complications. Early ambulation will increase incidence of muscle hemorrhage and prolong hospitalization. Physical therapy should be discouraged as it can cause extension of infarction 11. Several reports illustrated the development of muscle hemorrhage following an exercise program after excisional biopsy of the affected muscle 22,23 so biopsy should preferably be avoided. The concomitant use of antiplatelet and anti-inflammatory drugs may accelerate recovery time from 8 to 5 weeks 24. Optimal glucose control is also recommended. Surgery is mandatory for decompression in patients with compartment syndrome 25. For those who are unsafe to return home because of difficult mobilization, arrangements should be made for short-term placement and rehabilitation with 24 h assistance during the recuperation period 26. The disease is self-limited and has an excellent short-term prognosis with expected recovery within 1–2 months. Up to 50% of cases may have a recurrence, mostly within the previously affected muscle 11. Unfortunately, the long-term prognosis is poor with a mean 2-year mortality as high as 10%, which is mainly due to macrovascular events 10 and most patients die within 5 years of diagnosis.
Knee effusion in diabetic myonecrosis
In the presented case, musculoskeletal examination revealed a left knee effusion, with pain and limitation of range of motion at the knee joint. Synovial fluid analysis was noninflammatory with absence of crystals and negative Gram stain and culture. Knee effusion has been previously reported in patients with diabetic myonecrosis. Penglis et al. 27 reported a case of diabetic muscle necrosis with noninflammatory knee effusion in a patient who was HLA B-27 positive. Yoo et al.28 reported a case of diabetic muscle necrosis presenting with knee arthralgia. The knee effusion can possibly be explained as being reactive to the surrounding muscle infarction. Diabetic myonecrosis should therefore be considered in the differential diagnosis of unexplained knee arthralgia or effusion in patients with DM.
Diabetic myonecrosis needs to be differentiated from other causes of a swollen tender extremity. Differentials include deep venous thrombosis, cellulitis, necrotizing fasciitis, pyomyositis, soft tissue abscess, muscle strain/rupture, and diabetic amyotrophy. Main characteristics of necrotizing fasciitis include the profound signs of systemic infection and less intense pain, unlike diabetic myonecrosis where severe pain is a key feature but the patient is nontoxic looking. Diabetic amyotrophy presents with weight loss and severe thigh pain with neurogenic weakness of the anterior thigh muscles that develops over months and is characterized by normal imaging. Pyomyositis is a bacterial abscess of the muscle usually due to staphylococcus. It is distinguished by the evident signs of systemic infection and its appearance on T2-weighted MRI as a high intensity mass. The absence of history of trauma and a negative vascular ultrasound can simply exclude other etiologies. In conclusion, diabetic myonecrosis should be considered when a diabetic patient presents with a swollen tender muscle after other more common causes are excluded.
Conflicts of interest
There are no conflicts of interest.
1. Angervall L, Stener B. Tumoriform focal muscular degeneration in two diabetic patients. Diabetologia. 1965;1:39–42
2. Kattapuram TM, Suri R, Rosol MS, Rosenberg AE, Kattapuram SV. Idiopathic and diabetic skeletal muscle necrosis: evaluation by magnetic resonance imaging. Skeletal Radiol. 2005;34:203–209
3. Umpierrez G, Stiles RG, Kleinbart J, Krendel DA, Watts NB. Diabetic muscle infarction. Am J Med. 1996;101:245–250
4. Banker BQ, Chester CS. Infarction of thigh muscle in the diabetic patient. Neurology. 1973;23:667–677
5. Chester CS, Banker BQ. Focal infarction of muscle in diabetics. Diabetes Care. 1986;9:623–629
6. Lauro GR, Kissel JT, Simon SR. Idiopathic muscular infarction in a diabetic patient. Report of a case. J Bone Joint Surg Am. 1991;73:301–304
7. Bjornskov EK, Carry MR, Katz FH, Lefkowitz J, Ringel SP. Diabetic muscle infarction: a new perspective on pathogenesis and management. Neuromusc Disord. 1995;5:39–45
8. Assimes T. An unusual diabetic complication in a young patient. Clin Invest Med. 1997;20(Suppl):S43
9. Singer S, Rosenberg AE. Case records of the Massachusetts General Hospital: weekly clinicopathological exercises. Case 29–1997. A 54-year-old diaberic woman with pain and swelling of leg. N Engl J Med. 1997;337:839–845
10. Kapur S, Brunet JA, McKendry RJ. Diabetic muscle infarction: case report and review. J Rheumatol. 2004;31:190–194
11. Wintz RL, Pimstone KR, Nelson SD. Detection of diabetic myonecrosis: complication is often-missed sign of underlying disease. Postgrad Med. 2006;119:66–69
12. Maclsaac RJ, Jerums G, Scurrah L. Diabetic muscle infarction. Med J Australia. 2002;177:323–324
13. Chason DP, Fleckenstein JL, Burns DK, Rojas G. Diabetic muscle infarction: radiologic evaluation. Skelet Radiol. 1996;25:127–132
14. Khoury NJ, El-Khoury GY, Kathol MH. MRI diagnosis of diabetic muscle infarction: report of two cases. Skeletal Radiol. 1997;26:122–127
15. Barohn RJ, Kissel JT. Case of the month. Painful thigh mass in a young woman: diabetic muscle infarction. Muscle Nerve. 1992;15:850–855
16. Nagdev A, Murphy M, Sisson C. Bedside ultrasound for the detection of diabetic myonecrosis. Am J Emerg Med. 2008;26:969.e3–e4
17. Delaney-Sathy LO, Fessell DP, Jacobson JA, Hayes CW. Sonography of diabetic muscle infarction with MR imaging, CT, and pathologic correlation. Am J Roentgenol. 2000;174:165–169
18. Aboulafia AJ, Monson DK, Kennon RE. Clinical and radiological aspects of idiopathic diabetic muscle infarction. Rational approach to diagnosis and treatment. J Bone Joint Surg. 1999;81:323–326
19. Jelinek JS, Murphey MD, Aboulafia AJ, Dussault RG, Kaplan PA, Snearly WN. Muscle infarction in patients with diabetes mellitus: MR imaging findings. Radiology. 1999;211:241–247
20. Trujillo-Santos AJ. Diabetic muscle infarction: an underdiagnosed complication of long-standing diabetes. Diabetes Care. 2003;26:211–215
21. Keller DR, Erpelding M, Grist T. Diabetic muscle infarction: preventing morbidity by avoiding excisional biopsy. Arch Int Med. 1997;157:1611–1612
22. Rocca PV, Alloway JA, Nashel DJ. Diabetic muscular infarction. Semin Arthritis Rheum. 1993;22:280–287
23. Hinton A, Heinrich SD, Craver R. Idiopathic diabetic muscular infarction: the role of ultrasound, CT, MRI, and biopsy. Orthopedics. 1993;16:623–625
24. Woolley SL, Smith DRK. Acute compartment syndrome secondary to diabetic muscle infarction: case report and literature review. Eur J Emerg Med. 2006;13:113–116
25. Pamoukian VN, Rubino F, Iraci JC. Review and case report of idiopathic lower extremity compartment syndrome and its treatment in diabetic patients. Diabetes Metab. 2000;26:489–492
26. Ankam NS, Kancherla V, Jacobs SR Rehabilitation of a patient with diabetic myonecrosis: a case report. Poster presentation. PR_176
27. Penglis PS, Scott G, Cleland LG. Diabetic muscle infarction presenting as a knee effusion. Semin Arthritis Rheum. 1999;28:421–422
28. Yoo WH, Kim CH, Park JH, Kim HK, Kim JR, Park TS, Baek HS. Case report: diabetic muscle infarction presenting as knee arthralgia. Rheumatol Int. 2001;21:36–39