Article, Urology

Renal infarction in the ED: 10-year experience and review of the literature

Unlabelled imageRenal infarction in the ED: 10-year expe”>American Journal of Emergency Medicine (2012) 30, 1055-1060

Original Contribution

Renal infarction in the ED: 10-year experience and review of the literature?

Meir Antopolsky MD a,?, Natalia Simanovsky MD b, Ruth Stalnikowicz MD a,

Shaden Salameh MD a, Nurith Hiller MD b

aDepartment of Emergency Medicine, Hadassah Mount ScopusHebrew University Medical Center, POB 24035,

Jerusalem 91240, Israel

bDepartment of Radiology, Hadassah Mount ScopusHebrew University Medical Center, Jerusalem, Israel

Received 11 May 2011; revised 28 June 2011; accepted 30 June 2011

Abstract

Objective: We aimed to describe clinical and radiologic features of Acute renal infarction (RI). Methods: Clinical, computed tomography (CT), and laboratory findings were retrospectively reviewed for patients diagnosed from 1999 to 2009 with CT proof of acute RI. Possible etiology of infarction was recorded. All available published series of RI were reviewed.

Results: Thirty-eight patients with acute RI met inclusion criteria; 127 cases of RI from 7 previous series were pooled for analysis. The most common symptoms were abdominal pain, flank pain, nausea, and vomiting. Leukocytosis (N10 x 109/L) and elevated lactate dehydrogenase levels (N620 IU/L) were the most prominent laboratory findings. Computed tomography features included wedge-shaped hypodensities in the renal parenchyma in 35 (92%) and global renal ischemia in 3 (8%) patients; 13 patients (34%) had concomitant splenic infarction. The most common etiology was atrial fibrillation. Computed tomography determined the specific cause for RI in 5 patients (13%) and a possible etiology in 17 (45%). Exact correlation with previous series was limited by methodological diversity.

Conclusion: Renal infarction should be considered in the differential diagnosis of a patient presented to the emergency department with abdominal or flank pain. Laboratory workup should include lactate dehydrogenase levels. After ruling out stone disease, contrast-enhanced CT examination is essential for the diagnosis.

(C) 2012

Introduction

Infarction of the kidney is a rather uncommon condition that can result from obstruction or decrease of renal arterial flow. Renal infarction (RI) is seen in patients with intrinsic

? The authors received no outside funding to support this research, and have no actual or potential conflicts of interest to disclose.

* Corresponding author. Tel.: +972 50 8573960; fax: +972 2 5844176.

E-mail address: [email protected] (M. Antopolsky).

aortorenal vascular pathology such as atherosclerosis, aneurysm, dissection, fibromuscular dysplasia, and vasculi- tis; with existing renal disease including Nephrotic syndrome and glomerulonephritis; with thromboembolism caused by atrial fibrillation, endocarditis, hypercoagulable conditions, or catheterization; and in trauma.

The diagnosis of RI is usually not obvious because it can mimic many other pathologic states, including pyelonephri- tis, renal colic, acute abdomen, pulmonary embolus, rupture of an aortic aneurysm, and more. It is critical to raise a

0735-6757/$ – see front matter (C) 2012 doi:10.1016/j.ajem.2011.06.041

suspicion for this condition and to proceed with an optimal diagnostic workup for kidney salvage. A delayed diagnosis may result in impaired renal function or even death [1,2].

During the last 35 years, few series on RI have been published [1-7], and there were substantial differences in methodology between studies. We present a 10-year experience with this pathology in patients admitted to the emergency departments (ED) of 2 academic medical centers with a combined annual admission of approximately 130 000 patients, and we then pool our findings with reports from leading studies [1-7]. The purposes of our study were to investigate the typical and unusual presenting clinical manifestations of RI and to describe radiologic findings that help establish the diagnosis, with the goal of improving diagnostic acumen for RI in the ED.

Patients and methods

This is a retrospective case series of patients who presented through the ED and were hospitalized with the diagnosis of RI between January 1, 1999, and December 31, 2009. The hospital institutional review board approved this study, and informed consent was waived.

Included were all patients with a computed tomography (CT)-proven diagnosis of RI. Patients with traumatic injury to the kidney, those in whom the diagnosis of RI was diagnosed later during the hospitalization period, and cases in which the infarction was a direct consequence of recent vascular surgery or percutaneous transluminal coronary angioplasty were excluded.

A total of 38 patients met inclusion criteria. Patient files were reviewed for age at admission, medical history, and clinical signs and symptoms at presentation. Laboratory findings, including liver and renal function tests, blood glucose, white blood cell count, and urinalysis results, were recorded.

computed tomography scans were retrospectively reviewed by a senior radiologist to detect features charac- teristic of RI, including diffuse or focal areas of hypoatte- nuation of the renal parenchyma. In addition, the condition of

Table 1 Comparison of methods among all series

the renal arteries was assessed, and a search for a possible etiology of the infarction was performed. Computed tomographic examinations were performed in all patients after administration of an intravenous contrast agent. In 27 patients (71%), CT examination included arterial as well as hepatoportal phases, allowing angiographic demonstration of the renal arteries. In 11 patients (29%), a standard, 1-phase protocol at the hepatoportal phase was performed. Standard scan parameters for abdominal CT were used including 120 kVp, automated mA, 5 to 3.75 mm slice thickness,

1.3 pitch, large field of view (FOV), and a 512 x 512 matrix. A PubMed search was conducted using the terms of renal infarct(-ion), and data from all RI patient series that could be found in the medical literature since 1978 were reviewed [1-8]. We pooled findings on age, sex, presenting symptoms, laboratory tests, diagnostic imaging modalities, and etiology, including findings from our study. Case reports

were excluded from the analysis.

Data were analyzed using descriptive statistics.

Results

Findings for 38 patients in our series and 127 patients from 7 previous articles were “pooled,” for a total of 165 patients reviewed in this study. Mean age was 51.7 years (range, 24-80 years) in our series and 61.4 years in the pooled study. Female patients comprised 51.7% of our series and 55% of the entire group.

As shown in Table 1, there are differences in the inclusion criteria and methods used for diagnosis of RI in pooled articles, limiting comparisons for some parameters. One study included only patients with atrial fibrillation and RI [1]; all other series included patients with varying etiologies for RI.

In our study, the diagnosis was established by contrast- enhanced abdominal CT in all patients. In previously published series, CT was the sole diagnostic means in 4 of 7 articles [3-5,7]. In the pooled data, 114 (69%) of 165 patients were diagnosed by CT, angiography was performed in 20 patients (16%), and isotope scans were performed in

Current study

Lessman et al [2]

Wong et al [7]

Lumerman et al [6]

Domanovits et al [3]

Korzets et al [5]

Hazanov et al [1]

Huang et al [4]

Total

Year

2011

1978

1984

1999

1999

2002

2004

2007

No. of patients

38

17

12

7

17

10

44

20

165

Inclusion criteria

All RI

All RI

All RI

All RI

All RI

All RI

RI and AF only

All RI

Diagnosis by CT

All

0

All

5

All

All

12

All

114

Diagnosis by angiography

0

8

7

2

0

0

10

0

27

Diagnosis by scintigraphy

0

6

0

0

0

0

36

0

42

Clinical and PM

0

3

0

0

0

0

0

0

3

PM indicates postmortem examination; AF, atrial fibrillation.

42 cases (25%). In 3 patients, the diagnosis was based on the clinical picture and postmortem examination.

Symptoms

Patient symptoms are summarized in Table 2. The most common symptoms in both our group and the pooled data set were abdominal pain, flank pain, and nausea and vomiting. Abdominal or flank pain was recorded in 33 (86%) of our patients. Of these 5 patients, 2 had endocarditis and 3 had multiple emboli, causing strong pain in other locations (chest, leg). Fever, defined as body temperature greater than 37.5oC, was recorded in 6 patients (16%) in our group and

51 cases (31%) in the pooled data.

Laboratory

Laboratory findings are summarized in Table 2. Leuko- cytosis (N10 x 109/L) and elevated LDH levels (N620 IU/L)

Table 2 Symptoms, laboratory data, and etiology of renal infarcts

ND indicates no data provided in this study; AST, aspartate aminotransferase; ALT, alanine aminotransferase; LDH, lactate dehydrogenase.

a Some studies provide numbers, and some, only average levels, so total numbers are not provided in this table.

b Without work of Hazanov et al–57/121 (47%). Explanation in the text.

were the most prominent results. Prevalence of LDH elevation occurred in 26 (70%) of 37 patients in our group vs 97 (95%) of 102 in 5 studies in the pooled group that provide these data [1-4,6]. In our series, leukocyte count was elevated in 26 patients (68%), and the mean WBC count was

12.4. Other studies that provide data on WBC count report similar results [4,5], with higher leukocytosis reported in one study [2].

Kidney function on admission was almost uniformly normal in our series. There was mild elevation in creatinine (N1.2 mmol/L) in 3 patients (8%), with a mean level of

0.86 mmol/L. Transaminases (aspartate aminotransferase N50 IU/L or alanine aminotransferase N60 IU/L) were ele- vated in 37% of patients. Usually, both enzymes were only mildly elevated. Urinalysis findings were available only in 20 (53%) of our patients and included hematuria in 10 patients (50%) and proteinuria in 7 (35%). Both findings are less com- mon in our study group as compared with the pooled data.

D-dimer was determined only in 5 patients and was positive in 3.

Current study

Lessman et al [2]

Wong et al [7]

Lumerman et al [6]

Domanovits et al [3]

Korzets et al [5]

Hazanov et al [1]

Huang et al [4]

Total

No. of patients

38

17

12

7

17

10

44

20

165

Symptoms

Flank pain

21 (55%)

7 (41%)

7

17

10

44

20

153

Abdominal pain

23 (60%)

7 (41%)

3 (43%)

11 (65%)

9 (90%)

14 (32%)

17 (85%)

82 (54%)

Nausea/vomiting

16 (42%)

8 (47%)

3 (43%)

9 (53%)

3 (30%)

33 (75%)

17 (85%)

92 (60%)

Fever

6 (16%)

10 (59%)

4 (57%)

0

4 (40%)

43 (98%)

3 (15%)

78

Confusion

2 (5%)

0

3 (43%)

0

5 (50%)

18 (41%)

1 (5%)

43

Diarrhea

3 (8%)

0

1

0

0

0

0

3

Dyspnea

3 (8%)

0

0

0

0

0

0

3

Chest pain

2 (5%)

0

0

0

0

0

0

3

Laboratory

Elevated LDH

26/37

14/14

7/7 (ND)

16/17 (ND)

ND

41/44

19/20

97/102

data

(mean, IU/mL)

(1572)

(ND)

(1570)

(1100)

(812) (95%)

Elevated AST

12/35

10/15

4/7 (ND)

ND

ND

ND (99)

ND (60) a

(mean, IU/mL)

(82)

(ND)

Elevated ALT

12/35

5/6 (ND)

4/7 (ND)

ND

ND

ND (91)

ND (50) a

(mean, IU/mL)

(65)

Elevated WBC

26/38

17/17

4/7 (ND)

ND

ND

ND

ND a

(mean, IU/mL)

(12.4)

(19.9)

(12.9)

(12.9)

Elevated creatinine

3/38

15/17

1/7 (ND)

ND

3/10

29/44

ND (0.1) 51/116

(mean, mmol/L)

(0.86)

(ND)

(ND)

(ND)

Hematuria

Proteinuria

10/20

7/20

12/14

14/14

4/7

3/7

14/15

12/15

10/10

ND

21/39

17/38

9/20 80/125 (64%)

7/20 60/124

Etiology

Atrial fibrillation

14 (37%)

10 (59%)

2 (17%)

4 (57%)

11 (65%)

6 (60%)

44

(48%)

10 (50%) 101 b

(100%)

(61%)

Hypercoagulation

6 (16%)

0

0

0

1 (6%)

1 (10%)

0

2 (10%) 10

Endocarditis

3 (8%)

0

1 (8%)

1 (14%)

0

0

0

0 5

Others

10 (26%)

0

8 (67%)

2 (28%)

0

0

0

6 (30%) 26

Unknown

5 (13%)

7

1 (8%)

0

5 (29%)

3 (30%)

0

2 (10%) 23

Imaging findings

In our study, 35 patients (92%) had the typical CT presentation of a wedge-shaped area of decreased attenuation within renal parenchyma that was otherwise of normal appearance (Fig. 1). Three patients (8%) had global infarction of 1 kidney. Perirenal fat stranding was evident in 11 patients (29%). Renal infarction involved the right kidney in 17 patients (45%), left kidney in 15 (39%), and both kidneys in 6 cases (16%). Among patients with bilateral RI, 4 had atrial fibrillation, and 1 had proven coagulopathy. One infarct was visible in 21 patients (55%), 14 (37%) had multiple infarcts, and 3 (8%) had global RI. Infarcts were located in the lower aspect of the kidney in 13 patients (34%), upper pole in 4 (10.5%), central aspect in 4 (10.5%) cases, scattered in the renal parenchyma in 14 (37%), and involved the whole kidney in 3 patients (8%). There was concomitant acute infarction of the spleen in 13 patients (34%), including 8 (21%) with atrial fibrillation, 1 (3%)

with endocarditis, 1 (3%) with cardiomyopathy, 1 (3%) with coagulopathy, and 1 patient (3%) with a prosthetic valve. In 1 patient, no definite etiology for infarction was established. One patient with atrial fibrillation also had splenic and hepatic infarcts, and 1 young patient with patent foramen ovale and paradoxical emboli from Deep vein thrombosis had also an acute pulmonary infarction. Old renal infarcts, presenting as peripheral focal retraction and renal tissue scarring, were found in 5 patients (13%), including 3 with atrial fibrillation. An old splenic infarct was present in 1 patient with atrial fibrillation.

Computed tomography was able to depict the exact etio- logy for infarction in the renal arteries in 5 patients. Athero- sclerosis and severe stenosis of the renal artery were revealed

Fig. 1 Enhanced CT image at the level of the kidneys demonstrating a large hypodense region in the right kidney consistent with an acute infarct (black arrow) in a 57-year-old woman. Note mild thickening of the perirenal fascia (white arrow).

in 2 cases (5%). Obstructing thrombus was found in 1 patient (3%) with antiphospholipid antibodies syndrome, throm- bosed renal artery aneurysm was found in 1 patient (3%) with Ehlers-Danlos syndrome, and marked beading of the intrarenal arteries was seen in a patient with vasculitis (3%). These vascular pathologies were revealed in CT angiography performed on a 16-slice CT machine using volumetric acquisition in thin slices and multiplanar reconstruction.

Other relevant CT findings that may help in revealing the cause for thrombosis included large suprarenal soft aortic plaques in 6 patients (16%), including 2 (5%) with

coagulopathy, 2 (5%) with atrial fibrillation, and 2 (5%) with unknown etiology. In addition, there was enlargement of the left atrium in 9 patients (24%), including 7 (18%) with atrial fibrillation, 1 (3%) with endocarditis, and 1 (3%) with RI of unknown etiology. In 1 patient (3%) with atrial fibrillation and a prosthetic mitral valve, there was thrombus in the left atrium.

Etiologic factors

Etiologic factors for RI in our series and those cited in previous studies are presented in Table 2. Findings from Hazanov et al [1] must be considered in light of the fact that atrial fibrillation was an inclusion criteria in their study. Still, atrial fibrillation was found to be the most frequent etiologic factor for RI in our series and in other previous reports. The proportion of patients with atrial fibrillation in our series (37%) was lower compared with the pooled population, even after patients of Hazanov et al [1] were excluded (47%). Other frequent causes were bacterial endocarditis and coagulopathy. Other factors that were noted in our patients, including Ehlers-Danlos syndrome with thrombotic aneu- rysm of the renal artery, abdominal aorta aneurysm with mural thrombi, previous valve replacement with coagulo- pathy, polyarthritis nodosa, soft atheromas of the suprarenal aorta, patent foramen ovale with embolus caused by DVT, cardiogenic shock, and cardiomyopathy.

Discussion

We present our 10-year experience with RI in 38 patients admitted to the departments of emergency medicine in 2 centers of our academic medical organization and compare our findings with all published series of RI patients, including a total of 165 patients reported in a period spanning over 30 years. Although our results are similar to those reported in previous series in some respects, several important differences resulting from changes in diagnostic imaging over the last 30 years can be seen. In earlier studies, angiography and isotope scans were the main diagnostic tools. However, the use of CT has increased dramatically since the 1990s, and in last decade, CT has become the gold standard for the diagnosis of RI.

Patient demography has not changed greatly. There was no difference in sex distribution. Our patients are somewhat younger as compared with those of previous series.

As in other reports, most patients presented with flank or/and abdominal pain and somewhat less frequently nausea and vomiting. Only 5 patients from our group did not have abdominal or flank pain, including 3 who complained of strong pain caused by emboli in other locations and 2 of them found to have bacterial endocar- ditis. Fever was much less frequent in our group, although it was present in almost a third of patients in the pooled group. This can be a result of early detection and treatment of RI in our patients, resulting in less Ischemic necrosis and reduced inflammation.

Other symptoms were rare, occurring in less than 10% of patients, and included chest pain, diarrhea, confusion, dyspnea, and shoulder pain. It is hard to determine whether these symptoms are related to RI, underlying conditions that caused the RI, or concomitant emboli to other organs.

Regarding the laboratory data, liver and renal function tests were less disturbed in our patients in comparison with results reported in the literature. It seems probable that with current diagnostic technology, we are able to diagnose much smaller infarcts with a much milder clinical picture. This assumption is supported by our observation that substantial change in laboratory findings was encountered over 48 hours in almost every patient in our series with serial LDH levels.

Most patients with RI are referred by the ED physician to abdominal CT for evaluation of abdominal or flank plain. The index of suspicion for RI is often low. When CT is performed with no intravenous contrast injection, to rule out urolithiasis, the diagnosis of RI can be missed. In cases where there is no explanation for flank pain on noncontrast CT, it is strongly advised to consider further evaluation with contrast-enhanced CT, which has been shown to be very effective to diagnose or rule out RI [8].

Parenchymal appearance depends on the degree and location of the Arterial occlusion, and its age. Acute infarcts typically appear as wedge-shaped areas of decreased attenuation within an otherwise normal-appearing kidney [8]. Computed tomography acquisition should be performed with a sufficient delay after contrast injection to achieve effective enhancement of the renal parenchyma. For this reason, focal RI may be overlooked on CT angiography, where only the renal cortex is enhanced.

The radiologic appearance of RI may be more complex and misleading when an inflammatory reaction after vascular insult to the kidney causes mass effect that can be mistaken for other disorders, especially focal pyelone- phritis and tumor [8]. Perinephric fat stranding and edema are less common in RI as compared with acute pyelonephritis, and in certain clinical settings, the lack of perirenal stranding can assist in reaching the correct diagnosis. Specific causative vascular pathologies are not commonly defined on CT in patients with RI [9]. This can

be attributed to the fact that emboli are the most common etiologic factor. In addition, in most patients admitted with abdominal or flank pain, a standard CT protocol that does not provide good angiographic demonstration of the renal vasculature is used.

When the radiologist suspects renal infarction, there should be a heightened sensitivity to heart abnormalities, including a large atrium, mural thrombus, valve pathology, and others, as well as to vascular changes of the aorta and renal arteries in particular and evidence for infarction in other organs such as the spleen, bowel, liver, and lungs. Meticulous evaluation of the CT study can reveal the extent of disease and sometimes reveal the exact etiologic factor, allowing more targeted and efficient treatment.

Our study carries several major limitations. Because of the retrospective nature of this study, the data are not uniform. Laboratory data are only partially reported or missing in some patients. The etiology of RI could be determined in only some patients. Underlying conditions most likely continued to evolve after this hospitalization and patients may have been lost of follow-up. Computed tomographic examinations were performed through a decade of tremendous change in CT capabilities and protocols. As a result, there is a change in the quality of data associated with earlier cases compared with more recent patients. The ability to precisely compare our results with the literature is limited because of marked discrepancies in methodology and lack of necessary information.

In conclusion, when a patient presents to the ED with an acute flank and/or abdominal pain, especially in patients with risk factors for emboli such as atrial fibrillation or bacterial endocarditis, RI should be considered in the differential diagnosis. The laboratory survey should include liver and renal function tests, with increased LDH strongly supporting this diagnosis. If unenhanced CT is negative for urinary calculi, contrast-enhanced CT of the abdomen may also be needed for early diagnosis of RI and to rule out other conditions with similar clinical presentation. Alternatively, to minimize long-term radiation risks and maximize the chances of picking up the more serious RI, a combination of ultrasound followed by contrast-enhanced CT of the abdomen might be the initial diagnostic regimen of choice in high-risk individuals with flank pain.

Acknowledgments

We thank Shifra Fraifeld for her editorial contribution to the manuscript.

References

  1. Hazanov N, Somin M, Attali M, et al. Acute renal embolism. Forty-four cases of renal infarction in patients with atrial fibrillation. Medicine (Baltimore) 2004;83(5):292-9.
  2. Lessman RK, Johnson SF, Coburn JW, et al. Renal artery embolism: clinical features and long-term follow-up of 17 cases. Ann Intern Med 1978;89(4):477-82.
  3. Domanovits H, Paulis M, Nikfardjam M, et al. Acute renal infarction. Clinical characteristics of 17 patients. Medicine (Baltimore) 1999;78(6): 386-94.
  4. Huang CC, Lo HC, Huang HH, et al. ED presentations of acute renal infarction. Am J Emerg Med 2007;25(2):164-9.
  5. Korzets Z, Plotkin E, Bernheim J, et al. The clinical spectrum of acute renal infarction. Isr Med Assoc J 2002;4(10):781-4.
  6. Lumerman JH, Hom D, Eiley D, et al. Heightened suspicion and rapid evaluation with CT for early diagnosis of partial renal infarction. J Endourol 1999;13(3):209-14.
  7. Wong WS, Moss AA, Federle MP, et al. Renal infarction: CT diagnosis and correlation between CT findings and etiologies. Radiology 1984;150(1):201-5.
  8. Suzer O, Shirkhoda A, Jafri SZ, et al. CT features of renal infarction. Eur J Radiol 2002;44(1):59-64.
  9. Kawashima A, Sandler CM, Ernst RD, et al. CT evaluation of renovascular disease. Radiographics 2000;20(5):1321-40.