237

Significance of Elevated Levels of Serum Creatine Phosphokinase in Febrile Diseases: A Prospective Study Ohad Cohen, Leonard Leibovici, Felix Mor, and Arjeh J. Wysenbeek

From the Department of Internal Medicine "B," Beilinson Medical Center, Petah Tiqva; and Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel

Creatine phosphokinase (CPK) is an 80-kilodalton protein abundant in many tissues throughout the body, with relatively high concentrations in striated muscles, cardiac muscle, the brain, and the gastrointestinal tract [1]. CPK occurs as a dimer; the two types of subunit give rise to three main isoenzymes, as detected by electrophoresis. These enzymes are designated according to their main source: MM (muscle), BB (brain), and MB (heart). Other types of CPK exist [1, 2] but usually do not contribute to overall serum CPK levels. Elevated serum CPK concentrations are detected in many pathologic states and are considered to result from tissue damage. These cardiac and extracardial conditions are described elsewhere [1]. Febrile diseases have been thought to cause CPK elevation only in rare cases, most of which are attributable to specific bacteria [3-13], viruses [3, 14-17], and protozoa [18]. It is also known that serum CPK levels may be increased in conditions associated with extreme elevations of body temperature, such as heat stroke, malignant neuroleptic syndrome, and malignant hyperthermia [1, 19, 20]. We therefore designed a prospective study of febrile patients to determine the prevalence of CPK elevation in febrile diseases, the relevance of CPK elevation to the etiology of fever or underlying disease, and the prognostic significance of high CPK levels in the febrile patient.

Methods Patients. Included in the study were all patients who were hospitalized between January 1986 and January 1987 because

Received 6 November 1989; revised 7 June 1990. Reprints and correspondence: Dr. L. Leibovici, Department of Internal Medicine "B," Beilinson Medical Center, Petah Tiqva 49 100, Israel. Reviews of Infectious Diseases 1991;13:237-42 © 1991 by The University of Chicago. All rights reserved. 0162-0886/91/1302-0006$02.00

of a febrile disease and whose temperature was >38.0°C during the first 12 hours of hospitalization. The admitting ward was a 50-bed internal medicine department in a universityaffiliated facility serving an urban population of 200,000. The patients made up one-fourth of the total number of admissions to the medical wing and were randomly assigned to one of the five wards. Exclusion criteria were trauma and/or myocardial infarction during the 10 days prior to admission; two patients were excluded on these grounds. A detailed medical history was obtained, a physical examination done, and performance status assessed according to the Karnofsky scale [21]. The following laboratory tests were performed within 12 hours of admission and thereafter every other day for 10 days: determinations of serum CPK and CPK isoenzymes, serum aldolase, and urine myoglobin; a complete blood count; and a blood chemistry panel. Blood and urine for culture were obtained at admission from all patients. In cases with elevated CPK levels and without clear bacteriologic etiology, the following additional tests were performed on the third day: assays of free thyroxine, thyroid-stimulating hormone, antinuclear factor, rheumatoid factor, and complement; repeated serologic tests for Epstein-Barr virus, cytomegalovirus, hepatitis A and B viruses, herpesviruses, coxsackievirus, influenza virus, adenovirus, rubella virus, mumps virus, echoviruses, Mycoplasma, Legionella, Leptospira, Toxoplasma, Brucella, Coxiella burnetii, Chlamydia, and Rickettsia; and the Vidal, Weil-Felix, and Venereal Disease Research Laboratory tests. The study of patients ended at discharge or death. All data were recorded on a uniform questionnaire. To show that elevated CPK concentrations are in fact related to febrile diseases and are not simply a common feature of elderly hospitalized populations with a low functional status, we studied a control group of hospitalized patients. Included were all patients hospitalized in the department of medicine over a period of 2 months except for febrile patients, patients

Downloaded from http://cid.oxfordjournals.org/ at Michigan State University on June 29, 2015

The incidence and significance of elevated serum levels of creatine phosphokinase (CPK) in febrile diseases were studied prospectively in all patients admitted with fever to a department of medicine during 1 year. High serum CPK levels were detected in 70 (28%) of 247 febrile patients but in only six (6%) of 105 afebrile control patients (P = .0001). Elevated CPK levels were not related to any specific diagnosis. Logistic regression analysis identified five factors that correlated both significantly and independently with elevation of CPK values: increased blood urea nitrogen level, low serum phosphate level, a stuporous or comatose state, tremor, and muscle tenderness. Myoglobinuria, detected in 14 patients, was predictive of a fatal outcome, but a high CPK level by itself was not an independent correlate of mortality. In summary, CPK elevation is not uncommon in febrile diseases, but because it does not reflect a specific etiology it does not necessarily indicate that an extensive diagnostic work-up is required.

238

Cohen et at.

Results

Composition ofthe study and controlgroups. During the study period, 247 patients (128 men and 119 women) 18-98 years of age (median age, 71 years) were hospitalized because of fever and were enrolled in the study protocol. Seventeen patients were referred from nursing homes; others were referred by the family physician or presented themselves to the hospital's emergency room. Included in the control group were 105 patients (59 men and 46 women) 26-90 years of age (median age, 71 years). The main causes of hospitalization in the control group were chest pain (29%), cerebrovascular accidents (17%), exacerbation of obstructi ve lung disease (13%), and arrhythmia (7%). The study and control groups were similar as regards age, functional status, level of consciousness, gender, domicile, and underlying disorders. Patterns ofCPKelevation. Seventy patients (28%) in the study group and six patients (6%) in the control group had elevated CPK levels in serum (P = .0001). The mean CPK value for the studygroup was211 ± 91 U/L (range, 10-10,000 U/L), and that for the control group was 81 ± 68 U/L (range, 20-420 U/L) (P = .0001). Myoglobinuria was noted in 14 patients in the study group and in none of the control patients.

ioooo 5000 2500

1000

< ;:)

~ Q.

250

+

o

+ 100

+

50 25

o

2

4

6

DAY Figure 1. CPK values in 51 patients with high initial levels during a follow-up period of 6 days. The range is denoted by the straight vertical lines, the 25th and 75th percentiles by the box, the median by the horizontal line inside the box, and the mean by the cross.

CPK levels in the study group decreased with the duration of hospitalization. The serum CPK levels for 51 patients with high initial values are box-plotted for a period of 6 days in figure 1. The percentage of patients in the study group with elevatedCPK levelsdecreased from 28.3% on day 1 to 16.4%, 13.7%, 7.5%, and 4.0% on days 3, 5, 7, and 9, respectively. The elevated isoenzyme in all patients with high CPK levels was MM. Sources offever. The source of fever was determined for 88% of patients (figure2), with no statistical differencesfound between the normal and high CPK groups. The most common cause - urinary tract infection- was diagnosed in 27.1 % of the high CPK group and 24.5% of the normal CPK group (P value not significant); lower respiratory tract infectionwas diagnosed in 30 % and 17.3% of these two groups, respectively (P value not significant). Bacteremia was significantly more prevalent in the high CPK group (35.7% vs. 9.7%; P < .0001). The percentage of patients with elevated CPK was similar for gram-negative and gram-positive bacteremia. Three of the 70 patients with high CPK levels were diagnosed serologically as having influenza A. No other distinct bacteriologic or virologic agent could be imputed as a cause of CPK elevation in the febrile episodes. Only one patient had a low level of free thyroxine (0.4 mgldL) and a high level of thyroid-stimulating hormone (14mIU/mL). The median free

Downloaded from http://cid.oxfordjournals.org/ at Michigan State University on June 29, 2015

with an infectious disease, and patients with trauma or myocardial infarctionin the 10 daysprior to admission. This group was surveyed according to the protocol detailed for the study group. Laboratory analysis. Serum CPK levels were determined with a standard kit (CK NAC-activated; Boehringer, Mannheim, Germany) in which a coupled enzyme systemmeasures adenosine triphosphate formation by the hexokinase-glucose6-phosphate dehydrogenase nicotinamide-adenine dinucleotide phosphate(NADP) system[22]. The normal values(37°C) are 24-195 U/L for males and 24-170 U/L for females. CPK isoenzymes were assessed by electrophoresis on cellulose acetate [23] (Helena Laboratories, Beaumont, TX). CPK bands were quantitativelyevaluatedby means of the coupled enzyme systemdescribed above and measurement of the fluorescence ofNADP on a densitometer. Normal values (expressed as a percentage of total CPK) are as follows: CPKMM, 96%-100%; CPK-MB, 0 to 4%; and CPK-BB, O. For other laboratory studies, standard clinical chemistry and microbiologic techniques were used. Statistical analysis. Contingency tables were tested for statistical significance by the X2 test. As most continuous variables were not normally distributed, we used the Wilcoxon rank-sum test to compare continuous variables between two classes. For multivariate regression analysis, we used a logistic regression procedure(LOGIST) [24]. O'Brien's nonparametric test for association with censored data [25] was used in an assessment of which variables were significantly and independently associated with death during hospitalization.

RID 1991;13 (March-April)

RID 1991;13 (March-April)

CPK in Febrile Diseases

N-Normal CPK H-High CPK

N

H

[§J

80 70

N

Table 1. Clinical characteristics of febrile patients with and without elevated serum levels of CPK.

DAlive

H

•

Value for indicated group of patients

Dead

N H

Normal CPK (n == 177)

Elevated CPK (n = 70)

61.2 ± 20.8 (14-98)

68.8 ± 15.9 (16-88)

.01

83 (46.9)

45 (64.3)

.04

38.6 ± 0.6 (38.0-40.2)

38.7 ± 0.6 (38.0-40.5)

Duration of acute disease (d)*

3.4 ± 4.2 (0-30)

2.4 ± 2.5 (0-14)

.03

Duration of hospitalization (d)*

7.8 ± 5 (2-35)

9.7 ± 8.3 (1-46)

NS

Characteristic Age (y)*

Figure 2. Proportion of patients with high and normal CPK levels and of patients who died during hospitalization, according to diagnosis. UTI = urinary tract infection; LRTI = lower respiratory tract infection; Bac. Diarr. = bacterial diarrhea; Neutro Fever = febrile patients with neutropenia; Bac. Menin. = bacterial meningitis; and Undiag. = undiagnosed.

Gender (male)" Temperature on admission (0C)*

P

NSf

4 (5.7)

.02

24 (13.5)

18 (25.7)

.04

Karnofsky score (functional status) t 10-30 40-60 70-90 100

11 (6.2) 19 (10.7) 59 (33.3) 88 (49.7)

11 (15.7) 17 (24.3) 20 (28.6) 22 (31.4)

Permanent cathetert§

19 (10.7)

21 (30)

Dementia t Level of consciousnesst Coma Stupor Somnolence Normal Dehydrationt

8 (4.5) I (0.6) 2 (1.1) 34 (19.2) 140 (79.1) 9 (5.0)

8 (11.4) 4 8 23 35

(5.7) (11.4) (32.9) (50.0)

21 (30.0)

.001

.03 .04

.0001

.0001

* Continuous variables, with values given as mean ± so (range). t Discrete variable. with values given as number of patients (percentage). :j: NS = not significant. § Catheter inserted at least 2 weeks before admission.

rum creatinine, blood urea nitrogen, and serum sodium as well as lower levels of serum phosphate in the elevated CPK group, with no significant difference in potassium, calcium, or chloride level between the groups. Blood counts revealed higher hemoglobin and hematocrit levels in the high CPK group, though the elevation in hematocrit was not statistically significant. The lymphocyte count was lower in the high CPK group, whereas the granulocyte count did not differ significantly between the groups. Serum levels of aldolase, lactate dehydrogenase, and aspartate aminotransferase - enzymes that can originate from damaged muscle tissue-were all significantly elevated in the high CPK group, while the bilirubin level was not elevated significantly. On logistic regression analysis, the only variables significantly and independently associated with high CPK levels were elevated blood urea nitrogen level (adjusted odds ratio [AOR], 1.1; 95 % confidence interval [CIl, 1.0-1.1 for an increment

Downloaded from http://cid.oxfordjournals.org/ at Michigan State University on June 29, 2015

1 (0.6)

Tremorf

Muscle tenderness t

thyroxine level was 1.1 mg/dl, (range, 0.4-2.1 mg/dl.), and the median thyroid-stimulating hormone level was 2 rnIU/mL (range, 0.5-14 rnIU/mL). Five patients had an antinuclear titer of 1:40, and two had a titer of 1:80. The additional tests detailed in the Methods section yielded no other pathologic values in the 70 cases of high CPK. . Clinicaland laboratory characteristics ofpatients. Univariate analyses of clinical and laboratory data according to the presence or absence of CPK elevation are presented in tables 1 and 2. In the elevated CPK group, the mean age was higher (69 years vs. 61 years; P = .01) and the duration of the acute illness shorter (2.4 days vs. 3.4 days; P = .03). There was a numerical preponderance of men in the high CPK group (64.3% vs. 41.6%; P = .01), without significant differences in weight. We noted a small but statistically significant seasonal variation in the proportion of patients with high CPK: 59 % of these patients and 43 % of patients with normal CPK were admitted during the summer (P = .03). The two groups of patients did not differ significantly in terms of temperature or blood pressure, either at admission or during hospitalization. Functional status, assessed by the Kamofsky scale and other determinants of physical and mental status, was significantly poorer in the high CPK group (table 1). More patients in the latter group had clinically diagnosed dehydration, decubitus ulcers, or a permanent catheter that had been inserted at least 2 weeks prior to admission. On physical examination we found tremor and muscle tenderness to be much more common in patients with elevated CPK. Of patients in the high CPK group, 25.6% were identified as having diabetes mellitus by history and fasting blood glucose values. This prevalence is significantly higher than the 10% found for the normal CPK group (P = .0002). No other underlying diseases were more prevalent in one group than the other. Neither prior medications nor recent intramuscular injections were associated with CPK elevation. Laboratory studies (table 2) revealed higher levels of se-

239

240

RID 1991;13 (March-April)

Cohen et al.

Table 2. Laboratory values for febrile patients with and without elevated serum levels of CPK. Values for indicated group of patients* Parameter

Normal CPK (n = 177)

Elevated CPK (n = 70)

73 ± 42 (10-203)

820 ± 1,507.2 (191-10,000)

Serum CPK, day 1 (U/L)

p

2.8 ± 1.8 (1.1-10.5)

9.2 ± 7.6 (1.2-41.7)

Blood hemoglobin (g/dL)

12.4 ± 2.2 (6.7-20.1)

13.0 ± 2.2 (8.2-16.9)

Hematocrit (%)

37.2 ± 6.3 (20.0 - 56.0)

38.9 ± 6.0 (27.0-52.0)

1.2 ± 0.8 (0.4-9.1)

1.8 ± 1.85 (0.4-9.9)

.002

19.2 ± 11.5 (5.0-99.0)

28.9 ± 21.4 (8.0-99.0)

.0001

138.8 ± 3.8 (114.0-147.0)

140.4 ± 5.8 (116.0-156.0)

.01

3.0 ± 0.7 (1.6-6.0)

2.7 ± 0.8 (1.0-5.5)

.004

27.7 ± 17.8 (24.0-130.0)

55.3 ± 58.4 (26.0-300.0)

.00001

Serum lactate dehydrogenase (U/L)

240 ± 76 (110-532)

333 ± 186 (112-:>99)

.00001

Serum uric acid (mg/dl.)

5.6 ± 1.9 (1.5-10.8)

6.4 ± 2.4 (1.6-11.7)

.02

Serum albumin (g/dL)

3.7 ± 0.5 (2.4-5.0)

3.6 ± 0.5 (2.4-4.8)

.05

Serum aldolase, day 1 (UlL)

Serum creatinine (rng/dl.)

Serum sodium (mEq/L) Serum phosphate (mg/dL) Serum aspartate aminotransferase (U/L)

.05

* Values are given as mean ± SD (range). NS = not significant.

t

of 1 mg/dl.), low serum phosphate level (AOR, 0.3; 95% Cl, 0.2-0.4 for an increment of 1 mg/dL), decreased consciousness (AOR, 4.9; 95% Cl, 1.3-6.5 for an increment of one level of consciousness), tremor (AOR, 2.8; 95 % cr, 1.2-3.8), and muscle tenderness (this variable had a limited dispersion, and thus the AOR was very large). ElevatedCPKlevels and myoglobinuria. Fourteen patients had myoglobinuria in addition to CPK elevation. Their CPK levels were significantly higher than those of the 56 patients who had elevated CPK without myoglobinuria (1,459.9 ± 2,480.0 U/L [range, 337.0-10,000 U/Ll vs. 470.8 ± 392.5 U/L [range, 191-1,986 U/Ll; P = .03). This subgroup ofpatients (11 male and three female, aged 64.4 ± 19.3 years [range, 18-98 years]) had significantly lower scores on the functional performance scale; moreover, 57.1% had an abnormal consciousness level (vs. 28.4 % of febrile patients without myoglobinuria), and 42.9% had clinical dehydration. Muscle tenderness, decubitus ulcers, tremor, and rigidity were more prevalent in myoglobinuric patients. Mortality was very high in this group (42.9% vs. 7.7% among nonmyoglobinuric febrile patients).

Elevated CPK and myoglobinuria as prognostic factors in febrile diseases. Twenty-five (l 0.1%) of the 247 patients died in the course of their febrile disease. The mean duration of hospitalization ± SD before death was 8.9 ± 9.5 days (median, 6 days; range, 1-35 days). As a group, the patients who died were older (72.5 ± 12.3 years vs. 62.5 ± 20.2 years; P = .02) and had higher CPK levels (969.8 ± 165.6 U/L vs. 212.8 ± 702.2 U/L; P = .0001) than those who survived. Myoglobinuria was noted in 24 % of the patients who died. Other characteristics of these 25 patients were a lower level of consciousness, lower functional scores, clinically diagnosed dehydration (36 %), permanent catheterization (56 %), neutropenia, and laboratory evidence of dehydration and bacteremia. Using O'Brien's log-rank procedure, we found the following variables to be significantly and independently associated with mortality: a low functional assessment score, permanent catheterization, decreased consciousness on admission, dehydration, low weight, increased blood urea nitrogen, low serum albumin, and myoglobinuria. Elevated CPK was not found to be independently associated with mortality. Furthermore,

Downloaded from http://cid.oxfordjournals.org/ at Michigan State University on June 29, 2015

Blood urea nitrogen (mg/dL)

.0001

RID 1991;13 (March-April)

CPK in Febrile Diseases

when myoglobinuria was omitted from the analysis, serum CPK levels were closer to significance but were still not an independent factor in mortality (z = 1.71; P = .09).

Discussion

consciousness level, and underlying disorders. It seems that elevated CPK is a feature of febrile disease per se. Nevertheless, no specific agents causing the febrile episode could be linked to CPK elevation, nor were rare and specific pathogens known to cause rhabdomyolysis found (e.g., Trichinella, Leptospira, or Francisella tularensis). We did find influenza A infection in three patients and significantly higher rates of bacteremia in the high CPK group, confirming previous reports [3, 9, 26]; however, bacteremia was not included in the final logistic model as an independent correlate with higher CPK levels. CPK-MM was the isoenzyme elevated in all patients with high total CPK levels. Injury to striated muscle is considered the main source of this isoenzyme, but the pathogenesis underlying this injury in infectious disease is not clear. Some investigators report direct invasion of muscle by the pathogen [6, 8, 14, 27]. Others, on the basis of biopsy material, consider different mechanisms, such as immune-mediated muscular damage and vasculitis [5] or endotoxin-mediated injury [4]. Indirect effects of infection on muscular metabolism, with a decrease in muscular glycolytic activity leading to muscular injury, have also been reported [11]. Interleukin 1 stimulates proteolysis and production of prostaglandin E2 in muscle [28]. Extreme elevations of body temperature, as in heat stroke, malignant neuroleptic syndrome, or malignant hyperthermia, can cause CPK elevation, though primary muscular disorders are considered to underlie the latter syndrome. No cases of alcoholism were found among the study patients, but this diagnosis must be considered for febrile patients with elevated CPK in populations where alcoholism is prevalent. In conclusion, in our urban population of elderly febrile patients, CPK elevation was not a rare finding; however, since this alteration did not point to a specific diagnosis, it did not necessitate an extensive work-up for a unique etiology. The data presented here are compatible with a nonspecific elevation of CPK in febrile diseases, the association being more noticeable in debilitated patients. Elevated levels of CPK associated with fever usually return to normal within 5 days after admission to the hospital. CPK elevation is associated with a low level of consciousness, dehydration, muscle tenderness, and tremor. Significantly more patients with high CPK had bacteremia or died during hospitalization. However, those two associations were not independent of other variables. Myoglobinuria was significantly and independently correlated with death and time to death.

References

1. Bais R, Edwards ra Creatine kinase. CRC Crit Rev Clin Lab Sci 1982;16:291-335 2. Ingwall IS, Kramer MF, Fifer MA, Lorell BH, Shemin R, Grossman W, Allen PD. The creatine kinase system in normal and diseased human myocardium. N Engl J Med 1985;313:1050-4 3. Gabow PA, Kaehny WD, Kelleher SP. The spectrum of rhabdomyolysis. Medicine 1982;61:141-52

Downloaded from http://cid.oxfordjournals.org/ at Michigan State University on June 29, 2015

This is, to our knowledge, the first study to prospectively define the extent of elevated CPK in febrile diseases and to correlate this elevation with prognostic and clinical parameters. Previous studies and case reports [3-18] describe CPK elevation in the context of severe or uncommon infections or as a rare finding in a study population. In the present study, CPK elevation was noted in 28.3% of 247 patients with fever who were admitted during the survey year. In 14 patients CPK elevation was associated with myoglobinuria. Febrile patients with elevated CPK levels were significantly worse off, in terms of many clinical and laboratory parameters, than those with normal CPK values. Patients with high CPK values were older, had lower functional performance scores, more often had dementia, and had lower consciousness levels than those with normal CPK values; all of these factors can account for the higher rate of decubitus ulcers, permanent catheters, and clinical and laboratory signs of dehydration in this group. Of the many variables significantly associated with elevated CPK levels on univariate analysis, only five were independently correlated with high CPK levels: decreased level of consciousness, decreased serum phosphate level, increased blood urea nitrogen level, muscle tenderness, and tremor. Most of these variables are associated with more fulminant disease or a more debilitated state. Mortality among febrile patients with elevated CPK concentrations was higher than that among febrile patients with normal CPK levels (21.4 % vs. 5.6 %). This difference is in accordance with the report [10]associating elevated CPK with poor outcome in tularemia. In a subgroup of patients with both elevated CPK and myoglobinuria, the likelihood of a poor outcome was greater, with a mortality of 42.9 % in the course of the febrile illness. Indeed, in a multivariate regression analysis, myoglobinuria was found to be an independent factor in mortality while CPK elevation came close to but did not reach statistical significance (z = 1.71; P = .09). Patients with myoglobinuria are believed to suffer from more extensive muscle damage [3]. This greater damage could be due to a more serious febrile illness but, as was seen in our study, the premorbid state of these patients was also worse. The association between the premorbid state and muscle damage during the febrile disease has yet to be explained. Serial measurement of CPK levels in relation to the febrile period strongly supports the etiologic role of the acute febrile disease in the elevation of these enzyme levels. Furthermore, the percentage of patients with elevated CPK was significantly higher in a febrile group than in a control group of hospitalized, nonfebrile patients matched for age, functional status,

241

242

Cohen et al.

18. Miller KD, White NI, Lott lA, Roberts 1M, Greenwood BM. Biochemical evidence of muscle injury in African children with severe malaria. 1 Infect Dis 1989;159:139-42 19. Kew MC, Bersohn I, Peter 1, Wyndham CH, Seftel HC. Preliminary observations on the serum and cerebrospinal fluid enzymes in heatstroke. S Afr Med J 1967;41:530-2 20. Guze BH, Baxter LR Ir. Neuroleptic malignant syndrome. N Engl 1 Med 1985;313:163-6 21. Karnofsky DA, Abelmann WH, Craver LF, Burchenal IH. The use of the nitrogen mustards in the palliative treatment of carcinoma with particular reference to bronchogenic carcinoma. Cancer 1948;1:634-56 22. Szasz G, Gruber W, Bernt E. Creatine kinase in serum. I. Determination of optimum reaction conditions. Clin Chern 1976;22:650-6 23. Rosalki SD. Creatine phosphokinase isoenzymes. Nature 1965;207:414 24. Harrell F. The LOGIST procedure. In: SAS supplemental user's guide. Cary, NC: SAS Institute, 1980:83-107 25. O'Brien Pc. A nonparametric test for association with censored data. Biometrics 1978;34:243-50 26. Kiyasu IY. Detection of septicemia by the "shot-gun" method for assaying creatine kinase isoenzyme MB in serum [letter]. Clin Chern 1978;24:2064-5 27. Gamboa ET, Eastwood AB, Hays AP, Maxwell 1, Penn AS. Isolation of influenza virus from muscle in myoglobinuric polymyositis. Neurology 1979;29:1323-35 28. Baracos V, Rodemann HP, Dinarello CA, Goldberg AL. Stimulation of muscle protein degradation and prostaglandin Ez release by leukocytic pyrogen (interleukin-l). A mechanism for the increased degradation of muscle proteins during fever. N Engl 1 Moo 1983;308:553-8

Downloaded from http://cid.oxfordjournals.org/ at Michigan State University on June 29, 2015

4. Posner MR, Caudill MA, Brass R, Ellis E. Legionnaires' disease associated with rhabdomyolysis and myoglobinuria. Arch Intern Med 1980;140:848-50 5. Hautekeete ML, Berneman ZN, Bieger R, Stevens WJ, Bridts C, Buyssens N, Peetermans ME. Purpura fulminans in pneumococcal sepsis. Arch Intern Med 1986;146:497-9 6. Solbrig MV, Sher IH, Kula RW. Rhabdomyolysis in leptospirosis (Weil's disease). J Infect Dis 1987;156:692-3 7. Johnson WD Jr, Silva IC, Rocha H. Serum creatine phosphokinase in leptospirosis. JAMA 1975;233:981-2 8. Lannigan R, Austin TW, Vestrup 1. Myositis and rhabdomyolysis due to Staphylococcus aureus septicemia. 1 Infect Dis 1984;150:784 9. Provenza 1M, Klotz SA, Penn RL. Isolation of Francisella tularensis from blood. J Clin Microbiol 1986;24:453-5 10. Penn RL, Kinasewitz GT. Factors associated with a poor outcome in tularemia. Arch Intern Med 1987;147:265-8 11. Marino PL, Nahass GT, Novick W. Bacteremic pneumococcal pneumonia and myoglobinuric renal failure. Am 1 Med 1986;80:521-2 12. Armstrong IH. Tropical pyomyositis and myoglobinuria. Arch Intern Med 1978;138:1145-6 13. Kallen PS, Louie IS, Nies KM, Bayer AS. Infectious myositis and related syndromes. Semin Arthritis Rheum 1982;11:421-39 14. Berlin BS, Simon NM, Bovner RN. Myoglobinuria precipitated by viral infection. lAMA 1974;227:1414-5 15. Josselson 1, Pula T, Sadler JH. Acute rhabdomyolysis associated with an echovirus 9 infection. Arch Intern Med 1980;140:1671-2 16. Leibovici L, Sharir T, Kalter-Leibovici 0, Alpert G, Epstein LM. An outbreak of measles among young adults. Clinical and laboratory features in 461 patients. 1 Adolesc Health Care 1988;9:203-7 17. Antony IH. Creatine phosphokinase measurements in febrile children. Arch Neurol 1979;36:323

RID 1991;13 (March-April)