Introduction

pH1N1 (2009 pandemic influenza A) presented with substantial pulmonary involvement; however, extrapulmonary complications were not uncommon and represented an additional contribution to mortality [1–3]. Risk factors for poor outcome included multiple-organ dysfunction syndrome and requirement of mechanical ventilation or renal replacement therapies [4–6]; indeed, stage III acute kidney injury (AKI) has proved to be an independent risk factor for mortality [7].

Association between rhabdomyolysis and influenza A and B virus infection has been previously described [8, 9]. Proposed pathogenic mechanism includes viral invasion, following viral toxin and host immune-mediated cytokine production, which might be involved in muscular injury [9, 10]. However, to date, true pathogenesis remains unknown.

The purpose of this study is to assess if increased creatine kinase (CK) was related to worse global, renal, and respiratory outcomes in critically ill patients with pH1N1 infection, and if it could serve as a biomarker of severity. The primary endpoint was intensive care unit (ICU) mortality. Secondary endpoints were correlation between CK levels and ICU length of stay (LOS), mechanical ventilation (MV), and requirement of renal replacement techniques (RRT). Our hypothesis was that mild elevations of CK (below those usually seen with rhabdomyolysis) in pH1N1 infection were associated with AKI and severe respiratory failure, increasing ICU LOS and ICU mortality.

Materials and methods

The study was conducted in 148 Spanish intensive care units, from June 2009 to February 2010. CK breakpoints were 300, 500, 1,000, 2,500, 3,000, 4,000, and 10,000 UI/L. Given its observational structure, informed consent was waived. Data were obtained using a voluntary standardized registry by investigators of the Infectious Disease Working Group of the Spanish Society of Intensive Care Medicine (GTEI-SEMICYUC). The Ethics Committee of all hospitals approved this study.

We included all patients admitted to the ICU for severe acute respiratory insufficiency (SARI) with clinical suspicion of pH1N1 corresponding to the World Health Organization (WHO) case definitions [11] and microbiological confirmation. Clinical suspicion was defined by febrile (>38 °C) acute illness, respiratory symptoms (consistent with cough, sore throat), myalgia or influenza-like illness. Patients under 15 years were excluded. Microbiological confirmation was documented by real-time polymerase chain reaction (RT-PCR) (according to Centers for Disease Control and Prevention guidelines) [12] or viral culture, which were performed at the corresponding healthcare facility (pH1N1 testing was performed in each institution, or centralized in a reference laboratory when local resources were not available). Specimens were obtained from nasopharyngeal swabs or respiratory secretions in intubated patients. All patient management was determined by an attending physician, from the decision of ICU admission to ICU discharge.

Clinical data consisting of baseline demographics, comorbidities, pulmonary involvement, laboratory findings, and severity were registered at time of ICU admission, whereas organ involvement, treatment received, and prognostic data were registered during ICU and hospital stay. Demographic data included age and sex. Comorbidities data included presence or absence of diabetes mellitus, chronic kidney disease, cardiac heart failure, chronic obstructive pulmonary disease, asthma, pregnancy, body mass index (BMI) >30 kg/m2, BMI >40 kg/m2, HIV/AIDS, autoimmune disorder, hematological disorder, and neuromuscular disorder. Pulmonary data included primary diagnosis at time of ICU admission, pneumonia (viral or bacterial), chronic obstructive pulmonary disease (COPD) or asthma exacerbation, and bacterial co-infection. Laboratory data included creatine kinase, urea, and creatinine. Severity of the disease was assessed at ICU admission by Sequential Organ Failure Assessment (SOFA) [13].

Normal CK levels found in medical references range in men from 52–55 to 170–294 UI/L and in women from 39–45 to 135–238 UI/L [14, 15]. Therefore, we defined elevated CK values as those greater than or equal to 300 UI/L. Organ involvement was classified into two categories: renal and respiratory. Respiratory compromise was defined by the number of affected quadrants on chest radiograph, the number of patients who required mechanical ventilation, and duration of MV. Definition of pneumonia was established based on the Infectious Disease Society of America/American Thoracic Society Consensus Guidelines on the Management of Community-Acquired Pneumonia in Adults [16]. Primary viral pneumonia was defined as illness presenting with acute respiratory distress and unequivocal alveolar opacities involving two or more lobes with negative respiratory and blood bacterial cultures during the acute phase of influenza virus. Infections occurring later were considered nosocomial. Renal dysfunction (AKI) was defined as increase of serum creatinine of 50 % or 0.3 mg/dL from baseline values within 48 h [in accordance with Acute Kidney Injury Network (AKIN) definitions] [17], or by the requirement of renal replacement therapies, which included either dialysis or continuous veno-venous hemofiltration (CVVHF).

Statistical analysis

Due to the non-Gaussian distribution of the data, discrete variables are reported as frequency (%) and continuous variables as median (interquartile range, IQR). Odds ratio (OR) is presented with 95 % confidence interval (CI). Statistical analysis was performed using SPSS for Mac, version 18 (SPSS, Chicago, IL, USA). Univariate analysis was performed using Pearson’s χ 2 to calculate the two-tailed p value. Mann–Whitney U test was performed on quantitative variables. Statistical significance was defined by p < 0.05.

Results

Five hundred and five patients were evaluated. Patients were predominantly young (43 years), with no gender predominance. The majority had at least one underlying disease: 20 % were overweight, and between 10 and 16 % had type 2 diabetes mellitus, COPD, or asthma. Less than 5 % had chronic kidney disease. Median SOFA was 4 points. Normal CK values were documented in 64.7 %. CK was higher than 300, 500, 1,000, 2,500, 3,000 and 4,000 in 35.2, 23.8, 17.8, 4.8, 4.6, and 2.8 %, respectively. Only 1.2 % had CK levels above 10,000 UI/L at ICU admission (Table 1; Fig. 1). Median baseline values of renal function ranged within or were slightly above normal values. The majority (84 %) presented with a severe acute respiratory infection secondary to pneumonia at ICU admission. See Table 2 for clinical and laboratory characteristics of the patients. Global ICU mortality was 17.2 % (90 patients), with no difference when analyzed at the 300 and 500 UI/L breakpoints (p = 0.5 and 0.49, respectively; Table 3).

Table 1 Demographics, clinical, and laboratory characteristics of 505 patients with confirmed pH1N1 infection at time of ICU admission
Full size table
Fig. 1
figure 1

Abnormal creatine kinase levels in 505 patients with confirmed pH1N1 infection at time of ICU admission

Full size image
Table 2 Creatine kinase levels in 505 patients with confirmed pH1N1 infection at time of ICU admission
Full size table
Table 3 ICU mortality of patients with confirmed pH1N1 during ICU stay
Full size table

In patients with CK ≥500 UI/L, AKI was more frequent [26.1 versus 17.1 %, p < 0.05; OR = 1.7 (95 % CI 1.04–2.77)], with twofold requirement for RRT [9.3 versus 4.5 %, p < 0.05; OR = 2.17 (95 % CI 1.00–4.78)] (Table 4). For CK level ≥2,500 UI/L, a threefold higher requirement for renal replacement techniques was documented [17.4 versus 6.4 %; OR = 3.1 (95 % CI = 0.99–9.68)], although this did not reach statistical significance (p < 0.07).

Table 4 Renal involvement of patients with confirmed pH1N1 during ICU stay
Full size table

Greater respiratory impairment was seen in patients with CK ≥1,000 UI/L. They were more likely to have two or more quadrants affected on chest radiograph (63.2 versus 40.25 %, p < 0.01), significantly longer duration of mechanical ventilation [median (IQR) 15 (8.5–27) days versus 11 (6–17.8) days (p < 0.01)], and more frequent intubation (73.9 versus 56.7 %), although this finding did not reach statistical significance (p = 0.07) (Table 5). Patients with CK ≥1,000 UI/L also had significantly (p = 0.01) longer median ICU length of stay [13 (6–25) days versus 8 (4–16) days] and median hospital length of stay [15 (8–25) days versus 20 (13–34) days] (Table 6).

Table 5 Respiratory involvement of patients with confirmed pH1N1 during ICU stay
Full size table
Table 6 Length of stay of patients with confirmed pH1N1
Full size table

Discussion

The results of our study suggest that CK level is a biomarker of severe pH1N1 infection. Interestingly, even though no difference in mortality was observed, we found that slight elevations were associated with increased pulmonary and kidney complications, likewise increased length of stay (both ICU and hospital). Furthermore, starting from CK >500 UI/L, a breakpoint considerably close to normal upper values, significant associations with worse renal outcomes were documented. These included higher occurrence of AKI, as well as more frequent need for RRT, which progressively increased with rising CK values. Greater pulmonary injury was observed from CK >1,000 UI/L, with 50 % more involvement in chest radiograph in addition to more prolonged mechanical ventilation (Fig. 2).

Fig. 2
figure 2

Kaplan–Meier survival curves according to creatine kinase levels. Continuous line: CK <300 UI/L; dashed line: CK = 301–1,000 UI/L, dotted line: CK > 1,000 UI/L. Log-rank (Mantel–Cox) 0.12

Full size image

Although some isolated case reports of pH1N1-associated rhabdomyolysis have been published [18–22], uncertainty still remains regarding the clinical implications of elevated creatine kinase levels. While the association of elevated CK with requirement for RRT has been documented at CK level around 11,000 UI/L in several pH1N1 case reports [23, 24], others have failed to demonstrate a need for RRT, even at levels of CK of 16,000 UI/L and above [22, 25, 26]. A series of case reports showed elevated levels (1,000–5,000 UI/L) of CK in severe presentations of pH1N1 infection [27], and one case report showed AKI due to pH1N1-induced rhabdomyolysis (with CK levels above 40,000 UI/L) [28]. To our knowledge, this is the first large series that evaluates prospectively the relationship between CK levels and clinical outcomes in a large cohort of patients with pH1N1 infection.

Moreover, there is not a validated CK breakpoint for diagnosis of rhabdomyolysis. Literature [29–32] defines rhabdomyolysis by elevation of CK values above 10,000 UI/L (almost exclusively of the skeletal muscle fraction, MM) which is usually accompanied by elevation of other enzymes (aldolase, lactate dehydrogenase (LDH), aspartate aminotransferase, and alanine aminotransferase). Likewise, medical literature lacks specific information and evidence for CK values regarding rhabdomyolysis and its clinical course. The term hyperCKemia [33] refers to a less severe form of rhabdomyolysis, secondary to chronic or intermittent muscle destruction characterized by mild elevation of creatine kinase, which is usually asymptomatic, and without renal impairment. Kidney failure has been associated with levels of CK above 20,000 UI/L in patients without concomitant significant disease (13–50 % of cases), however, lower cutoff points (5,000 UI/L) have been reported in patients with dehydration, sepsis or acidosis. In our study, slight CK elevations within the range of hyperCKemia at ICU admission (which usually are not taken into consideration in the management of patients) were associated with kidney and lung complications and worse outcomes. This raises the question of whether the standard concept of pathological values of CK associated with rhabdomyolysis may not apply in other contexts and whether the definition of new risk-based breakpoints would have greater relevance. Also of interest, we found an association between CK levels and severity of respiratory failure and duration of mechanical ventilation.

Limitations are inherent to the design of this study, particularly the lack of measurement of CK levels during ICU stay to assess the exact threshold that might cause AKI in pH1N1 infection and therefore establish which CK levels were related to increased risk for developing AKI and which levels are crucial for AKI. Decisions for intubation or renal replacement therapy were not standardized, and this weakness may limit generalization. Also, creatine kinase MB fraction was not determined, so cardiac origin of CK elevation could not be ruled out. Likewise, other biomarkers such as aldolase or LDH were not recorded.

In summary, our findings suggest that CK at ICU admission serves as a biomarker of severity in pH1N1 infection and may provide clinical insight into which patients are at higher risk of renal and respiratory artificial support, and require increased healthcare resources.