acfe

ezGuciJoalpanircu

Available online 28 March 2013

Introduction amongst others (Chertow et al., 2005). However, excessive immune ac-

Experimental and Molecular Pathology 94 (2013) 486492

Contents lists available at SciVerse ScienceDirect

Experimental and M

j ourna l homepage: www.e lsDuring March of 2009, an unusual outbreak of inuenza-like ill-ness was observed in Mexico later found to be caused by a new strainof the swine-origin inuenza virus, now called pandemic A/H1N12009 virus (WHO, 2009). The new virus was able to produce acuterespiratory distress syndrome (ARDS) requiring mechanical venti-lation, acute kidney injury (AKI), and death (Perez-Padilla et al., 2009).AKI is uncommon in patients with seasonal inuenza A virus infec-tion, whereas a high incidence of AKI (33.6%51%) was reported inpatients with severe forms of pandemic A/H1N1 2009 virus infection,and presents a mortality rate between 36.6 and 51% (Nin et al., 2011;Pettil et al., 2011). The AKI associated to inuenza virus infection hasbeen mostly described as being due to rhabdomyolisis (Cunninghamet al., 1979). The development of AKI in this subset of patients has

tivation and a dysregulated production of inammatorymediators, typ-ical of severe disease due to pandemic A/H1N1 virus infection, maycontribute to the pathogenesis of early AKI (La Gruta et al., 2007). Inaddition, angiogenic factors may play an important role in the patho-genesis of AKI and/or in the tissue repair mechanisms induced byischemia associated AKI (Villanueva et al., 2006), but it is unclear ifthey also contribute to the development of AKI in ARDS patients dueto the pandemic A/H1N1 virus infection.

In the present study, we compared the concentration of angiogenicfactors as well as pro-inammatory cytokines and chemokines in bloodfrom patients with ARDS with and without AKI.

Materials and methodsbeen linked to ischemic damage, drug toxicity

Correspondence to: E. Bautista, Intensive Care Unit, InstRespiratorias Ismael Coso Villegas, Mexico City, Mexico. Correspondence to: J. Ziga, Department of ImmuEnfermedades Respiratorias Ismael Coso Villegas, TlalpaCity, Mexico.

E-mail addresses: drbautista40@iner.gob.mx (E. Bau(J. Ziga).

0014-4800/$ see front matter 2013 Elsevier Inc. Allhttp://dx.doi.org/10.1016/j.yexmp.2013.03.007Keywords:Acute kidney failureA/H1N1ARDSChemokinesMCP-1VEGFA/H1N1 virus infection. The prole of angiogenic and inammatory factors in ARDS patients may be relevantfor AKI. We analyzed the serum levels of several angiogenic factors, cytokines, and chemokines in 32 patientswith A/H1N1 virus infection (17 with ARDS/AKI and 15 ARDS patients who did not developed AKI) and in 18healthy controls. Signicantly higher levels of VEGF, MCP-1, IL-6, IL-8 and IP-10 in ARDS/AKI patients weredetected. Adjusting by confusing variables, levels of MCP-1 150 pg/mL (OR = 12.0, p = 0.04) and VEGF225 pg/mL (OR = 6.4, p = 0.03) were associated with the development of AKI in ARDS patients. Higherlevels of MCP-1 and IP-10 were signicantly associated with a higher risk of death in patients with ARDS(hazard ratio (HR) = 10.0, p = 0.02; HR = 25.5, p = 0.03, respectively) even taking into account AKI.Patients with inuenza A/H1N1 infection and ARDS/AKI have an over-production of MCP-1, VEGF andIP-10 possibly contributing to kidney injury and are associated to a higher risk of death.

2013 Elsevier Inc. All rights reserved.Article history:Received 6 November 2012Acute kidney injury (AKI) is often associated to acute respiratory distress syndrome (ARDS) due to inuenzaa b s t r a c ta r t i c l e i n f oAngiogenic and inammatory markers inrenal injury associated to A/H1N1 virus in

Edgar Bautista a,, Magali Arcos a, Lus Jimenez-AlvarMara E. Vzquez a, Erika Pea a, Anjarath Higuera a,Alfredo Cruz-Lagunas a, Sara A. Garca-Moreno a, FranRicardo Correa-Rotter c, Jos Rogelio Prez-Padilla a,a Instituto Nacional de Enfermedades Respiratorias Ismael Coso Villegas, Tlalpan 4502, Tlab Facultad de Ciencias, Universidad Nacional Autnoma de Mxico, Av. Universidad 3000, Cc Instituto Nacional de Ciencias Mdicas y Nutricin Salvador Zubirn. Mexico City, Mexico, and interstitial nephritis

ituto Nacional de Enfermedades

nology, Instituto Nacional den 4502, Tlalpan 14080, Mexico

tista), joazu@yahoo.com

rights reserved.ute respiratory distress syndrome andctiona, Ma. Cecilia Garca-Sancho a,stavo Ramrez a, Rosario Fernndez-Plata a,sco Urrea a, Remedios Ramrez b,qun Ziga a,, 14080, Mexico City, Mexicoito Exterior S/N, Coyoacn, 04510, Mexico City, Mexico

olecular Pathology

ev ie r .com/ locate /yexmpSubjects

A total group of 32 patients with ARDS associated to pandemicA/H1N1 inuenza virus infection were studied. The diagnosis of ARDSwasbased on standard denitions (Bernard et al., 1994) and thepatientswere classied in those who developed AKI (ARDS/AKI: N = 17), andwho did not develop AKI (ARDS: N = 15). We excluded patients whodeveloped renal failure after 3 days of admission to reduce the risk

counterstained with hematoxylin. The primary antibody was replacedby non-immune serum for negative controls.

Data collection and statistical analysis

Demographic characteristics, vital signs, clinical outcomes, treat-ments, mechanical ventilation and gas exchange parameters, labora-tory tests, and APACHE II scores were determined at days 1, 2, and 3after the admission to the ICU. All measures summarized were thenexpressed as means and standard deviation, or as frequencies andproportions. Comparisons of the continuous variables among groupswere performed using the MannWhitneyWilcoxon test, whereasthe X2 test was used to compare frequencies. Cox logistic regressionmodels were constructed to explore the factors associated with deathdue to ARDS/AKI. The independent variables tested in the models were:demographic and anthropometric data, lung function, laboratory nd-ings, and comorbid conditions. Any variables that proved signicantin the univariate analysis were included in the multivariate models.The signicance level was set at p b 0.05. All analyses were performedusing a commercially available statistical package (Stata v. 10.0 StataCorp,College Station, TX, USA). Differences in the serum levels of cytokines/chemokines/growth factors between controls, ARDS/AKI, and ARDSpatients were evaluated by the MannWhitneyWilcoxon test, andgraphics were generated using the Graph Pad Prism software version5.04 (Graph Pad Software, La Jolla, CA).

Results

The clinical and demographic characteristics of the ARDS/AKIand ARDS patients are presented in Table 1. All patients required

Table 1Demographic and clinical characteristics of patients with ARDS-AKI and ARDS associatedto A/H1N1 infection.

Variable ARDS/AKI (N = 17) ARDS (N = 15) Pa

487E. Bautista et al. / Experimental and Molecular Pathology 94 (2013) 486492of including patients with late renal failure, which could be attributedto other pathogenic mechanisms. The diagnosis of renal failure wasestablished according to the Acute Kidney Injury Network (AKIN)criteria: 1) Rapid time course (less than 48 h), 2) Reduction of kidneyfunction, 3) Rise in serum creatinine, 4) Absolute increase in serumcreatinine of 0.3 mg/dl (26.4 mol/l), 5) Percentage increase inserum creatinine of 50%, and 6) Reduction in urine output, denedas b0.5 ml/kg/h for more than 6 h.

We simultaneously recruited a control group of 18 asymptomatichealthy controls (mean age SD 35 13 years, 8 males and 10 fe-males) thatwere in close contactwithA/H1N1-conrmed critical patientsbut did not developed severe respiratory acute disease. Their serum sam-ples were obtained at the same time as the critical patients' samples.

The Institutional ReviewBoard of theNational Institute of RespiratoryDiseases (INER) reviewed and approved the protocol for biomarkerstudies (Protocol number B27-10) under which all subjects wererecruited in accordance with the ethical standards of the Declarationof Helsinki in 1964. All subjects provided written informed consentfor these studies, and they authorized the storage of their serum andplasma samples at INER repositories for this and future studies. Afterobtaining the signed informed consent letters from patients and con-trols, serum samples were obtained.

A/H1N1 virus detection

Nasopharyngeal swab samples were obtained from hospitalizedpatients and processed at the INER Microbiology Laboratory for RNAisolation using the viral RNA mini kit (Qiagen Westburg, Leusden,The Netherlands). Detection of inuenza A/H1N1 virus in respiratoryspecimens was assessed by real time RT-PCR according to the CDCand World Health Organization guidelines.

Measurement of angiogenic factors, cytokines, and chemokines

A set of various angiogenic factors, vascular endothelial growthfactor A (VEGF-A), broblast growth factor 2 (FGF-2), platelet derivedgrowth factor (PDGF), monocyte chemotactic protein-1 (MCP-1), andplasminogen activator inhibitor-1 (PAI-1) as well as pro-inammatorycytokines/chemokines (IL-6, IL-8, IFN-, IP-10 and TNF-) were mea-sured in serum samples obtained from all studied patients and controlsby Luminex (Bio-Rad Laboratories, Inc., CA, USA). The concentrations ofangiogenic factors, cytokines, and chemokines, were calculated usingBio-Plex v 4.1 software (Bio-Rad Laboratories, Inc., Hercules, CA, USA).

Muscle biopsies

Muscle sampleswere obtained from the quadriceps (vastus lateralis)by openmuscle biopsy fromAKI patients and stainedwith hematoxylinand eosin (HE) and Masson's trichrome stains, and examined by lightmicroscopy to rule out muscle destruction and rhabdomyolysis.

VEGF determination by immunohistochemistry

Renal sections from deceased patients who developed AKI associ-ated to severe pneumonia by the A/H1N1 infection and from controlindividuals without renal disease, were obtained from the pathologytissue bank of the INER. Fixed tissue sections were treated as previ-ously described (Selman et al., 2000). Antigen retrieval was carriedout in citrate buffer 10 mM, pH 6.0 for 6 min in a microwave. Slideswere incubated with anti-VEGF monoclonal antibody (5 g/ml;ab-46154; Abcam, Cambridge, UK) at 4 C overnight and thenwith a sec-ondary biotinylated anti-immunoglobulin antibody followed by horse-radish peroxidase-conjugated streptavidin (BioGenex, San Ramon, CA,USA). They were used according to the manufacturer instructions,using 3-amino-9-ethyl-carbazole (AEC, BioGenex) as a substrate in

acetate buffer containing 0.05% H2O2. The tissue sections wereAge years 41.4 (11.8) 39.6 (11.5) 0.7Male (%) 13/17 (76.5) 9/15 (60) 0.3Obesity (BMI 30 kg/m2) 13/17 (76.5) 7/15 (46.7) 0.08Oseltamivir delay (days) 10.5 (4.9) 8.8 (2.5) 0.3Mechanical ventilation (days) 14.2 (5.8) 16.1 (9.1) 0.5APACHE II score 21.4 (5.3) 18.1 (2.8) 0.05Hospitalization (days) 22 (8.3) 28.7 (16.1) 0.1Co-morbidities (%) 13/17 (76.5) 5/15 (80) 0.8Current smoking (%) 10/17 (58.8) 6/15 (40) 0.3High blood pressure (%) 6/17 (35.3) 3/15 (20) 0.3Prior corticosteroid treatment (%) 1/17 (6) 1/15 (6.7) 0.9Deaths (%) 6/17 (35.2) 1/15 (0.06) 0.3Laboratory ndingsPlatelets (x103) 134 (470) 169 (78) 0.1Glucose (mg/dL) 154.5 (52.5) 134.5 (43.19) 0.07Creatinine (mg/dL) 1.4 (0.7) 0.8 (0.2) 0.03Albumine (gr/dl) 3.4 (0.4) 2.9 (0.6) 0.04Diuresis (L/day) 1.2 (0.8) 3.5 (2.5) 0.003BUN (mg/dl) 19.6 (12) 15.1 (4.9) 0.2LDH (IU/L) 1553 (2420) 877 (399) 0.5CPK (IU/L) 1085 (1942) 331 (370) 0.07Na (mEq/L) 137 (4.3) 139 (4) 0.09PaO2 (Torr) 66.8 (23.4) 69.1 (31.3) 0.8PaCO2 (Torr) 32.1 (9.7) 34.4 (12.4) 0.6PaO2/FiO2 (Torr) 214 (161) 199 (135) 0.8PEEP (cm H20) 14.9 (5.7) 13.3 (4.2) 0.3Angiogenic/inammatory factorsVEGF (225 pg/mL) 12/17 (70.6) 4/15 (26.7) 0.01MCP-1 (150 pg/mL) 8/17 (47.1) 1/15 (6.7) 0.01IL-6 (50 pg/mL) 11/17 (64.7) 4/15 (26.7) 0.03IL-8 (40 pg/mL) 11/17 (64.7) 3/15 (20) 0.01IP-10 (6000 pg/mL) 13/17 (76.5) 6/15 (40) 0.03IFN- (100 pg/mL) 4/17 (23.5) 11/15 (73.3) 0.005

ARDS: Acute respiratory distress syndrome; AKI: acute kidney injury. Data are means standard deviation, or number and percentage.

a Comparisons of the continuous variables among groups were performed using the2MannWhitneyWilcoxon test, whereas the X test was used to compare frequencies.

mechanical ventilation and were admitted to the ICU at the INER inMexico City during the rst A/H1N1 outbreak in 2009. Six of the ARDSpatients who developed AKI (35.2%) had a fatal outcome. The meanage of the ARDS/AKI patients was 41.4 11.8 years, whereas in theARDS group was 39.6 11.5 years. No signicant differences between

ventilation period. Muscle biopsies from ARDS/AKI patients revealedslight lymphocyte inltration and muscle destruction characterized by

Fatal outcome in ARDS patients is associated to increased levels ofinammatory mediators and BMI.

Table 3 summarizes demographic and clinical variables accordingto vital status at the end of hospitalization. Deceased patients had, as

nd werumRDof A

488 E. Bautista et al. / Experimental and Molecular Pathology 94 (2013) 486492myositis and atrophy rather than rhadbomyolysis.

Increased levels of angiogenic and inammatory mediators wereobserved in ARDS/AKI patients

Table 1 depicts the levels of angiogenic and inammatory factorsin the two studied groups with signicantly higher levels of VEGF,MCP-1, IL-6, IL-8, and IP-10 in the ARDS/AKI group compared with theARDS group. In contrast, lower levels of IFN- were observed in theARDS group. The levels of the inammatory mediators (IL-6, IL-8, IP-10,TNF-, and IFN-) in patients were signicantly higher (particularlyin those who developed AKI) than in controls (see Fig. 1).

ARDS/AKI patients and ARDS patients exhibited markedly higherlevels of angiogenic mediators VEGF, MCP-1, PAI-1, and FGF, butnot PDGF-bb, compared with controls. Interestingly, lower levels ofPDGF-bb and FGF-2 were observed in ARDS/AKI patients when com-pared with the ARDS group (see Fig. 1). VEGF was expressed diffuselyin the cytoplasmof renal tubuleswith hydropic degeneration in patientswith AKI but not in controls (see Fig. 2).

After multivariate analysis controlling by confusion factors includ-ing age, gender, PaO2/FiO2, and IFN-, we found that MCP-1 levels150 pg/mL(OR = 12.0, p = 0.04) or VEGF 225 pg/mL (OR = 6.4,p = 0.03) were associated with a higher risk of developing AKI in ARDSpatients. In contrast, levels of IFN-100 pg/mL protected against thedevelopment of AKI (OR = 0.03, p = 0.01, see Table 2).More than halfof ARDS/AKI patients were obese and had a greater BMI than ARDSpatients, although this difference was not statistically signicant.

Fig. 1. Serum levels of inammatory and angiogenic factors in patients with ARDS with awere observed in ARDS/AKI patients compared with ARDS and controls. In contrast, sangiogenic factors (B), higher serum levels of VEGF, MCP-1, and PAI-1 were observed in AAKI (ARDS) and controls. In contrast, lower levels of FGF-2 were detected in the groupARDS/AKI and ARDS patients were observed in gender, oseltamivirtreatment delay, mechanical ventilation days, APACHE II score, lengthof hospital stay, and presence of co-morbidities such as obesity, tobaccouse, hypertension, and immunosuppression. The bodymass index (BMI)in the group of ARDS/AKI was higher (34.6 8.5) compared with29.3 7.5 in ARDS patients, with marginal signicance (p = 0.08),and controls (28.2 3.9 kg/m2) (p b 0.05). Regarding laboratoryndings, signicant differences were detected in levels of creatinine,albumin, and BUN. As expected, the diuresis was markedly lower in thegroup of ARDS AKI patients.

The main signs and symptoms at the start of the illness includedfever, myalgias, cough, and headaches, while chest pain, dyspnea,and cyanosis occurred after the third day. Oseltamivir dose and dura-tion of treatment were 150 mg/day during the period of requiredmechanical ventilation. In addition, all patients received ceftriaxoneand clarithromycin from the time of admission and treatment wassustained while the patients were on mechanical ventilation, but wasupgraded in case of fever, if leukocytosis developed again, if new inl-trates on chest X-rays appeared, or if antibiotic resistance was docu-mented in samples from bronchial aspirates. Blood, urine, and bronchialsecretion cultures performed at the time of hospitalization in the ICU(at the time when the samples for our experiments were also obtained)revealed neither bacterial, fungal, nor mycobacterial secondary infec-tions. Systemic corticosteroids were administered to all patients at adose of 1 mg/kg/day of methylprednisolone during the mechanicalwere signicantly lower than those observed in ARDS patients and controls. The results areexpected, evidence of severe ARDS and impairment of gas exchangeor multi organic failure with signicant differences in levels of glu-cose, creatinine, LDH, SaO2, PaCO2, PaO2/FiO2, and PEEP levels com-pared to ARDS patients who survived. Serum levels of MCP-1, IL-6,and IP-10 were higher in ARDS patients with a fatal outcome thanin survivors (see Table 3).

Patients with ARDS had a higher risk of death in the presence oflevels 150 pg/mL of MCP-1 (hazard rate (HR) = 10.0, p = 0.02) orlevels of IP-10 6000 pg/mL (HR = 25.5, p = 0.03), even adjustingby days of illness at admission, gender, PaO2/FiO2, age and obesity(BMI 30 kg/m2) in a Cox regression model (see Table 4).

Discussion

AKI is often associated to severe pneumonia due to inuenzaA/H1N1 virus infection, and has a high incidence of mortality amongcritical patients (Abdulkader et al., 2010; Nin et al., 2011). Our patientswith ARDS/AKI exhibited a marked increase in the levels of IL-6, IL-8,IP-10, and in the levels of pro-angiogenic factors VEGF and MCP-1,and lower levels of PDGF-bb and FGF-2 compared with the ARDSpatients who did not develop AKI. After adjustment by age, gender, andcomorbidities, we found that high levels of MCP-1 and VEGF were asso-ciatedwith the development of AKI in ARDS patients.We also found thathigh levels of MCP-1 and IP-10 were associated with an increased riskof death in ARDS patients, even taking into account kidney damage.Although gas exchange was more deteriorated in non-survivors com-pared to survivors, it did not differwhether or not the patients developedAKI and therefore, although AKI was present in patients with severedisease and ARDS, it was not appearing simply in those with the mostsevere lung damage.

The development of AKI due to ischemia reperfusion injury resultsin a signicant reduction of the glomerular ltration rate (Isaac et al.,2007), causing damage in the proximal tubule, alterations in renalvascular resistance, and injury to the renal microvasculature (Bonventreand Weinberg, 2003). Hypoxia-inducible factors (HIFs) are activatedin response to low oxygen supply, which stimulate the expression ofpro-angiogenic genes (Haase, 2006).

Inammation and angiogenesis are two closely related processes.Several studies have documented that besides the contribution ofcytokines, chemokines, and growth factors in the development ofinammatory responses, these molecules also exert pro or anti-angiogenic activities (Dimberg, 2010). MCP-1 is a key factor involvedin the trafcking and activation of monocytes and macrophages, con-tributing to the inammatory state in response to pathogens or in-ammatory stimuli (Fuentes et al., 1995). In addition, pro-angiogenicproperties of MCP-1 have been described (Goede et al., 1999). Studiesin experimental models have demonstrated that MCP-1 induces theexpression of HIF-1 by endothelial cells and VEGF gene expression,promoting angiogenesis (Hong et al., 2005).

Here we describe that increased levels of MCP-1 may increase therisk of death in patients with inuenza and ARDS, as its expressionfacilitates acute ischemic and toxic kidney injury. Functional studiesin both human and animal models have suggested that MCP-1 mightbe useful as an early biomarker of AKI, which provides complementaryinformation to other AKI-classical biomarkers such as NGAL (Munshi

ithout AKI and healthy controls. A marked increase in the levels of IL-6, IL-8, and IP-10concentrations of TNF- and IFN- were higher in the ARDS group (A). Regarding

S patients who developed AKI (ARDS/AKI) when compared with ARDS patients withoutRDS/AKI when compared to ARDS group. The levels of PDGF-bb in ARDS/AKI patients

shown as means SEM.

489E. Bautista et al. / Experimental and Molecular Pathology 94 (2013) 486492

ctioubulesbas

490 E. Bautista et al. / Experimental and Molecular Pathology 94 (2013) 486492Fig. 2. Immunohistochemistry assays of VEGF expression. Representative renal tissue sewith acute tubular damage characterized by extense hydropic degeneration, epithelial tperformed on parafn-embedded renal sections to determine VEGF expression. The tubuexpression. In patients with renal damage, an intense cellular focal positivity (apical andlower in the distal tubule compared with the expression in ARDS/AKI patients.et al., 2011). Interestingly, we also observed a signicant increase ofVEGF in ARDS/AKI patients; however, in the multivariable analysisthis association did not reach statistical signicance. From the biologicalpoint of view, this nding is relevant because of the well-described roleof VEGF in the pathogenesis of renal disease. The role of VEGF in renaldisease in experimental models and humans is complex and contro-versial (Schrijvers et al., 2004), as progression of chronic glomerularlesions has been associated to low levels of VEFG expression (Ereminaet al., 2003). On the other hand, over-expression of specic isoformsof VEFG (VEGF-164) have been associated with marked alterations inthe glomerular ltration rate (Veron et al., 2010). High levels of VEGFobserved in early stages of AKI associated to ARDS might reect the in-duction of protective and tissue repair mechanisms, although a betterunderstanding of the role of VEGF in renal disease associated to severeinuenza is needed.

In this study we also found a signicant association between highlevels of IP-10, a chemokine with anti-angiogenic activity, and anincreased risk of mortality in ARDS patients. IP-10 (CXCL10) is a CXCchemokine that binds CXCR3 receptor to induce chemotaxis, apoptosis,cell growth, and angiostasis (Angiolillo et al., 1995). Urinary IP-10 levelsmay predict renal function and mortality in ischemic AKI (Kwon et al.,2010). The mechanism(s) of MCP-1- and/or IP-10-mediated kidneyinjury are not well understood, but thesemolecules may inuence traf-cking of inammatory cells including neutrophils, monocytes/macro-phages, and activated T cells to the renal tissue, contributing to thedevelopment of pathogenic inammatory responses.

On the other hand, lower plasma levels of PDGF-bb and FGF-2, twokey factors involved in renal growth and regeneration in both humansand experimental models of renal damage (Alpers, 1992), weredetected in patients with ARDS/AKI compared with ARDS patients.Studies in experimental models of ischemia/reperfusion have demon-strated that expression of PDGF-bb is enhanced in kidneys withns obtained from ARDS/AKI fatal patients showed histopathological changes consistentlar shedding, and necrosis with loss of cellular nuclei. Immunohistological analysis waswith hydropic degeneration associated to hypoxia exhibit a diffuse cytoplasmatic VEGFal) was observed in the preserved tubules. In control renal tissue, VEGF expression wasischemic tubular injury. The experimental inhibition of PDGF-bb/PDGFRs axis in rats with the injured kidney was associated with moresevere renal damage and a higher mortality rate (Nakagawa et al.,1999). In contrast, in experimental lipopolysaccharide-induced acutekidney injury the over-expression of FGF-2 in renal tissuesmay contrib-ute to renal damage (Mattison et al., 2012).

Hyper-activation of the inammatory response and high levelsof IL-6, IL-8, and MCP-1 are often observed in patients with A/H1N1infection who develop severe pneumonia (Bermejo-Martin et al.,2009; Lee et al., 2011; Ziga et al., 2011). The overproduction ofpro-inammatory cytokines and chemokines, also called cytokinestorm, induce antiviral responses by recruiting neutrophils andmacro-phages to the infected tissues, but may induce a systemic inammatorystate promoting the development of organ damage secondary to the se-vere inuenza A virus infection. Evidence suggests that IL-6 and IL-8may be early biomarkers of AKI in children undergoing cardiac surgery.In addition, high levels of IL-6 and IL-8 were predictive of prolongedmechanical ventilation in these patients (Liu et al., 2009). Studies inexperimental models have demonstrated that IL-6 decient mice areresistant to the development of AKI when compared to wild type mice(Nechemia-Arbely et al., 2008). Furthermore, increased plasminogenactivator inhibitor-1 (PAI-1) and IL-6 serum levels have been associatedwith AKI in critical patients with acute lung injury (ALI), suggesting thatthe combined alterations in coagulation and inammation participatein the pathogenesis of AKI in patients with ALI (Liu et al., 2007). Weobserved a signicant increase in IL-6 levels in ARDS patients withAKI, suggesting a pathogenic role.

Finally, even though a higher BMI was not signicantly associatedwith renal involvement and with mortality in ARDS patients, animportant percentage of patients included in this study had BMI30 kg/m2. The BMI in the control group (28.2 3.9 kg/m2) waslower than in patients with signicant differences in the levels of

inammatory mediators, suggesting that obesity may contribute, atleast in part, to the inammatory phenotype in critical disease.

In summary, high levels of MCP-1 and VEGF were found in indi-viduals with ARDS secondary to inuenza that developed AKI. Like-wise, high levels of MCP-1 and IP-10 were associated with higherrisk of death in these patients. Whether the increased levels of theseangiogenic/pro-inammatory mediators are just a consequence ofextensive damage by inuenza and part of homeostatic or regulatorymechanisms, or contribute to the kidney injury and multiple organfailure, as some animal experiments suggest, is currently unknown.The described proles of angiogenic and pro-inammatory mediatorsmay help to identify individuals with a higher risk of adverse out-comes in A/H1N1 associated ARDS.

Conict of interest statement

The authors declare that there are no conicts of interest.

Isaac, J., et al., 2007. Extent of glomerular tubularization is an indicator of the severity ofexperimental acute kidney injury in mice. Nephron Experimental Nephrology 105,e33e40.

Table 2Factors associated to AKI in ARDS patients after logistic regression analysis.

Variable ARDS/AKI(N = 17)

ARDS(N = 15)

Adjusted OR(95%CI)

p

MCP-1 (150 pg/mL) 8/17 (47.1) 1/15 (6.7) 12.0 (1.1130) 0.04VEGF (225 pg/mL) 12/17 (70.6) 4/15 (26.7) 6.4 (1.135.7) 0.03PaO2/FiO2 (Torr) 214 (161) 199 (135) 1.002 (0.991.008) 0.5IFN- (100 pg/mL) 4/17 (23.5) 11/15 (73.3) 0.03 (0.0010.55) 0.01

Data are expressed as numbers and percentages and mean standard deviation (SD).

Table 3Demographic and clinical characteristics of patients with ARDS who died versuspatients with ARDS who survived associated to A/H1N1 infection.

Obesity (BMI 30 kg/m2) 5/7 (71.4) 15/25 (60) 0.6

491E. Bautista et al. / Experimental and Molecular Pathology 94 (2013) 486492Oseltamivir delay (days) 8.1 (2.1) 10.1 (4.3) 0.3Mechanical ventilation (days) 16.7 (6.5) 14.6 (7.8) 0.5APACHE II score 20.1 (4.9) 19.8 (4.5) 0.8Hospitalization (days) 17.1 (7.1) 27.4 (13.3) 0.06Co-morbidities (%) 5/7 (71.4) 20/25 (80) 0.6Current Smoking (%) 5/7 (71.4) 11/25 (44) 0.2High blood pressure (%) 1/7 (14.3) 8/25 (32) 0.4Laboratory ndingsPlatelets (103/ul) 136.3 (64.4) 155.1 (65.5) 0.5Glucose (mg/dL) 187 (64.6) 133.2 (36.5) 0.006Creatinine 3 (mg/dL) 2.9 (2.9) 1.2 (1.1) 0.01Albumine (gr/dl) 2.9 (0.2) 3.2 (0.6) 0.2Diuresis (L) 1981 (1998) 2280 (2157) 0.8BUN (mg/dl) 21.6 (15.1) 16.3 (7.4) 0.2LDH (IU/L) 2603 (3637) 853 (381) 0.01CPK (IU/L) 1568 (3030) 498 (447) 0.08Na (mEq/L) 140 (4.3) 137 (4.1) 0.07Variable ARDS death(N = 7)

ARDS alive(N = 25)

pa

Age years 39.4 (8.9) 40.9 (12.2) 0.8Male (%) 8/7 (85.7) 16/25 (64) 0.3SaO2 (%) 70.1 (21.9) 83.7 (11.0) 0.02PaO2 (Torr) 71.8 (17.4) 66.7 (29.3) 0.7PaCO2 (Torr) 43 (14.5) 30.5 (8.1) 0.005PaO2/FiO2 (Torr) 112 (99) 233 (149) 0.01PEEP 19.6 (2.3) 12.6 (4.5) 0.0005Angiogenic/inammatory factorsb

MCP-1 (150 pg/mL) 5/7 (71.4) 4/25 (16) 0.004IP-10 (6000 pg/mL) 6/7 (85.7) 10/25 (40) 0.03IL-6 (50 pg/mL) 6/7 (85.7) 9/25 (36) 0.02

ARDS: Acute respiratory distress syndrome; AKI: acute kidney injury. Data are means standard deviation, or number and percentage.

a Comparisons of the continuous variables among groups were performed using theMannWhitneyWilcoxon test, whereas the X2 test was used to compare frequencies.

b Only markers with p values b 0.05 are shown.Acknowledgments

We thank Dr. Steven Mink for his valuable critical review. We alsothank the group of Respirologists from the INER and ABC hospitals:Arturo Corts Telles, Mariano Arturo Daz Orta, Fabiola FernndezFigueroa, Enrique Monares Zepeda and Jesus A. Cedillo Jurez for theircollaboration in the collection of clinical data.

This study was supported by grants of the National Council ofScience and Technology of Mexico (CONACYT grants numbers: 127002and 126698).

References

Abdulkader, R.C., et al., 2010. Characteristics of acute kidney injury in patients infectedwith the 2009 inuenza A (H1N1) virus. Clinical Journal of the American Society ofNephrology 5, 19161921.

Alpers, C.E., 1992. Developmental patterns of PDGF B-chain, PDGF-receptor, and a-actinexpression in human glomerulogenesis. Kidney International 42, 390399.

Angiolillo, A.L., et al., 1995. Human interferon-inducible protein 10 is a potent inhibitorof angiogenesis in vivo. The Journal of Experimental Medicine 182, 155162.

Bermejo-Martin, J.F., et al., 2009. Th1 and Th17 hypercytokinemia as early hostresponse signature in severe pandemic inuenza. Critical Care 13, R201.

Bernard, G.R., et al., 1994. The AmericanEuropean consensus conference on ARDS:denitions,mechanisms, relevant outcomes and clinical trials coordination. AmericanJournal of Respiratory and Critical Care Medicine 149, 818824.

Bonventre, J.V., Weinberg, J.M., 2003. Recent advances in the pathophysiology of ischemicacute renal failure. Journal of the American Society of Nephrology 14, 21992210.

Chertow, G.M., et al., 2005. Acute kidney injury, mortality, length of stay, and costs inhospitalized patients. Journal of the American Society of Nephrology 16, 33653370.

Cunningham, E., Kohli, R., Venuto, R.C., 1979. Inuenza-associated myoglobinuric renalfailure. JAMA: The journal of the American Medical Association 242, 24282429.

Dimberg, A., 2010. Chemokines in angiogenesis. Current Topics in Microbiology andImmunology 341, 5980.

Eremina, V., et al., 2003. Glomerular-specic alterations of VEGF-A expression lead todistinct congenital and acquired renal diseases. The Journal of Clinical Investigation111, 707716.

Fuentes, M.L., et al., 1995. Controlled recruitment of monocytes and macrophages tospecic organs through transgenic expression of monocyte chemoattractantprotein-1. Journal of Immunology 155, 57695776.

Goede, V., et al., 1999. Induction of inammatory angiogenesis by monocytechemoattractant protein-1. International Journal of Cancer 82, 765770.

Haase, V.H., 2006. Hypoxia-inducible factors in the kidney. American Journal of Physi-ology. Renal Physiology 291, F271F281.

Hong, K.H., et al., 2005. Monocyte chemoattractant protein-1-induced angiogenesis ismediated by vascular endothelial growth factor-A. Blood 105, 14051407.

Table 4Factors associated with fatal outcome in patients with ARDS due to A/H1N1 infection,by Cox regression analysis with independent variable days of evolution and death orsurvival as dependent variable.*

Variable ARDS deaths(N = 7)

ARDS survived(N = 25)

Adjusted HR(95%CI)

P

Male (%) 8/7 (85.7) 16/25 (64) 1.7 (0.214.9) 0.6Obesity [BMI (30 kg/m2)] 5/7 (71.4) 15/25 (60) 0.1 (0.0061.6) 0.1IP-10 (6000 pg/mL) 6/7 (85.7) 10/25 (40) 25.5 (1.2553) 0.03MCP-1 (150 pg/mL) 5/7 (71.4) 4/25 (16) 10.0 (1.6273) 0.02PaO2/FiO2 (Torr) 112 (99) 233 (149) 0.99 (0.971.00) 0.2

Data are expressed as numbers and percentages and mean standard deviation (SD).Kwon, O., et al., 2010. Simultaneous monitoring of multiple urinary cytokines maypredict renal and patient outcome in ischemic AKI. Renal Failure 32, 699708.

La Gruta, N.L., et al., 2007. A question of self-preservation: immunopathology in inuenzavirus infection. Immunology and Cell Biology 85, 8592.

Lee, N., et al., 2011. Cytokine response patterns in severe pandemic 2009 H1N1 andseasonal inuenza among hospitalized adults. PloS One 6, e26050.

Liu, K.D., et al., 2007. Predictive and pathogenetic value of plasma biomarkers for acutekidney injury inpatientswith acute lung injury. Critical CareMedicine 35, 27552761.

Liu, K.D., et al., 2009. Serum interleukin-6 and interleukin-8 are early biomarkers ofacute kidney injury and predict prolonged mechanical ventilation in childrenundergoing cardiac surgery: a case-control study. Critical Care 13, R104.

Mattison, P.C., et al., 2012. Role of circulating broblast growth factor-2 inlipopolysaccharide-induced acute kidney injury in mice. Pediatric Nephrology 27,469483.

Munshi, R., et al., 2011. MCP-1 gene activation marks acute kidney injury. Journal of theAmerican Society of Nephrology 22, 165175.

Nakagawa, T., et al., 1999. Role of PDGF B-chain and PDGF receptors in rat tubularregeneration after acute injury. The American Journal of Pathology 155, 16891699.

Nechemia-Arbely, Y., et al., 2008. IL-6/IL-6R axis plays a critical role in acute kidneyinjury. Journal of the American Society of Nephrology 19, 11061115.

Nin, N., et al., 2011. Acute kidney injury in critically ill patientswith 2009 inuenza A (H1N1)viral pneumonia: an observational study. Intensive Care Medicine 37, 768774.

Perez-Padilla, R., et al., 2009. Pneumonia and respiratory failure from swine-origininuenza A (H1N1) in Mexico. The New England Journal of Medicine 361, 680689.

Pettil, V., et al., 2011. Acute kidney injury in patients with inuenza A (H1N1) 2009.Intensive Care Medicine 37, 763767.

Schrijvers, B.F., et al., 2004. The role of vascular endothelial growth factor (VEGF) inrenal pathophysiology. Kidney International 65, 20032017.

Selman, M., et al., 2000. TIMP-1, -2, -3, and -4 in idiopathic pulmonary brosis: aprevailing nondegradative lung microenvironment? American Journal of Physiology.Lung Cellular and Molecular Physiology 279, L562L574.

Veron, D., et al., 2010. Overexpression of VEGF-A in podocytes of adult mice causesglomerular disease. Kidney International 77, 989999.

Villanueva, S., et al., 2006. Ischemic acute renal failure induces the expression of a widerange of nephrogenic proteins. American Journal of Physiology. Regulatory, Integrativeand Comparative Physiology 290, 861870.

WHO, Global Alert and Response (GAR), . Inuenza-like illness in the United States andMexico, April 24, 2009http://www.who.int/csr/don/2009_04_24/en/index.html.

Ziga, J., et al., 2011. Inammatory proles in severe pneumonia associated with thepandemic inuenza A/H1N1 virus isolated inMexico City. Autoimmunity 44, 562570.

492 E. Bautista et al. / Experimental and Molecular Pathology 94 (2013) 486492

Angiogenic and inflammatory markers in acute respiratory distress syndrome andrenal injury associated to A/H1N1 virus infectionIntroductionMaterials and methodsSubjectsA/H1N1 virus detectionMeasurement of angiogenic factors, cytokines, and chemokinesMuscle biopsiesVEGF determination by immunohistochemistryData collection and statistical analysis

ResultsIncreased levels of angiogenic and inflammatory mediators were observed in ARDS/AKI patientsFatal outcome in ARDS patients is associated to increased levels of inflammatory mediators and BMI.

DiscussionConflict of interest statementAcknowledgmentsReferences