Listen to this manuscript’s

audio summary by

JACC Editor-in-Chief

Dr. Valentin Fuster.

J O U R N A L O F T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y V O L . 7 1 , N O . 2 5 , 2 0 1 8

ª 2 0 1 8 B Y T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y F O U N D A T I O N

P U B L I S H E D B Y E L S E V I E R

Bioimpedance-Guided Hydration for thePrevention of Contrast-InducedKidney InjuryThe HYDRA Study

Mauro Maioli, MD,a Anna Toso, MD,a Mario Leoncini, MD,a Nicola Musilli, MD,a Gabriele Grippo, MD,a

Claudio Ronco, MD,b Peter A. McCullough, MD, MPH,c,d,e,f Francesco Bellandi, MDa

ABSTRACT

ISS

Fro

Ins

Me

Va

rel

Ma

BACKGROUND Intravascular volume expansion plays a major role in the prevention of contrast-induced acute kidney

injury (CI-AKI). Recommended standard amounts of fluid infusion before procedures do not produce homogeneous re-

sponses in subjects with different initial hydration status.

OBJECTIVES The goal of this study was to compare the effect of standard and double intravenous (IV) infusion vol-

umes in patients with low body fluid level, assessed by using bioimpedance vector analysis (BIVA), on the incidence of CI-

AKI after elective coronary angiographic procedures.

METHODS A total of 303 patients with low BIVA level on admission were randomized to receive standard volume saline

(1 ml/kg/h for 12 h before and after the procedure) or double volume saline (2 ml/kg/h). Patients (n ¼ 715) with an

optimal BIVA level received standard volume saline and were included in a prospective registry. The saline infusion was

halved in all patients with an ejection fraction <40%. BIVA was repeated immediately before the angiographic procedure

in all patients. CI-AKI was defined as an increase in levels of cystatin C $10% above baseline at 24 h after contrast

administration.

RESULTS The incidence of CI-AKI was significantly lower (11.5% vs. 22.3%; p ¼ 0.015) in patients receiving double

volume saline than in those receiving standard volume saline, respectively. Before the angiographic procedure, 50% of

the double volume patients achieved the optimal BIVA level compared with only 27.7% in the standard group

(p ¼ 0.0001). The findings were consistent in all the pre-specified subgroups excluding patients with a left ventricular

ejection fraction <40% (p for interaction ¼ 0.01).

CONCLUSIONS Evaluation of BIVA levels on admission in patients with stable coronary artery disease allows

adjustment of intravascular volume expansion, resulting in lower CI-AKI occurrence after angiographic procedures.

(Personalized Versus Standard Hydration for Prevention of CI-AKI: A Randomized Trial With Bioimpedance Analysis;

NCT02225431) (J Am Coll Cardiol 2018;71:2880–9) © 2018 by the American College of Cardiology Foundation.

E xpansion of intravascular volume with intra-venous (IV) infusion of fluids before adminis-tration of contrast media is a proven

protective strategy against contrast-induced acutekidney injury (CI-AKI) because it favors renal bloodflow and glomerular filtration and increases urine

N 0735-1097/$36.00

m the aDivision of Cardiology, Santo Stefano Hospital, Prato, Italy; bDepa

titute (IRRIV), S. Bortolo Hospital, Vicenza, Italy; cDepartment of Interna

dical Center, Dallas, Texas; dBaylor Heart and Vascular Institute, Dallas

scular Hospital, Dallas, Texas; and fThe Heart Hospital Baylor Plano, Plano

ationships relevant to the contents of this paper to disclose.

nuscript received January 18, 2018; revised manuscript received March 3

flow rates, thus reducing the concentration and dura-tion of the nephrotoxic agent in the tubule (1–3).International guidelines recommend isotonicsolution infusion for 6 to 12 h before and after admin-istration of contrast medium (4–6). However, fixedamounts of fluid infusion do not affect all patients

https://doi.org/10.1016/j.jacc.2018.04.022

rtment of Nephrology, International Renal Research

l Medicine, Division of Cardiology, Baylor University

, Texas; eBaylor Jack and Jane Hamilton Heart and

, Texas. The authors have reported that they have no

0, 2018, accepted April 3, 2018.

AB BR E V I A T I O N S

AND ACRONYM S

BIVA = bioimpedance vector

analysis

CAD = coronary artery disease

CI-AKI = contrast-induced

acute kidney injury

IV = intravenous

LVEF = left ventricular

ejection fraction

R/H = resistance/height ratio

J A C C V O L . 7 1 , N O . 2 5 , 2 0 1 8 Maioli et al.J U N E 2 6 , 2 0 1 8 : 2 8 8 0 – 9 BIVA-Tailored Hydration Against CI-AKI

2881

in the same way. Optimal fluid administration regi-mens should be based on pre-defined parameters,adjusting volumes according to patient hydration sta-tus and current clinical condition, especially kidneyand heart function (7,8). Studies have shown thattailored peri-procedural infusion of fluids based onindividual hemodynamic parameters (e.g., centralvenous pressure, left ventricular end-diastolic pres-sure) or urinary flow, measured by using invasivetechniques, can reduce the incidence of CI-AKI inhigh-risk patients (9–12).

SEE PAGE 2890

In a previous study, we showed that total bodyfluid level evaluated by using bioimpedance vectoranalysis (BIVA) immediately before administration ofcontrast medium is a valid predictor of CI-AKI risk inpatients undergoing elective coronary angiographicprocedures (angiography and/or angioplasty). In fact,patients with low BIVA status immediately beforesuch procedures who received standard volumes of IVfluids presented with a significantly higher risk ofdeveloping CI-AKI than patients with optimal BIVAstatus (13). We therefore decided to investigatetailoring fluid administration based on individualcurrent BIVA status. Specifically, we evaluated theeffect of double volumes of IV saline infusioncompared with standard volumes in preventing CI-AKI after coronary angiographic procedures in pa-tients with stable coronary artery disease (CAD) whopresent with low total body fluid levels (BIVA evalu-ation) on hospital admission.

METHODS

STUDY POPULATION. The 2,324 consecutive patientsadmitted to our Prato Hospital cardiology division andscheduled for coronary angiographic procedures fromNovember 2013 to April 2016 were screened for inclu-sion in the study. Exclusion criteria were as follows:1) patients requiring urgent (within 6 to 12 h) oremergency (n ¼ 746) procedures; 2) patients admittedto other divisions (n ¼ 238); 3) administration ofcontrast medium in the previous 10 days (n ¼ 107);4) overt congestive heart failure with ascites or pleuro-pericardial effusion (n ¼ 77); 5) patients undergoingdialysis (n ¼ 52); and 6) patient refusal of consent(n¼ 86). Thus, a total of 1,018 patients were enrolled inthis study and underwent on-admission total bodyfluid level evaluation by using BIVA (Figure 1).

Patients were divided into 2 preliminary groupsbased on admission BIVA status: optimal total bodyfluid level and low total body fluid level (13). Patientswith optimal body fluid level (Registry group, n ¼ 715)

received standard volume IV isotonic salineinfusion (0.9% sodium chloride, 1 ml/kg/h for12 h before and after the procedure) (4–8,13).Patients with low body fluid level (n ¼ 303)were randomized for the standard (1 ml/kg/hfor 12 h before and after the procedure) ordouble (2 ml/kg/h) amount of IV saline infu-sion. The rate, and thus total volume, of IVsolution infusion was halved in all patientswith a left ventricular ejection fraction(LVEF) <40% or clinical signs of heart failure(New York Heart Association functional clas-

ses III and IV).

A second BIVA assessment was performed imme-diately before the angiographic procedure (i.e., beforecontrast medium administration) in 1,000 patients ofthe total cohort. They were analyzed for the primaryendpoint of the study: 704 in the Registry group and296 in the randomized group (Figure 1).

LVEF was evaluated by echocardiographic exami-nation performed on admission in all patients. Theangiographic and revascularization procedures wereperformed according to standard protocols, usuallythe day after admission. All patients received thesame nonionic, dimeric iso-osmolar contrast medium(Iodixanol [Visipaque), GE Healthcare Ltd., Amer-sham, United Kingdom) using an automated injectorsystem (ACIST CVi Contrast Delivery System-ACISTMedical Systems, Eden Prairie, Minnesota). Serumcystatin C levels were measured on admission and onday 1 after contrast medium administration; serumcreatinine levels were measured on admission and ondays 1, 2, and 3 after the procedure. After discharge,adverse clinical events were assessed and registeredby the attending physician at the 12-month follow-up.

All clinical, angiographic, and biochemical datawere recorded in a dedicated database. Randomiza-tion was performed after on-admission BIVA evalua-tion by computerized open-label assignment using anelectronic spreadsheet with blocks of 20 patientseach. The protocol was approved by the hospitalethics committee, and all patients provided writteninformed consent.

BIOIMPEDANCE ANALYSIS. BIVA is performed byusing tetrapolar impedance plethysmography (EFGElectroFluidGraph, Akern SRL, Florence, Italy). It isbased on the principle that the body acts as a circuitwith a given resistance (R, opposition of current flowthrough intracellular and extracellular solutions) andreactance (Xc, the capacitance of cells to store energy)(14). The volume of the body fluid component islargely reflected in the resistance, whereas reactancemight represent cell membrane integrity. Resistance

FIGURE 1 Flow Diagram of the Study

Patients referred for angiographicprocedures (n = 2,324)

On admission BIVA evaluation (n = 1,018)

Nonhydrated (n = 7)

Pre-angiographic BIVAevaluation (n = 704)

Standard Hydration(n = 708)

No BIVA control (n = 4)

Pre-angiographic BIVAevaluation (n = 148)

Pre-angiographic BIVAevaluation (n = 148)

Standard i.v. infusionvolume (n = 152)

Double i.v. infusionvolume (n = 151)

No BIVA control (n = 3)No BIVA control (n = 4)

Registry Randomized

Optimal body fluid level(n = 715)

Low body fluid level(n = 303)

1,306 patients excluded for:- Urgent or emergency procedures (n = 746)- BIVA machine unavailability (n = 238)- Administration of contrast medium within the previous 10 days (n = 107)- Overt congestive heart failure (n = 77)- End-stage renal disease requiring dialysis (n = 52)- Patient consent refusal (n = 86)

In the cohort of 1,018 eligible patients who underwent bioimpedance vector analysis (BIVA) evaluation on admission, 715 had optimal body fluid levels (Registry group)

and 303 had low body fluid levels (randomization group). The latter were randomized to receive standard or double intravenous (i.v.) infusion volume. Of these, 296

(148 per group) underwent a second BIVA evaluation, immediately before the angiographic procedure, and were evaluated for the primary endpoint.

Maioli et al. J A C C V O L . 7 1 , N O . 2 5 , 2 0 1 8

BIVA-Tailored Hydration Against CI-AKI J U N E 2 6 , 2 0 1 8 : 2 8 8 0 – 9

2882

and reactance values are measured by using an elec-tric alternating current flux of 800 microampere andan operating frequency of 50 kHz (15). Total bodyfluid volume is calculated based on the resistance/height (R/H) and capacitance/height (Xc/H) ratios(16,17). Further technical details are presented in ourprevious study (13). The earlier study defined optimaltotal body fluid values as an R/H ratio #380 U/m infemale subjects and #315 U/m in male subjects andlow total body fluid levels as an R/H ratio >380 U/m infemale subjects and >315 U/m in male subjects.

ENDPOINTS AND DEFINITIONS. The primaryendpoint was CI-AKI occurrence, defined as an in-crease in serum cystatin C level$10% over the baselinevalue within 24 h after administration of contrast

medium (18). Additional endpoints were: 1) CI-AKIoccurrence, defined by an increase in serum creati-nine level $0.3 mg/dl within 48 h (19) or $0.5 mg/dlwithin 72 h after administration of contrast medium(4); 2) CI-AKI occurrence in pre-specified risk sub-groups: age (<75 and$75 years), sex, diabetesmellitus,anemia, creatinine clearance (<60 and $60 ml/min),LVEF (<40% or $40%), CI-AKI risk score (#11 or >11),contrast medium volume administered (#200 or>200ml), and BIVA status immediately before contrastmedium administration; and 3) adverse cardiovascularand renal events at 12 months, defined as dialysis, all-cause mortality, myocardial infarction, and/or stroke.Creatinine clearance was calculated by using theCockcroft-Gault formula (20).

TABLE 1 Baseline Demographic, Clinical and BIVA Parameters of All Patients Grouped by

Optimal and Low Total Body Fluid Level According to BIVA

Optimal BIVA Level(n ¼ 704 [70.4%])

Low BIVA Level(n ¼ 296 [29.6%]) p Value

Age, yrs 68 � 10 71 � 10 0.001

Age $75 yrs 219 (31.1) 122 (41.2) 0.002

Female 189 (26.8) 94 (31.8) 0.12

Diabetes mellitus 197 (28.0) 73 (24.7) 0.31

Hypertension 470 (66.8) 205 (69.3) 0.46

NYHA functional class II or higher 36 (5.1) 17 (5.7) 0.76

Anemia 78 (11.1) 70 (23.6) 0.0001

Diuretic therapy 106 (15.1) 81 (27.4) 0.0001

PCI 419 (59.5) 176 (59.5) 0.98

LVEF, % 50 � 10 48 � 10 0.001

LVEF <40% 98 (13.9) 53 (18) 0.13

Serum creatinine, mg/dl 1.02 � 0.35 1.01 � 0.41 0.73

Creatinine clearance, ml/min 73 � 27 68 � 25 0.005

Creatinine clearance <60 ml/min 236 (33.5) 124 (41.9) 0.013

Serum cystatin C, mg/ml 0.96 � 0.3 1.01 � 0.4 0.04

Volume of contrast medium, ml 135 � 94 131 � 87 0.52

Volume of contrast medium

<100 ml 315 (44.7) 137 (46.2)

101–200 ml 255 (36.3) 102 (34.5) 0.83

201–300 ml 96 (13.6) 41 (13.9)

>300 ml 38 (5.4) 16 (5.4)

Contrast volume-to-creatinineclearance ratio

2.1 � 1.5 2.2 � 1.7 0.14

Contrast nephropathy risk score 4.7 � 3.2 5.4 � 3.6 0.003

Contrast nephropathy risk score

#5 446 (63.4) 170 (57.4) 0.017

6–10 217 (30.8) 97 (32.8)

11–16 40 (5.7) 27 (9.1)

$17 1 (0.1) 2 (0.7)

On-admission BIVA evaluation:mean resistance/height ratio, U/m

Female subjects 322 � 40 423 � 35 0.0001

Male subjects 266 � 32 350 � 35 0.0001

Values are mean � SD or n (%).

BIVA ¼ bioimpedance vector analysis; LVEF ¼ left ventricular ejection fraction; NYHA ¼ New York HeartAssociation; PCI ¼ percutaneous coronary intervention.

J A C C V O L . 7 1 , N O . 2 5 , 2 0 1 8 Maioli et al.J U N E 2 6 , 2 0 1 8 : 2 8 8 0 – 9 BIVA-Tailored Hydration Against CI-AKI

2883

The risk score for development of CI-AKI wasdefined as specified by Mehran et al. (21). Anemia wasdefined by using World Health Organization criteria:baseline hematocrit value <39% for men and <36%for women (22). The terms “hydration” and “dehy-dration” refer to changes in total body fluid level thataffects only the osmolality of body fluids, (i.e., theserum sodium level). The terms “extracellular fluidvolume expansion” and “contraction” refer tochanges in intravascular volume that are caused bychanges in both salt and water content (23).

STATISTICAL ANALYSIS. We established the samplesize based on CI-AKI incidence in patients with stableCADwith low BIVA-estimated fluid status enrolled in aprevious study with cystatin C data at both baselineand 24 h (13). In that study subgroup (220 of 900 pa-tients), the cumulative incidence of CI-AKI (defined asfor the aforementioned primary endpoint) was 25.2%in patients with low BIVA status, meaning that at least280 patients (140 patients per treatment group) wereneeded to detect a 50% reduction in CI-AKI incidencein the double volume saline group compared with thestandard volume group, with an 80% statistical powerand 2-sided type 1 error of 5%. Our protocol requiredthat each treatment group comprise at least 150 pa-tients to allow for dropouts and/or incomplete data.Categorical variables were summarized as percent-ages; continuous variables are summarized as mean �SD or median with interquartile range.

The association between categorical variables andinfusion volume groups were investigated by usingthe chi-square test or the Fisher exact test. A para-metric unpaired Student’s t-test (with Satterthwaite’scorrection for df when necessary) or the analogousnonparametric test (Wilcoxon-Mann Whitney U test)was applied to evaluate differences for continuousvariables between the standard and double IV infu-sion volume groups. Unconditional logistic analysiswas performed to evaluate the efficacy of doubleinfusion volume on CI-AKI prevention, adjusting forvarious potential prognostic and confounding factors(age, sex, diabetes, anemia, baseline estimatedcreatinine clearance, LVEF, CI-AKI risk score, contrastvolume, and immediately pre-procedural BIVAstatus).

The association between infusion volume strategyand CI-AKI was expressed as odds ratio (OR), and the95% confidence interval (CI) was also reported. Theefficacy of the double infusion volume was evaluatedin the pre-specified risk subgroups. The adjusted ef-fect was controlled for the same prognostic and con-founding factors, but the specific variable relative toeach individual subgroup was excluded from the

model for that subgroup. A p value <0.05 wasconsidered significant (2-sided). The Kaplan-Meieranalysis was used to evaluate 12-month event-freesurvival of the standard and double IV infusion vol-ume group, and the differences between groups wereassessed by using the log-rank test. All the analyseswere conducted by using STATA 12 software (2012,StataCorp, College Station, Texas).

RESULTS

CLINICAL AND PROCEDURAL DATA. Table 1 presentsbaseline clinical, angiographic, and BIVA data of all1,000 patients divided according to on-admissionBIVA status. The 296 patients (29.6%) in the lowBIVA level group were generally older, more anemic,undergoing treatment with diuretic agents, and

TABLE 2 Baseline Demographic, Clinical, and BIVA Parameters of Low BIVA Level

Patients Randomized to Receive Standard or Double IV Infusion Volume

Standard IVInfusion Volume

(1 ml/kg/h) (n ¼ 148)

Double IVInfusion Volume

(2 ml/kg/h) (n ¼ 148)p

Value

Mean age, yrs 71 � 11 70 � 10 0.16

Age $75 yrs 65 (43.9) 57 (38.5) 0.41

Female 52 (35.1) 42 (28.4) 0.26

Diabetes mellitus 39 (26.4) 34 (23.0) 0.60

Hypertension 104 (70.3) 100 (68.2) 0.80

NYHA functional class II or higher 9 (6.1) 8 (5.4) 0.80

Anemia 38 (25.7) 32 (21.6) 0.49

Diuretic therapy 35 (23.6) 46 (31.1) 0.19

PCI 91 (61.5) 85 (57.4) 0.55

Mean LVEF, % 47 � 11 49 � 10 0.30

LVEF <40% 32 (21.6) 21 (14.2) 0.09

Serum creatinine, mg/dl 1.03 � 0.5 0.98 � 0.3 0.21

Mean creatinine clearance, ml/min 67 � 25 69 � 24 0.25

Creatinine clearance <60 ml/min 63 (42.6) 61 (41.2) 0.90

Mean serum cystatin C, mg/ml 1.02 � 0.4 0.99 � 0.4 0.40

Mean volume of contrast medium, ml 131 � 85 131 � 90 0.97

Volume of contrast medium 0.69

<100 ml 65 (43.9) 72 (48.6)

101–200 ml 56 (37.8) 46 (31.1)

201–300 ml 18 (12.3) 23 (15.5)

>300 ml 9 (6.1) 7 (4.8)

Contrast volume-to-creatinineclearance ratio

2.33 � 1.8 2.12 � 1.5 0.25

Mean contrast nephropathy risk score 5.6 � 3.8 5.3 � 3.5 0.35

Contrast nephropathy risk score 0.55

#5 82 (55.3) 88 (59.5)

6–10 51 (34.5) 46 (31.1)

11–16 14 (9.5) 13 (8.8)

$17 1 (0.7) 1 (0.7)

On-admission BIVA evaluation:mean resistance/height ratio, U/m

Female subjects 425 � 35 420 � 30 0.50

Male subjects 348 � 39 352 � 42 0.33

Pre-angiographic BIVA evaluation:mean resistance/heightratio, U/m

Female subjects 420 � 61* 391 � 60‡ 0.02

Male subjects 333 � 51† 316 � 53§ 0.03

Mean total hydration volume, ml

All patients 1,476 (961–1,680) 3,216 (2,522–3,600) 0.0001

Patients with LVEF $40% 1,560 (1,422–1,733) 3,320 (2,976–3,648) 0.0001

Patients with LVEF <40% 876 (732–1,080) 2,082 (1,656–3,240) 0.0001

Values are mean � SD, n (%), or median (interquartile range). Pre-angiographic vs. on-admission BIVA values:*p ¼ 0.44. †p ¼ 0.001. ‡p ¼ 0.001. §p ¼ 0.0001.

IV ¼ intravenous; other abbreviations as in Table 1.

TABLE 3 Incidence of CI-AKI in the 2 Randomized Low BIVA

Level Groups

Standard IVInfusionVolume

(1 ml/kg/h)(n ¼ 148)

Double IVInfusionVolume

(2 ml/kg/h)(n ¼ 148)

pValue

Primary endpoint

Cystatin C $10% within 24 h 33 (22.3) 17 (11.5) 0.013

Additional endpoints

Serum creatinine level $0.3mg/dl within 48 h

16 (10.8) 7 (4.7) 0.08

Serum creatinine level $0.5mg/dl within 72 h

11 (7.4) 5 (3.4) 0.19

Value are n (%). Statistical analysis was performed by using the chi-square orFisher exact test.

CI-AKI ¼ contrast-induced acute kidney injury; other abbreviations as inTables 1 and 2.

Maioli et al. J A C C V O L . 7 1 , N O . 2 5 , 2 0 1 8

BIVA-Tailored Hydration Against CI-AKI J U N E 2 6 , 2 0 1 8 : 2 8 8 0 – 9

2884

presented with worse renal and cardiac function andsignificantly higher Mehran CI-AKI risk scores thanthe optimal BIVA level group.

Randomized assignment of the low BIVA level pa-tients produced 2 statistically homogeneous groups,one scheduled to receive standard IV infusion volumeand the other to receive double IV infusion volume(Table 2).

CI-AKI INCIDENCE IN THE RANDOMIZATION

GROUPS. The overall incidence of CI-AKI was 16.9%in the 2 randomized low BIVA level groups: 22.3% inpatients who received standard IV infusion volume(33 of 148) and 11.5% in patients receiving the doubleinfusion volume (17 of 148) with a crude OR of 0.45(95% CI: 0.24 to 0.85; p ¼ 0.015) (Table 3). Whenadjusted for confounding factors, this OR remainedsignificant (adjusted OR: 0.48; 95% CI: 0.24 to 0.94;p ¼ 0.033). Double IV infusion volume was associatedwith a consistent reduction in the incidence of CI-AKIalso with the nonprimary endpoint CI-AKI criteria. Apositive impact of the double IV infusion volumewas evident in all pre-specified subgroups exceptpatients with reduced baseline LVEF (<40%) (p forinteraction ¼ 0.01) (Figure 2).

PRE-ANGIOGRAPHIC BIVA STATUS AND CI-AKI IN

THE RANDOMIZATION GROUPS. BIVA evaluationimmediately before angiography (after the first 12 h ofIV saline infusion) showed that patients receivingdouble IV infusion volumes presented with signifi-cantly higher total body fluid levels (lower BIVA R/Hratio) than patients receiving standard volumes(Table 2). From baseline to pre-angiographic BIVAevaluation, the total body fluid levels increased inboth groups, as shown by the decrease in the R/Hratio, which was greater in the double IV group(p values are specified in the legend of Table 2).Moreover, patients receiving double IV infusionvolume were re-classified as optimally hydrated in50% of cases compared with only 27.7% of patientsassigned to receive the standard IV infusion volume(p < 0.0001) (Figure 3).

The incidence of CI-AKI was highest in patientsassigned to the standard IV infusion volume who

FIGURE 2 Occurrence of CI-AKI in the Pre-Specified Risk Subgroups

0.01 0.1Double Hydration

Volume BetterStandard Hydration

Volume Better

1 10

0.45

0.18

0.24

0.01

0.15

0.92

0.41

0.68

0.57

0.59 (0.24-1.43), p = 0.240.31 (0.11-0.95), p = 0.04

0.58 (0.26-1.29), p = 0.210.18 (0.03-0.94), p = 0.043

0.86 (0.22-3.38), p = 0.830.38 (0.17-0.85), p = 0.019

0.28 (0.12-0.64), p = 0.0031.61 (0.22-11.5), p = 0.63

1.11 (0.25-4.97), p = 0.890.33 (0.14-0.75), p = 0.009

0.45 (0.16-1.24), p = 0.120.44 (0.16-1.15), p = 0.094

0.74 (0.19-2.87), p = 0.660.39 (0.17-0.91), p = 0.028

0.44 (0.20-0.97), p = 0.0430.53 (0.12-2.28), p = 0.39

0.43 (0.20-0.91), p = 0.0270.62 (0.06-6.25), p = 0.69

OR adj (95% CI)

16/8317/65

21/9612/52

8/3925/109

28/1165/32

8/3825/110

16/8517/63

5/4128/107

25/1218/27

27/1336/15

Eventsp for Interaction

11/91

Standard HydrationDouble Hydration

6/57

14/1063/42

6/3411/114

11/1276/21

6/3211/116

8/879/61

7/7410/74

11/1186/30

13/1344/14

Age (yrs)

Gender

Diabetes

Baseline EF (%)

Anemia

Creatinine Clearance (ml/min)

Pre-Procedural BIVA

Contrast Volume (ml)

CI-AKI Risk Score

<75≥75

MalesFemales

YesNo

≥40<40

YesNo

≥60<60

OptimalLow

≤200>200

<11≥11

The graph shows the incidence of CI-AKI in the pre-specified risk subgroups in relation to double versus standard i.v. infusion volume. Estimates are odds ratios (ORs) with

95% confidence intervals (CIs). The unconditional logistic model was adjusted for age, sex, diabetes, anemia, baseline estimated creatinine clearance, left ventricular

ejection fraction (EF), CI-AKI (contrast-induced acute kidney injury) risk score, contrast volume, and immediately pre-procedural BIVA (bioimpedance vector analysis)

status. Abbreviations as in Figure 1.

FIGURE 3 Pre-Angiographic BIVA Status (Low/Optimal) Distribution in Double and

Standard Volume Groups After 12 h of IV Saline Infusion

100%

80

60

40

20

72.3

27.7

50

Standard i.v. Infusion Volume

p < 0.0001*

Double i.v. Infusion Volume

50

0

Low-Fluid Status Optimal-Fluid Status

The percentage of low-hydrated patients randomized to treatment who achieved an

optimal BIVA status measured immediately before the angiographic procedure was

significantly higher in the double i.v. infusion volume group than in the standard group

(50% vs 27.7%; p < 0.0001). *Chi-square analysis. Abbreviations as in Figure 1.

J A C C V O L . 7 1 , N O . 2 5 , 2 0 1 8 Maioli et al.J U N E 2 6 , 2 0 1 8 : 2 8 8 0 – 9 BIVA-Tailored Hydration Against CI-AKI

2885

continued to have a low BIVA level immediatelybefore the angiographic procedure (Figure 4).

CLINICAL EVENTS. Protocols for IV saline adminis-tration, both standard and double volume, werecompleted in all patients without adverse clinicalevents. In particular, no patients developed signs ofcongestive heart failure or pulmonary edema duringsaline infusion. Follow-up at 12 months wascompleted for all patients.

Table 4 presents the adverse cardiovascularand renal events at 12 months. No significantbetween-group differences were found in either sin-gle or cumulative event rates. Kaplan-Meier analysisshowed no between-group differences in 12-monthevent-free survival (p ¼ 0.624).

CI-AKI OCCURRENCE IN THE REGISTRY GROUP.

Only 9.4% (66 of 704) of the patients with optimalon-admission BIVA levels (not included in therandomization group; standard IV infusion volume)developed CI-AKI as defined by using the primaryendpoint criteria. CI-AKI occurrence was 8.4% (51 of606) in patients with LVEF $40% and 15.3% (15 of 98)in those with LVEF <40% (p ¼ 0.047).

FIGURE 4 CI-AKI Occurrence and Pre-Angiographic Body Fluid Status (Low/Optimal)

in Double and Standard Volume Groups

%

CI-A

KI In

cide

nce

Standard i.v. Infusion Volume

26.2

12.2

p = 0.07

p = 0.04

p = 0.65

p = 0.61

13.5

9.5

Double i.v. Infusion Volume

30

20

10

0

Low-Fluid Status Optimal-Fluid Status

The occurrence of CI-AKI was lower in patients assigned to receive double i.v. infusion

volume who achieved the optimal BIVA level measured immediately before the angio-

graphic procedure after 12 h of saline infusion. Abbreviations as in Figures 1 and 2.

TABLE 4 Adverse Cl

Dialysis

Myocardial infarction

Stroke

All-cause deaths

Cumulative adverse ev

Values are n (%). *The Fish

Maioli et al. J A C C V O L . 7 1 , N O . 2 5 , 2 0 1 8

BIVA-Tailored Hydration Against CI-AKI J U N E 2 6 , 2 0 1 8 : 2 8 8 0 – 9

2886

DISCUSSION

The present study shows that in patients with stableCAD with low on-admission BIVA-estimated fluid sta-tus (insufficiently hydrated patients), a regimen ofdouble IV infusion volume is more efficacious thanstandard infusion volume in reducing the incidence ofCI-AKI (Central Illustration). This effect was consistentin all the pre-specified high-risk subgroups exceptpatients with a baseline LVEF <40%. Double IV infu-sion volume led to optimal pre-angiographic hydrationlevel (second BIVA evaluation) in a significantly higherpercentage of patients than standard volume infusion.

CI-AKI is a common complication of diagnostic andtherapeutic procedures using iodinated contrast me-dia. This condition may be either not recognized orunderestimated because it often follows a subclinicalcourse (7,19). However, it is well known that evenslight, transient contrast-induced renal damage can

inical Events at 12 Months

Standard IntravenousInfusion Volume

(1 ml/kg/h) (n ¼ 148)

Double IntravenousInfusion Volume

(2 ml/kg/h) (n ¼ 148) p Value*

2 (1.4) 1 (0.7) 0.98

1 (0.7) 4 (2.7) 0.37

– 1 (0.7) –

8 (5.5) 3 (2.1) 0.21

ents 11 (3.6) 9 (6.2) 0.82

er exact test was applied apart from for overall adverse events (chi-square test).

have a negative impact on short- and long-term car-diovascular and renal prognoses (24–26). Thus,application of preventive measures is mandatory,especially in high-risk patients (e.g., those with dia-betes or chronic kidney disease) (6). Periproceduralintravascular volume expansion is a staple of CI-AKIprevention, and short-term administration of IVfluids exercises an important protective actionagainst both intrarenal hemodynamic alterations anddirect renal tubular toxic effects caused by contrastmedium (1,2,7). Fluid administration, particularlywhen it corrects volume depletion and expands IVvolume, can be expected to increase the clearance ofcontrast medium, decrease the concentration ofcontrast medium in the tubule lumen and vasa recta,and counteract activation of neurohormonal systemsthat lead to medullary vasoconstriction (3).

Current international guidelines recommendintravascular volume expansion with IV administra-tion of normal saline (1.0 to 1.5 ml/kg/h) for at least6 h before and after injection of contrast medium (4).Protocols specifying standard infusion volumes areeasy to apply but do not guarantee adequate volumeexpansion for many high-risk patients. Various indi-vidual, clinical, and therapeutic factors influenceintravascular volume status and response to volemicload (7,27,28). Studies using tailored fluid protocolsbased on hemodynamic parameters (central venouspressure or left ventricular end-diastolic pressure) orurinary flow (requiring invasive techniques) showthat an individualized approach reduces the inci-dence of CI-AKI in high-risk patients (9–12).

BIVA allows a rapid, accurate, and noninvasivedetermination of body hydration status (total bodywater); it is currently used in clinical settings to guidefluid-related therapies in critically ill patients (un-dergoing dialysis or with heart failure) (29–31). Astudy using BIVA evaluation in patients subjected toangiographic procedures found that low BIVA hy-dration level was associated with a 3 times higher riskof CI-AKI development than the optimal BIVA hy-dration level (13).

The present study found that w30% of patientswith stable CAD scheduled for angiographic proced-ures and submitted to BIVA measurement are insuf-ficiently hydrated at hospital admission. Thiscondition signals high risk for CI-AKI. It is signifi-cantly more common in patients who are already athigh risk for CI-AKI, in particular the elderly, patientswith low LVEF, and/or low creatinine clearance,anemia, and undergoing treatment with diureticagents (4,8). Moreover, more than two thirds of ouron-admission insufficiently hydrated patientsremained insufficiently hydrated immediately before

CENTRAL ILLUSTRATION BIVA-Tailored Hydration Against Contrast-Induced Acute Kidney Injury (CI-AKI)

Maioli, M. et al. J Am Coll Cardiol. 2018;71(25):2880–9.

The graphic shows how bioimpedance vector analysis (BIVA) measures are obtained and the results for the randomization group (patients with low total body fluid

levels). The graphic on the left illustrates a BIVA instrument that measures total body fluid level based on the resistance/height (R/H) ratio, which differs in men and

women. The electrodes are applied to the hand and foot on the same side of the body. The randomization group (low total body fluid levels on admission: R/H ratio

>315 U/m in male subjects and >380 U/m in female subjects) was divided into 2 treatment arms: double and standard intravenous volume infusion. Mean R/H ratios

for both sexes on admission and immediately before the angiographic procedure are reported below the stylized figures, which also give a visual indication of body

fluid quantities. The CI-AKI incidence in the 2 arms is indicated on the far right.

J A C C V O L . 7 1 , N O . 2 5 , 2 0 1 8 Maioli et al.J U N E 2 6 , 2 0 1 8 : 2 8 8 0 – 9 BIVA-Tailored Hydration Against CI-AKI

2887

angiography during standard hydration protocols,leaving them at high risk for CI-AKI.

A strategy of IV volume expansion with higher-than-standard amounts of saline solution (meanvalues 3,216 ml vs. 1,476 ml) in the same routine timeintervals resulted in a significantly lower incidence ofCI-AKI. The beneficial effect of double IV infusionvolume remained statistically significant even afteradjustment for specific prognostic and confoundingfactors.

It is interesting to note that of the pre-specifiedcategories of risk for CI-AKI, elderly patients are

among those who benefit most from increased infu-sion volumes. Dehydration is very common in theelderly, both those living in institutions and at home.Age-related physiological changes, concomitantmedical disorders, polypharmacy, and social andenvironmental factors render the elderly more sus-ceptible to dehydration and intravascular fluid vol-ume contraction (32). This condition in turn increasesthe risk of development of various diseases, as well asCI-AKI, thus complicating treatment. Adequateintravascular volume expansion is crucial, especiallyin vulnerable elderly patients; it helps prevent

PERSPECTIVES

COMPETENCY IN PATIENT CARE AND

PROCEDURAL SKILLS: Fluid administration guided

by individual current BIVA is more efficacious than

standard fluid infusion in reducing CI-AKI in low-

hydrated patients undergoing elective coronary

angiography.

TRANSLATIONAL OUTLOOK: Further studies are

needed to establish the safety and efficacy of BIVA for

prevention of nephropathy in underhydrated patients

with a history of heart failure or reduced LVEF un-

dergoing contrast angiography.

Maioli et al. J A C C V O L . 7 1 , N O . 2 5 , 2 0 1 8

BIVA-Tailored Hydration Against CI-AKI J U N E 2 6 , 2 0 1 8 : 2 8 8 0 – 9

2888

contrast-induced acute renal injury and hinders theprogressive reduction of glomerular filtration ratetypical in aging. Subgroup analysis based on ageshowed that in our low-hydrated elderly patients,peri-procedural double IV volume administrationresulted in a 15% reduction in CI-AKI. Larger focusedstudies are needed to better define the impact of thisstrategy in the elderly.

Patients with reduced LVEF are another particu-larly high-risk group for CI-AKI also because theyoften do not receive adequate fluids given concernsregarding overhydration and onset of congestiveheart failure. Guidelines generically recommendlowering infusion amounts in these patients (4,7),and our routine preventive strategy calls for IV infu-sion of half volumes for patients with LVEF <40%. Inour patients with low on-admission BIVA (insuffi-ciently hydrated) and LVEF <40%, this double vol-ume strategy was well tolerated; however, it did notresult in significant reductions in CI-AKI occurrence.The small number of patients with LVEF <40%enrolled in the present study and the exclusion ofpatients with overt signs of congestive heart failuremake it impossible to draw any conclusions regardingadequate IV infusion volumes in this patient cate-gory. In particular, it is difficult to tell if the inefficacydepends on the amounts administered (too much ortoo little) or the particular mechanisms involved inthe acute renal damage in these patients who mayrequire other specific preventive measures.

STUDY LIMITATIONS. This single-center, open-labelstudy had a relatively small sample size. Too-smallsample size means further limitations in statisticalsignificance of subgroup data. Another limitation isthat BIVA evaluates overall body fluid volume, whichmay include eventual compartmentalized fluids,although our patients were screened to exclude caseswith clinically overt fluid retention (pleural, pericar-dial, or peritoneal effusion) and/or congestive heartfailure. Moreover, BIVA measurements were per-formed only at admission and after 12 h (immediatelybefore contrast medium injection) of the IV infusionof saline solution; they were not repeated upon

completion of the IV infusion. Finally, incorrect bodyposition or placement of electrodes can distort theresults.

CONCLUSIONS

BIVA is easy to perform, noninvasive, and yields im-mediate results. It allows immediate assessment andcontinuous monitoring of hydration status in patientsscheduled for angiographic procedures even duringIV fluid administration, particularly important inpatients at high risk for developing CI-AKI. On-ad-mission low BIVA status indicates patients who arethe best candidates to benefit from double IV volumefusion in the prevention of post-angiographic CI-AKI.This method is another step forward in the field ofprecision medicine.

ACKNOWLEDGMENTS The authors thank NiccolòRisaliti and the hemodynamic team nurses, Divisionof Cardiology, S. Stefano Hospital, Prato, Italy, for thecareful monitoring and recording of data.

ADDRESS FOR CORRESPONDENCE: Dr. Anna Toso,Division of Cardiology, Santo Stefano Hospital, ViaSuor Niccolina, 20 59100, Prato, Italy. E-mail: anna.toso@libero.it. Twitter: @McCulloughBHVH,@bswhealth, @bcmhouston.

RE F E RENCE S

1. Weisbord SD, Palevsky PM. Prevention ofcontrast-induced nephropathy with volumeexpansion. Clin J Am Soc Nephrol 2008;3:273–80.

2. Rojkovskiy I, Solomon R. Intravenous and oralhydration. Approaches, principles, and differingregimens. Intervent Cardiol Clin 2014;3:393–404.

3. Persson PB, Hansell P, Liss P. Pathophysiologyof contrast medium-induced nephropathy. KidneyInt 2005;68:14–22.

4. Stacul F, van der Molen AJ, Reimer P, et al.Contrast induced nephropathy: updated ESURContrast Media Safety Committee guidelines. EurRadiol 2011;21:2527–41.

5. KDIGO Clinical Practice Guideline for acuteKidney Injury. Section 4: contrast induced AKI.Kidney Int 2012;Suppl 2:80–6.

6. Windecker S, Kolh P, Alfonso F, et al. Guidelineon myocardial revascularization. The Task Force onMyocardial Revascularization of the European So-ciety of Cardiology (ESC) and the European

J A C C V O L . 7 1 , N O . 2 5 , 2 0 1 8 Maioli et al.J U N E 2 6 , 2 0 1 8 : 2 8 8 0 – 9 BIVA-Tailored Hydration Against CI-AKI

2889

Association for Cardio-Thoracic Surgery (EACTS).Section 11.2. Eur Heart J 2014;35:2541–619.

7. McCullough PA, Choi JP, Feghali GA, et al.Contrast-induced acute kidney injury. J Am CollCardiol 2016;68:1465–73.

8. Fähling M, Seeliger E, Patzak A, Persson PB. Un-derstanding and preventing contrast-induced acutekidney injury. Nat Rev Nephrol 2017;13:169–80.

9. Qian G, Fu Z, Guo J, Cao F, Chen Y. Preventionof contrast-induced nephropathy by centralvenous pressure-guided fluid administration inchronic kidney disease and congestive heart failurepatients. J Am Coll Cardiol Intv 2016;9:89–96.

10. Brar SS, Aharonian V, Mansukhani P, et al.Haemodynamic-guided fluid administration for theprevention of contrast-induced acute kidneyinjury: the POSEIDON randomised controlled trial.Lancet 2014;383:1814–23.

11. Briguori C, Visconti G, Focaccio A, et al. RenalInsufficiency After Contrast Media AdministrationTrial II (REMEDIAL II): RenalGuard System in high-risk patients for contrast-induced acute kidneyinjury. Circulation 2011;124:1260–9.

12. Marenzi G, Ferrari C, Marana I, et al. Preventionof contrast nephropathy by furosemide withmatched hydration: the MYTHOS (Induced DiuresisWith Matched Hydration Compared to StandardHydration for Contrast Induced Nephropathy Pre-vention) trial. J Am Coll Cardiol Intv 2012;5:90–7.

13. Maioli M, Toso A, Leoncini M, et al. Pre-pro-cedural bioimpedance vectorial analysis of fluidstatus and prediction of contrast-induced acutekidney injury. J Am Coll Cardiol 2014;63:1387–94.

14. Kushner RF. Bioelectrical impedance analysis:a review of principles and applications. J Am CollNutrition 1992;11:199–209.

15. Bioelectrical impedance analysis in bodycomposition measurement. Proceedings of aNational Institutes of Health Technology

Assessment Conference. Bethesda, Maryland,December 12–14, 1994. Am J Clin Nutr 1996;64Suppl:387S–532S.

16. Piccoli A, Rossi B, Pillon L, Bucciante G. A newmethod for monitoring body fluid variation bybioimpedance analysis: the RXc graph. Kidney Int1994;46:534–9.

17. Piccoli A, Nigrelli S, Caberlotto A, et al. Bivar-iate normal values of the bioelectrical impedancevector in adult and elderly populations. Am J ClinNutr 1995;61:269–70.

18. Briguori C, Visconti G, Rivera NV, et al. CystatinC and contrast-induced acute kidney injury. Cir-culation 2010;121:2117–22.

19. Mehta RL, Kellum JA, Shah SV, et al. AcuteKidney Injury Network: report of an initiative toimprove outcomes in acute kidney injury. Crit Care2007;11:R31.

20. Cockcroft DW, Gault MH. Prediction of creat-inine clearance from serum creatinine. Nephron1976;16:31–41.

21. Mehran R, Aymong ED, Nikolsky E, et al.A simple risk score for prediction of contrast-induced nephropathy after percutaneouscoronary intervention: development and initialvalidation. J Am Coll Cardiol 2004;44:1393–9.

22. Nutritional Anemias: Report of a WHO Scien-tific Group. Geneva, Switzerland: World HealthOrganisation, 1968.

23. Mange K, Matsuura D, Cizman B, et al. Lan-guage guiding therapy: the case of dehydrationversus volume depletion. Ann Intern Med 1997;127:848–53.

24. Solomon R, Mehran R, Natarajan MK, et al.Contrast-induced nephropathy and long-termadverse events: cause and effect? Clin J Am SocNephrol 2009;4:1162–9.

25. James MT, Ghali WA, Knudtson ML, et al.Associations between acute kidney injury and

cardiovascular and renal outcomes after coronaryangiography. Circulation 2011;123:409–16.

26. Maioli M, Toso A, Leoncini M, Gallopin M,Musilli N, Bellandi F. Persistent renal damage aftercontrast-induced acute kidney injury: incidence,evolution, risk factors, and prognosis. Circulation2012;125:3099–107.

27. Gupta R, Moza A, Cooper CJ. Intravenous hy-dration and contrast-induced acute kidney injury:too much of a good thing? J Am Heart Assoc 2016;5:e003777.

28. McCullough PA, Zhang J, Ronco C. Volumeexpansion and contrast-induced acute kidneyinjury. Lancet 2017;389:1277–8.

29. Basso F, Berdin G, Mason G, et al. Fluid man-agement in the intensive care unit: bioelectricalimpedance vector analysis as a tool to assess hy-dration status and optimal fluid balance in criti-cally ill patients. Blood Purif 2013;36:192–9.

30. Valle R, Aspromonte N, Milani L, et al. Opti-mizing fluid management in patients with acutedecompensated heart failure (ADHF): theemerging role of combined measurement of bodyhydration status and brain natriuretic peptide(BNP) levels. Heart Fail Rev 2011;16:519–29.

31. Di Somma S, Lalle I, Magrini L, et al. Additivediagnostic and prognostic value of bioelectricalimpedance vector analysis (BIVA) to brain natri-uretic peptide ’grey-zone’ in patients with acuteheart failure in the emergency department. EurHeart J Acute Cardiovasc Care 2014;3:167–75.

32. Schols JM, De Groot CP, van der Cammen TJ,Olde Rikkert MG. Preventing and treating dehy-dration in the elderly during periods of illness andwarmweather. J Nutr Health Aging 2009;13:150–7.

KEY WORDS angiography, bioimpedancevector analysis, contrast-induced acutekidney injury