The Nephro-Protective Trial

(1)

Online-Only Supplement:

The effect of intravenous amino acid supplementation on mortality in ICU patients may be dependent on kidney function:

Post hoc subgroup analyses of a multi-centre randomized trial

Gordon S. Doig and Ran Zhu (

^朱然

)

for the Nephro-Protective Trial Investigators Group.

The Nephro-Protective Trial NHMRC Project Grant Number 632614

Australian and New Zealand Clinical Trials Registry Number

Corresponding Author:

Dr. Gordon S. Doig,

Royal North Shore Hospital, Intensive Care Unit,

St. Leonards, NSW Australia 2065

www.EvidenceBased.net/NephroProtect

24 Jan 2018

(2)

© 2018 Gordon S Doig, University of Sydney. All rights reserved. This publication is protected by

copyright. No part of it may be reproduced for commercial purposes or distributed electronically

without prior written permission of the publisher. Reproduction for personal or educational use is

acceptable.

(3)

eTable 7. Patient demographics and baseline balance in the subgroup of patients with baseline kidney dysfunction or risk of progression of AKI ... 27 eTable 8. Mortality, Length of Stay and Health-Related Quality of Life, subgroup of patients with baseline kidney dysfunction and/or risk of progression of AKI ... 29 eTable 9. Secondary outcomes, adjusted for time at risk (ICU stay) in the subgroup of patients with baseline kidney dysfunction or risk of progression of AKI. ... 30 Subgroup 3: Patients with no baseline kidney dysfunction and no risk of progression of AKI ... 32 eTable 10. Additional baseline balance in the subgroup of patients with no baseline kidney dysfunction and no risk of progression of AKI ... 32 eFigure 6. Study process measures in the subgroup of patients with no baseline kidney dysfunction and no risk of progression of AKI. ... 34 eAppendix 1. https://research.EvidenceBased.net/nephroCALC Screen Capture... 35 Note: Additional information can be found at the study web site:

(5)

THE NEPHRO-PROTECTIVE TRIAL INVESTIGATORS

Management Committee: Gordon Doig (Chair), Fiona Simpson, Elizabeth Sweetman, Phillipa

Heighes, Rinaldo Bellomo, Carol Pollock, Douglas Chesher, Michael Reade, Andrew Davies, Peter Harrigan and John Botha. Protein Dosing Sub-Committee: Gordon Doig (Chair), Rinaldo Bellomo, Fiona Simpson. Data Quality and Management: Jennifer L. Hannam (Northern Clinical School Intensive Care Research Unit, University of Sydney, Australia). Statistical analysis: Gordon S. Doig.

Independent Data and Safety Monitoring Committee: Christopher Doig (Chair), University of

Calgary, Canada.

Participating Sites: (Auckland City) Shay McGuiness, Rachael Parke, Eileen Gilder, Lianne

McCarthy, Jodi Brown, Anna Whitley, (Austin) Rinaldo Bellomo, Glenn Eastwood, Leah Peck, Helen Young, (Blacktown) Kalpesh Gandhi, Treena Sara, Kiran Nand, (Frankston) John Botha, Sachin Gupta, Sharon Allsop, David Lewis, (Geelong) Neil Orford, Claire Cattigan, Allison Bone, Tania Elderkin, Tania Salerno, Melissa Fraser, (Gosford) Rob Cameron, Sheridan Hatter, (John Hunter

Hospital) Peter Harrigan, Miranda Hardie, Emma Pollock, Paul Carless, (Mater Newcastle) Katrina

Ellem, Alan Rashid, Irene Bailey, (Middlemore) Tony Williams, Judy Tai, Anna Tilsley, Rima Song, Laura Rust, Chantal Hogan, Marilyn Beggs, (Nepean) Ian Seppelt, Leonie Weisbrodt, Maria Nikas, Rebecca Gresham, Phoebe Palejs, Anne Ritchie, Sarah Whereat, (Northern) Graeme Duke, Tanya Gilliver, Amy Sutherland, Emily Dynon, Michael Reade, (RNSH) Liz Hickson, Gwen Hickey, Kirilee Matters, Deirdre Mathai, (St George) Theresa Jacques, Deborah Inskip, Rebecca Sidoli, (St Vincents Melbourne) John Santamaria, Jenny Holmes, Roger Smith, Espedito Farone, (St Vincents Sydney) Priya Nair, Claire Reynolds, Serena Knowles, Karen Storer, (Wollongong) Michael Davis, Martin Sterba, Bronwyn Johnson, Wenli Geng.

Visiting Research Fellow: Ran Zhu (^朱然

),

1^st Affiliated Hospital of China Medical University, Shenyang, Liaoning, PRC.

(6)

Complete eligibility criteria:

To be applied within the first two days of study ICU admission.

Inclusion Criteria

Patients will be considered eligible for the trial if all of the following inclusion criteria are met at the time of screening:

See next page for Exclusion Criteria.

1) Is the patient currently on their First or Second calendar day of admission to the study ICU?

2) Is the patient expected to remain in the study ICU today and tomorrow?

3) Does the patient have a working central venous access line through which the study intervention could be delivered?

4) Is the patient able to tolerate at least 1L of fluid volume per day?

[The study intervention may involve the delivery of up to one litre of amino acids per day. This volume of amino acids can replace equivalent volumes from other fluids being received.]

5) Is the patient 18 years of age or older?

(7)

Exclusion Criteria

Patients will be considered ineligible for the trial if any of the following exclusion criteria are met at the time of screening: (Answer NO to all questions)

Note: The Exclusion criteria are continued over two pages:

1) Is the patient currently receiving an NSAID with specific emphasis on COX-2 inhibitors?

['Currently receiving' means the patient will continue to receive the NSAID or COX-2 inhibitor during their ICU stay. If the patient does NOT plan to continue receiving one of these drugs daily, they may be enrolled:

the COX-2 inhibitors Celecoxib, Paracoxib, Etoricoxib or > 160mg Aspirin or any dose of Ibuprofen, Naproxen, Piroxicam, Indomethacin, Meloxicam, Mefenamic Acid, Diclofenac, Prioxicam, Ketoprofen, Tiaprofenic Acid, Ketorolac.]

2) Is the patient currently enrolled into a clinical trial evaluating a nitric oxide (NO) inhibitor?

3) Is the patient currently receiving Acetazolamide, which is a diuretic that works on the proximal tubule?

4) Is the patient's current serum creatinine greater than the allowable age and gender adjusted maximum reported in the Maximum Creatinine Table? (see page 19)

5) Does the patient have severe Acute Kidney Injury, defined as:

Current serum creatinine (SCr) increased 3 times pre-acute illness value OR SCr >350 µmol/L with recent increase greater than 44 µmol/L?

[Note: If pre-acute illness creatinine values are unknown, assume upper limit of normal: 90 µmol/L for females and 110 µmol/L for males.]

6) Is the patient currently receiving or scheduled for dialysis / renal replacement therapy?

7) Has the patient ever had a kidney transplant?

8) Is the patient expected to receive palliative care only and is not expected to survive ICU or hospital discharge?

9) Is the patient moribund and not expected to survive 24 hours?

10) Is the patient brain dead or suspected to be brain dead?

Continued on next page...

(8)

Exclusion Criteria continued

Patients will be considered ineligible for the trial if any of the following exclusion criteria are met at the time of screening: (Answer NO to all questions)

11) If the patient has been admitted to the study ICU directly from another ICU, is the total number of calendar days from other ICU admission until today greater than two calendar days?

12) Does the patient require treatment of a burn injury to greater than 20% of total body surface area?

13) Has the patient been taking Nardil (phenelzine) within the last 6 weeks?

[Nardil (phenelzine) is an uncommonly used anti-depressant. It may be used in people who are resistant to other treatment for major depression or anxiety disorders. Because of its unique side-effect profile, and explicit drug interactions (meperidine, epinephrine, norepinephrine), a history of Nardil (phenelzine) use is usually found clearly documented in the patient's charts.]

14) Has the patient previously been enrolled and randomised into this study?

15) Does the patient have a documented contraindication to the study intervention (IV amino acids), as listed on the TGA product licensing document?

Australian Therapeutic Goods Association Product Licensing Contraindications, Synthamin 17 EF:

(Answer NO to all questions)

15A. Is the patient known to be pregnant or currently breastfeeding?

15B. Does the patient have severe liver disease (Biopsy proven cirrhosis, or documented portal hypertension with a known past history of either upper GI bleeding attributed to portal hypertension or of hepatic failure leading to encephalopathy / coma)?

15C. Does the patient have a documented hypersensitivity (known allergy) to one or more of the included amino acids?

[Known allergies will be clearly documented in the Patient's Charts. See Product Details for complete list of Ingredients.]

15D. Does the patient have a documented inborn error of amino acid metabolism?

[Ex. Phenylketonuria (PKU), Maple Syrup Urine Disease. These disorders are actively screened for during childhood. If present, they will be clearly documented in the patient's charts and will likely be accompanied by strict directions to consult a Dietitian prior to providing nutritional support. Other less common inborn errors include: atypical phenylketonuria, hereditary tyrosinaemia, biotinidase deficiency, methylmalonic aciduria, glutaric acidemia, methylglutaconic acidemia.]

<end exclusion criteria>

(9)

Maximum Creatinine Table

Age and Gender adjusted maximum allowable creatinine for enrolment.

See Exclusion Criteria 4.

Females

Age (years) 18 19

serum Creatinine (µmol/L)* 284 282

Age (years) 20 21 22 23 24 25 26 27 28 29

serum Creatinine (µmol/L) 279 277 274 272 270 268 266 265 263 261

Age (years) 30 31 32 33 34 35 36 37 38 39

Age (years) 40 41 42 43 44 45 46 47 48 49

Age (years) 50 51 52 53 54 55 56 57 58 59

Age (years) 60 61 62 63 64 65 66 67 68 69

Age (years) 70 71 72 73 74 75 76 77 78 79

Age (years) 80 81 82 83 84 85 86 87 88 89

Age (years) 90 91 92 93 94 95 96 97 98 99

Age (years) 100 101 102 103 104 105 106 107 108 >109

serum Creatinine (µmol/L) 210 210 209 209 209 208 208 208 207 203 Males

Age (years) 18 19

serum Creatinine (µmol/L) 368 365

Age (years) 20 21 22 23 24 25 26 27 28 29

Age (years) 30 31 32 33 34 35 36 37 38 39

Age (years) 40 41 42 43 44 45 46 47 48 49

Age (years) 50 51 52 53 54 55 56 57 58 59

Age (years) 60 61 62 63 64 65 66 67 68 69

Age (years) 70 71 72 73 74 75 76 77 78 79

Age (years) 80 81 82 83 84 85 86 87 88 89

Age (years) 90 91 92 93 94 95 96 97 98 99

Age (years) 100 101 102 103 104 105 106 107 108 >109

*To convert μmol/L to mg/dL multiply by 0.0113

(10)

Study intervention: Dosing algorithm for supplementary amino acids

If randomised to the intervention arm, the patient received a continuous infusion of a standard mixture of L-amino acids (Synthamin 17 Electrolyte Free, Baxter Healthcare, Australia) delivered at a rate to achieve a total daily protein intake of approximately 2.0 g/kg/day.

The initial infusion was begun at approximately 100 g/day. If the patient was receiving any form of enteral or parenteral nutritional support, the infusion rate of the study L-amino acid intervention (study Synthamin 17 EF) was reduced such that the total protein intake from all sources (nutrition and study intervention) was approximately 2.0g/kg/day.

The study intervention was discontinued at discharge from the study ICU, or death, or when the patient’s central venous catheter was removed.

Study nephroCALC web tool and Synthamin 17 EF infusion protocol

(https://Research.EvidenceBased.net/nephroCALC, see eAppendix 1 for screen capture)

The nephroCALC study web tool calculated study Synthamin 17 EF infusion rates based on a patient’s current protein intake from Enteral and Parenteral nutrition and the patient’s weight. Protein intake calculations for overweight (BMI > 25 kg/m²) patients were based on their ideal body weight (ideal BMI set at 23 kg/m²).

The majority of patients commenced study Synthamin 17 EF at an infusion rate of 42 ml/h, which provided 100g amino acids per day.

The protocol reduced the Synthamin 17 EF infusion to a lower rate only if total protein from EN, PN and study Synthamin 17 EF reached 2.5g/kg.

If total protein from EN, PN and study Synthamin 17 EF reached 2.5g/kg, the Protocol reduced the patient’s study Synthamin 17 EF infusion rate such that a total protein intake of 2.0g/kg from EN, PN and study Synthamin 17 EF was achieved.

The nephroCALC study web tool was used to conduct all calculations. Results were printed or written on to a blank form to create a permanent record of calculations. It was the site investigator's responsibility to ensure that the appropriate study Synthamin 17 EF infusion rates, as calculated by the nephroCALC web tool, were effectively communicated to, and achieved by, the bedside health care team.

If a patient switched to a different brand of EN or a different type of PN, or if EN or PN was started or discontinued, the nephroCALC study web tool was used to calculate a new study Synthamin 17 EF infusion rate.

The nephroCALC study web tool was not password protected. It could be accessed by the bedside nurse on the weekend / at night if required.

(11)

eFigure 1: CONSORT 2010 Patient Recruitment Flow Diagram, entire study population and baseline kidney function

Recorded as enrolled on study web site (n = 474)

Allocated to Standard Care (n = 235) Allocated to amino acid supplement (n = 239) Allocation

Enrolment 21% of patients enrolled on 1^st

calendar day of ICU admission 79% of patients enrolled on 2^nd

calendar day of ICU admission

Allocated to Standard Care:

n = 235

46 with abnormal baseline kidney function 189 with normal baseline kidney function

Allocated to amino acid supplement:

n = 239

Follow-Up, Day 90 mortality Withdrew consent (n=0)

Unable to contact for Day 90 Interview (n=0) Withdrew consent for Day 90 Interview (n=1) Unable to contact for Day 90 Interview (n=2) Note: All three had normal kidney function at baseline.

Day 90 outcomes:

n = 235

Day 90 outcomes:

n = 236

(12)

eFigure 2. Study process measures, entire study population

0 0.5 1 1.5 2 2.5

g/kg (IBW used if BMI > 25)

Protein delivered per patient (including amino acid supplement)

1 2 3 4 5 6 7 Days in ICU after study enrolment

200 400 600 800 1000 1200 1400 1600

mean kcals/patient

Energy delivered per patient (not including amino acid supplement)

600 800 1000 1200 1400

Energy delivered per patient (including amino acid supplement)

Amino acid supplement (n=239)

Standard Care (n=235) Error bars indicate 95%

confidence intervals around differences between groups.

IBW: ideal body weight;

BMI: Body Mass Index

0 0.5 1 1.5 2 2.5

g/kg (IBW used if BMI > 25)

Protein delivered per patient (including amino acid supplement)

200 400 600 800 1000 1200 1400

Energy delivered per patient (not including amino acid supplement)

400 600 800 1000 1200 1400

Energy delivered per patient (including amino acid supplement)

(13)

Subgroup 1: Baseline kidney dysfunction

Definition of subgroup: Baseline kidney dysfunction

Baseline kidney dysfunction: Creatinine at time of enrolment > 168µmol/L.

Covariate adjusted analysis within each subgroup

Covariate adjusted analysis of the primary outcome within each subgroup was not undertaken because test of treatment x subgroup interaction did not meet pre-specified threshold. See manuscript Table 1.

Subgroup 2: Baseline risk of progression of AKI

Definition of subgroup:

At risk of progression of AKI: Creatinine increased over the previous 24h by at least 20% to over 120µmol/L.

Covariate adjusted analysis within each subgroup was undertaken because test of treatment x subgroup interaction did met pre-specified threshold. See manuscript Table 1.

Subgroup of patients at risk of progression of AKI (n = 72)

After adjustment for potential confounding due to APACHE II score, gender and baseline urea, there was no significant effect of the study intervention (Risk difference 3.2% in favour of standard care, 95% CI -16.1 to 22.5, P=0.75).

Subgroup of patients not at risk of progression of AKI (n=402)

After adjustment for potential confounding due to APACHE II score, Source of ICU admission and Admission Diagnosis, there was no significant effect of the study intervention (Risk difference -6.9% in favour of study intervention, 95% CI -14.0 to 0.2, P=0.058).

Detailed analysis of other outcomes within each subgroup

(14)

eFigure 3: Estimated Glomerular Filtration Rate (CKD-EPI

creatinine

) by study day, in the subgroup of patients with baseline risk of progression of AKI

CKD-EPIcreatinine: Glomerular filtration rate estimated from creatinine using the equations developed by Levey et al; AKI: Acute Kidney Injury; ICU: Intensive Care Unit; Note: p = 0.0796 treatment x time interaction from repeated measures ANOVA. Error bars indicate 95% confidence intervals around differences between groups at each time point.

20 25 30 35 40 45 50 55 60

0 1 3 5 7

e G F R

Days in Study ICU after enrolment

Estimated Glomerular Filtration Rate (CKD-EPI) by Study Day, in the subgroup of patients with risk of progression of AKI

Standard Care (n=35)

Standard Care

N = 35 34 32 29 25 22 18 Amino acid supplement

N = 36 35 33 29 28 26 24 Patients in ICU at end of each study day

(15)

eFigure 4: Estimated Glomerular Filtration Rate (CKD-EPI

creatinine

) by study day, in the subgroup of patients with no baseline risk of progression of AKI

CKD-EPIcreatinine: Glomerular filtration rate estimated from creatinine using the equations developed by Levey et al¹⁹; AKI: Acute Kidney Injury; ICU: Intensive Care Unit; Note: p = 0.0079 treatment x time interaction from repeated measures ANOVA. Error bars indicate 95% confidence intervals around differences between groups at each time point.

65 70 75 80 85 90 95

0 1 3 5 7

e G F R

Estimated Glomerular Filtration Rate (CKD-EPI) by Study Day, in the subgroup of patients with no risk of progression of AKI

Standard Care

(16)

eTable 1. Patient demographics and baseline balance in the subgroup of patients with baseline risk of progression of AKI

Patient characteristics

Standard Amino Acid Care Supplement 35 patients 37 patients

Age, in years (mean ± SD) 63.0 ± 15.0 66.9 ± 11.4

Gender, n (%) females** 13 (37.1) 4 (10.8)

APACHE II score† (mean ± SD)* 24.1 ± 6.6 27.4 ± 8.3

Mechanically ventilated, n (%) 8 (22.9) 6 (16.2)

BMI, kg/m² (mean ± SD)** 33.1 ± 7.8 28.4 ± 5.3

BMI < 18, n (%) 0 (0.0) 0 (0.0)

SGA muscle wasting score, (mean ± SD) 1.10 ± 0.28 1.21 ± 0.48

SGA fat loss score, (mean ± SD)* 1.10 ± 0.28 1.24 ± 0.49

Currently receiving EN or PN, n (%)* 10 (28.6) 18 (48.7) Calories, kcal over previous 24 h, all patients, (mean ± SD) * 87.8 ± 226 277.1 ± 528 Protein, g over previous 24 h, all patients, (mean ± SD)* 3.4 ± 8.2 13.7 ± 28.4

Clinically meaningful organ failure

^{n (%)}

Cardio-vascular failure

(systolic pressure < 90, not fluid responsive) 32 (91.4) 29 (78.4) Coagulation failure (platelets < 81 x 10⁹/L) 5 (14.3) 5 (13.5) Respiratory failure (PaO₂:FiO₂ ratio < 301) 31 (88.6) 35 (94.6) Hepatic failure (total bilirubin > 32.5 μmol/L) 6 (17.1) 6 (16.2)

Measures of kidney function

Creatinine, ^µmol/L, (mean ± SD) 182.1 ±37.1 189.9 ± 55.0 Estimated GFR, ml/min/1.73m², (mean ± SD) 32.3 ± 9.6 33.1 ± 10.1

Urea, mmol/L, (mean ± SD)** 11.7 ± 3.9 14.6 ± 6.2

Urine output, mls over previous 24 h, (mean ± SD) 1429 ± 1060 1404 ± 822

Known risk factors for AKI

^{n (%)}

Vasoactive drugs, over previous 24 h 35 (100) 34 (91.9) Nephrotoxic agent or pigment, over previous 24 h 13 (37.1) 16 (43.2) Proximal or distal loop diuretic, over previous 24 h 13 (37.4) 11 (29.7)

Oliguria, over previous 24 h 12 (34.3) 7 (18.9)

Pre-acute illness CKD > 1, or history of CKD 6 (17.1) 5 (13.5) Massive transfusion, over previous 24 h 1 (2.9) 0 (0) Obstructive uropathy, over previous 24 h 1 (2.9) 0 (0)

Chronic health states

^{n (%)}^§

Immuno-compromised 5 (14.3) 4 (10.8)

Respiratory disease 2 (5.7) 2 (5.4)

Cardiovascular disease 0 (0) 0 (0)

Hepatic cirrhosis 0 (0) 0 (0)

Chronic dialysis 0 (0) 0 (0)

Source of admission to ICU

^{n (%)}

Operating Room 10 (28.6) 11 (29.7)

Emergency Department 7 (20.0) 10 (27.0)

Hospital Ward 9 (25.7) 8 (21.6)

(17)

Patient characteristics

Standard Amino Acid Care Supplement 35 patients 37 patients

Surgical admission

^{n (%)}

Emergency Surgery 7 (11.4) 7 (18.9)

Elective Surgery 3 (8.5) 4 (10.8)

APACHE III admission diagnosis

^{n (%)}

Respiratory 8 (22.9) 5 (13.5)

Cardiovascular / vascular 8 (22.9) 10 (27.3)

Gastrointestinal 6 (17.1) 6 (16.2)

Neurological 1 (2.9) 0 (0)

Sepsis 9 (25.7) 12 (32.4)

Trauma 1 (2.9) 2 (5.4)

Metabolic 0 (0) 0 (0)

Hematological 0 (0) 1 (2.7)

Orthopedic surgery 1 (2.9) 0 (0)

Kidney 0 (0) 0 (0)

Other 1 (2.9) 1 (2.7)

** Indicates potential confounders due to baseline imbalance with balance p-values < 0.05. * Indicates potential confounders due to baseline imbalance with balance p-values < 0.15. All other baseline imbalance p-values > 0.15. APACHE: Acute Physiology and Chronic Health Evaluation, scores range from 0 to 71.

APACHE scores have a non-linear relationship with the risk of death. Higher scores indicate more severe disease, associated with a higher risk of death. Scores in excess of 37 have been associated with a greater than 99.9% risk of subsequent death in-hospital. ^§ Defined using APACHE II criteria. † Nine patients (1.9%) had one or more missing APACHE II physiology variables, which were imputed with average values. No other data were imputed. SGA: Subjective Global Assessment, EN: enteral nutrition, PN: parenteral nutrition, SD: Standard Deviation, ICU: Intensive Care Unit, BMI: Body Mass Index, AKI: Acute kidney injury, CKD: Chronic Kidney Disease.

(18)

eTable 2. Mortality, Length of Stay and Health-Related Quality of Life, subgroup of patients with baseline risk of progression of AKI

Primary Outcome

Standard Care 35 patients

Amino acid supplement

37 patients

Risk Difference

(95% CI) P –value

Deaths before Study Day 90, % (n/N) 17.1%

(6/35) 35.1%

(13/37) 3.2%^§

(-16.1 to 22.5) 0.75^§

Quality of Life and Physical Function

(reported by survivors at Day 90 Interview)

Standard Care 29 survivors

24 survivors

Difference

(95% CI) P -value

RAND-36 General Health, mean (SD)

(n responses available for analysis) 46.4 (28.6)

(n=28) 50.0 (26.6)

(n=24) 3.6

(-11.8 to 19.0) 0.64 ECOG Performance Status, mean (SD)

(n=28) 1.20 (1.1)

(n=24) 0.14

(-0.45 to 0.75) 0.62 RAND-36 Physical Function, mean (SD),

(n=28) 50.8 (36.9)

(n=24) -0.06

(-20.2 to 20.1) 1.00 Requiring RRT at Day 90,

% (n/N responses available for analysis) 0%

(0/38) 0%

(0/43) 0.0 1.00**

Discharge status and Length of stay

Standard Care 35 patients

37 patients

Difference

(95% CI) P-value

ICU stay (days), mean (95% CI) 11.9

(10.0 to 14.2) 14.4

(12.0 to 17.1) 2.4 days

(-1.8 to 8.5) 0.30 ICU discharge mortality, % (n/N) 14.3%

(5/35) 21.6%

(8/37) 7.3%

(-16.1 to 29.5 0.54**

Hospital stay (days), mean (95% CI) 29.7

(24.7 to 35.7) 31.5

(26.2 to 37.9) 1.8 days

(-7.9 to 15.9) 0.75 Hospital discharge mortality, % (n/N) 14.3%

(5/35) 35.1%

(13/37) 20.9%

(-2.4 to 42.4) 0.057**

** Exact P-value and test-based 95% CI.

§Covariate adjusted model controlled for confounding due to APACHE 2, gender and baseline Urea.

CI: confidence interval; ICU: intensive care unit; RAND: Public domain version of the Short Form 36; ECOG:

Eastern Collaborative Oncology Group; RRT: renal replacement therapy.

(19)

eTable 3. Secondary outcomes, adjusted for time at risk (ICU stay) in the subgroup of patients

with baseline risk of progression of AKI.

Secondary Renal Outcomes

^Standard^Care

35 patients

37 patients

difference

( 95% CI) p-value Volume received, mean (SD)

mLs per ICU day 2471 (988) 2674 (950) 202 mls

(-253 to 658) 0.37 Urine output, mean (SD)

(-407 to 576) 0.73 Net fluid balance, mean (SD)

(-438 to 673) 0.67 Oliguria days, per 10 days at risk, (95% CI)

< 100 mLs for 6 consecutive h 1.96

(1.21 to 3.18) 1.40

(0.86 to 2.27) -0.56

(-1.42 to 1.71) 0.49 Diuretic days, per 10 days at risk, (95% CI)

distal tubule diuretics 4.02

(3.38 to 4.79) 3.87

(3.25 to 4.60) -0.16

(-1.29 to 1.46) 0.83 Patients receiving RRT,

%, (n/N) 22.9%

(8/35) 35.1%

(13/37) 12.3%

(-10.7% to 34.7%) 0.31 Days of RRT,

per 10 patient•ICU days 1.47

(0.82 to 2.63) 1.89

(1.05 to 3.39) 0.42

(-0.88 to 4.61) 0.67 Patients with renal dysfunction,

(creatinine > 168 μmol/L), %, (n/N) 74.3%

(26/35) 66.6%

(24/36) -7.62%

(-30.9% to 14.7%) 0.60 Patients with eGFR < 60 mLs/min/1.73m²

%, (n/N) 88.6%

(31/35) 94.4%

(34/36) 5.9%

%, (n/N) 65.7%

(23/35) 58.3%

(21/36) -7.4%

%, (n/N) 25.7%

(9/35) 22.2%

(8/36) -3.5%

(-25.6% to 20.2%) 0.79 Days with eGFR < 60 mLs/min/1.73m²

(6.63 to 8.68) 6.81

(5.95 to 7.78) -0.79

(-2.39 to 1.31) 0.42 Days with eGFR < 30 mLs/min/1.73m²

(2.80 to 5.27) 3.17

(2.32 to 4.35) -0.67

(0.49 to 2.35) 0.98

(0.45 to 2.17) -0.08

(-0.86 to 3.69) 0.91

Organ system failure,

adjusted for time at risk of failure (ICU stay) failure days per 10 patient•ICU days

mean

(95% CI) mean

(95% CI) difference

(95% CI) p-value Kidney dysfunction,

(creatinine > 168 μmol/L) 4.14

(3.16 to 5.45) 3.99

(3.04 to 5.24) -0.16

(-1.83 to 2.73) 0.89 Respiratory failure,

PaO₂:FiO₂ ratio < 301 9.13

(8.43 to 9.89) 8.56

(7.90 to 9.28) -0.57

(-1.84 to 0.91) 0.43 Hepatic failure,

total bilirubin > 32.5 μmol/L 1.88

(1.04 to 3.41) 1.77

(0.98 to 3.21) -0.11

(-1.34 to 3.92) 0.91 Coagulation failure,

platelets < 81 x10⁹/L 0.85

(0.37 to 1.93) 0.83

(0.36 to 1.88) -0.02

(-0.69 to 3.40) 0.96 Cardiovascular failure

systolic blood pressure < 90mmHg, not fluid responsive

1.18

(0.89 to 1.57) 1.28

(0.97 to 1.70) 0.10

(-0.45 to 1.08) 0.78 MODS

two or more organ system dysfunctions on the same day

4.62

(3.84 to 5.55) 4.27

(3.55 to 5.12) -0.35

(-1.66 to 1.54) 0.67

Number of organ failures 1.53 1.62 0.09

(20)

Secondary Renal Outcomes

^Standard^Care

35 patients

37 patients

difference

( 95% CI) p-value

Concomitant therapies and tertiary outcomes,

adjusted for time at risk (ICU stay), days per 10 patient•ICU days Days of mechanical ventilation 7.81

(7.28 to 8.40) 8.10

(7.54 to 8.70) 0.28

-0.80. to 1.53) 0.63 Pressure ulcer treatment days

Treatment for Stage 1 or greater 0.70

(0.39 to 1.28) 1.85

(1.02 to 3.37) 1.15

(-0.14 to 5.40) 0.11 Low serum albumin days

< 25 g/L 5.23

(4.34 to 6.32) 6.76

(5.60 to 8.16) 1.52

(-0.59 to 4.60) 0.18

Systemic antibiotic days 7.36

(6.78 to 7.98) 7.82

(7.21 to 8.49) 0.46

(-0.71 to 1.85) 0.46 CI: confidence interval; ICU: intensive care unit; SD: standard deviation; CI: confidence interval; ICU:

intensive care unit; RRT: renal replacement therapy; eGFR: estimated glomerular filtration rate using CKD- EPI creatinine equations, MODS: multiple organ dysfunction syndrome.

eTable4. Patient demographics and baseline balance, subgroup of patients not at baseline risk of progression of AKI.

Patient characteristics

Standard Amino Acid Care Supplement

200 patients 202 patients

Age,

in years (mean ± SD)

62.7

± 17.0

62.7

± 16.0

Gender,

n (%) females

75

(37.5)

77

(38.1)

APACHE II score†

(mean ± SD)*

19.6

± 6.7

20.7

± 7.1

Mechanically ventilated,

n (%)

167

(83.5)

166

(82.2)

BMI,

kg/m² (mean ± SD)

28.8

± 6.3

29.0

± 7.3

BMI < 18

, n (%)

2

(1.0)

5

(2.5)

SGA muscle wasting score,

(mean ± SD)

1.35

± 0.67

1.29

± 0.61

SGA fat loss score,

(mean ± SD)

1.36

± 0.67

1.31

± 0.61

Currently receiving EN or PN,

n (%)

88

(44.0)

91

(45.1)

Calories,

kcal over previous 24 h, all patients

248

(450)

209

(370)

Protein,

g over previous 24 h, all patients

11.3

(21.5)

9.2

(16.5)

Clinically meaningful organ failure

^{n (%)}

Kidney dysfunction at study entry

(creatinine > 168 μmol/L)**

11

(5.5)

23

(11.4)

Cardio-vascular failure

(systolic pressure < 90, not fluid responsive)

126

(63.0)

131

(64.9)

Coagulation failure

(platelets < 81 x 10⁹/L)

16

(8.0)

10

(5.0)

Respiratory failure

^(PaO2:FiO₂ ratio < 301)

174

(87.0)

168

(83.2)

Hepatic failure

(total bilirubin > 32.5 μmol/L)

18

(9.0)

20

(9.9)

Measures of kidney function

Creatinine,

^µmol/L, mean (SD)

95.6

(39.6)

101.5

(51.3)

Estimated GFR,

ml/min/1.73m², mean (SD)

74.2

(28.5)

72.2

(30.3)

(21)

Patient characteristics

Standard Amino Acid Care Supplement

200 patients 202 patients

Known risk factors for AKI

^{n (%)}

Vasoactive drugs

, over previous 24 h

145

(72.5)

143

(70.8)

Nephrotoxic agent or pigment

76

(38.0)

74

(36.6)

Proximal or distal loop diuretic

41

(20.5)

46

(22.8)

20% Creatinine Rise to > 120

μmol/L, over previous 24 h

0

(0)

0

(0)

Oliguria

19

(9.5)

18

(8.9)

Pre-acute illness CKD > 1, or history of CKD 8

(4.0)

14

(6.9)

Massive transfusion,

over previous 24 h

5

(2.5)

8

(4.0)

Obstructive uropathy,

over previous 24 h

0

(0)

0

(0)

Chronic health states

^{n (%)}^§

Immuno-compromised 14

(7.0)

11

(5.5)

Respiratory disease 9

(4.5)

11

(5.5)

Cardiovascular disease 6

(3)

2

(1)

Hepatic cirrhosis 0

(0)

0

(0)

Chronic dialysis 0

(0)

0

(0)

Source of admission to ICU

^{n (%)}

Operating Room (2) 66

(33.0)

69

(34.2)

Emergency Department (1) 59

(29.5)

68

(33.7)

Hospital Ward(3) 40

(20.0)

28

(13.9)

Other hospital (6) 31

(15.5)

32

(15.8)

Transfer from ICU (4) 4

(2.0)

2

(1.0)

ICU readmission (5) 0

(0)

3

(1.5)

Surgical admission

^{n (%)}

Emergency Surgery 45

(22.5)

44

(21.8)

Elective Surgery 24

(11.0)

25

(12.4)

APACHE III admission diagnosis

^{n (%)}

Respiratory 55

(27.5)

55

(27.2)

Gastrointestinal 43

(21.5)

35

(17.3)

Cardiovascular / vascular 40

(20.0)

42

(20.8)

Sepsis 16

(8.0)

23

(11.4)

Neurological 17

(8.5)

29

(14.4)

Trauma 16

(8.0)

13

(6.4)

Metabolic 5

(2.5)

1

(0.5)

Hematological 4

(2.0)

2

(0.5)

Orthopedic surgery 0

(0)

1

(0.5)

Renal 1

(0.5)

0

(0)

Other 3

(1.5)

2

(1.0)

Potential confounders due to imbalance are indicated by a single or double asterix (* Imbalance p-

value <0.15, **Imbalance p-value <0.05).

APACHE: Acute Physiology and Chronic Health

Evaluation, scores range from 0 to 71. APACHE scores have a non-linear relationship with the risk

of death. Higher scores indicate more severe disease, associated with a higher risk of death.

(22)

APACHE II physiology variables, which were imputed with average values. No other data were imputed. SGA: Subjective Global Assessment, EN: enteral nutrition, PN: parenteral nutrition, SD:

Standard Deviation, ICU: Intensive Care Unit, BMI: Body Mass Index, AKI: Acute kidney injury, CKD:

Chronic Kidney Disease.

eTable 5. Mortality, Length of Stay and Quality of Life, subgroup of patients not at baseline risk of progression of AKI.

Primary Outcome

Standard Care

200 patients

202 patients

Risk Difference

^§

(95% CI)

P –value

^§

Deaths before Study Day 90, %

(n/N)

20.5%

(41/200)

14.6%

(29/199)* -6.9%

(-14.0 to 0.2) 0.058

Quality of Life and Physical Function

(reported by survivors at Day 90 Interview)

159 survivors

Nephro- Protect

170 survivors

Difference

(95% CI)

P -value

RAND-36 General Health,

mean (SD)

(n responses available for analysis)

53.9

(25.3)

(n=152)

50.6

(27.2)

(n=168) -3.3

(-9.1 to 2.4) 0.26

ECOG Performance Status,

mean (SD)

(n responses available for analysis)

1.14

(1.0)

(n=153)

1.32

(1.0)

(n=168) 0.18

(-0.4 to 0.40) 0.11

RAND-36 Physical Function,

mean

(SD), (n responses available for analysis)

53.7

(32.6)

(n=152)

47.3

(33.3)

(n=168) -6.3

(-13.6 to 0.90) 0.086

Requiring RRT at Day 90,

% (n/N responses available for analysis)

0.7%

(1/154)

0.0%

(0/168) -0.7

(-11.5% to 10.3%) 0.48**

Discharge status and Length of stay

200 patients

202 patients

Difference

(95% CI)

P-value

ICU stay (days),

mean (95% CI)

10.5

(9.7 to 11.3)

11.1

(10.3 to 12.0) 0.62 days

(-0.97 to 2.47) 0.46

ICU discharge mortality,

% (n/N)

12.5%

(25/200)

9.9%

(20/202) 2.06%

(-7.0 to 12.5) 0.43**

Hospital stay (days),

mean (95% CI)

23.9

(22.1 to 25.8)

25.0

(23.2 to 27.1) 1.2 days

(-2.5 to 5.4) 0.55

Hospital discharge mortality,

^{% (n/N)}

19.0%

(38/200)

11.9%

(24/202) 7.1%

(0.0 to 14.0) 0.054**

*1 patient withdrew consent for Day 90 interview and 2 patients could not be contacted (lost to

follow-up). These patients were considered ‘missing at random’ for Day 90 mortality analysis.

** Exact P-value and test-based 95% CI.

§

Covariate adjusted model controlled for confounding due to APACHE II score, Source of ICU

admission and Admission Diagnosis.

(23)

eTable 6. Clinically Significant Organ Dysfunction and Concomitant Interventions, adjusted for time at risk (ICU stay), subgroup of patients not at baseline risk of progression of AKI.

Secondary Renal Outcomes

^Standard^Care

200 patients

Amino acid supplement 202 patients

difference

( 95% CI) p-value Volume received, mean (SD)

(241 to 580) <0.0001 Urine output, mean (SD)

(180 to 501) <0.0001 Net fluid balance, mean (SD)

(-106 to 254) 0.42 Oliguria days, per 10 days at risk, (95% CI)

< 100 mLs for 6 consecutive h 0.58

(0.44 to 0.77) 0.66

(0.52 to 0.91) 0.11

(-0.19 to 0.62) 0.55 Diuretic days, per 10 days at risk, (95% CI)

distal tubule diuretics 3.52

(3.16 to 3.92) 3.34

(3.00 to 3.72) -0.18

(-0.81 to 0.62) 0.64 Patients receiving RRT,

%, (n/N) 2.5%

(5/200) 5.9%

(12/202) 3.4%

(-6.5% to 13.2%) 0.14 Days of RRT,

(0.04 to 0.16) 0.27

(0.13 to 0.53) 0.19

(-0.01 to 0.97) 0.088 Patients with renal dysfunction,

(creatinine > 168 μmol/L), %, (n/N) 13.5%

(27/200) 16.9%

(34/201) 3.4%

%, (n/N) 42.5%

(85/200) 37.8%

(76/201) -4.7%

%, (n/N) 11.5%

(23/200) 15.9%

(32/201) 4.4%

%, (n/N) 1.5%

(3/200) 4.0%

(8/201) 2.5%

(-7.3% to 12.2%) 0.22 Days with eGFR < 60 mLs/min/1.73m²

(2.18to 3.26) 2.14

(1.75 to 2.61) -0.53

(0.27 to 0.56) 0.63

(0.44 to 0.91) 0.24

(0.01 to 0.05) 0.11

(0.05 to 0.23) 0.08

(-0.00 to 0.48) 0.067

Organ system failure,

adjusted for time at risk of failure (ICU stay) failure days per 10 patient•ICU days

mean

(95% CI) mean

(95% CI) difference

(95% CI) p-value Kidney dysfunction,

(creatinine > 168 μmol/L) 0.42

(0.30 to 0.60) 0.68

(0.48 to 0.96) 0.26

(-0.08 to 0.93) 0.18 Respiratory failure,

PaO₂:FiO₂ ratio < 301 8.46

(8.11 to 8.82) 7.82

(7.51 to 8.16) -0.63

(--1.26 to -0.05) 0.067 Hepatic failure,

total bilirubin > 32.5 μmol/L 0.54

(0.43 to 0.69) 0.54

(0.42 to 0.69) 0.00

(-0.21 to 0.34) 0.98 Coagulation failure,

platelets < 81 x10⁹/L 0.36

(0.26 to 0.51) 0.42

(0.30 to 0.59) 0.06

(-0.15 to 0.46) 0.66 Cardiovascular failure

systolic blood pressure < 90mmHg, not fluid responsive

1.04

(0.87 to 1.25) 0.81

(0.68 to 0.97) -0.23

(-0.48 to 0.12) 0.16 MODS

two or more organ system dysfunctions on the same day

1.59

(1.36 to 1.86) 1.45

(1.24 to 1.69) -0.14

(-0.53 to 0.38) 0.55

Number of organ failures 0.89 0.88 -0.01

(24)

Secondary Renal Outcomes

^Standard^Care

200 patients

Amino acid supplement 202 patients

difference

( 95% CI) p-value

Concomitant therapies and tertiary outcomes,

adjusted for time at risk (ICU stay), days per 10 patient•ICU days Days of mechanical ventilation 7.21

(6.89 to 7.54) 7.16

(6.84 to 7.49) -0.05

(-0.66 to 0.62) 0.90 Pressure ulcer treatment days

Treatment for Stage 1 or greater 1.14

(0.87 to 1.49) 0.94

(0.72 to 1.23) -0.19

(-0.58 to 0.46) 0.48 Low serum albumin days

< 25 g/L 4.01

(3.62 to 4.45) 4.06

(3.66 to 4.50) 0.04

(-0.71 to 0.98) 0.93

Systemic antibiotic days 7.85

(7.54 to 8.16) 7.95

(7.65 to 8.27) 0.10

(-0.49 to 0.74) 0.75 CI: confidence interval; ICU: intensive care unit; SD: standard deviation; CI: confidence interval; ICU:

intensive care unit; RRT: renal replacement therapy; eGFR: estimated glomerular filtration rate using CKD- EPI creatinine equations, MODS: multiple organ dysfunction syndrome.

(25)

Subgroup 3: Baseline kidney dysfunction and/or baseline risk of progression of AKI

Definition of subgroup: Baseline kidney dysfunction or risk of progression of AKI

Creatinine at time of enrolment > 168µmol/L and/or creatinine increased over the previous 24h by at least 20% to over 120µmol/L.

Covariate adjusted analysis within each subgroup was undertaken because test of treatment x subgroup interaction did met pre-specified threshold. See manuscript Table 1.

Subgroup of patients with baseline renal dysfunction and/or risk of progression of AKI (n = 106)

After adjustment for potential confounding due to APACHE II score, gender and baseline urea, there was no significant effect of the study intervention (Risk difference -0.6% in favour of study interventrion, 95% CI - 16.2 to 15.2, P=0.95).

Subgroup of patients no baseline kidney dysfunction and no risk of progression of AKI (n=368)

After adjustment for potential confounding due to Age, BMI, APACHE 2, Source of ICU admission, Admission Diagnosis and presence of a chronic immune compromised state there was significant effect of the study intervention (Risk difference -7.9% in favour of study intervention, 95% CI -15.1 to -0.7, P=0.0341).

Detailed analysis of other outcomes within this subgroup.

Note: Complete details of the analysis conducted within the subgroup with no baseline renal dysfunction and no risk of progression of AKI are presented in the main manuscript.

(26)

eFigure 5. Estimated Glomerular Filtration Rate (CKD-EPI

creatinine

) by study day, in the subgroup of patients with baseline kidney dysfunction or risk of progression of AKI

CKD-EPIcreatinine: Glomerular filtration rate estimated from creatinine using the equations developed by Levey et al¹⁹; AKI: Acute Kidney Injury; ICU: Intensive Care Unit; Note: p = 0.3560 treatment x time interaction from repeated measures ANOVA. Error bars indicate 95% confidence intervals around differences between groups at each time point.

20 25 30 35 40 45 50 55 60

0 1 3 5 7

e G F R

Estimated Glomerular Filtration Rate (CKD-EPI) by Study Day, in the subgroup of patients with baseline kidney dysfunction or risk of progression of AKI

Standard Care

The Nephro-Protective Trial

Online-Only Supplement:

The effect of intravenous amino acid supplementation on mortality in ICU patients may be dependent on kidney function:

Post hoc subgroup analyses of a multi-centre randomized trial

Gordon S. Doig and Ran Zhu (

)

for the Nephro-Protective Trial Investigators Group.

The Nephro-Protective Trial NHMRC Project Grant Number 632614

Australian and New Zealand Clinical Trials Registry Number

Dr. Gordon S. Doig,

Royal North Shore Hospital, Intensive Care Unit,

St. Leonards, NSW Australia 2065

www.EvidenceBased.net/NephroProtect

24 Jan 2018

© 2018 Gordon S Doig, University of Sydney. All rights reserved. This publication is protected by

copyright. No part of it may be reproduced for commercial purposes or distributed electronically

without prior written permission of the publisher. Reproduction for personal or educational use is

acceptable.

Contents

THE NEPHRO-PROTECTIVE TRIAL INVESTIGATORS

Calgary, Canada.

),

Complete eligibility criteria:

See next page for Exclusion Criteria.

Maximum Creatinine Table

Study intervention: Dosing algorithm for supplementary amino acids

eFigure 1: CONSORT 2010 Patient Recruitment Flow Diagram, entire study population and baseline kidney function

eFigure 2. Study process measures, entire study population

Subgroup 1: Baseline kidney dysfunction

Subgroup 2: Baseline risk of progression of AKI

eFigure 3: Estimated Glomerular Filtration Rate (CKD-EPI

) by study day, in the subgroup of patients with baseline risk of progression of AKI

eFigure 4: Estimated Glomerular Filtration Rate (CKD-EPI

) by study day, in the subgroup of patients with no baseline risk of progression of AKI

eTable 1. Patient demographics and baseline balance in the subgroup of patients with baseline risk of progression of AKI

Patient characteristics

Clinically meaningful organ failure

Measures of kidney function

Known risk factors for AKI

Chronic health states

Source of admission to ICU

Patient characteristics

Surgical admission

APACHE III admission diagnosis

eTable 2. Mortality, Length of Stay and Health-Related Quality of Life, subgroup of patients with baseline risk of progression of AKI

Primary Outcome

Quality of Life and Physical Function

Discharge status and Length of stay

eTable 3. Secondary outcomes, adjusted for time at risk (ICU stay) in the subgroup of patients

Secondary Renal Outcomes

Organ system failure,

Secondary Renal Outcomes

Concomitant therapies and tertiary outcomes,

eTable4. Patient demographics and baseline balance, subgroup of patients not at baseline risk of progression of AKI.

Patient characteristics

Age,

62.7

62.7

Gender,

75

77

APACHE II score†

19.6

20.7

Mechanically ventilated,

167

166

BMI,

28.8

29.0

BMI < 18

2

5

SGA muscle wasting score,

1.35

1.29

SGA fat loss score,

1.36

1.31

Currently receiving EN or PN,