• 
• 

Three Part Question

In [patients undergoing cardiac surgery], is the use of [thyroxine] associated with [improved cardiac output, better recovery or fewer complications].

Clinical Scenario

You are anaesthetising a high risk CABG patient. Before coming off bypass, the surgeon requests that you give some thyroxine. You have never heard of this strategy before and thus while you give the thyroxine, you decide to review the literature to see if there is any evidence to back up this strategy.

Search Strategy

Medline 1966 to November 2005 using OVID interface, EMBASE 1980 to 2005 Week 52.
[exp Cardiopulmonary Bypass OR CABG.mp. OR exp Thoracic Surgery OR exp Cardiac surgical procedures OR Coronary art$ bypass.mp. OR Cardiopulmonary bypass.mp. OR exp Cardiopulmonary Bypass/ OR exp Cardiovascular Surgical Procedures/ OR exp Thoracic Surgical Procedures/ OR exp Coronary Artery Bypass/ OR cardiac transplantation.mp. OR exp Heart Transplantation/] AND [exp THYROXINE/ OR Thyroxine.mp. OR exp TRIIODOTHYRONINE/ OR triiodothyronine.mp. OR thyronine.mp. OR exp Thyronines/] AND [exp Cardiac Output/ph, de, su, th OR exp Hemodynamic Processes/ OR haemodynamic processes.mp. OR postoperative care.mp. OR exp Postoperative Care/ OR exp ANESTHESIA RECOVERY PERIOD/ OR exp "RECOVERY OF FUNCTION"/ OR exp Intraoperative Complications/ OR exp Postoperative Complications/ OR surgical complications.mp. OR Postoperative Complications.mp OR Length of Stay.mp ]

Search Outcome

A total of 86 papers on Medline and 113 on Embase were identified using the reported search. A further relevant paper was identified by hand-searching reference of lists. Thirteen papers represented the best evidence on the subject and are summarised below.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Sirlak et al, 2004, Turkey	80 pts undergoing elective CABG surgery with LVEF<40% Exclusions - valvular surgery - IABP -emergency surgery - thyroid disease - thyroid replacement therapy Standard anaesthetic / surgical / CPB management Treatment group - oral T3 125mcg/day for 7 days preop and from 1st postop day until discharge Control group - placebo	PRCT (level 1b)	TSH levels at multiple time points	- Pre-CPB levels significantly different between T3 vs. control grp (p<0.001). - significant decrease in both groups with institution of CPB (p<0.001). -mean TSH decrease greater in T3 treated grp compared to baseline than in placebo grp (p<0.001) in first 5 hrs. - significant increase in both grps in 6th hr post CPB institution (p<0.001)	Thyroid hormone assay data displayed graphically with few absolute values quoted Dose based on that used previously to treat obese normothyroid patients
			Plasma free T3 (fT3) levels at multiple time points	- significantly higher in T3 vs. control group at induction of anaesthesia. - significant fT3 decrease in both groups after initiation of CPB (p<0.001). - recovery of fT3 level after 24hrs to baseline but significant fT3 decrease in control group continued to end of study period - mean fT3 decrease between grps significant from 2nd hr to end of study (p<0.001)
			Plasma free T4 levels at multiple time points	- fT4 significantly lower in T3 grp (p<0.001). - significant fT4 decrease in both groups with institution of CPB sustained for 3hrs (p<0.001). - significant decrease in fT4 level after 6 hrs in control but not T3 grp
			Perioperative haemodynamic data at multiple time points	Cardiac Index - T3 vs. Control grp significantly higher pre-induction (p<0.001) & at all time points thereafter (p<0.01) SVR - T3 vs. Control similar pre-induction but significantly less in T3 grp after 12 hours (p<0.001) PVR - T3 vs. Control similar pre-induction but significantly less in T3 grp after 12 hours (p<0.05) Mean PAP - no signif difference except at 24hrs when MPAP signif lower in T3 grp (p<0.05) PCWP - T3 vs. Control values higher preinduction (p<0.05) but no signif difference postop CVP - T3 vs. Control values higher preinduction (p<0.01) but no signif difference postop Mixed venous SvO2 higher in T3 vs. control grp pre-induction & at all time points (p<0.001)
			Inotrope requirements	Mean Dobutamine T3 vs. control grp 5+/-1.2mcg/kg/hr vs. 8+/-1.1mcg/kg/hr (p<0.001) Mean Dopamine T3 vs. Control 3+/-0.1mcg/kg/hr vs. 6+/-2.1mcg/kg/hr (p<0.001). 8/40 (20%) T3 vs. 17/40 (42.5%) Control pts required IABP (p<0.05)
			Morbidity	ITU stay significantly shorter in T3 grp (p<0.001). No significant difference in myocardial ischaemia / infarction, atrial fibrillation or death
Novitzky et al 1989(1) USA	Study I - 24 CABG pts with LVEF < 30% - IV T3 0.1mcg/kg @ cross-clamp removal; 0.075mcg/kg before coming off CPB; 0.05mcg/kg @ 4 & hrs post-CPB vs. placebo Study II - 24 CABG pts with LVEF >40% - T3 protocol - 0.20 mcg/kg @ cross-clamp removal (time 0); 0.15 mcg/kg @ +4hrs; 0.10mcg/kg @ +8hrs; 0.05 mcg/kg @ +12hrs; & 0.05mcg/kg @ +20hrs - doses given derived from kinetic studies obtained from Study I Standardised periop care for both groups with exception of trial intervention Standardised protocols for use of Inotropes & Diuretics	PRCT (level 1b)	T3 levels Haemodynamic measurements Requirement for inotrope agents Requirement for diuretics - Urine output	Study I - 12/24 received T3 - 12/24 received placebo - Plasma free-T3 levels significantly higher in T3 vs. placebo group immediately post-CPB (p=0.001) but below normal range within 12hrs. Placebo group consistently below normal range. - No significant difference in haemodynamic parameters T3 vs. placebo. - Need for Inotropes significantly less in T3 group vs. placebo at time points 15min-8hrs (p<0.05), 8-16hrs (p<0.02) & 16-24hrs (p<0.01) - Need for diuretics significantly less in T3 group vs. placebo at time points first 15 mins post-CPB (p=0.03) & 15mins-8hrs (p=0.05). - No difference in urine output T3 vs. placebo. - No signif difference in deaths - 2/12 T3 vs. 1/12 placebo grp. - No adverse effects related to T3 Study II - 13/24 received T3. - 11/24 received placebo. - plasma free-T3 levels significantly raised in T3 vs. placebo group immediately post-CPB (p=0.001) & subsequently little different to pre-CPB levels. Placebo T3 level fell significantly post-CPB. - Cardiac output significantly rose in T3 vs. preop & controls post-CPB (p=0.008-0.05 at various time points within first 24hrs post-CPB). - SVR & PVR lower in both T3 grp post-CPB vs. pre-CPB (p<0.0001 & p<0.01 respectively), and placebo grp (p<0.01 & p<0.05 respectively) - No signif difference in Inotrope or Diuretic use T3 group vs. placebo - No significant difference in Urine output T3 group vs. placebo - No adverse effects related to T3	Small studies PRCT preceded by observational animal studies, and clinical observation of T3 as rescue therapy in pts unable to wean from CPB or with poor post-CPB function (Novitsky, 1989 (2)) Timing & (higher) dosing of T3 in Study II modified on basis of kinetic studies obtained from Study I therefore interventions not comparable between groups I & II Results for diuretic & inotrope use displayed graphically as total doses rather than number of patients treated
Klemperer et al, 1995 USA	142 pts with EF<40% undergoing CABG surgery Excluded if > 85 years, for valve procedure, on Amiodarone or Thyroxine therapy, thyroid disease or preop Inotropes / IABP Stratified on basis of LV dysfunction (EF<25% or EF 25-40%) T3 therapy vs. placebo Active grp received IV triodothyronine 0.8mcg/kg at cross-clamp removal, then 0.113mcg/kg/hr for 6hrs, then tapered to off Standard Anaesthetic / Surgical management Unit protocols for admin of Inotropes, Vasodilators & Vasopressors 71/142 (50%) T3 vs. 71/142 (50%) placebo	PRCT (level 1b)	Triiodothyronine (T3) levels	Serum T3 low normal pre-CPB, decreased by 40% 30mins into CPB in both grps (p<0.001 compared to baseline) T3 treated grp developed supra-normal T3 values during treatment , returning to normal after 24hrs T3 levels in both treated & placebo grp significantly different from baseline at 0.5hrs & 6hrs into protocol (p<0.001)	No information re dose rationale Arrhythmia data quoted as percentages, not absolute numbers Absolute data for CI & SVR 4 & 6hrs post-cross clamp removal not quoted directly to check statistics
			Haemodynamic response - heart rate, BP, CVP, PCWP, cardiac output / index, SVR	Cardiac Index significantly higher in T3 vs. placebo grp at 2,4 & 6hrs post-cross clamp removal (p=0.007) SVR significantly lower in T3 vs. placebo grp at 2,4 & 6hrs post-cross clamp removal (p=0.003) No other signif differences in haemodynamics
			Incidence of dysrhythmia	71% T3 vs. 66% placebo grp had supraventricular arrhythmia (incl sinus tachycardia) in first 6hrs of protocol & 40% T3 vs. 34% placebo grp in subsequent 18hrs. 39% T3 vs. 34% placebo grp required pharmacol therapy 42% T3 vs. 44% placebo grp had ventricular arrhythmia in first 6hrs of protocol & 24% T3 vs. 24% placebo grp in subsequent 18hrs. 22% T3 vs. 21% placebo required pharmacolog therapy
			Need for / quantity of inotrope / vasopressor drugs &/or mechanical support	No significant difference between groups in interventions (IABP, pacing, inotropes, vasopressors) required to come off CPB No significant difference between groups in requirement for / total doses of post-op inotrope agents
			Incidence of other major morbidity &/or mortality	No difference between grps in terms of duration of postop IPPV, ITU / hospital length of stay or perioperative morbidity / mortality
Guden et al, 2002, Turkey	60 CABG pts assigned to two groups of 30 Exclusions - over 75 years old - on thyroid replacement therapy - emergency surgery - reoperation / concomitant procedures Standardised anaesthetic / surgical / CPB technique Group A - 0.8mcg/kg IV triiodothyronine bolus at removal of cross-clamp followed by infusion 0.113mcg/kg/hr Group B (control) - placebo	PRCT (level 1b)	Serum triiodothyronine (T3) levels	Serum T3 levels Group B - normal preop but decreased significantly during surgery & remained low for 12hrs Group A - normal preop but increased above normal during infusion and returned to wards normal at 12hrs	Serum T3 data obtained from subgroup - 10 T3 & 10 placebo pts Choice of T3 dose not justified although same as that used by Klemperer et al [1995]
			Haemodynamic parameters	Cardiac Index T3 grp vs. control - immediately post-CPB: 2.75+/-0.52 vs. 2.63+/-0.6 (ns) - 6hrs post-CPB: 2.7+/-0.2 vs. 2.6+/-0.1 (ns) - 12hrs post-CPB: 2.7+/-0.1 vs. 2.7+/-0.2 (ns) SVR T3 grp vs. control - immediately post-CPB: 1040+/-220 vs. 1350+/-420 (p<0.001) - 6hrs post-CPB: 1100+/-100 vs. 1280+/-190 (ns) - 12hrs post-CPB: 1190+/-100 vs. 1200+/-90 (ns) No significant difference in heart rate, MAP, CVP or PCWP between groups
			Inotrope requirement	75% T3 grp vs. 78% control required inotropes immediately post-CPB (42% vs. 46% at 6hrs). Not significant
			Incidence of AF	3/30 (10%) T3 grp and 4/30 (13%) control developed AF. Not significant
			Overall morbidity / mortality	No significant differences in incidence of arrhythmia, need for inotropic support, ITU stay, morbidity & mortality between groups
Bennett-Guerrero et al, 1996, USA	211 pts at high risk of requiring inotropic drug support undergoing CABG Inclusions - at least one of age>65yrs, EF <=40% or cardiac reoperation Exclusions - pre-existing thyroid disease, thyroid hormone therapy, preoperative inotropes / IABP, renal impairment, recent MI Comparison of T3 therapy (0.8mcg/kg bolus + 0.12mcg/kg for 6 hrs, then weaned over 5 hrs) vs. Dopamine (5mcg/kg/min for 6hrs) vs. Placebo commencing at aortic cross-clamp release Standard Anaesthetic / Surgical / CPB management 66/205 received T3; 68/205 Dopamine; 71/205 Placebo	PRCT (level 1b)	Serum T3 concentrations	Free T3 (fT3) levels just below normal range in all three grps at baseline and decrease to about 25% of baseline value by time of aortic cross-clamp release (p<0.001 compared to baseline) Dopamine & Placebo grp fT3 levels remained signif depressed 1st postop day (p<0.001) T3 grp fT3 significantly increased 1hr after commencement of study drug (p<0.001 compared to baseline & other grps at same time point) with reduction to within normal range on postop day 1 (p<0.001 compared to baseline, other grps at same time point & same grp at 1hr time point) Total T3 levels mirrored fT3 levels Free T4 (fT4) & total T4 levels remained within normal range throughout study period	Blinded safety monitor reviewed all cases of death and serious morbidity and designated two complications as "possibly" related to study drug admin (one T3; one Dopamine) Dose selected with aim of avoiding contradictory results seen in studies using lower doses represented highest dose used in preceding safety trial by manufacturers in cardiac surgery patients 6/211 (2.8%) randomisation codes broken for clinical reasons - these sets of data excluded from analysis
			Perioperative haemodynamic measurements	T3 admin had no signif effect on any haemodynamic variable Significant increase in heart rate in Dopamine grp (p<0.001) 40% T3 vs. 20% Dopamine vs. 46% Placebo grp required pacing in first 6 hrs (p<0.001)
			Inotropic support requirements	53/66 (80.3%) T3 vs. 38/68 (55.9%) Dopamine vs. 45/71 (63.4%) Placebo pts required additional inotrope in first 6 hrs - not significant (although trend to decrease use in Dopamine grp)
			Other morbidity / mortality	6 pts in each grp required IABP - not significant No signif difference in time to extubation, time in ITU or time to hosp discharge between grps No signif difference in incidence of postop AF, ventricular dysrhythmias, stroke, MI or death between grps
Mullis-Jansson et al, 1999, USA	177 pts undergoing elective CABG surgery Excluded if thyroid disease present, if on thyroxine therapy, or if preop IABP required Pts randomised to receive either triodothyronine bolus+ infusion, or placebo at release of aortic cross-clamp for 6 hrs Standardised Anaesthetic / Surgical technique Standard criteria used for diagnosis of myocardial ischaemia 81/170 received T3 vs. 89/170 placebo	PRCT (level 1b)	Haemodynamic data	T3 grp cardiac indexes significantly higher than placebo grp in first 12hrs post-CPB (p=0.0001; multivariate longitudinal testing) – however this result not confirmed in female only cohort – male gender independent predictor of improved postop CI No signif difference in heart rate or SVR T3 vs. placebo grp in first 12hrs post-CPB	Male: female ~4:1 Confusion between parts of text e.g. different dosing regimes described in abstract and paper – 0.4mcg/kg bolus then 0.1mck/kg over 6hrs in abstract; 1mcg/kg bolus then 1mcg/kg over 6hrs in paper, and other disagreement / confusion between parts paper 2 pts from original 177 removed from study "at discretion of cardiologist" and excluded from analysis. No further information given Placebo grp mean age 4 yrs more than T3 grp (66 vs. 62.1; p=0.03) although all statistical analyses adjusted to compensate
			Inotrope / vasopressor use	5% T3 grp receiving Dobutamine on arrival in CITU vs. 9% placebo grp 11% T3 grp required Dopamine in first 48hrs vs. 18% placebo grp Inotrope requirements / dependence less in T3 than placebo grp at all time points (p=0.04) T3 grp had significantly decreased prevalence of Dopamine (p=0.01) but not Dobutamine (p=0.43) use when corrected for other influencing factors (age and LV function)
			Perioperative morbidity	3/81 (3.7%) T3 vs. 16/89 (18%) placebo grp developed myocardial ischaemia (p=0.0019). No signif difference in incidence of atrial fibrillation or myocardial infarction between groups 11/81 (13.6%) T3 vs. 22/89 (24.7%) placebo grp became pacemaker dependant (p=0.128) 0/81 T3 grp vs. 7/89 (7.9%) placebo grp required mechanical support – IABP x4; LVAD x3 (p=0.01)
			Mortality	Deaths – 0/81 T3 grp vs. 2/89 (2.2%) placebo grp (p=0.23)
Klemperer et al, 1996, USA	142 pts with EF<40% undergoing CABG surgery Excluded if > 85 years, for valve procedure, on Amiodarone or Thyroxine therapy, thyroid disease or preop Inotropes / IABP T3 therapy vs. placebo Active grp received IV triodothyronine 0.8mcg/kg at cross-clamp removal, then 0.113mcg/kg/hr for 6hrs, then tapered to off Pts monitored for first 5 days postop with bedside monitors, telemetry and daily 12-lead ECG's	PRCT (level 1b)	Development of arrhythmia in first 5 days postop	No significant differences in incidence of ST, SVT, PAC, PVC or VT in first 24hrs Significant differences in incidence of AF between groups in subsequent 96 hrs - 29 episodes of AF in 16/66 (24%) T3 pts vs. 69 episodes in 30/65 (46%) placebo pts (p=0.009 for incidence of AF; p=0.003 for episodes of AF) No signif difference in presence of AF at discharge	Same study population as in Klemperer's NEJM paper [1995] but extended to look specifically at postoperative dysrhythmia in first 5 postop days 89% T3 & 84% placebo pts received prophylactic Digoxin & 80% T3 & 79% placebo grp given Beta-blockade in first 48hrs - may be confounding factor 5/71 (7%) T3 & 6/71 (8.4%) placebo pts excluded from analysis due to preop AF
			Treatment interventions required	15/16 (94%) T3 vs. 28/30 (93%) placebo grp required treatment to control rate, convert rhythm, or to anticoagulate (p=0.02) 6/15 T3 vs. 16/28 on Procainamide at discharge (p=0.019) 2/15 T3 vs. 10/28 placebo required in-hospital anticoagulation (p=0.013) 0/15 T3 vs. 6/28 placebo required Cardioversion (p=0.012)
Cimochowski et al, 1997, USA	111 consecutive pts with severe LV dysfunction Combination of metabolic & mechanical support used to optimise LV function post-CPB Metabolic support - Triiodothyronine 0.4mcg/kg post-induction + 0.8mcg/kg 10 mins before end of CPB + 0.4mcg/kg/24hrs given the next day - Glucose / Insulin Potassium given concomitantly with T3 - Aspartate / Glutamate cardioplegia - Warm-cold-warm / antegrade-retrograde-antegrade cardioplegia Mechanical support - liberal IABP usage - ultrafiltration to remove myocardial depressants - occlusive retrograde cardioplegia catheter - delayed sternal closure Standard surgical / CPM / inotrope management	Prospective cohort (level 2b)	Comparisons of pt data & actual outcomes to those on Society of Thoracic Surgeons (STS) database over same period	Mean EF 27.9+/-5.4% cohort vs. 26.3+/-7.2% STS database - ns 12/111 (10.8%) had acute or chronic renal failure & 12/111 (10.8%) emergency procedures (not statistically different to STS group) 56/111 (50.4%) cohort vs. 15903/2666 (59.6%) STS group required inotropes (p=0.04). No signif difference in IABP usage 1/111 (0.9%) vs. 3964/26661 STS group had stroke (p=0.00) 0/111 (0%) cohort vs. 1200/26661 (4.5%) STS group developed new renal failure (p=0.02) 2/111 (1.8%) cohort developed mediastinitis and 4/111 (3.6%) re-explored for bleeding (not significantly different from STS group) No complications related to interventions ITU stay 2.2 +/-0.9 days. Hospital stay 13.7+/-22.1 days cohort vs. 10.2+/-16.7 days STS group ITU stay (p=0.02) 2/11 (1.8%) cohort vs. 2035/26661 (7.6%) STS group operative mortality (p=0.02) 1/108 (0.9%) cohort vs. 1557/23207 (6.7%) STS operative mortality without reoperation (p=0.01)	Male: female ~4:1 Significant differences between cohort & STS subjects for number of factors therefore groups not comparable Study not designed specifically to look at effects of T3 - instead designed to study a combination of factors incl thyroxine therapy. Therefore cannot state that improvements due to T3 use alone
Teiger et al, 1993, France	20 clinically euthyroid pts undergoing CPB T3 or placebo given at removal of aortic cross-clamp Pts excluded with dysrhythmias or if on Thyroxine or Amiodarone T3 protocol - 0.20 mcg/kg @ cross-clamp removal (time 0); 0.15 mcg/kg @ +4hrs; 0.10mcg/kg @ +8hrs; 0.05 mcg/kg @ +12hrs; & 0.05mcg/kg @ +20hrs Standardised CPB technique 10/20 (50%) pts received T3 vs. 10/20 (50%) placebo	PRCT (level 1b)	Thyroid hormone levels	Post-CPB TT3 signif decrease compared to preop levels (p<0.001). Duration of TT3 fall longer in placebo grp fT3 signif decrease at aortic cross-clamp removal vs. preop in treated vs. control grps (p<0.05) but not significant if corrected for haemodilution fT3 signif different for treated vs. control grp at 48hrs sampling time when corrected for haemodilution (p<0.05) fT3 levels at 48hrs in placebo grp lower than preop levels but not statistically significant Transient increase in fT4 levels at cross-clamp removal in both treated & control grps (p<0.05 compared to respective preop values) No other significant differences	Small study size Non-standardised surgical procedure Same dosing regime as used by Novitzky et al in part of previous study [1989 (2)] Bmax & Kd of â-adrenoreceptors data on 4 active & 4 placebo pts only Unable to compare Bmax data for T3 grp vs. placebo grp [3]
			Haemodynamic parameters	No evidence of T3 induced tachycardia No significant difference in PCWP, MAP, LVSWI between groups at any time point
			Density (Bmax) & affinity (Kd) of â-adrenoreceptors (subgroup of 4 active & 4 placebo pts)	No signif difference in Bmax values preop vs. postop for either T3 or placebo group No significant difference in Kd in T3-treated grp No signif difference in Kd values preop vs. postop for either active or placebo group No adverse events related to T3 therapy
Novitzky et al 1989(2) South Africa	10 patients with difficulty weaning for CPB (5) or with poor LV function post-CPB (5) despite inotrope or IABP support IV T3 dose 4-10mcg	Case series (level 4)	T3 levels	T3 levels decrease from mean 1.03mcg/ml to 3.56mcg/ml within 1hr of T3 admin 5/5 (100%) CPB-dependant cases off bypass after 1 hour	Small case-series No control group No overall outcome data (mortality, length of stay etc)
			Clinical outcome	2 IABP dependant pts no longer dependant after 3hrs Significant increases in MAP & heart rate, & decreases in LAP & CVP at 1hr post-T3 admin (p<0.0001) with significant decrease / discontinuation of inotropes
Novitzky et al, 1996, USA	68 high risk patients undergoing cardiac surgery (26.5% preop IABP 14.7% preop inotropes) 32 CABG, 9 valve, 27 combined & one heart transplant procedure 29 elective; 7 urgent & 32 emergent procedures Four T3 regimes used - 2-3mcg/kg at cross-clamp release - 1-2mcg/kg at cross-clamp release whenever pt became CPB-dependant - 1mcg/kg postop to pts with low cardiac output despite high IABP support - 8/68 (11.8%) given 1mcg/kg pre-CPB then 1mcg/kg at cross clamp release - all received 0.5mcg/kg/24hr for 1-3 days postop Standard periop / CPB management	Case series (level 4)	Comparison of mortality to predicted (using New Jersey Risk Stratification model)	Preoperative mean expected mortality (EM) 29.67% (range 3-58%, New Jersey Risk Assessment) 56/68 (82.4%) weaned from CPB 12/68 (17.6%) remained CPB dependant 44/68 (64.7%) given T3 at cross-clamp release 12/68 (17.6%) given T3 at cross-clamp release whenever pt became CPB-dependant 4/68 (5.9%) given T3 postop for low cardiac output despite high IABP support 8/68 (11.8%) given T3 pre-CPB then at cross clamp release Predicted mortality for overall group 26/68 (38.2%). Actual mortality 7/68 (10.3%). P<0.007 4 cardiac; 3 non-cardiac deaths Predicted mortality from CPB-dependant group 8+/12 (>67%); actual mortality 1/12 (8.3%)	64 male: 4 female No justification of doses chosen Variable dosing regimes make comparison difficult No controls
			Free T3 (FT3) levels	Significant decrease FT3 levels between preop (2.01+/-0.01pg/ml) & pre-T3 (0.8+/-0.07pg/ml) admin time points (p<0.0001) FT3 levels elevated for 2hrs
Malik et al, 1999, USA	10 consecutive patients with severe systolic failure / cardiogenic shock awaiting cardiac transplantation Pts unresponsive to conventional pharmacological / IABP support IV levothyroxine 20mcg bolus, then 10mcg/hr titrated to BP (max 20mcg/hr)	Case series (level 4)	Thyroid function tests	10/10 had normal TSH levels. 7/10 had normal Throxine levels. 3/10 had low Thyroxine levels	Small study Male : female 4:1 Inhomogeneous group of patients (5 dilated cardiomyopathies, one valvular heart disease; 4 undefined pathologies) Uncontrolled Dose derived from that recommended for myxoedema coma
			Haemodynamic parameters in first 36hrs post-admin - RAP - PCWP- MAP- mean PAP- cardiac index (CI)	Significant haemodynamic changes after 24hrs infusion - CI increase from 1.75+/-0.23 to 2.3+/-0.28 l/min/m2 after 24hrs (p=0.0004) & 2.44+/-0.28l/min/m2 after 36 hrs (p<0.05) - MAP increase from 65+/-7 to 74+/-6 mmHg after 36hrs (p<0.05) - Mean PAP decrease from 59+/-6 to 48+/-5.6mmHg after 36hrs (p<0.05) - PCWP decrease from 29.3+/-2.2 to 24.2+/-4 after 24hrs (p=0.001) & 20.8+/-2.3mmHg after 36hrs (p<0.05) - no signif differences in SVR, RAP and heart rate
			Success at bridging to transplant / LVAD placement	Inotropes decrease in 33% pts & not escalated in others 9/10 (90%) pts showed sustained improvement - 5/10 (50%) LVAD - 4/10 (40%) transplanted Transplanted pts - weaned from Thyroxine in 2+/-1 day LVAD pts - weaned from Thyroxine in 4+/-2 days
			Long-term outcomes to 6 & 12 months	4/5 (80%) LVAD pts transplanted 38+/-8 days post cessation of thyroxine therapy 9/10 (90%) cohort survived to 6 months 8/10 (80%) cohort survived to 12 months
Carrel et al, 2002, Switzerland	54 adults & 7 paediatric pts with severe low cardiac output syndrome 8 pts awaiting cardiac transplant - given T3 2-4mcg/kg over 1hr then 2-5mcg/hr for 60+/-12hrs 32 brain-dead multi-organ donors in ITU - T3 1-2mcg/hr + Hydrocortisone leading up to organ procurement 5/60 pts with immediate allograft dysfunction - LVAD insertion vs. T3 admin (2mcg bolus + 0.6mcg/hr for 12hrs) 11/1780 pts with severe compromised post-CPB haemodynamics - T3 2-3mcg bolus + 0.2-0.6mcg/hr for 6-12 hrs 7/495 neonates / children with low output states following surgical correction of congenital disease utilising CPB - T3 0.5mcg bolus + 0.1mcg/hr for 24-48hrs	Case series (level 4)	Haemodynamic function	8/8 Pts awaiting cardiac transplant - decreased inotrope requirements. No deaths Brain dead group - 20/32 on moderate-high dose inotrope pre-T3 - Inotropes significantly decreased in 2/20 (10%) & stopped in 12/20 (60%) post T3 admin Allograft dysfunction grp- 2/5 had LVAD inserted - both died - 3/5 given T3 - survived Compromised post-CPB haemodynamic grp -11/11 had inotropes + IABP - 2/11 died - low cardiac output - 5/11 had inotropes decrease within 48hrs - 2/11- IABP remained for 48-72hrs - 2/11- no information Paediatric group - 2/7 died - 5/7 improved over 48-96hrs & recovered well	Lack of a clear homogeneous patient type, even within individual groups Many pts merely parts of other un-controlled case series No clear management protocol / uniformity of thyroxine dosages No information as to why particular doses given Steroids may confound data in "brain-dead" group No formal statistical analyses of data

Comment(s)

It is well recognised that decreased thyroxine hormone levels accompanying cardiopulmonary bypass (CPB) can be reversed by thyroxine or triiodothyronine (T3) administration before or during bypass [Sirlak, Novitzky (1989,2), Klemperer (1995), Guden, Bennett-Guerrero ]. But whilst some studies have shown that this strategy can improve cardiac output there is little evidence that recovery is enhanced and complications decreased. Several authors have identified some short-term haemodynamic benefits following T3 or thyroxine administration during cardiac surgery. Sirlak et al gave oral T3 both preoperatively and postoperatively and found significant differences in cardiac index (CI) and systemic vascular resistance (SVR) with decreased inotrope and IABP requirements. But whilst ITU stay was less, there was no significant difference in myocardial ischaemia / infarction rate, atrial fibrillation or death. Novitsky (1989,2) et al found T3 administration on bypass decreased inotrope and diuretic requirement in patients with ejection fraction (EF) less than 30% although there was no difference in haemodynamic data. Their T3 regime was subsequently modified on the basis of data from this study and repeated in patients with higher ejection fractions. Whilst there was no difference in inotrope or diuretic use in this latter group, cardiac output (CO) was significantly higher and SVR and pulmonary vascular resistance (PVR) lower in their T3-treated group. Mullis-Jansson et al reported similar improvements in CI in their male cohort following T3 administration at cross-clamp release although both sexes experienced decreased myocardial ischaemia, inotrope dependence / requirement and need for mechanical support. However their control group was significantly older than their T3-group possibly confounding some of their findings. Finally Klemperer (1995) et al noted that T3 administration in patients with impaired LV function significantly increased CI and decreased SVR for the first 6hrs post-cross clamp removal but found no significant differences in interventions required to support separation from CPB, postoperative inotrope use, duration of postoperative ventilation, ITU or hospital stay, or overall morbidity and mortality. However whilst this study demonstrated comparable incidences of supraventricular and ventricular dysrhythmias in the first 18 hours a subsequent paper by the same authors reported decreased prevalence and incidence of atrial fibrillation (AF) in the T3-treated cohort monitored beyond 24 hours. Significantly fewer required rate and rhythm control, or anticoagulation therapy although there was no significant difference in presence of AF at discharge. Cimochowski et al administered T3 as part of a complex combination of metabolic and mechanical support strategies in patients with severe LV dysfunction. Inotrope requirement, operative mortality, stroke and renal failure were significantly lower in their intervention group although it is clearly impossible to separate T3-effects from other interventions. Other studies have failed to demonstrate improvements in cardiac output let alone other outcome markers. Guden et al found no significant differences in post-CPB haemodynamic parameters, dysrhythmia incidence, need for inotropic support, ITU stay and morbidity or mortality in patients receiving T3. Bennett-Guerrero et al could demonstrate no beneficial effects of T3 on perioperative haemodynamics compared to Dopamine or placebo in patients at high risk of requiring post-CPB inotrope support. Requirement for pacing in the first 6 hours was significantly greater in the T3 and placebo patients but requirement for IABP or inotropes between groups was not significant. There was no difference in time to extubation, time in ITU or to hospital discharge, or other morbidity and mortality between groups. Finally, Teiger et al were unable to demonstrate significant differences in haemodynamic parameters in a small group given T3 at cross-clamp removal. A number of different doses of T3 or Thyroxine are described. Whilst some authors justify their dose rationale [Sirlak, Novitzky (1989, 2), Bennett-Guerrero, Teiger, Malik], others do not [Klemperer (1995), Guden, Mullis-Jansson, Novitzky 1996, Carrel] and it may well be that the optimal dosing regime has not been described. It may also be noteworthy that most studies have been performed in patients with LV impairment [Sirlak, Novitzky (1989,2), Klemperer (1995), Bennett-Guerrero, Klemperer 1996, Cimochowski, Teiger, Novitzky (1989,1), Malik, Carrel]. One study has reported sex-dependant haemodynamic responses to T3 although outcome measurements were similar [Mullis-Jansson]. T3 has also been described as rescue therapy in various case series involving high risk cases undergoing cardiac surgery. In 1989 Novitzky(1) et al gave T3 to 10 patients difficult to wean from CPB and demonstrated significant improvements in cardiac function with decreased IABP or inotrope support. The same author subsequently reported another 68 cases given various T3 doses with similar decreases in mortality compared to predicted not only in their overall group but also in a subset of 12 patients (1989, 2)remaining "unexpectedly" CPB-dependant. Malik et al successfully bridged 9 out of 10 patients with severe cardiogenic shock to LVAD or heart transplantation using IV Thyroxine. Finally Carrel et al reported various T3 regimes in post-CPB, brain dead or pre-transplant patients showing apparent improvements in cardiac function. However their inhomogeneous series lacked consistent management protocols / uniformity of dosage regimes and their data was not subjected to statistical analysis.

Clinical Bottom Line

Thyroxine levels decrease significantly during Cardiopulmonary bypass. However there is conflicting evidence that prophylactic perioperative thyroxine / triiodothyronine is a useful adjunct in adult patients undergoing cardiac surgery. Whilst some papers report improved haemodynamic parameters in the immediate post-bypass period, there is no standardised dose or evidence that its use decreases postoperative morbidity, mortality or length of stay in the routine patient. It may however still be useful as rescue therapy in weaning some high risk cases from CPB or bridging them to LVAD or transplant.

References

1. Sirlak M. Yazicioglu L. Inan MB. Eryilmaz S. Tasoz R. Aral A. Ozyurda U. Oral thyroid hormone pretreatment in left ventricular dysfunction. European Journal of Cardio-Thoracic Surgery 2004: 26;720-5.
2. Novitzky D. Cooper DK. Barton CI. Greer A. Chaffin J. Grim J. Zuhdi N. Triiodothyronine as an inotropic agent after open heart surgery. Journal of Thoracic & Cardiovascular Surgery. 1989: 98;972-7.
3. Klemperer JD. Klein I. Gomez M. Helm RE. Ojamaa K. Thomas SJ. Isom OW. Krieger K. Thyroid hormone treatment after coronary-artery bypass surgery. New England Journal of Medicine 1995: 333;1522-7.
4. Guden M. Akpinar B. Saggbas E. Sanisoglu I. Cakali E. Bayindir O. Effects of intravenous triiodothyronine during coronary artery bypass surgery. Asian Cardiovascular & Thoracic Annals 2002:10;219-22.
5. Bennett-Guerrero E. Jimenez JL. White WD. D'Amico EB. Baldwin BI. Schwinn DA. Cardiovascular effects of intravenous triiodothyronine in patients undergoing coronary artery bypass graft surgery. A randomized, double-blind, placebo- controlled trial. Duke T3 study group JAMA 1996:275;687-9.
6. Mullis-Jansson SL. Argenziano M. Corwin S. Homma S. Weinberg AD. Williams M. Rose EA. Smith CR. A randomized double-blind study of the effect of triiodothyronine on cardiac function and morbidity after coronary bypass surgery. Journal of Thoracic & Cardiovascular Surgery 1999:117;1128-34.
7. Klemperer JD. Klein IL. Ojamaa K. Helm RE. Gomez M. Isom OW. Krieger KH. Triiodothyronine therapy lowers the incidence of atrial fibrillation after cardiac operations. Annals of Thoracic Surgery 1996:61;1323-7.
8. Cimochowski GE. Harostock MD. Foldes PJ. Minimal operative mortality in patients undergoing coronary artery bypass with significant left ventricular dysfunction by maximization of metabolic and mechanical support. Journal of Thoracic & Cardiovascular Surgery 1997:113;655-64.
9. Teiger E, Menasche P, Mansier P, Chevalier B, Lajeunie E, Bloch G, Piwnica A. Triiodothyronine therapy in open-heart surgery: from hope to disappointment. European Heart Journal 1993;14:629-33.
10. Novitzky D, Cooper DKC, Swanepoel A. Inotropic effect of triiodothyronine (T3) in low cardiac output following cardioplegic arrest and cardiopulmonary bypass: an initial experience in patients undergoing open heart surgery. European Journal of Cardio-Thoracic Surgery 1989;3:140-145.
11. Novitzky D. Fontanet H. Snyder M. Coblio N. Smith D. Parsonnet V. Impact of triiodothyronine on the survival of high-risk patients undergoing open heart surgery. Cardiology 1996:87;509-15.
12. Malik FS. Mehra MR. Uber PA. Park MH. Scott RL. Van Meter CH. Intravenous thyroid hormone supplementation in heart failure with cardiogenic shock. Journal of Cardiac Failure 1999;5:31-37.
13. Carrel T. Eckstein F. Englberger L. Mury R. Mohacsi P. Thyronin treatment in adult and pediatric heart surgery: Clinical experience and review of the literature. European Journal of Heart Failure 2002:4;577-582.