Economy & Finance

Bayesian networks for pathological scarring due to burn injuries

Description
Bayesian networks for pathological scarring due to burn injuries
Published
of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Low triiodothyronine serum levels as a predictor of poorprognosis in burn patients § Ezio Nicola Gangemi a,1 , Francesca Garino b,1 , Paola Berchialla c  , Morena Martinese b ,Federico Arecco b , Fabio Orlandi b , Maurizio Stella a, *  a Department of Plastic and Reconstructive Surgery–Burn Center, Traumatological Hospital, Turin, Italy b Section of Endocrinology, Department of Clinical and Biological Sciences, University of Turin, Italy c Department of Public Health and Microbiology, University of Turin, Turin, Italy 1. Introduction Extensiveburnsleadtoaseveresystemicmetabolicresponseconsisting in an early ‘‘ebb’’ phase and a later ‘‘flow’’phase.The ‘‘flow’’ phase is characterized by hypercatabolism,lastingat least9 months[1]. Thisconditionisthoughttobe an adaptive mechanism promoting survival evoked by theendocrine system [2]. In fact the ‘‘flow’’ phase is based on activation of gluconeogenesis and protein degradation, which burns 34 (2008) 817–824 a r t i c l e i n f o  Article history: Accepted 23 October 2007 Keywords: BurnsEuthyroid sick syndromeTriiodothyroninePrognostic index a b s t r a c t Objective:  Euthyroid sick syndrome is a common finding in critically ill patients withnonthyroidal illness, characterized by low serum levels of free triiodothyronine (fT3) witha peculiar increase in reverse T3 (rT3) and normal-to-low free thyroxine (fT4) as well asthyroid-stimulatinghormone(TSH)levels.Thisconditionhasbeenproposedasaprognosticfactor of worse outcome in critically ill patients, while no conclusive data are available inburns. Methods:  Since thyroid function testing is contained in our baseline laboratory tests atadmission, we retrospectively evaluated fT3, fT4 and TSH in 295 consecutive burn patientsadmitted to the Burn Center of Turin from January 2002 to December 2006, comparing hormone levels in survivors and non-survivors. Results:  fT3 and TSH levels were significantly lower (  p  0.0002) in non-survivors comparedto survivors, while no significant difference between the two populations was found in fT4concentrations. Excluding from the study 20 patients who received dopamine administra-tion for more than 21 h, serum fT3 levels fell further still (  p  = 0.0003). In addition, fT3concentrations showed a significant correlation with burn severity expressed by the Roiscore (Spearman’s correlation coefficient   0.37 with  p < 0.00001). Conclusion:  Low fT3 levels are associated with poor outcome in burn patients. Hence, fT3measurement could be proposed as a strong and cost-effective tool of poor prognosis. # 2007 Elsevier Ltd and ISBI. All rights reserved. § The paper was partially presented on September 15, 2007 at the 12th Congress of the European Burns Association (EBA) held inBudapest (Hungary) and on June 15, 2007 at the 32nd Congress of Italian Society of Endocrinology held in Verona (Italy).*  Corresponding author at : Department of Plastic and Reconstructive Surgery–Burn Center, C.T.O. Hospital, Via Zuretti 29, 10126 Turin,Italy. Tel.: +39 011 6933435; fax: +39 011 6933425.E-mail address: Stella.Maurizio@libero.it (M. Stella). 1 These authors made equal contribution to this work and should be considered first authors. available at www.sciencedirect.comjournal homepage: www.elsevier.com/locate/burns 0305-4179/$34.00 # 2007 Elsevier Ltd and ISBI. All rights reserved.doi:10.1016/j.burns.2007.10.002  may function to provide the metabolic substrates forinflammation, host defence mechanisms and wound healing processes [2]. In this context,pro-inflammatorycytokines and high levels of stress-induced hormones, such as glucagon,cortisol and catecholamines, may play a major role inmaintaining the catabolic state [3], while thyroid hormone setting seems to be irrelevant [4].The ‘‘ebb’’ phase is an acute reaction lasting 2–3 days andrepresentsaresponsewithimpairedcardiacoutputalongwithadropinbasalmetabolicrate.Thismetabolicsettingissimilarto that occurring in starvation, in which low-T3 serum levelsaresomehowinvolved.Aspreviouslyreported,theoccurrenceof any severe injury or chronic illness may result in aconsistent decrease in thyroid function, configuring theclinical feature of euthyroid sick syndrome (ESS) [2].ESS is a clinical condition characterized by low serumlevelsoffreetriiodothyronine(fT3)withapeculiarincreaseof reverse-T3 (rT3) and normal-to-low serum concentrations of free thyroxine (fT4), as well as of thyroid-stimulating hormone (TSH). Such thyroid hormone changes have beeninterpreted as an adaptive compensatory response to reduceenergy expenditure due to the hyperstimulation of thesympathetic nervous system [4–5] or, alternatively, as aninsufficientneuroendocrineadaptationto stress,resultinginan adverse effect on clinical outcome. In this last case, theoccurrence of the ESS has been proposed as a negativeprognostic factor, showing an association with a poorprognosis for survival in different severe diseases [6–18].However, no conclusive data about low-T3 concentrationsand worst outcome of burn patients are available to date(Table 1).The aim of this study was to investigate serum thyroidhormone in a large population of burned patients admittedto the Turin Burn Center, in order to verify whether thisclinical condition resulted in some fT3, fT4 and TSHalterations, and if measurement of these variables repre-sented a good index of poor prognosis in burn patients. 2. Patients and methods 2.1. Patients FT3,fT4andTSHcirculatingconcentrationsof314consecutiveburn patients (107 women; 207 men; aged as median 55 years)admitted to the Turin Burn Center from January 2002 toDecember 2006, were retrospectively assessed, since thyroidfunction setting is routinely included in our baseline labora-tory tests. Burn Center referral criteria follow an internalprotocol built according to the guidelines of the AmericanBurnAssociation.Atotalof19patientswereexcludedfromthestudy because of pregnancy in 3 patients and previouslydocumented thyroid disease in 16 patients (2 thyrotoxicosisand 14 autoimmune or iatrogenic hypothyroidisms in L-T4replacement therapy). Out of the 295 remaining patients, 238were survivors, 57 patients were non-survivors. Clinical andpathological features for each patient are summarized inTable 2, including Roi score, as index of illness severity [19]. 2.2. Blood sampling Blood sampleswerecollected from1 to 3 daysafteradmissionin the Burn Center. On the basis of the elapsed time from theburn to the serum sampling, surviving and non-surviving patients were further subdivided into two groups ( < 48 h and > 48 h). FT3, fT4 and TSH were measured by a completelyautomatedAIA600 system(TosohCorporation, Tokyo,Japan).The reference intervals, according to the manufacturer’sinstructions, were as follows: fT3, 3.1–7.5 pmol/L; fT4, 10.6–21.0 pmol/L and TSH, 0.41–4.01 mIU/L.In order to evaluate possible interferences on thyroidhormone secretion, patients were also clustered on the basisof dopamine administration. If dopamine was not used or itwas infused after the laboratory test then we considered thatcase as ‘‘Non-Dopamine-Treated’’. On the contrary, the‘‘Dopamine-Treated’’ cases were stratified, according tothe data of the literature, on the basis of the time from thebeginning of dopamine infusion to the blood sampling, ( < 4 h,4–21 h,  > 21 h) [16,20,21]. The thyroid function setting of thesubdivided groups are shown in Table 3. 2.3. Statistical analysis Continuous variables describing patients were expressed asmedian with interquartile range. A nonparametric test forunpaired-samples (Mann–Whitney test) for continuous vari-ables and a Pearson’s Chi-square test for categorical variableswere used for preliminary correlations. The median with 95%confidence interval (CI) was calculated for fT3, fT4 and TSHand direct comparisons were performed with the Mann–Whitney test. A sequential multivariate modelling strategywas conducted whereby crude association estimates of fT3and TSH levels to the outcome were initially adjusted fordemographic information only (gender and age), thenadjusted for total burn surface area burnt [TBSA], presence/absence of comorbidities and inhalation injury.Correlation between the variables and Roi score wasassessed by Spearman’s correlation test. Two-sided  p -valuesless than 0.05 were considered statistically significant. Table 1 – Review of clinical studies considering thethyroid function serum parameters as statistical signifi-cant predictors of poor prognosis Reference Issue  n  Predictor [6] Respiratory failure 32 T3[7] Septic shock 20 TSH[8] Heart failure 84 fT3 index/rT3 ratio[9] Critical illness 206 T3[10] Critical illness 86 T4[11] Complex trauma 20 T3[12] Critical illness 42 T3, T4[13] Bone marrowtransplantation78 T3,T4[14] Liver cirrhosis 28 T3, TSH[15] Emergency operation 66 T3[16] Meningococcal sepsis 69 T4[17] Heart failure 573 T3[18] Critical illness 451 rT3, T3/rT3 ratio n : number of patients considered; T3: triiodothyronine; T4:thyroxine; TSH: thyroid-stimulating hormone; rT3: reverse T3. burns 34 (2008) 817–824 818  AnalyseswereperformedusingAnalyse-it TM SoftwareLtd.,Version 1.73 (Leeds, UK).The study was conducted according to the principlesestablished in Helsinki. Written informed consent wasobtained and the work was approved by the C.T.O. Hospitalreview board. 3. Results The study was composed of 238 survivors and 57 non-survivors. As shown in Table 2 the two groups significantlydiffer each other by age, Roi index, total burn surface area,presence of inhalation injury and comorbidities. In the sameway serum fT3 levels were significantly lower (  p  = 0.0002) innon-survivors (median 2.71 pmol/L, 95% CI 2.36–3.08) than insurvivors (median 3.30 pmol/L, 95% CI 3.12–3.45). Also serumTSHconcentrationsresultedsignificantlylower(  p < 0.0001)innon-survivors(median0.61 mIU/L,95%CI0.39–0.80)comparedwith survivors (median 1.32 mIU/L, 95% CI 1.05–1.51). On thecontrary, no significant difference for serum fT4 levels wasobserved between the two groups.The statistical significance of the difference in fT3 and TSHlevelsbetween the two groupsofpatientsdidnot substantiallychange when we divided the study population in differentsubgroups on the basis of the time from injury to the bloodsamplecollection( < 48 hand > 48 h)(Table4).ThedifferencesinFT3 serum levels and TSH concentrations among the twogroups showed similar statistical significance when adjustedfordemographicvariables,aswellascorrectedforcomorbidityand inhalation injury. However the association was no longerstatistically significant after adjustment for TBSA (Table 5). Table 2 – Clinical and pathological features of the study population n  Survivors ( n , 238) Non-survivors ( n , 57) Combined ( n , 295)  p Gender, % ( n ) 295 0.744Male 69% (164) 67% (38) 68% (202)Female 31% (74) 33% (19) 32% (93)Age, median (IQR) (years) 295 52 (36–71) 70 (50–82) 55 (38–74)  <0.001 Roi, median (IQR) 295 0.09 (0.01–0.28) 0.85 (0.58–0.97) 0.15 (0.02–0.45)  <0.001 TBSA a , median (IQR) (%) 295 15 (10–20) 40 (20–65) 16 (10–28)  <0.001 Inhalation injury, % ( n ) 295  <0.001 Yes 29% (69) 74% (42) 38% (111)No 71% (169) 26% (15) 62% (184)Comorbidities, % ( n ) 295  0.04 Yes 36% (86) 51% (29) 39% (115)No 64% (152) 49% (28) 61% (180)Aetiology, % ( n ) 293 0.479Chemical 1% (2) 0% (0) 1% (2)Contact 0% (1) 0% (0) 0% (1)Flame 74% (175) 88% (50) 76% (225)Scald 11% (27) 4% (2) 10% (29)Pressure 1% (2) 0% (0) 1% (2)Flash 11% (26) 7% (4) 10% (30)Steam 1% (3) 2% (1) 1% (4)fT3, median (IQR) (pmol/L) 295 3.3 (2.72–3.8) 2.71 (2.02–3.28) 3.2 (2.51–3.75)  <0.001 fT4, median (IQR) (pmol/L) 295 13.5 (11.41–15.63) 13.02 (10.58–16.08) 13.4 (11.24–15.68)  0.396 TSH, median (IQR) (mIU/L) 295 1.33 (0.63–2.22) 0.61 (0.29–1.31) 1.15 (0.52–2.07)  <0.001 Continuous variables are expressed as median with interquartile range (IQR). Categorical variables are expressed as frequencies.  p  value refersto the nonparametric unpaired-samples Mann–Whitney test for continuous variables and to Pearson’s Chi-square test for categoricalvariables.  P < 0.05 is considered statistically significant. The variable comorbidities has to be intended for at least one associated illness.  n :number of patients considered; TBSA: total burn surface area; fT3: free triiodothyronine; fT4: free thyroxine; TSH: thyroid-stimulating hormone. a Evaluated after 48 h from the burn trauma. Table 3 – Thyroid function features in function of the time from the beginning of dopamine administration to the bloodsample collection n  fT3 (pmol/L) fT4 (pmol/L) TSH (mIU/L) Non-dopamine-treated 215 3.38 (2.85–3.87) 13.76 (11.79–15.80) 1.54 (0.85–2.49)Dopamine-treated 80 2.36 (1.81–3.10) 11.7 (8.96–14.5) 0.35 (0.23–0.67) < 4 h 19 3.18 (2.72–3.96) 13.7 (11.47–16.33) 0.46 (0.3–1.37)4–21 h 41 2.33 (1.87–3.02) 12.01 (9.39–14.37) 0.4 (0.28–0.67) > 21 h 20 1.69 (1.48–2.18) 8.82 (7.34–10.87) 0.18 (0.13–0.26)All data are expressed as median with interquartile range (IQR).  n : number of patients considered; fT3: free triiodothyronine; fT4: freethyroxine; TSH: thyroid-stimulating hormone. burns 34 (2008) 817–824  819  In addition, a highly significant correlation was observedbetween fT3 and TSH concentrations, and the parameter of burn severity expressed by the Roi score (Spearman’scorrelation coefficient   0.37 with  p < 0.0001 for fT3 and  0.25 with  p < 0.0001 for TSH) (Fig. 1A and B).As dopamine infusion is concerned, no significant differ-ence was found in non-survivor fT3 levels between patientsreceiving dopamine administration for less than 21 h and thedopamineuntreatedpatients(  p  = 0.1051).Onthecontrary,theprolonged dopamine infusion time ( > 21 h) resulted in a Table 4 – Thyroid hormone levels in survivors and non-survivors in function of the time from burn trauma to the bloodsample collection n  Survivors Non-survivors  p Total study group 295  n , 238  n , 57fT3 (pmol/L) 3.30 (3.12–3.45) 2.71 (2.36–3.08)  0.0002 fT4 (pmol/L) 13.50 (12.95–14.00) 13.02 (11.90–14.21) 0.3960TSH (mIU/L) 1.32 (1.05–1.51) 0.61 (0.39–0.80)  <0.0001 Group with TBC < 48 h 192  n , 148  n , 44fT3 (pmol/L) 3.22 (3.03–3.37) 2.86 (2.45–3.20)  0.0168 TSH (mIU/L) 1.03 (0.90–1.30) 0.60 (0.39–0.72)  0.0013 Group with TBC > 48 h 103  n , 90  n , 13fT3 (pmol/L) 3.46 (3.16–3.57) 2.36 (1.86–2.90)  0.0013 TSH (mIU/L) 1.92 (1.48–2.26) 0.74 (0.18–2.01)  0.0227 All data are expressed as median with 95% confidence interval.  p  value refers to the nonparametric unpaired-samples Mann–Whitney test. P < 0.05 is considered statistically significant.  n : number of patients considered; fT3: free triiodothyronine; fT4: free thyroxine; TSH: thyroid-stimulating hormone; TBC: time from the burn trauma to the blood sample collection. Table 5 – Sequential multivariate modelling Effect a  p  Effect b  p  Effect c  p fT3 0.51 (0.34–0.76) 0.001 0.57 (0.37–0.87) 0.0099 1.1 (0.63-1.93)  > 0.05fT3 d 0.52 (0.35–0.76) 0.0008 0.55 (0.37–0.82) 0.0036 0.96 (0.56–1.64)  > 0.05TSH 0.38 (0.22–0.66) 0.0006 0.43 (0.24–0.78) 0.005 0.69 (0.38–1.28)  > 0.05All data are expressed as odds ratio (95% confidence interval). fT3: free triiodothyronine; TSH: thyroid-stimulating hormone. a Gender and age adjusted odds ratio. b Gender, age, inhalation injury and comorbidities adjusted odds ratio. c Gender, age, inhalation injury, comorbidities and TBSA adjusted odds ratio. d fT3 analysis independently from influence of dopamine Infusion. Table 6 – Analysis of fT3 and TSH in survivors and non-survivors in function of dopamine infusion n  Dopamine-treated Non-dopamine-treated  p Non-survivorsfT3 (pmol/L) 51 2.61 (2.02–3.08) a 3.00 (2.45–3.44) 0.1051fT3 (pmol/L) 26 2.15 (1.22–3.08) b 3.00 (2.45–3.44)  0.0207 SurvivorsfT3 (pmol/L) 224 2.81 (2.26–3.25) a 3.45 (3.30–3.55)  0.0012 fT3 (pmol/L) 209 1.52 (1.43–2.20) b 3.45 (3.30–3.55)  <0.0001 Non-survivorsTSH (mIU/L) 28 0.57 (0.09–2.23)c 1.69 (1.11–2.49)  0.0419 TSH (mIU/L) 49 0.31 (0.28–0.58)d 1.69 (1.11–2.49)  <0.0001 SurvivorsTSH (mIU/L) 206 0.42 (0.23–1.99) c 1.54 (1.36–1.79)  0.0049 TSH (mIU/L) 227 0.27 (0.23–0.55) d 1.54 (1.36–1.79)  <0.0001 All data are expressed as median with 95% confidence interval.  p  value refers to the nonparametric unpaired-samples Mann–Whitney test. P < 0.05 is considered statistically significant.  n : number of patients; fT3: free triiodothyronine; TSH: thyroid-stimulating hormone. a Dopamine infusion  < 21 h. b Dopamine infusion  > 21 h. c Dopamine infusion  < 4 h. d Dopamine infusion  > 4 h. burns 34 (2008) 817–824 820  significant decrease of fT3 values in both non-survivors(  p  = 0.0207) and survivors (  p < 0.0001) (Table 6).ConcerningTSHlevels,asignificantdecreasewasobservedin the dopamine-treated group, independently from theinfusion time (less or more than 4 h) (Table 6).In order to avoid any possible dopamine effect, for the fT3analysis we excluded the 20 patients treated with dopamineinfusion for more than 21 h. In the remaining 275 patients fT3concentrations were still lower in non-survivors than insurvivors (median 2.86 pmol/L vs. 3.33 pmol/L,  p  = 0.0003)(Fig. 2 A and Table 7), maintaining statistical significance in the multivariate modelling (Table 5) and in correlation withRoi score (Spearman’s correlation coefficient   0.35 with  p < 0.0001) (Fig. 1 C).On the other hand, excluding all the 80 dopamine-treatedpatientsfortheTSHanalysis,intheremaining215patientswedid not confirm any significant difference in TSH concentra-tions between survivors and non-survivors (median 1.69 mIU/L vs. 1.54 mIU/L,  p  = 0.9774) (Fig. 2 B and Table 7). 4. Discussion Mechanisms and mediators involved in the pathogenesis of the ESS are still poorly understood, even if the reduction of serum thyroid hormone levels is now generally agreed to be aresultofconcomitantimpairmentofbothperipheralhormonemetabolism and central regulation [22]. Particularly, thehallmark of the ESS consists in increased levels of biologicallyinactive rT3 associated to decreased levels of fT3, due tochanges in peripheral deiodination by means of induction of type 3 deiodinase [16]. Increased levels of cytokines [23–25], endogenous or exogenous glucocorticoids [26], as well ascathecolamines [15], are implicated in the dysregulation of thyroid hormones that occurs in critical illness. Moreover,elevated levels of free fatty acids and bilirubin, found indifferent pathological conditions, have been proposed asadjunctive factors contributing to the onset of the ESS, byindirectly promoting the reduction of hormone binding protein synthesis and the inhibition of fT3 binding to itsreceptor [27,28].Although for many years ESS has been considered as atransient biological adaptive process, there is increasing evidence that an induced hypothyroid-like state may worsenpatient’s clinical outcome [29]. Low fT3 levels have also beenhypothesised to promote the feeding-resistant catabolic stateof prolonged critical illness [2].Several studies showed that the low T3 concentration maybe considered a strong predictor of death in patients affectedby different acute and chronic diseases (Table 1) [6–18]. Nevertheless, it remains speculative whether low fT3 serumconcentration has a direct impact on the outcome andprognosis of burn patients. In fact, low fT3 levels have beensuggested to be involved in the genesis of sepsis, respiratoryfailure and multi-organ dysfunction syndrome which are also Fig. 1 – Spearman’s correlation coefficients. (A) Significant relationship between plasma levels of fT3 and Roi score inthe total study population. (B) Significant relationship between plasma levels of TSH and Roi score in the totalstudy population. (C) Significant relationship betweenplasma levels of fT3 and Roi score in the study populationexcluding the 20 burn patients treated with dopamineinfusion for more than 21 h. burns 34 (2008) 817–824  821
Search
Similar documents
View more...
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x