The American Journal of Surgery (2013) - -ndash-

Procalcitonin ratio as a predictor of successful surgicaltreatment of severe necrotizing soft tissue infections

Jan Friederichs MDa Martin Hutter MDa Christian Hierholzer MDaAlexander Novotny MDb Helmut Friess MDb Volker Buhren MDaSven Hungerer MDa

aTrauma Center Murnau Prof-Kuntscher-Str 8 Murnau 82

418 Germany bDepartment of Surgery Klinikum rechts derIsar Technische Universitat Munchen Munich Germany

KEYWORDSNecrotizing soft tissueinfectionNecrotizing fasciitisProcalcitonin

The authors declare no conflicts of i

Corresponding author Tel149-884

E-mail address JanFriederichsbgu

Manuscript received August 5 2012

2012

0002-9610$ - see front matter 2013

httpdxdoiorg101016jamjsurg20

AbstractBACKGROUND Necrotizing soft tissue infections often are characterized by fulminant presentation

and lethal outcomes Besides critical care support and antibiotic therapy aggressive surgical treatmentis important for the therapy of necrotizing fasciitis The aim of this study was to develop a procalcitonin(PCT) ratio indicating successful surgical intervention

METHODS The study group consisted of 38 patients treated with clinical signs of sepsis caused by anecrotizing soft tissue infection All patients received radical surgical treatment and serum levels ofPCT and C-reactive protein were monitored postoperatively The ratio of day 1 to day 2 was calculatedand correlated with the successful elimination of the infectious source and clinical recovery

RESULTS An eradication of the infectious focus was successfully performed in 84 of patientsaveraging 19 operations (range 1 to 6) to achieve an elimination of the infectious source The PCT ratiowas significantly higher in the group of patients with successful surgical intervention (1665 vs 9P 001) A ratio higher than the calculated cutoff of 114 indicated successful surgical treatment witha sensitivity of 833 and a specificity of 714 The positive predictive value was 758 and the neg-ative predictive value was 800

CONCLUSIONS The PCT ratio of postoperative day 1 to day 2 following major surgical proceduresfor necrotizing soft tissue infections represents a valuable clinical tool indicating successful surgicaleradication of the infectious focus 2013 Elsevier Inc All rights reserved

Necrotizing soft tissue infections (NSTIs) are rare butfulminant and life-threatening bacterial infections Sincethe primary infection affects predominantly the fascia andsubcutaneous tissue the term necrotizing fasciitis was de-fined by Wilson1 in 1952 Characteristics of this highly

nterest

1-484731 fax149-8841-484678

-murnaude

revised manuscript November 2

Elsevier Inc All rights reserved

1211024

lethal infection include clinical symptoms of pain out ofproportion edema blisters and erythema with a systemicinflammation syndrome as well as the histopathologic fea-tures of fascial necrosis vasculitis and thrombosis of per-forating veins2 Synergistic polymicrobial infections areabout as common as monomicrobial fasciitis caused by amultitude of aerobic and anaerobic organisms3ndash5 The loca-tion of infection involves mainly the extremities in 50 to80 of the cases and the trunk and the perineal region(Fournierrsquos gangrene) account for the rest256 Severalclassification systems have been used to describe this

Table 1 Demographic and clinical data of 38 patients withnecrotizing soft tissue infections and sepsis

Demographics and clinical categories

SexMale 2738 (71)Female 1138 (29)

Age 58 years (29ndash85)Location of InfectionUpper limb 638 (16)Lower limb 1838 (47)Thoracic 238 (5)Pelvic 1238 (32)

Cause of InfectionTrauma 1138 (29)Chronic ulcer 438 (10)Injection 638 (16)Postoperative 338 (8)

2 The American Journal of Surgery Vol - No - - 2013

heterogeneous disease based on microbial subtype or loca-tion but no validated classification scheme has been corre-lated with outcome6 Mortality of necrotizing fasciitis hasdropped from up to 32 in study collectives between1980 and 1998 to a range of 88 to 25 in recent reportsThis is most likely due to more aggressive surgical treat-ment more effective antibiotic therapy and adjunctivetreatment36 because advances are based on observationalstudies and randomized controlled trials are lacking

Many prognostic predictors of mortality or limb losshave been published Predictors of outcome include ad-vanced age6ndash9 female sex7 advanced Acute Physiologyand Chronic Health Evaluation II (APACHE II) score610

evidence of sepsis and shock upon admission711 comor-bidities3611 and malignancies1213 Several hematologicvalues at the point of hospital admission have been shownto predict mortality such as lactate7914 and sodiumlevels14 White blood cell count creatinine and hematocritin combination with clinical parameters such as age heartrate and temperature have been used to predict mortalityas well as limb loss in cases of infected extremities15

This seems extremely important as prompt aggressive sur-gical treatment is a mainstay of successful therapy of nec-rotizing fasciitis A delay of surgical therapy results inincreased mortality269 However despite radical initial sur-gical debridement and a limb amputation rate up to 206

surgical control of the infectious source is often notachieved Tools in deciding whether additional radicaltreatment such as a proximal amputation is necessary aremissing and the decision is only based on clinical param-eters and the experience of the surgeon

For patients with abdominal sepsis a marker of suc-cessful surgical treatment has been recently described16

Based on the finding that procalcitonin (PCT) in combina-tion with a clinical score (APACHE II) is highly sensitiveand specific as a prognostic score for the septic course ofabdominal sepsis Novotny et al17 identified a cutoff levelof the PCT ratio of postoperative day 1 to day 2 indicatinga successful surgical treatment as an aid in deciding if fur-ther laparotomies were necessary A PCT ratio below 103represented an unsuccessful operative treatment of the sep-tic focus with a sensitivity of 95

The aim of this study was to evaluate postoperative PCTlevels as a parameter for the detection of progressing diseasein patients with NSTIs to facilitate the decision-makingprogress for reoperations or major amputations

Insect bite 238 (5)Unknown 1238 (32)

Clinical DataHyperbaric oxygen -therapy 3238 (84)Focus eradication 3238 (84)Critical care (days) 22 (1ndash112)Hospitalization (days) 50 (1ndash173)Amputation 1724 (71)Mortality 1138 (29)

Age duration of critical care and hospitalization are expressed as

mean with range in parentheses

Methods

The study population primarily consisted of 38 patientstreated with the diagnosis of NSTI in the intensive care unitof the Trauma Center Murnau Germany from November2006 until December 2010 The initial diagnosis of necro-tizing fasciitis was based on clinical characteristics2ndash5

and intraoperative findings in all patients the diagnosiswas confirmed by histopathologic examination All patients

included in this study initially required critical care supportbecause of acute sepsis and met the established criteria ofsepsis18 Patients who did not meet septic criteria were ex-cluded from the study The study group consisted of 27(71) male and 11 (29) female patients and the medianage was 585 years (range 29 to 85 years) A detailed profileof the clinical characteristics of the study group is presentedin Table 1

All patients underwent surgery at least once For lateranalysis surgical interventions were defined as major andminor procedures Major procedures included amputationsexarticulations and extensive debulking with resection ofnecrotic and infected soft tissue whereas minor procedurescomprised of vacuum-assisted closures without resectionand wound closures For this study successful surgicaleradication of the septic focus was defined by negativeresults in at least 2 consecutive microbiologic examina-tions the absence of macroscopic infectious tissue in allfurther operations and an uneventful recovery (uneventfulhealing of all wounds no reappearance of local infectioussigns and no systemic inflammation) For microbiologicexaminations tissue samples and cultures were obtainedduring each operative procedure All patients receivedcritical care support as needed antibiotic therapy and inthe absence of contraindications hyperbaric oxygentherapy

Medical records of all treated patients were retrospec-tively reviewed The variables examined in this study

J Friederichs et al Procalcitonin ratio and severe necrotizing soft tissue infections 3

included age sex body mass index comorbidities durationof hospitalization and intensive care site and cause ofinfection number and type of surgical procedures rate ofamputation surgical eradication of the septic focus andmortality C-reactive protein (CRP) and PCT were mea-sured at least daily and statistical evaluation was performedwith values of postoperative days 1 and 2 If patientsreturned to the operating room before day 2 or a PCT valuewas missing for other reasons no ratio was calculatedSerum PCT levels were determined using an immunolumi-nometric assay (LIAISON BRAHMS AG HennigsdorfGermany) and serum CRP levels were measured with theDIMENSION VISTA System (Siemens Health Care Mar-burg Germany) both according to the manufacturersrsquoinstructions

Statistical analysis was performed using the StatisticalPackage for the Social Sciences version 190 (SPSSChicago IL) Data were analyzed for normal distributionusing the Kolmogorov-Smirnov test Comparisons betweengroups were performed using the nonparametric Wilcoxonand the MannndashWhitney U tests The probability signifi-cance level was set at less than 05 To determine the cutofflevel of the PCT ratio a chi-square test was used as de-scribed elsewhere17 Briefly PCT ratios were sorted in as-cending values and every value was used as a potentialcutoff assigning all patients with a cutoff less than thisvalue to the persisting infection group and all patientswith ratios above this cutoff to the eradicated groupThis assignment was then compared with the real statuscalculating the corresponding chi-square statistics All re-sulting chi-square statistics were compared and the cutoffvalue with the highest chi-square statistic was used as theoptimal cutoff To investigate the prognostic impact ofPCT and CRP ratios of postoperative day 1 to day 2 re-ceiver operating characteristic curves were plotted and theareas under the curve were assessed

Results

A total of 38 patients presenting in a septic state andtreated for the diagnosis of NSTI were included in thisstudy The study group consisted of 27 (71) male and 11

Table 2 Microbiologic findings in 38 patients with necrotizing soft

Monomicrobial infection (n 5 22)

Streptococcus group A 7 (32)Streptococcus group B 1 (45)Staphylococcus aureus 9 (41)Staphylococcus others 2 (9)Pseudomonas aeruginosa 1 (45)Anaerobes 2 (9)

(29) female patients The median age was 585 years(average 58 range 29 to 85 years) the median body massindex was 275 (range 19 to 60) The most common pre-existing comorbidity was diabetes mellitus occurring in 17of the 38 patients (45) while there was no trendobservable for other comorbidities The average stay inthe intensive care unit was 22 days (range 1 to 112 days)the average duration of hospitalization was 50 days (range1 to 173 days) The location and etiology of all infectionsare summarized in Table 1

All patients underwent surgery on the day of admissionand were operated on at least once A median of 6 oper-ations (average 7 range 1 to 25) was performed perpatient 68 (256) of which were classified as majorsurgical procedures In this selected patient collective theamputation rate was rather high with 17 of 24 (71)patients with NSTIs of the extremities Adjuvant hyper-baric oxygen therapy was performed in 84 of patientsAn eradication of the infectious focus was achieved in84 of the patients averaging 19 operations (range 1 to6) to achieve an elimination of the infectious source Theoverall mortality in our study group was 11 of 38 patients(29) with a mean survival of 18 days Five patients diedof nonseptic complications in the postoperative periodafter a successful eradication of the infectious focus hadbeen achieved

Of the 38 patients treated for NSTI 22 (58) suffereda monomicrobial infection and in 16 patients (42)a polymicrobial infection was detected In the groupof monomicrobial infections b-hemolyzing streptococci(32) and Staphylococcus aureus (41) were the mostfrequent causative organisms including 1 case of a methi-cillin-resistant Staphylococcus aureus Anaerobes weredetected as causative species in only 2 cases (9) Inthe group of 16 patients with a polymicrobial NSTI thenumber of organisms isolated in each patient variedfrom 2 to 5 species with an average of 25 The predom-inant group of organisms in polymicrobial infections werestreptococcal and staphylococcal species However a totalof 17 different species were detected as summarized inTable 2

PCT levels after 63 major operative procedures in thestudy group of 38 patients were measured on postoperative

tissue infections

Polymicrobial infection (n 5 16)

Streptococcus group A 3 (8)Streptococcus group B 4 (105)Staphylococcus aureus 4 (105)Staphylococcus epidermidis 5 (13)Staphylococcus others 5 (13)Enterococcus species 5 (13)Escherichia coli 4 (105)other gram-negative rods 4 (105)Anaerobes 3 (8)Candida species 1 (3)

Figure 2 Cutoff value for the procalcitonin (PCT) ratio of post-operative day 1 to day 2 predicting a successful surgical eradica-tion of the infectious focus using classification and regression treeanalysis Ratios higher than 114 indicate a successful eliminationof the infectious source

4 The American Journal of Surgery Vol - No - - 2013

day 1 and of 61 major surgical procedures on postoperativeday 2 resulting in a total of 58 procedures with calculatedPCT ratios An elimination of the infectious focus wasachieved in 30 of these 58 operations (52) and apersistence of the infectious source was observed in 28(48) PCT levels on days 1 and 2 after major surgery weresignificantly increased above normal with a mean of 143ngmL (standard deviation [SD] 322 ngmL) for postoper-ative day 1 and 127 ngmL (SD 321 ngmL) for postop-erative day 2 The change of PCT levels was then expressedby the calculation of the corresponding ratios (day 1day 2)The PCT ratios of surgical procedures with a successfuleradication of the infectious focus were significantly higher(mean 167 SD 65) than in the group of 28 operations withpersisting infection (mean 9 SD 36 P 001) In con-trast the comparison of the CRP ratios of day 1 with thoseof day 2 did not reach the significance level (P 5 63) Ad-ditionally there were no significant differences when pri-mary major surgical procedures were compared with latermajor operations (data not shown) The results are illus-trated in Fig 1 where corresponding receiver operatingcharacteristic curves with an area under the curve of 872for the PCT ratio and an area under the curve of 617 forthe CRP ratio were calculated The optimal cutoff valuewas calculated at 114 for the PCT ratio between postoper-ative days 1 and 2 Ratios higher than 114 suggested a suc-cessful eradication of the infectious focus as comparedwith a persisting infection if the ratio was lower than the

Figure 1 Receiver operating characteristic (ROC) curves of pro-calcitonin (PCT) and C-reactive protein (CRP) ratio of postopera-tive day 1 and 2 after a major surgical procedure in 38 patientswith necrotizing soft tissue infection The area under the curve(AUC) ROC for successful elimination of the infectious focuswas 872 for the postoperative PCT ratio and 617 for the CRPratio

cutoff as demonstrated in Fig 2 Using this cutoff a pre-diction of a successful elimination was possible with aspecificity of 714 and a sensitivity of 833 resultingin a positive predictive value of 758 and a negative pre-dictive value of 800

Comments

NSTIs are life-threatening infections with a rapid courseof inflammation and often sepsis and shock Although thereported incidence of 4 cases per 100000 population1319

makes it a rare disease and the treatment of necrotizing fas-ciitis has improved its mortality remains high with up to25 to 32 in recent studies671013 Compared with recentstudies on NSTIs the mortality rate of 29 found in ourstudy group appears to be rather high A possible explana-tion for this finding is that our study included only patientswho required critical care support because of acute sepsisand met the established criteria of sepsis18 In conclusionthis study group represented only a third of all our patientstreated for the diagnosis of NSTI as an overall of 131 pa-tients received therapy for necrotizing fasciitis These find-ings correlate with the recently published findings of Kaoet al6 where in a collective of 296 patients vasopressorswere only required in 11 to 27 and intensive care in75 of the patients and the mean length of stay on the in-tensive care unit was only 0 to 6 days compared with amean of 22 days in our study group Another recently pub-lished article describes signs of sepsis including tachycar-dia or hypotension in 25 to 4013 Thus the selectionof only septic patients in need of critical care results in ahigher mortality as compared with other studies Thesame explanation may be valid when looking at the rateof amputation Limb loss is reported to reach 17 to

J Friederichs et al Procalcitonin ratio and severe necrotizing soft tissue infections 3

included age sex body mass index comorbidities durationof hospitalization and intensive care site and cause ofinfection number and type of surgical procedures rate ofamputation surgical eradication of the septic focus andmortality C-reactive protein (CRP) and PCT were mea-sured at least daily and statistical evaluation was performedwith values of postoperative days 1 and 2 If patientsreturned to the operating room before day 2 or a PCT valuewas missing for other reasons no ratio was calculatedSerum PCT levels were determined using an immunolumi-nometric assay (LIAISON BRAHMS AG HennigsdorfGermany) and serum CRP levels were measured with theDIMENSION VISTA System (Siemens Health Care Mar-burg Germany) both according to the manufacturersrsquoinstructions

Statistical analysis was performed using the StatisticalPackage for the Social Sciences version 190 (SPSSChicago IL) Data were analyzed for normal distributionusing the Kolmogorov-Smirnov test Comparisons betweengroups were performed using the nonparametric Wilcoxonand the MannndashWhitney U tests The probability signifi-cance level was set at less than 05 To determine the cutofflevel of the PCT ratio a chi-square test was used as de-scribed elsewhere17 Briefly PCT ratios were sorted in as-cending values and every value was used as a potentialcutoff assigning all patients with a cutoff less than thisvalue to the persisting infection group and all patientswith ratios above this cutoff to the eradicated groupThis assignment was then compared with the real statuscalculating the corresponding chi-square statistics All re-sulting chi-square statistics were compared and the cutoffvalue with the highest chi-square statistic was used as theoptimal cutoff To investigate the prognostic impact ofPCT and CRP ratios of postoperative day 1 to day 2 re-ceiver operating characteristic curves were plotted and theareas under the curve were assessed

Results

A total of 38 patients presenting in a septic state andtreated for the diagnosis of NSTI were included in thisstudy The study group consisted of 27 (71) male and 11

Table 2 Microbiologic findings in 38 patients with necrotizing soft

Monomicrobial infection (n 5 22)

Streptococcus group A 7 (32)Streptococcus group B 1 (45)Staphylococcus aureus 9 (41)Staphylococcus others 2 (9)Pseudomonas aeruginosa 1 (45)Anaerobes 2 (9)

(29) female patients The median age was 585 years(average 58 range 29 to 85 years) the median body massindex was 275 (range 19 to 60) The most common pre-existing comorbidity was diabetes mellitus occurring in 17of the 38 patients (45) while there was no trendobservable for other comorbidities The average stay inthe intensive care unit was 22 days (range 1 to 112 days)the average duration of hospitalization was 50 days (range1 to 173 days) The location and etiology of all infectionsare summarized in Table 1

All patients underwent surgery on the day of admissionand were operated on at least once A median of 6 oper-ations (average 7 range 1 to 25) was performed perpatient 68 (256) of which were classified as majorsurgical procedures In this selected patient collective theamputation rate was rather high with 17 of 24 (71)patients with NSTIs of the extremities Adjuvant hyper-baric oxygen therapy was performed in 84 of patientsAn eradication of the infectious focus was achieved in84 of the patients averaging 19 operations (range 1 to6) to achieve an elimination of the infectious source Theoverall mortality in our study group was 11 of 38 patients(29) with a mean survival of 18 days Five patients diedof nonseptic complications in the postoperative periodafter a successful eradication of the infectious focus hadbeen achieved

Of the 38 patients treated for NSTI 22 (58) suffereda monomicrobial infection and in 16 patients (42)a polymicrobial infection was detected In the groupof monomicrobial infections b-hemolyzing streptococci(32) and Staphylococcus aureus (41) were the mostfrequent causative organisms including 1 case of a methi-cillin-resistant Staphylococcus aureus Anaerobes weredetected as causative species in only 2 cases (9) Inthe group of 16 patients with a polymicrobial NSTI thenumber of organisms isolated in each patient variedfrom 2 to 5 species with an average of 25 The predom-inant group of organisms in polymicrobial infections werestreptococcal and staphylococcal species However a totalof 17 different species were detected as summarized inTable 2

PCT levels after 63 major operative procedures in thestudy group of 38 patients were measured on postoperative

tissue infections

Polymicrobial infection (n 5 16)

Streptococcus group A 3 (8)Streptococcus group B 4 (105)Staphylococcus aureus 4 (105)Staphylococcus epidermidis 5 (13)Staphylococcus others 5 (13)Enterococcus species 5 (13)Escherichia coli 4 (105)other gram-negative rods 4 (105)Anaerobes 3 (8)Candida species 1 (3)

Figure 2 Cutoff value for the procalcitonin (PCT) ratio of post-operative day 1 to day 2 predicting a successful surgical eradica-tion of the infectious focus using classification and regression treeanalysis Ratios higher than 114 indicate a successful eliminationof the infectious source

4 The American Journal of Surgery Vol - No - - 2013

day 1 and of 61 major surgical procedures on postoperativeday 2 resulting in a total of 58 procedures with calculatedPCT ratios An elimination of the infectious focus wasachieved in 30 of these 58 operations (52) and apersistence of the infectious source was observed in 28(48) PCT levels on days 1 and 2 after major surgery weresignificantly increased above normal with a mean of 143ngmL (standard deviation [SD] 322 ngmL) for postoper-ative day 1 and 127 ngmL (SD 321 ngmL) for postop-erative day 2 The change of PCT levels was then expressedby the calculation of the corresponding ratios (day 1day 2)The PCT ratios of surgical procedures with a successfuleradication of the infectious focus were significantly higher(mean 167 SD 65) than in the group of 28 operations withpersisting infection (mean 9 SD 36 P 001) In con-trast the comparison of the CRP ratios of day 1 with thoseof day 2 did not reach the significance level (P 5 63) Ad-ditionally there were no significant differences when pri-mary major surgical procedures were compared with latermajor operations (data not shown) The results are illus-trated in Fig 1 where corresponding receiver operatingcharacteristic curves with an area under the curve of 872for the PCT ratio and an area under the curve of 617 forthe CRP ratio were calculated The optimal cutoff valuewas calculated at 114 for the PCT ratio between postoper-ative days 1 and 2 Ratios higher than 114 suggested a suc-cessful eradication of the infectious focus as comparedwith a persisting infection if the ratio was lower than the

Figure 1 Receiver operating characteristic (ROC) curves of pro-calcitonin (PCT) and C-reactive protein (CRP) ratio of postopera-tive day 1 and 2 after a major surgical procedure in 38 patientswith necrotizing soft tissue infection The area under the curve(AUC) ROC for successful elimination of the infectious focuswas 872 for the postoperative PCT ratio and 617 for the CRPratio

cutoff as demonstrated in Fig 2 Using this cutoff a pre-diction of a successful elimination was possible with aspecificity of 714 and a sensitivity of 833 resultingin a positive predictive value of 758 and a negative pre-dictive value of 800

Comments

NSTIs are life-threatening infections with a rapid courseof inflammation and often sepsis and shock Although thereported incidence of 4 cases per 100000 population1319

makes it a rare disease and the treatment of necrotizing fas-ciitis has improved its mortality remains high with up to25 to 32 in recent studies671013 Compared with recentstudies on NSTIs the mortality rate of 29 found in ourstudy group appears to be rather high A possible explana-tion for this finding is that our study included only patientswho required critical care support because of acute sepsisand met the established criteria of sepsis18 In conclusionthis study group represented only a third of all our patientstreated for the diagnosis of NSTI as an overall of 131 pa-tients received therapy for necrotizing fasciitis These find-ings correlate with the recently published findings of Kaoet al6 where in a collective of 296 patients vasopressorswere only required in 11 to 27 and intensive care in75 of the patients and the mean length of stay on the in-tensive care unit was only 0 to 6 days compared with amean of 22 days in our study group Another recently pub-lished article describes signs of sepsis including tachycar-dia or hypotension in 25 to 4013 Thus the selectionof only septic patients in need of critical care results in ahigher mortality as compared with other studies Thesame explanation may be valid when looking at the rateof amputation Limb loss is reported to reach 17 to

Figure 2 Cutoff value for the procalcitonin (PCT) ratio of post-operative day 1 to day 2 predicting a successful surgical eradica-tion of the infectious focus using classification and regression treeanalysis Ratios higher than 114 indicate a successful eliminationof the infectious source

4 The American Journal of Surgery Vol - No - - 2013

day 1 and of 61 major surgical procedures on postoperativeday 2 resulting in a total of 58 procedures with calculatedPCT ratios An elimination of the infectious focus wasachieved in 30 of these 58 operations (52) and apersistence of the infectious source was observed in 28(48) PCT levels on days 1 and 2 after major surgery weresignificantly increased above normal with a mean of 143ngmL (standard deviation [SD] 322 ngmL) for postoper-ative day 1 and 127 ngmL (SD 321 ngmL) for postop-erative day 2 The change of PCT levels was then expressedby the calculation of the corresponding ratios (day 1day 2)The PCT ratios of surgical procedures with a successfuleradication of the infectious focus were significantly higher(mean 167 SD 65) than in the group of 28 operations withpersisting infection (mean 9 SD 36 P 001) In con-trast the comparison of the CRP ratios of day 1 with thoseof day 2 did not reach the significance level (P 5 63) Ad-ditionally there were no significant differences when pri-mary major surgical procedures were compared with latermajor operations (data not shown) The results are illus-trated in Fig 1 where corresponding receiver operatingcharacteristic curves with an area under the curve of 872for the PCT ratio and an area under the curve of 617 forthe CRP ratio were calculated The optimal cutoff valuewas calculated at 114 for the PCT ratio between postoper-ative days 1 and 2 Ratios higher than 114 suggested a suc-cessful eradication of the infectious focus as comparedwith a persisting infection if the ratio was lower than the

Figure 1 Receiver operating characteristic (ROC) curves of pro-calcitonin (PCT) and C-reactive protein (CRP) ratio of postopera-tive day 1 and 2 after a major surgical procedure in 38 patientswith necrotizing soft tissue infection The area under the curve(AUC) ROC for successful elimination of the infectious focuswas 872 for the postoperative PCT ratio and 617 for the CRPratio

cutoff as demonstrated in Fig 2 Using this cutoff a pre-diction of a successful elimination was possible with aspecificity of 714 and a sensitivity of 833 resultingin a positive predictive value of 758 and a negative pre-dictive value of 800

Comments

NSTIs are life-threatening infections with a rapid courseof inflammation and often sepsis and shock Although thereported incidence of 4 cases per 100000 population1319

makes it a rare disease and the treatment of necrotizing fas-ciitis has improved its mortality remains high with up to25 to 32 in recent studies671013 Compared with recentstudies on NSTIs the mortality rate of 29 found in ourstudy group appears to be rather high A possible explana-tion for this finding is that our study included only patientswho required critical care support because of acute sepsisand met the established criteria of sepsis18 In conclusionthis study group represented only a third of all our patientstreated for the diagnosis of NSTI as an overall of 131 pa-tients received therapy for necrotizing fasciitis These find-ings correlate with the recently published findings of Kaoet al6 where in a collective of 296 patients vasopressorswere only required in 11 to 27 and intensive care in75 of the patients and the mean length of stay on the in-tensive care unit was only 0 to 6 days compared with amean of 22 days in our study group Another recently pub-lished article describes signs of sepsis including tachycar-dia or hypotension in 25 to 4013 Thus the selectionof only septic patients in need of critical care results in ahigher mortality as compared with other studies Thesame explanation may be valid when looking at the rateof amputation Limb loss is reported to reach 17 to

J Friederichs et al Procalcitonin ratio and severe necrotizing soft tissue infections 5

50 of patients with necrotizing fasciitis located on an ex-tremity3131520 In our study 24 patients suffered an NSTIof an extremity and the amputation rate was as high as71 Again this high rate of amputation is based on thespecific collective of patients who were admitted to our in-stitution with all signs of shock and sepsis and thereforediffers from the overall amputation rate of 34 in our hos-pital for patients with necrotizing fasciitis of a limb Ac-cording to that shock has been described as anindependent prognostic factor for limb loss15

Several prognostic factors for a favorable outcome ofNSTIs have been published Immediate operative treatmentappears to have an important impact on mortality and rateof amputation as several publications report a delay ofsurgery being an independent negative prognostic fac-tor3720 Wong et al3 observed a decline in the cumulativesurvival rate with progressing time between admissionand surgery Our study group consisted entirely of septicpatients and all patients underwent emergency surgery onthe day of admission to our hospital Although our datado not permit us to determine an optimized time point ofoperation we believe that immediate surgical treatment isessential for successful therapy

The importance of radical surgical debridement resec-tion of all necrotic tissue based on the aggressiveness of theinfection and early amputation of a limb has been de-scribed in previous studies Limited incisions drainage as amonotherapy and onetime debridements have been shownto be ineffective and do not decrease mortality7 Howeveraggressive surgical resections are sometimes not radicalenough and extensive reoperations or a more proximal am-putation in the further course of disease may be necessaryOn the other hand lsquolsquoovertreatmentrsquorsquo resulting in a loss ofnoninfected tissue should be avoided These decisionsare often based on the experience of the surgeon and post-operatively on the clinical state of the patient Many factorsincluding all septic parameters and laboratory results suchas leucocyte count CRP interleukin-621 lactate base def-icite14 and PCT16 can be taken into account for the deci-sion to increase or decrease surgical aggressivenessHowever in most cases the decision of the surgeon willbe based on the clinical condition of the patient includingall available prognostic markers

Conclusions

Our study identified the PCT ratio of postoperative day1 to day 2 as a valuable tool to decide whether the aim offocus eradication has been achieved or if the initialtreatment was not radical enough As described by Novotnyet al PCT represents an appropriate marker as it can easilybe measured within less than 2 hours with a commerciallyavailable assay It has favorable kinetics with a rapidincrease and a half-life of approximately 20 to 24 hourswithout prolongation with impaired kidney function172122

In our study a strong decrease of serum PCT was observed

from postoperative day 1 to day 2 after a major surgicalprocedure if the source of infection was eliminated whilethe decrease was not significant in patients with persistinginfection This relationship is best expressed by the PCT ra-tio between postoperative day 1 and day 2 and it was sig-nificantly higher after successful surgery (P 001) Thecalculated cutoff which was 114 in our study seems tobe a practical aid for clinical decisions as it is able to dis-tinguish between successful and unsuccessful surgical erad-ication of the infectious focus with a high sensitivity andspecificity In the clinical setting a ratio below the cutoffshould especially raise suspicion for persistence of the in-fectious focus and together with other parameters maylead to an earlier and more radical reoperation or an earlierlife-saving amputation

We are aware that this cutoff only represents an aid for thesurgeon adding one piece of information to the overall clinicalimpression and that good surgical judgment still representsthe leading factor in decision making A large prospectivestudy to validate this PCT ratiowould be necessary to prove itsclinical value although only a multicenter design can provideenough patients with a fulminant course of necrotizingfasciitis However we believe that septic patients with thediagnosis of NSTI may benefit from the proposed method

References

1 Wilson B Necrotizing fasciitis Am Surg 195218416

2 Herr M Grabein B Palm HG et al Necrotizing fasciitis 2011 update

Unfallchirurg 2011114197ndash216

3 Wong CH Chang HC Pasupathy S et al Necrotizing fasciitis clinical

presentation microbiology and determinants of mortality J Bone Joint

Surg 2003851454ndash60

4 Elliott D Kufera JA Myers RA The microbiology of necrotizing soft

tissue infections Am J Surg 2000179361ndash6

5 Fontes Jr RA Ogilvie CM Miclau T Necrotizing soft-tissue infec-

tions J Am Acad Orthop surg 20008151ndash8

6 Kao LS Lew DF Arab SN et al Local variations in the epidemiology

microbiology and outcome of necrotizing soft-tissue infections a

multicenter study Am J Surg 2011202139ndash45

7 Elliott DC Kufera JA Myers RAM Necrotizing soft tissue infections

Risk factors for mortality and strategies for management Ann Surg

199622672ndash83

8 Mulla ZD Gibbs SG Aronoff DM Correlates of length of stay cost of

care and mortality among patients hospitalized for necrotizing fascii-

tis Epidemiol Infect 2007135868ndash76

9 Malangoni MA Necrotizing soft tissue infections are we making any

progress Surg Infect Larchmt 20012145ndash50

10 Gunter OL Guillamondegui OD May AK et al Outcome of necrotiz-

ing skin and soft tissue infections Surg Infect Larchmt 20089443ndash50

11 Dworkin MS Westercamp MD Park L et al The epidemiology of

necrotizing fasciitis including factors associated with death and ampu-

tation Epidemiol Infect 20091371ndash6

12 Golger A Ching S Goldsmith CH et al Mortality in patients with

necrotizing fasciitis Plast Reconstr Surg 20071191803ndash7

13 Hsiao CT Weng HH Yuan YD et al Predictors of mortality in pa-

tients with necrotizing fasciitis Am J Emerg Med 200826170ndash5

14 Yaghoubian A de Virgilia C Dauphine C et al Use of admission se-

rum lactate and sodium levels to predict mortality in necrotizing soft-

tissue infections Arch Surg 2007142840ndash6

15 Anaya DA McMahon K Nathens AB et al Predictors of mortality and

limb loss in necrotizing soft tissue infections Arch Surg 2005140151ndash7

6 The American Journal of Surgery Vol - No - - 2013

16 Novotny AR Emmanuel K Hueser N et al Procalcitonin ratio indicates

successful surgical treatment of abdominal sepsis Surgery200914520ndash6

17 Novotny AR Emmanuel K Matevossian E et al Use of procalcitonin

for early prediction of lethal outcome of postoperative sepsis Am J

Surg 200719435ndash9

18 Bone RC Balk RA Cerra FB et al Definitions for sepsis and organ

failure and guidelines for the use of innovative therapies in sepsis

The ACCPSCCM Consensus Conference Committee American Col-

lege of Chest PhysiciansSociety of Critical Care Medicine Chest

19921011644ndash55

19 File TM Tan JS DiPersio JR Group A streptococcal necrotizing fas-

ciitis Diagnosing and treating the lsquolsquoflesh eating bacteria syndromersquorsquo

Cleve Clin J Med 199865241ndash9

20 McHenry CR Piotrowski JJ Petrinic D et al Determinants of mortal-

ity for necrotizing soft-tissue infections Ann Surg 1995221558ndash65

21 Brunkhorst FM Heinz U Forycki ZF Kinetics of procalcitonin in iat-

rogenic sepsis Intensive Care Med 199824888ndash92

22 Meisner M Schmidt J Huettner H et al The natural elimination rate

of procalcitonin in patients with normal and impaired renal function

Intensive Care Med 200026212ndash6