- Email: [email protected]
Preoperative thrombocytopenia and outcomes of hepatectomy for hepatocellular carcinoma
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Preoperative thrombocytopenia and outcomes of hepatectomy for hepatocellular carcinoma Raghunandan Venkat, MD, MPH,a Jack R. Hannallah, MD, MPH, MBA,a Robert S. Krouse, MD, FACS,a,b and Felipe B. Maegawa, MD, FACSa,b,* a b
Department of Surgery, University of Arizona, Tucson, Arizona Surgical Care Line, Southern Arizona VA Health Care System, Tucson, Arizona
article info
abstract
Article history:
Background: Platelet count is known to be an indirect indicator of portal hypertension but is
Received 31 May 2015
not a part of the model for end-stage liver disease (MELD) score or the ChildePugh score for
Received in revised form
risk stratification in hepatobiliary surgery.
25 July 2015
Methods: Data from 2097 hepatic resections for hepatocellular carcinoma (HCC) were
Accepted 21 August 2015
evaluated from 2005e2012 using the National Surgical Quality Improvement Program
Available online xxx
database. Patient demographics, morbidity, and mortality were evaluated. Results: Median age and body mass index were 64 y and 26.5 kg/m2, respectively. Majority of
Keywords:
the patients had American Society of Anesthesiologists 3 (78.1%) and median MELD score
Hepatectomy
was 7. On multivariate analysis, thrombocytopenia (platelet count <150/nL) and severe
Hepatocellular carcinoma
thrombocytopenia (platelet count <100/nL) were independently associated with an
Clinical outcomes
increased risk of mortality (odds ratio [OR], 1.79; P ¼ 0.024 and OR, 4.19; P < 0.001), car-
Thrombocytopenia
diopulmonary complications (OR, 1.61; P ¼ 0.009 and OR, 1.96; P ¼ 0.018), need for blood
Portal hypertension
transfusion (OR, 1.35; P ¼ 0.05 and OR, 1.60; P ¼ 0.05), septic complications (OR, 1.53; P ¼ 0.025 and OR, 1.96; P ¼ 0.016), reintubation (OR, 1.91; P ¼ 0.004 and OR, 2.64; P ¼ 0.003),
NSQIP
and renal insufficiency and/or failure (OR, 2.48; P ¼ 0.001 and OR, 4.96; P < 0.001), respectively. Conclusions: Thrombocytopenia, which is an indirect indicator for portal hypertension, is significantly associated with adverse outcomes after hepatectomy, independent of the MELD score. Platelet count should be integrated into the selection criteria for hepatic resections for HCC. ª 2015 Elsevier Inc. All rights reserved.
1.
Introduction
Hepatocellular carcinoma (HCC) with cirrhosis is the most common primary tumor of the liver and is the fifth most common malignancy worldwide [1,2]. In patients with cirrhosis, HCC has an annual incidence of 3%e5% [3]. In the United States, it is estimated that there will be 33,190 new
cases diagnosed in 2014 and 23,000 deaths due to this disease [4]. The surgical treatment options for HCC, which offer the potential for cure, include liver transplantation, hepatic resection, and percutaneous ablation [5e7]. Limited organ availability for transplantation [8], better oncological outcomes than percutaneous ablation [9], and improvement in perioperative care over the recent years [10,11] have made
Oral presentation at the scientific article sessions, American College of Surgeons 2014 Annual Clinical Congress, San Francisco. * Corresponding author. Surgical Care Line 2-112, Southern Arizona VA Health Care System, 3601 South 6th Avenue, Tucson, AZ 85723. Tel.: þ1 520 792 1450 (6156); fax: þ1 520 629 4603. E-mail addresses: [email protected], [email protected] (F.B. Maegawa). 0022-4804/$ e see front matter ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2015.08.038
2
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8
hepatic resection the mainstay of treatment for HCC, especially for patients with normal liver function or wellcompensated cirrhosis [12]. Furthermore, hepatectomy has been shown to be more cost-effective than liver transplantation [13]. Careful patient selection with assessment of the tumor burden and residual liver function are essential to optimize surgical outcomes after hepatic resection. However, the definition of selection criteria for hepatic resection is far from being established. Several tools can assess liver function. Traditionally, the ChildePugh score has been widely used due to its ease of application [14]. However, because it uses subjective parameters such as ascites and encephalopathy, there are limitations to its discriminatory abilities [15]. The model for end-stage liver disease (MELD) score, which was first introduced to predict survival after transjugular intrahepatic portosystemic shunts, has been shown to be an efficient method to evaluate hepatic function and predict outcomes after liver resection [16e19]. Another method to evaluate liver function includes measurement of indocyanine green (ICG) clearance, which is more frequently used in Korea and Japan [20]. The presence of portal hypertension (PHT) is also indicative of poor liver function and may be associated with postoperative liver insufficiency and increased morbidity [21]. Significant PHT is usually considered a contraindication to liver resection, as recommended by the European Association for Study of Liver (EASL) [22] and American Associations for Study of Liver Diseases (AASLD) [23] guidelines for HCC treatment. PHT can be identified either instrumentally by hepatic venous pressure gradient (HVPG) measurement or clinically by the presence of esophageal and gastric varices, and a low platelet count associated with splenomegaly. Chronic PHT results in a vascular congestive splenomegaly that triggers a platelet sequestration causing thrombocytopenia. Preoperative platelet count may thus serve as an inexpensive and noninvasive laboratory indicator of PHT. PHT clinically manifested by ascites and encephalopathy is a clear contraindication for major liver resection in most centers. However, in some patients the only sign of PHT is low platelet count, lacking the other clinical signs described previously. In this setting, the ChildePugh classification or MELD scores are within normal limits, misleading an adequate patient selection. In this context, the focus of this study was to evaluate the association of preoperative platelet count, as a noninvasive marker for PHT, with postoperative outcomes in patients undergoing hepatic resection for HCC, independent of their MELD score. We hypothesized that preoperative thrombocytopenia was independently associated with increased postoperative morbidity and mortality, whereas accounting for patient comorbidities, MELD score, and the extent of liver resection.
2.
Methods
2.1.
Data source
We used the 2005e2012 American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP)
database participant use file. The ACS-NSQIP obtains data on patients undergoing inpatient and outpatient surgical procedures from more than 250 university and private sector medical centers. Preoperative patient characteristics, intraoperative procedure characteristics, and 30-d postoperative mortality and complications, including more than 136 variables, are recorded in the data set and are well described elsewhere in the literature [24,25]. The University of Arizona Institutional Review Board determined that this study qualified for an exemption under the Department of Health and Human Studies regulations because of the use of deidentified administrative data. We identified 2097 patients who underwent hepatic resection for HCC. The hepatic resections were classified into partial lobectomy (current procedural terminology [CPT] 47120), total left lobectomy (CPT 47125), total right lobectomy (CPT 47130), and trisegmentectomy (CPT 47122). HCC was defined by the International Classification of Diseases, Ninth Revision diagnosis code for primary liver cancer (155.0). Patient characteristics were defined from reported NSQIP variables as follows: race and/or ethnicity was dichotomized to Caucasian, AfricaneAmerican, and other, American Society of Anesthesiologists (ASA) classification was dichotomized to 3 or higher or less than 3, and smoking status was dichotomized as current smoker within 1 y or not. Age and body mass index (BMI, calculated as weight in kilograms divided by height in meters squared) were evaluated as continuous variables as well as classified in categories based on prior research examining the effect of age [26] and BMI [27] on postoperative complications. Age (in y) was classified in four categories: 49, 50e64, 65e79, and 80. BMI (in kg/m2) was classified in five categories: <18.5, 18.5e24.9, 25.0e29.9, 30.0e34.9, and 35.0.
2.2.
Assessment of preoperative liver function
The MELD score was calculated from preoperative laboratory variables using the standard equation: MELD ¼ 9.57 loge (creatinine mg/dL) þ 3.78 loge(bilirubin mg/dL) þ 11.20 loge(INR) þ 6.43 [28]. Thrombocytopenia was defined as a platelet count <150/nL, and severe thrombocytopenia was defined as a platelet count of <100/nL. ChildePugh score was not calculated because of the lack of the degree of ascites and encephalopathy in the ACS-NSQIP data set.
2.3.
Definition of perioperative outcomes
Thirty-day outcomes were evaluated and included mortality, return to the operating room, postoperative complications, serious morbidity, and overall morbidity. Postoperative complications were categorized as previously described in the literature [29]: (1) wound complications (dehiscence, organ space surgical site infection [SSI], superficial SSI, and deep incisional SSI); (2) cardiopulmonary complications (cardiac arrest requiring cardiopulmonary resuscitation, myocardial infarction, ventilator dependence >48 h, pulmonary embolism (PE), unplanned intubation, and cerebrovascular accident); (3) postoperative transfusion; and (4) septic complications (septic shock and sepsis). The definition of serious morbidity was adapted from Ingraham et al. [30] as
3
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8
having documentation of at least one of the following ACSNSQIP complications: organ space SSI, wound dehiscence, neurologic event (cerebrovascular accident or coma lasting >24 h), cardiac arrest, myocardial infarction, PE, ventilator dependence >48 h, progressive or acute renal insufficiency, and sepsis or septic shock. Overall morbidity was defined as documentation of a serious morbidity or at least one of the following ACS-NSQIP complications: superficial SSI, deep SSI, pneumonia, unplanned intubation (without preoperative ventilator dependence), urinary tract infection, peripheral neurologic deficit, or deep vein thrombosis (DVT). Prolonged length of stay (LOS) was defined as an LOS above the 75th percentile, consistent with the convention used in other NSQIP studies regarding LOS [31].
2.4.
Statistical analysis
For unadjusted comparison between our groups, we used the t-test and/or analysis of variance or the Wilcoxon rank-sum test for continuous variables and the chi-squared or the Fisher exact test for categorical variables, as appropriate. Variables demonstrating association with each end point to the level of P 0.25 on univariate analysis were entered into a multivariate model. A P value of 0.05 in the final model was considered statistically significant. To account for missing data and minimize any potential bias due to data not being missing completely at random, multiple imputation using chained equations with 10 imputations was used for all regression analyses that included variables with missing data [32]. Statistical analysis was performed using Intercooled Stata, version 12.0 (StataCorp, College Station, TX).
3.
Results
3.1.
Operative and patient characteristics
A total of 2097 hepatic resections were performed for HCC during 2005e2012 (partial lobectomy: 1166 [55.6%], left hemihepatectomy: 262 [12.5%], right hemihepatectomy: 427 [20.4%], and trisegmentectomy: 242 [11.5%]). Median age was 64 y (interquartile range, 56e72 y). A total of 67.3% of the patients were male (n ¼ 1411), and the majority were white (n ¼ 1344, 64.1%). Median BMI was 26.5 kg/m2 (interquartile range, 23.4e30.3 kg/m2), and the majority of the patients were ASA class 3 or greater (n ¼ 1637, 78.1%). Laboratory data to calculate the MELD score were available in 1841 patients (256 missing values). The median MELD score was 7, and 19.2% patients had MELD 10. Data on platelet count were available in 2051 patients (46 missing values). Median platelet count was 205 per nL. Patients having thrombocytopenia were 22.7% (n ¼ 465), and 6.1% patients (n ¼ 125) had severe thrombocytopenia (Table 1).
3.2.
Perioperative outcomes
Outcomes were stratified by the presence of a preoperative platelet count <150/nL (thrombocytopenia) versus 150/nL and preoperative platelet count <100/nL (severe thrombocytopenia) versus 100/nL. On univariate analysis, patients with
Table 1 e Patient characteristics and comorbidities. Variable
n ¼ 2097
Age (y, median [IQR]) Age category 49 50e64 65e79 80 Sex Female Male Race White AfricaneAmerican Others ASA classification <3 3 BMI (kg/m2, median [IQR]) BMI category <18.5 18.5e24.9 25.0e29.9 30.0e34.9 35.0 Current smoker Current alcohol use Diabetes Hypertension CHF MI (last 6 mo) COPD Dialysis Surgery Partial lobectomy Left hemihepatectomy Right hemihepatectomy Trisegmentectomy Platelet count* (per nL, median [IQR]) 150 100e150 <100 MELD scorey (median [IQR]) MELD 10
64 [56e72] 265 839 864 129
(12.6) (40.0) (41.2) (6.2)
686 (32.7) 1411 (67.3) 1344 (64.1) 207 (9.9) 546 (26.0) 460 (21.9) 1637 (78.1) 25.6 [23.4e30.3] 59 759 721 361 197 470 75 561 1249 5 5 106 7
(2.8) (36.2) (34.4) (17.2) (9.4) (22.4) (3.6) (26.7) (59.6) (0.2) (0.2) (5.0) (0.3)
1166 (55.6) 262 (12.5) 427 (20.4) 242 (11.5) 205 [153e275] 1856 (77.3) 340 (16.6) 125 (6.1) 7 [6e9] 353 (19.2)
CHF ¼ congestive heart failure; COPD ¼ chronic obstructive pulmonary disease; IQR ¼ interquartile range; MI ¼ myocardial infarction; nL ¼ nanoliter. Data expressed in n (%) unless otherwise specified. * Platelet count was missing in 46 patients. y MELD score was missing in 256 patients.
thrombocytopenia had a significantly increased rate of septic complications (11.6% versus 8.5%, P ¼ 0.04), pneumonia (6.4% versus 4.0%, P ¼ 0.02), unplanned reintubation rate (7.7% versus 5.2%, P ¼ 0.04), ventilator dependence >48 h (6.7% versus 4.6%, P ¼ 0.05), renal insufficiency and/or failure (6.0% versus 3.0%, P < 0.01), and septic shock (5.6% versus 3.3%, P ¼ 0.03). There was no significant difference between the two groups (platelet count <150/nL and 150 nL) in the rates of 30-d mortality, overall morbidity, serious morbidity, wound complication, cardiopulmonary complication, need for postoperative blood transfusion, reoperation, prolonged LOS, DVT and/or PE, cardiac arrest and/or myocardial infarction, or sepsis (Table 2).
4
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8
To further evaluate the incremental effect of a low preoperative platelet count on postoperative outcomes, we analyzed the patients with platelet count <100/nL and 100/ nL. Severe preoperative thrombocytopenia was significantly associated with an increased risk of 30-d mortality (11.2% versus 4.0%, P < 0.01), septic complication (16% versus 8.8%, P < 0.01), pneumonia (8.0% versus 4.3%, P ¼ 0.05), unplanned reintubation rate (10.4% versus 5.5%, P ¼ 0.02), renal insufficiency and/or failure (12.0% versus 3.2%, P < 0.01), and septic shock (9.6% versus 3.5%, P < 0.01). There was no significant difference between the two groups (platelet count <100/nL and 100 nL) in the rates of overall morbidity, serious morbidity, wound complication, cardiopulmonary complication, need for postoperative blood transfusion, reoperation, prolonged LOS, ventilator dependence >48 h, DVT and/or PE, cardiac arrest and/or myocardial infarction, or sepsis (Table 3). A multivariate logistic regression analysis was performed, controlling for potential confounding factors of age, gender, BMI, ASA, MELD score, and the type of hepatectomy. Platelet count was seen to be independently and incrementally associated with poor postoperative outcomes (Tables 4 and 5). Platelet <150/nL was associated with a significant increase in the risk of 30-d mortality (adjusted odds ratio [AOR], 1.79; P ¼ 0.024), cardiopulmonary complications (AOR, 1.61; P ¼ 0.009), need for blood transfusions (AOR, 1.35; P ¼ 0.05), septic complications (AOR, 1.53; P ¼ 0.025), pneumonia (AOR, 1.71; P ¼ 0.032), unplanned reintubations (AOR, 1.91; P ¼ 0.004), ventilator dependence >48 h (AOR, 1.98; P ¼ 0.004), renal insufficiency and/or failure (AOR, 2.48; P ¼ 0.001), and septic shock (AOR, 2.36; P ¼ 0.002).
Table 2 e Univariate analysis comparing postoperative outcomes with preoperative platelet count. (‡150 per nL versus <150 per nL). Variable
Platelet count (per nL) 150 (n ¼ 1586)
Mortality Overall morbidity Serious morbidity Wound complication Cardiopulmonary complication Postoperative transfusion Septic complication Return to the operating room Prolonged LOS Postoperative complication Pneumonia Unplanned reintubation Vent dependence >48 h DVT/PE Renal insufficiency/failure Cardiac arrest/MI Sepsis Septic shock
<150 (n ¼ 465)
P value
66 392 282 194 147
(4.2) (24.7) (17.8) (12.2) (9.3)
26 123 84 45 52
(5.6) (26.4) (18.1) (9.7) (11.2)
0.19 0.45 0.89 0.13 0.22
248 135 64 377
(15.6) (8.5) (4.0) (23.8)
84 54 13 109
(18.1) (11.6) (2.8) (23.4)
0.21 0.04 0.22 0.88
63 83 73 61 48 35 87 53
(4.0) (5.2) (4.6) (3.8) (3.0) (2.2) (5.5) (3.3)
30 36 32 11 28 8 30 26
(6.4) (7.7) (6.7) (2.4) (6.0) (1.7) (6.4) (5.6)
MI ¼ myocardial infarction; vent ¼ ventilator. Data expressed in n (%) unless otherwise specified.
0.02 0.04 0.05 0.12 <0.01 0.52 0.43 0.03
Table 3 e Univariate analysis comparing postoperative outcomes with preoperative platelet count. (‡100 per nL versus <100 per nL). Variable
Platelet count (per nL) 100 (n ¼ 1926)
Mortality Overall morbidity Serious morbidity Wound complication Cardiopulmonary complication Postoperative transfusion Septic complication Return to the operating room Prolonged LOS Postoperative complication Pneumonia Unplanned reintubation Vent dependence >48 h DVT/PE Renal insufficiency/failure Cardiac arrest/MI Sepsis Septic shock
<100 (n ¼ 125)
P value
78 479 338 227 182
(4.0) (24.9) (17.5) (11.8) (9.4)
14 36 28 12 17
(11.2) (28.8) (22.4) (9.6) (13.6)
<0.01 0.33 0.17 0.46 0.13
308 169 73 452
(16.0) (8.8) (3.8) (23.5)
24 20 4 34
(19.2) (16.0) (3.2) (27.2)
0.34 <0.01 0.74 0.34
83 106 96 69 61 40 107 67
(4.3) (5.5) (5.0) (3.6) (3.2) (2.1) (5.6) (3.5)
10 13 9 3 15 3 10 12
(8.0) (10.4) (7.2) (2.4) (12.0) (2.4) (8.0) (9.6)
0.05 0.02 0.28 0.49 <0.01 0.81 0.25 <0.01
MI ¼ myocardial infarction; vent ¼ ventilator. Data expressed in n (%) unless otherwise specified.
On multivariate analysis of patients with a platelet count <100/nL, we observed a significant worsening of postoperative outcomes. There was an increased risk of 30-d mortality (AOR, 4.19; P < 0.001), cardiopulmonary complications (AOR, 1.96; P ¼ 0.018), need for blood transfusions (AOR, 1.60; P ¼ 0.05), septic complications (AOR, 1.96; P ¼ 0.016), unplanned reintubations (AOR, 2.64; P ¼ 0.003), renal insufficiency and/or failure (AOR, 4.96; P < 0.001), and septic shock (AOR, 2.36; P < 0.001). There was a trend toward an increased risk of postoperative pneumonia (AOR, 2.00; P ¼ 0.06), whereas there was no association with ventilator dependence >48 h. We then performed a subset multivariate analysis of 1488 patients with an MELD score <10 and evaluated the association of thrombocytopenia and severe thrombocytopenia with postoperative outcomes. A platelet count of <150/nL was associated with a significant increase in the need for postoperative blood transfusion (AOR, 1.49; 95% confidence interval [CI], 1.05e2.12), ventilator dependence >48 h (AOR, 1.90; 95% CI, 1.09e3.33), renal insufficiency and/or failure (AOR, 2.09; 95% CI, 1.07e4.08), and septic shock (AOR, 2.00; 95% CI, 1.02e3.91). We observed that in this patient population of MELD <10, a platelet count <100/nL was associated with a profound worsening of postoperative outcomes. There was a significant increase in the risk of 30-d mortality (AOR, 5.24; 95% CI, 2.13e12.89), cardiopulmonary complications (AOR, 2.51; 95% CI, 1.25e5.06), septic complications (AOR, 2.56; 95% CI, 1.25e5.27), need for postoperative transfusion (AOR, 1.82; 95% CI, 1.02e3.34), reintubation, (AOR, 3.17; 95% CI, 1.411e7.11), ventilator dependence >48 h (AOR, 2.57; 95% CI, 1.04e6.33), renal insufficiency and/or failure (AOR, 6.87; 95% CI, 2.86e16.52), and septic shock (AOR, 5.77; 95% CI, 2.34e14.21).
5
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8
Table 4 e Multivariate analysis evaluating postoperative outcomes with preoperative platelet count of <150 per nL.
Table 5 e Multivariate analysis evaluating postoperative outcomes with preoperative platelet count of <100 per nL.
Outcome measure
Outcome measure
Mortality Overall morbidity Serious morbidity Wound complication Cardiopulmonary complication Postoperative transfusion Septic complication Return to the operating room Prolonged LOS Postoperative complication Pneumonia Unplanned reintubation Vent dependence >48 h DVT/PE Renal insufficiency/failure Cardiac arrest/MI Sepsis Septic shock
AOR (95% CI; platelet count <150 versus 150) 1.79 1.15 1.14 0.70 1.61 1.35 1.53 0.79 1.05
(1.08e2.97)* (0.89e1.49) (0.85e1.54) (0.47e1.10) (1.12e2.32)* (1.01e1.83)* (1.05e2.21)* (0.41e1.52) (0.80e1.38)
1.71 1.91 1.98 0.56 2.48 0.85 1.07 2.36
(1.05e2.80)* (1.23e2.94)* (1.25e3.14)* (0.26e1.20) (1.49e4.15)* (0.38e1.90) (0.66e1.74) (1.39e4.00)*
MI ¼ myocardial infarction. AOR obtained after controlling for age, gender, BMI, ASA, MELD score, and type of hepatectomy. * P 0.05.
4.
Discussion
The outcome of hepatic resection in cirrhotic patients has improved significantly over the last decade. With improvement in surgical techniques and perioperative care, recently published large series of liver resections for HCC have reported good results, with a significant reduction in postoperative morbidity, mortality, and liver decompensation [11,33]. These results have largely been possible due to the importance given to careful patient selection. The optimal treatment for HCC is determined after consideration of the extent of disease as well as the overall condition of the patient, with particular attention given to baseline hepatic function. There are a number of methods used to make a determination of the liver function, which include the ChildePugh score, the MELD score, and ICG clearance. None of these tests have emerged as the universal best assessment of preoperative liver function. The discriminatory ability of the Child score is limited by its use of subjective parameters such as ascites and encephalopathy. The ICG clearance, which is utilized mostly in Europe and Asia, is rarely performed in the United States. Although the MELD score has been shown to correlate with postoperative outcomes after liver resection for HCC [16,18], there are limited studies that evaluate risk factors associated with poor outcomes in patients who have low MELD scores and undergo major hepatic resections. Moreover, none of these scoring systems incorporate an assessment of PHT, which has been shown to be associated with postoperative liver insufficiency and increased morbidity [21,34,35]. Significant PHT is usually considered a contraindication to liver resection, as recommended by the EASL [22] and AASLD [23] guidelines for HCC treatment.
Mortality Overall morbidity Serious morbidity Wound complication Cardiopulmonary complication Postoperative transfusion Septic complication Return to the operating room Prolonged LOS Postoperative complication Pneumonia Unplanned reintubation Vent dependence >48 h DVT/PE Renal insufficiency/failure Cardiac arrest/MI Sepsis Septic shock
AOR (95% CI; platelet count <100 versus 100) 4.19 1.18 1.41 0.63 1.96 1.60 1.96 0.77 1.16
(2.18e8.04)* (0.77e1.83) (0.88e2.26) (0.31e1.28) (1.12e3.43)* (1.02e2.60)* (1.13e3.41)* (0.23e2.51) (0.73e1.82)
2.00 2.64 1.71 0.27 4.96 1.65 1.20 3.88
(0.95e4.20)y (1.39e4.99)* (0.82e3.58) (0.03e2.00) (2.60e9.45)* (0.48e5.66) (0.56e2.59) (1.94e7.75)*
MI ¼ myocardial infarction. AOR obtained after controlling for age, gender, BMI, ASA, MELD score, and type of hepatectomy. * P 0.05, yP ¼ 0.06.
The gold standard for measuring PHT is direct measurement of HVPG, and a pressure 10 mm Hg has been shown to be associated with postoperative liver decompensation [21]. However, direct measurement of HVPG has some limitations, such as cost, invasiveness, and the need for specific training to conduct and interpret the test, which may not be available in all centers [36]. Thus, the Barcelona Clinic Liver Cancer staging [37], as well as the EASL and AASLD guidelines, state clinically significant PHT as a contraindication for liver resection. However, PHT in the absence of other clinical signs, such as ascites, encephalopathy, and esophageal varices, could be easily underestimated during the selection criteria for hepatic resection, especially in the setting of a normal ChildePugh classification or MELD scores. With the goal to improve the selection criteria for liver resection, the focus of this study was to evaluate the association of preoperative platelet count, as a widely available and noninvasive indicator of PHT, with postoperative outcomes in patients undergoing hepatic resection for HCC using the NSQIP database. Although the EASL and AASLD guidelines clearly define indications for hepatic resection for HCC, patients with single HCC and completely preserved liver function without PHT, this recommendation has been subject to considerable debate [38,39]. Cucchetti et al. [39] retrospectively analyzed 241 cirrhotic patients who underwent resection for HCC and used propensity score matching to overcome the bias due to patient comorbidities. They concluded that on one-to-one matching, patients with and without PHT had the same preoperative characteristics and showed the same intraoperative course, postoperative occurrence of liver failure, morbidity, length of inhospital stay, and survival rate. In their study, the only predictors of postoperative liver failure were MELD score and
6
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8
extent of hepatectomy. Sugimachi et al. [38] evaluated 15 of 341 patients who had severe thrombocytopenia and underwent hepatic resection for HCC. They concluded that severe thrombocytopenia was not a contraindication for hepatectomy. On the contrary, numerous studies have supported the AASLD and EASL guidelines including the recent metaanalyses by Tang et al. [34] and Choi et al. [35], which have reported higher rates of morbidity and mortality in patients with PHT after hepatic resection for HCC. Maithel et al. [40] performed a multi-institutional analysis of 231 patients treated with liver resection for HCC and concluded that a platelet count of <150/nL was independently associated with an increased risk of 60-d mortality (odds ratio [OR], 4.6; P ¼ 0.009), postoperative liver insufficiency, as defined by a peak bilirubin >7 mg/dL or development of ascites (OR, 4.0; P ¼ 0.008), and increased risk of major complications (OR, 2.8; P ¼ 0.026). In the present study, we evaluated 2097 patients who underwent hepatic resection for HCC using the NSQIP database. To our knowledge, this is the largest series of patient population studied to evaluate the association of preoperative platelet count, which may act as a placeholder for PHT and postoperative outcomes. We performed a multivariate analysis to evaluate the independent association of thrombocytopenia and to overcome the bias due to patient characteristics (age, sex, BMI, and ASA score), MELD score, and the extent of hepatic resection. A low platelet count was significantly associated with an increased risk of 30d mortality, cardiopulmonary complications, need for blood transfusion, septic complications, pneumonia, unplanned reintubations, renal insufficiency/failure, and septic shock. The increase in the risk of unplanned reintubations and renal insufficiency and/or failure may suggest worsening of liver function in the form of hepatopulmonary and hepatorenal syndrome. This further supports the findings of the study by Maithel et al. that thrombocytopenia is associated with increased incidence of postoperative liver insufficiency. In addition, platelet count was also seen to be incrementally associated with poor postoperative outcomes as observed by the increases in the magnitude of the AOR’s between a preoperative platelet count of <150/nL and <100/nL with morbidity and mortality. Even in patients with low MELD scores (<10), platelet count was independently and incrementally associated with poor outcomes. The predominant mechanism of the association between low platelet count and the increase in the incidence of postoperative morbidity and mortality is likely secondary to worsening PHT after resection [41,42]. Other studies have suggested that thrombocytopenia may greatly compromise the critical role that platelets have in the initiation of liver regeneration, as demonstrated in experimental rodent models [43,44]. Recently, Starlinger et al. [45] provided the first evidence in vivo that platelet-derived serotonin is relevant to liver regeneration in humans. Our study had several limitations related to its retrospective design and its inherent biases. Data on the ChildePugh score were unable to be calculated based on the variables available in the NSQIP database and hence were unable to be controlled for, as a potential confounder in the multivariate analysis. Data on the MELD score nevertheless were
calculated using preoperative laboratory parameters and used to account for the baseline liver function in the patients. There were missing variables encountered in the database. However, any potential bias due to data not being missing completely at random was accounted for by using multiple imputation for all regression analyses that included variables with missing data. The NSQIP is a large national database, the integrity of individual data points is dependent on numerous data-entry sites, and opportunity exists for error at more than one site. Quality of data from NSQIP is maintained by continuity of indepth personnel training and study-site monitoring as discussed previously. The relationship between postoperative complications of reintubation, renal insufficiency, and septic complications with thrombocytopenia is not completely clear. We can speculate that these are surrogate for postoperative liver insufficiency. Because the NSQIP database does not capture postoperative liver insufficiency as a separate variable, future studies are warranted to evaluate this association further.
5.
Conclusions
A low preoperative platelet count is significantly and incrementally associated with adverse outcomes after hepatic resections for HCC, independent of the MELD score. Platelet count should be integrated to the selection criteria for hepatic resections for HCC. In patients with low platelet count, especially those with <100/nL, other curative therapies, such as ablation and transplantation, might be warranted instead of hepatectomy.
Acknowledgment Authors’ contributions: R.V. and F.B.M. contributed to the concept and design. R.V., J.R.H., and F.B.M. did the data acquisition and analysis. R.V., R.S.K., and F.B.M. did the drafting of the article and critical revisions.
Disclosure The authors declare that they have no conflicts of interests.
references
[1] El-Serag HB, Davila JA, Petersen NJ, McGlynn KA. The continuing increase in the incidence of hepatocellular carcinoma in the United States: an update. Ann Intern Med 2003;139:817. [2] Thompson Coon J, Rogers G, Hewson P, et al. Surveillance of cirrhosis for hepatocellular carcinoma: systematic review and economic analysis. Health Technol Assess 2007;11:1. [3] Nordenstedt H, White DL, El-Serag HB. The changing pattern of epidemiology in hepatocellular carcinoma. Dig Liver Dis 2010;42(Suppl 3):S206. [4] American Cancer Society. Cancer facts and figures 2014. Atlanta, GA: American Cancer Society; 2014.
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8
[5] Llovet JM, Di Bisceglie AM, Bruix J, et al. Design and endpoints of clinical trials in hepatocellular carcinoma. J Natl Cancer Inst 2008;100:698. [6] Forner A, Reig ME, de Lope CR, Bruix J. Current strategy for staging and treatment: the BCLC update and future prospects. Semin Liver Dis 2010;30:61. [7] Llovet JM, Bruix J, Gores GJ. Surgical resection versus transplantation for early hepatocellular carcinoma: clues for the best strategy. Hepatology 2000;31:1019. [8] Farinati F, Giacomin A, Vanin V, et al. Liver transplantation for hepatocellular carcinoma in clinical practice: the lesson from a 20-year multicentre experience in Italy. Eur J Gastroenterol Hepatol 2012;24:195. [9] Fu C, Liu N, Deng Q, Li X, Ma K, Bie P. Radiofrequency ablation vs. surgical resection on the treatment of patients with small hepatocellular carcinoma: a system review and meta-analysis of five randomized controlled trials. Hepatogastroenterology 2014;61:1722. [10] Jarnagin WR, Gonen M, Fong Y, et al. Improvement in perioperative outcome after hepatic resection: analysis of 1,803 consecutive cases over the past decade. Ann Surg 2002; 236:397. discussion 406e407. [11] Poon RT, Fan ST, Lo CM, et al. Improving perioperative outcome expands the role of hepatectomy in management of benign and malignant hepatobiliary diseases: analysis of 1222 consecutive patients from a prospective database. Ann Surg 2004;240:698. discussion 708e710. [12] Schiffman SC, Woodall CE, Kooby DA, et al. Factors associated with recurrence and survival following hepatectomy for large hepatocellular carcinoma: a multicenter analysis. J Surg Oncol 2010;101:105. [13] Lim KC, Wang VW, Siddiqui FJ, et al. Cost-effectiveness analysis of liver resection versus transplantation for early hepatocellular carcinoma within the Milan criteria. Hepatology 2015;61:227. [14] Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. The Br J Surg 1973;60:646. [15] Lau H, Man K, Fan ST, Yu WC, Lo CM, Wong J. Evaluation of preoperative hepatic function in patients with hepatocellular carcinoma undergoing hepatectomy. The Br J Surg 1997;84: 1255. [16] Cucchetti A, Ercolani G, Vivarelli M, et al. Impact of model for end-stage liver disease (MELD) score on prognosis after hepatectomy for hepatocellular carcinoma on cirrhosis. Liver Transpl 2006;12:966. [17] Teh SH, Sheppard BC, Schwartz J, Orloff SL. Model for endstage liver disease score fails to predict perioperative outcome after hepatic resection for hepatocellular carcinoma in patients without cirrhosis. Am J Surg 2008;195: 697. [18] Delis SG, Bakoyiannis A, Biliatis I, Athanassiou K, Tassopoulos N, Dervenis C. Model for end-stage liver disease (MELD) score, as a prognostic factor for post-operative morbidity and mortality in cirrhotic patients, undergoing hepatectomy for hepatocellular carcinoma. HPB 2009;11:351. [19] Hsu KY, Chau GY, Lui WY, Tsay SH, King KL, Wu CW. Predicting morbidity and mortality after hepatic resection in patients with hepatocellular carcinoma: the role of model for end-stage liver disease score. World J Surg 2009;33:2412. [20] Lam CM, Fan ST, Lo CM, Wong J. Major hepatectomy for hepatocellular carcinoma in patients with an unsatisfactory indocyanine green clearance test. The Br J Surg 1999;86:1012. [21] Bruix J, Castells A, Bosch J, et al. Surgical resection of hepatocellular carcinoma in cirrhotic patients: prognostic value of preoperative portal pressure. Gastroenterology 1996; 111:1018.
7
[22] Bruix J, Sherman M, Llovet JM, et al. Clinical management of hepatocellular carcinoma. Conclusions of the Barcelona2000 EASL conference. European Association for the Study of the Liver. J Hepatol 2001;35:421. [23] Bruix J, Sherman M. Practice guidelines committee AAftSoLD. Management of hepatocellular carcinoma. Hepatology 2005;42:1208. [24] Khuri SF, Daley J, Henderson W, et al. The Department of Veterans Affairs’ NSQIP: the first national, validated, outcome-based, risk-adjusted, and peer-controlled program for the measurement and enhancement of the quality of surgical care. National VA Surgical Quality Improvement Program. Ann Surg 1998;228:491. [25] Khuri SF, Henderson WG, Daley J, et al. Successful implementation of the department of Veterans Affairs’ national surgical quality improvement program in the private sector: the patient safety in surgery study. Ann Surg 2008;248:329. [26] Al-Refaie WB, Parsons HM, Henderson WG, et al. Major cancer surgery in the elderly: results from the American College of Surgeons National Surgical Quality Improvement Program. Ann Surg 2010;251:311. [27] Merkow RP, Bilimoria KY, McCarter MD, Bentrem DJ. Effect of body mass index on short-term outcomes after colectomy for cancer. J Am Coll Surg 2009;208:53. [28] Malinchoc M, Kamath PS, Gordon FD, Peine CJ, Rank J, ter Borg PC. A model to predict poor survival in patients undergoing transjugular intrahepatic portosystemic shunts. Hepatology 2000;31:864. [29] Venkat R, Valdivia PL, Guerrero MA. Resident participation and postoperative outcomes in adrenal surgery. The J Surg Res 2014;190:559. [30] Ingraham AM, Cohen ME, Bilimoria KY, Pritts TA, Ko CY, Esposito TJ. Comparison of outcomes after laparoscopic versus open appendectomy for acute appendicitis at 222 ACS NSQIP hospitals. Surgery 2010;148:625. discussion 635e637. [31] Cohen ME, Bilimoria KY, Ko CY, Richards K, Hall BL. Variability in length of stay after colorectal surgery: assessment of 182 hospitals in the national surgical quality improvement program. Ann Surg 2009;250:901. [32] Rubin DB, Schenker N. Multiple imputation in health-care databases: an overview and some applications. Stat Med 1991;10:585. [33] Grazi GL, Ercolani G, Pierangeli F, et al. Improved results of liver resection for hepatocellular carcinoma on cirrhosis give the procedure added value. Ann Surg 2001;234:71. [34] Tang YH, Zhu WJ, Wen TF. Influence of clinically significant portal hypertension on hepatectomy for hepatocellular carcinoma: a meta-analysis. Asian Pac J Cancer Prev 2014;15: 1649. [35] Choi SB, Kim HJ, Song TJ, Ahn HS, Choi SY. Influence of clinically significant portal hypertension on surgical outcomes and survival following hepatectomy for hepatocellular carcinoma: a systematic review and metaanalysis. J hepato-biliary-pancreatic Sci 2014;21:639. [36] Hernandez-Gea V, Turon F, Berzigotti A, Villanueva A. Management of small hepatocellular carcinoma in cirrhosis: focus on portal hypertension. World J Gastroenterol 2013;19: 1193. [37] Llovet JM, Bru C, Bruix J. Prognosis of hepatocellular carcinoma: the BCLC staging classification. Semin Liver Dis 1999;19:329. [38] Sugimachi K, Ikeda Y, Tomikawa M, et al. Appraisal of hepatic resection in the treatment of hepatocellular carcinoma with severe thrombocytopenia. World J Surg 2008; 32:1077.
8
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8
[39] Cucchetti A, Ercolani G, Vivarelli M, et al. Is portal hypertension a contraindication to hepatic resection? Ann Surg 2009;250:922. [40] Maithel SK, Kneuertz PJ, Kooby DA, et al. Importance of low preoperative platelet count in selecting patients for resection of hepatocellular carcinoma: a multi-institutional analysis. J Am Coll Surg 2011;212:638. discussion 648e650. [41] Boleslawski E, Petrovai G, Truant S, et al. Hepatic venous pressure gradient in the assessment of portal hypertension before liver resection in patients with cirrhosis. The Br J Surg 2012;99:855. [42] Giannini EG, Savarino V, Farinati F. Influence of clinically significant portal hypertension on survival after hepatic
resection for hepatocellular carcinoma in cirrhotic patients. Liver Int 2013;33:1594. [43] Alkozai EM, Nijsten MW, de Jong KP, et al. Immediate postoperative low platelet count is associated with delayed liver function recovery after partial liver resection. Ann Surg 2010;251:300. [44] Murata S, Ohkohchi N, Matsuo R, Ikeda O, Myronovych A, Hoshi R. Platelets promote liver regeneration in early period after hepatectomy in mice. World J Surg 2007;31: 808. [45] Starlinger P, Assinger A, Haegele S, et al. Evidence for serotonin as a relevant inducer of liver regeneration after liver resection in humans. Hepatology 2014;60:257.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Preoperative thrombocytopenia and outcomes of hepatectomy for hepatocellular carcinoma Raghunandan Venkat, MD, MPH,a Jack R. Hannallah, MD, MPH, MBA,a Robert S. Krouse, MD, FACS,a,b and Felipe B. Maegawa, MD, FACSa,b,* a b
Department of Surgery, University of Arizona, Tucson, Arizona Surgical Care Line, Southern Arizona VA Health Care System, Tucson, Arizona
article info
abstract
Article history:
Background: Platelet count is known to be an indirect indicator of portal hypertension but is
Received 31 May 2015
not a part of the model for end-stage liver disease (MELD) score or the ChildePugh score for
Received in revised form
risk stratification in hepatobiliary surgery.
25 July 2015
Methods: Data from 2097 hepatic resections for hepatocellular carcinoma (HCC) were
Accepted 21 August 2015
evaluated from 2005e2012 using the National Surgical Quality Improvement Program
Available online xxx
database. Patient demographics, morbidity, and mortality were evaluated. Results: Median age and body mass index were 64 y and 26.5 kg/m2, respectively. Majority of
Keywords:
the patients had American Society of Anesthesiologists 3 (78.1%) and median MELD score
Hepatectomy
was 7. On multivariate analysis, thrombocytopenia (platelet count <150/nL) and severe
Hepatocellular carcinoma
thrombocytopenia (platelet count <100/nL) were independently associated with an
Clinical outcomes
increased risk of mortality (odds ratio [OR], 1.79; P ¼ 0.024 and OR, 4.19; P < 0.001), car-
Thrombocytopenia
diopulmonary complications (OR, 1.61; P ¼ 0.009 and OR, 1.96; P ¼ 0.018), need for blood
Portal hypertension
transfusion (OR, 1.35; P ¼ 0.05 and OR, 1.60; P ¼ 0.05), septic complications (OR, 1.53; P ¼ 0.025 and OR, 1.96; P ¼ 0.016), reintubation (OR, 1.91; P ¼ 0.004 and OR, 2.64; P ¼ 0.003),
NSQIP
and renal insufficiency and/or failure (OR, 2.48; P ¼ 0.001 and OR, 4.96; P < 0.001), respectively. Conclusions: Thrombocytopenia, which is an indirect indicator for portal hypertension, is significantly associated with adverse outcomes after hepatectomy, independent of the MELD score. Platelet count should be integrated into the selection criteria for hepatic resections for HCC. ª 2015 Elsevier Inc. All rights reserved.
1.
Introduction
Hepatocellular carcinoma (HCC) with cirrhosis is the most common primary tumor of the liver and is the fifth most common malignancy worldwide [1,2]. In patients with cirrhosis, HCC has an annual incidence of 3%e5% [3]. In the United States, it is estimated that there will be 33,190 new
cases diagnosed in 2014 and 23,000 deaths due to this disease [4]. The surgical treatment options for HCC, which offer the potential for cure, include liver transplantation, hepatic resection, and percutaneous ablation [5e7]. Limited organ availability for transplantation [8], better oncological outcomes than percutaneous ablation [9], and improvement in perioperative care over the recent years [10,11] have made
Oral presentation at the scientific article sessions, American College of Surgeons 2014 Annual Clinical Congress, San Francisco. * Corresponding author. Surgical Care Line 2-112, Southern Arizona VA Health Care System, 3601 South 6th Avenue, Tucson, AZ 85723. Tel.: þ1 520 792 1450 (6156); fax: þ1 520 629 4603. E-mail addresses: [email protected], [email protected] (F.B. Maegawa). 0022-4804/$ e see front matter ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2015.08.038
2
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8
hepatic resection the mainstay of treatment for HCC, especially for patients with normal liver function or wellcompensated cirrhosis [12]. Furthermore, hepatectomy has been shown to be more cost-effective than liver transplantation [13]. Careful patient selection with assessment of the tumor burden and residual liver function are essential to optimize surgical outcomes after hepatic resection. However, the definition of selection criteria for hepatic resection is far from being established. Several tools can assess liver function. Traditionally, the ChildePugh score has been widely used due to its ease of application [14]. However, because it uses subjective parameters such as ascites and encephalopathy, there are limitations to its discriminatory abilities [15]. The model for end-stage liver disease (MELD) score, which was first introduced to predict survival after transjugular intrahepatic portosystemic shunts, has been shown to be an efficient method to evaluate hepatic function and predict outcomes after liver resection [16e19]. Another method to evaluate liver function includes measurement of indocyanine green (ICG) clearance, which is more frequently used in Korea and Japan [20]. The presence of portal hypertension (PHT) is also indicative of poor liver function and may be associated with postoperative liver insufficiency and increased morbidity [21]. Significant PHT is usually considered a contraindication to liver resection, as recommended by the European Association for Study of Liver (EASL) [22] and American Associations for Study of Liver Diseases (AASLD) [23] guidelines for HCC treatment. PHT can be identified either instrumentally by hepatic venous pressure gradient (HVPG) measurement or clinically by the presence of esophageal and gastric varices, and a low platelet count associated with splenomegaly. Chronic PHT results in a vascular congestive splenomegaly that triggers a platelet sequestration causing thrombocytopenia. Preoperative platelet count may thus serve as an inexpensive and noninvasive laboratory indicator of PHT. PHT clinically manifested by ascites and encephalopathy is a clear contraindication for major liver resection in most centers. However, in some patients the only sign of PHT is low platelet count, lacking the other clinical signs described previously. In this setting, the ChildePugh classification or MELD scores are within normal limits, misleading an adequate patient selection. In this context, the focus of this study was to evaluate the association of preoperative platelet count, as a noninvasive marker for PHT, with postoperative outcomes in patients undergoing hepatic resection for HCC, independent of their MELD score. We hypothesized that preoperative thrombocytopenia was independently associated with increased postoperative morbidity and mortality, whereas accounting for patient comorbidities, MELD score, and the extent of liver resection.
2.
Methods
2.1.
Data source
We used the 2005e2012 American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP)
database participant use file. The ACS-NSQIP obtains data on patients undergoing inpatient and outpatient surgical procedures from more than 250 university and private sector medical centers. Preoperative patient characteristics, intraoperative procedure characteristics, and 30-d postoperative mortality and complications, including more than 136 variables, are recorded in the data set and are well described elsewhere in the literature [24,25]. The University of Arizona Institutional Review Board determined that this study qualified for an exemption under the Department of Health and Human Studies regulations because of the use of deidentified administrative data. We identified 2097 patients who underwent hepatic resection for HCC. The hepatic resections were classified into partial lobectomy (current procedural terminology [CPT] 47120), total left lobectomy (CPT 47125), total right lobectomy (CPT 47130), and trisegmentectomy (CPT 47122). HCC was defined by the International Classification of Diseases, Ninth Revision diagnosis code for primary liver cancer (155.0). Patient characteristics were defined from reported NSQIP variables as follows: race and/or ethnicity was dichotomized to Caucasian, AfricaneAmerican, and other, American Society of Anesthesiologists (ASA) classification was dichotomized to 3 or higher or less than 3, and smoking status was dichotomized as current smoker within 1 y or not. Age and body mass index (BMI, calculated as weight in kilograms divided by height in meters squared) were evaluated as continuous variables as well as classified in categories based on prior research examining the effect of age [26] and BMI [27] on postoperative complications. Age (in y) was classified in four categories: 49, 50e64, 65e79, and 80. BMI (in kg/m2) was classified in five categories: <18.5, 18.5e24.9, 25.0e29.9, 30.0e34.9, and 35.0.
2.2.
Assessment of preoperative liver function
The MELD score was calculated from preoperative laboratory variables using the standard equation: MELD ¼ 9.57 loge (creatinine mg/dL) þ 3.78 loge(bilirubin mg/dL) þ 11.20 loge(INR) þ 6.43 [28]. Thrombocytopenia was defined as a platelet count <150/nL, and severe thrombocytopenia was defined as a platelet count of <100/nL. ChildePugh score was not calculated because of the lack of the degree of ascites and encephalopathy in the ACS-NSQIP data set.
2.3.
Definition of perioperative outcomes
Thirty-day outcomes were evaluated and included mortality, return to the operating room, postoperative complications, serious morbidity, and overall morbidity. Postoperative complications were categorized as previously described in the literature [29]: (1) wound complications (dehiscence, organ space surgical site infection [SSI], superficial SSI, and deep incisional SSI); (2) cardiopulmonary complications (cardiac arrest requiring cardiopulmonary resuscitation, myocardial infarction, ventilator dependence >48 h, pulmonary embolism (PE), unplanned intubation, and cerebrovascular accident); (3) postoperative transfusion; and (4) septic complications (septic shock and sepsis). The definition of serious morbidity was adapted from Ingraham et al. [30] as
3
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8
having documentation of at least one of the following ACSNSQIP complications: organ space SSI, wound dehiscence, neurologic event (cerebrovascular accident or coma lasting >24 h), cardiac arrest, myocardial infarction, PE, ventilator dependence >48 h, progressive or acute renal insufficiency, and sepsis or septic shock. Overall morbidity was defined as documentation of a serious morbidity or at least one of the following ACS-NSQIP complications: superficial SSI, deep SSI, pneumonia, unplanned intubation (without preoperative ventilator dependence), urinary tract infection, peripheral neurologic deficit, or deep vein thrombosis (DVT). Prolonged length of stay (LOS) was defined as an LOS above the 75th percentile, consistent with the convention used in other NSQIP studies regarding LOS [31].
2.4.
Statistical analysis
For unadjusted comparison between our groups, we used the t-test and/or analysis of variance or the Wilcoxon rank-sum test for continuous variables and the chi-squared or the Fisher exact test for categorical variables, as appropriate. Variables demonstrating association with each end point to the level of P 0.25 on univariate analysis were entered into a multivariate model. A P value of 0.05 in the final model was considered statistically significant. To account for missing data and minimize any potential bias due to data not being missing completely at random, multiple imputation using chained equations with 10 imputations was used for all regression analyses that included variables with missing data [32]. Statistical analysis was performed using Intercooled Stata, version 12.0 (StataCorp, College Station, TX).
3.
Results
3.1.
Operative and patient characteristics
A total of 2097 hepatic resections were performed for HCC during 2005e2012 (partial lobectomy: 1166 [55.6%], left hemihepatectomy: 262 [12.5%], right hemihepatectomy: 427 [20.4%], and trisegmentectomy: 242 [11.5%]). Median age was 64 y (interquartile range, 56e72 y). A total of 67.3% of the patients were male (n ¼ 1411), and the majority were white (n ¼ 1344, 64.1%). Median BMI was 26.5 kg/m2 (interquartile range, 23.4e30.3 kg/m2), and the majority of the patients were ASA class 3 or greater (n ¼ 1637, 78.1%). Laboratory data to calculate the MELD score were available in 1841 patients (256 missing values). The median MELD score was 7, and 19.2% patients had MELD 10. Data on platelet count were available in 2051 patients (46 missing values). Median platelet count was 205 per nL. Patients having thrombocytopenia were 22.7% (n ¼ 465), and 6.1% patients (n ¼ 125) had severe thrombocytopenia (Table 1).
3.2.
Perioperative outcomes
Outcomes were stratified by the presence of a preoperative platelet count <150/nL (thrombocytopenia) versus 150/nL and preoperative platelet count <100/nL (severe thrombocytopenia) versus 100/nL. On univariate analysis, patients with
Table 1 e Patient characteristics and comorbidities. Variable
n ¼ 2097
Age (y, median [IQR]) Age category 49 50e64 65e79 80 Sex Female Male Race White AfricaneAmerican Others ASA classification <3 3 BMI (kg/m2, median [IQR]) BMI category <18.5 18.5e24.9 25.0e29.9 30.0e34.9 35.0 Current smoker Current alcohol use Diabetes Hypertension CHF MI (last 6 mo) COPD Dialysis Surgery Partial lobectomy Left hemihepatectomy Right hemihepatectomy Trisegmentectomy Platelet count* (per nL, median [IQR]) 150 100e150 <100 MELD scorey (median [IQR]) MELD 10
64 [56e72] 265 839 864 129
(12.6) (40.0) (41.2) (6.2)
686 (32.7) 1411 (67.3) 1344 (64.1) 207 (9.9) 546 (26.0) 460 (21.9) 1637 (78.1) 25.6 [23.4e30.3] 59 759 721 361 197 470 75 561 1249 5 5 106 7
(2.8) (36.2) (34.4) (17.2) (9.4) (22.4) (3.6) (26.7) (59.6) (0.2) (0.2) (5.0) (0.3)
1166 (55.6) 262 (12.5) 427 (20.4) 242 (11.5) 205 [153e275] 1856 (77.3) 340 (16.6) 125 (6.1) 7 [6e9] 353 (19.2)
CHF ¼ congestive heart failure; COPD ¼ chronic obstructive pulmonary disease; IQR ¼ interquartile range; MI ¼ myocardial infarction; nL ¼ nanoliter. Data expressed in n (%) unless otherwise specified. * Platelet count was missing in 46 patients. y MELD score was missing in 256 patients.
thrombocytopenia had a significantly increased rate of septic complications (11.6% versus 8.5%, P ¼ 0.04), pneumonia (6.4% versus 4.0%, P ¼ 0.02), unplanned reintubation rate (7.7% versus 5.2%, P ¼ 0.04), ventilator dependence >48 h (6.7% versus 4.6%, P ¼ 0.05), renal insufficiency and/or failure (6.0% versus 3.0%, P < 0.01), and septic shock (5.6% versus 3.3%, P ¼ 0.03). There was no significant difference between the two groups (platelet count <150/nL and 150 nL) in the rates of 30-d mortality, overall morbidity, serious morbidity, wound complication, cardiopulmonary complication, need for postoperative blood transfusion, reoperation, prolonged LOS, DVT and/or PE, cardiac arrest and/or myocardial infarction, or sepsis (Table 2).
4
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8
To further evaluate the incremental effect of a low preoperative platelet count on postoperative outcomes, we analyzed the patients with platelet count <100/nL and 100/ nL. Severe preoperative thrombocytopenia was significantly associated with an increased risk of 30-d mortality (11.2% versus 4.0%, P < 0.01), septic complication (16% versus 8.8%, P < 0.01), pneumonia (8.0% versus 4.3%, P ¼ 0.05), unplanned reintubation rate (10.4% versus 5.5%, P ¼ 0.02), renal insufficiency and/or failure (12.0% versus 3.2%, P < 0.01), and septic shock (9.6% versus 3.5%, P < 0.01). There was no significant difference between the two groups (platelet count <100/nL and 100 nL) in the rates of overall morbidity, serious morbidity, wound complication, cardiopulmonary complication, need for postoperative blood transfusion, reoperation, prolonged LOS, ventilator dependence >48 h, DVT and/or PE, cardiac arrest and/or myocardial infarction, or sepsis (Table 3). A multivariate logistic regression analysis was performed, controlling for potential confounding factors of age, gender, BMI, ASA, MELD score, and the type of hepatectomy. Platelet count was seen to be independently and incrementally associated with poor postoperative outcomes (Tables 4 and 5). Platelet <150/nL was associated with a significant increase in the risk of 30-d mortality (adjusted odds ratio [AOR], 1.79; P ¼ 0.024), cardiopulmonary complications (AOR, 1.61; P ¼ 0.009), need for blood transfusions (AOR, 1.35; P ¼ 0.05), septic complications (AOR, 1.53; P ¼ 0.025), pneumonia (AOR, 1.71; P ¼ 0.032), unplanned reintubations (AOR, 1.91; P ¼ 0.004), ventilator dependence >48 h (AOR, 1.98; P ¼ 0.004), renal insufficiency and/or failure (AOR, 2.48; P ¼ 0.001), and septic shock (AOR, 2.36; P ¼ 0.002).
Table 2 e Univariate analysis comparing postoperative outcomes with preoperative platelet count. (‡150 per nL versus <150 per nL). Variable
Platelet count (per nL) 150 (n ¼ 1586)
Mortality Overall morbidity Serious morbidity Wound complication Cardiopulmonary complication Postoperative transfusion Septic complication Return to the operating room Prolonged LOS Postoperative complication Pneumonia Unplanned reintubation Vent dependence >48 h DVT/PE Renal insufficiency/failure Cardiac arrest/MI Sepsis Septic shock
<150 (n ¼ 465)
P value
66 392 282 194 147
(4.2) (24.7) (17.8) (12.2) (9.3)
26 123 84 45 52
(5.6) (26.4) (18.1) (9.7) (11.2)
0.19 0.45 0.89 0.13 0.22
248 135 64 377
(15.6) (8.5) (4.0) (23.8)
84 54 13 109
(18.1) (11.6) (2.8) (23.4)
0.21 0.04 0.22 0.88
63 83 73 61 48 35 87 53
(4.0) (5.2) (4.6) (3.8) (3.0) (2.2) (5.5) (3.3)
30 36 32 11 28 8 30 26
(6.4) (7.7) (6.7) (2.4) (6.0) (1.7) (6.4) (5.6)
MI ¼ myocardial infarction; vent ¼ ventilator. Data expressed in n (%) unless otherwise specified.
0.02 0.04 0.05 0.12 <0.01 0.52 0.43 0.03
Table 3 e Univariate analysis comparing postoperative outcomes with preoperative platelet count. (‡100 per nL versus <100 per nL). Variable
Platelet count (per nL) 100 (n ¼ 1926)
Mortality Overall morbidity Serious morbidity Wound complication Cardiopulmonary complication Postoperative transfusion Septic complication Return to the operating room Prolonged LOS Postoperative complication Pneumonia Unplanned reintubation Vent dependence >48 h DVT/PE Renal insufficiency/failure Cardiac arrest/MI Sepsis Septic shock
<100 (n ¼ 125)
P value
78 479 338 227 182
(4.0) (24.9) (17.5) (11.8) (9.4)
14 36 28 12 17
(11.2) (28.8) (22.4) (9.6) (13.6)
<0.01 0.33 0.17 0.46 0.13
308 169 73 452
(16.0) (8.8) (3.8) (23.5)
24 20 4 34
(19.2) (16.0) (3.2) (27.2)
0.34 <0.01 0.74 0.34
83 106 96 69 61 40 107 67
(4.3) (5.5) (5.0) (3.6) (3.2) (2.1) (5.6) (3.5)
10 13 9 3 15 3 10 12
(8.0) (10.4) (7.2) (2.4) (12.0) (2.4) (8.0) (9.6)
0.05 0.02 0.28 0.49 <0.01 0.81 0.25 <0.01
MI ¼ myocardial infarction; vent ¼ ventilator. Data expressed in n (%) unless otherwise specified.
On multivariate analysis of patients with a platelet count <100/nL, we observed a significant worsening of postoperative outcomes. There was an increased risk of 30-d mortality (AOR, 4.19; P < 0.001), cardiopulmonary complications (AOR, 1.96; P ¼ 0.018), need for blood transfusions (AOR, 1.60; P ¼ 0.05), septic complications (AOR, 1.96; P ¼ 0.016), unplanned reintubations (AOR, 2.64; P ¼ 0.003), renal insufficiency and/or failure (AOR, 4.96; P < 0.001), and septic shock (AOR, 2.36; P < 0.001). There was a trend toward an increased risk of postoperative pneumonia (AOR, 2.00; P ¼ 0.06), whereas there was no association with ventilator dependence >48 h. We then performed a subset multivariate analysis of 1488 patients with an MELD score <10 and evaluated the association of thrombocytopenia and severe thrombocytopenia with postoperative outcomes. A platelet count of <150/nL was associated with a significant increase in the need for postoperative blood transfusion (AOR, 1.49; 95% confidence interval [CI], 1.05e2.12), ventilator dependence >48 h (AOR, 1.90; 95% CI, 1.09e3.33), renal insufficiency and/or failure (AOR, 2.09; 95% CI, 1.07e4.08), and septic shock (AOR, 2.00; 95% CI, 1.02e3.91). We observed that in this patient population of MELD <10, a platelet count <100/nL was associated with a profound worsening of postoperative outcomes. There was a significant increase in the risk of 30-d mortality (AOR, 5.24; 95% CI, 2.13e12.89), cardiopulmonary complications (AOR, 2.51; 95% CI, 1.25e5.06), septic complications (AOR, 2.56; 95% CI, 1.25e5.27), need for postoperative transfusion (AOR, 1.82; 95% CI, 1.02e3.34), reintubation, (AOR, 3.17; 95% CI, 1.411e7.11), ventilator dependence >48 h (AOR, 2.57; 95% CI, 1.04e6.33), renal insufficiency and/or failure (AOR, 6.87; 95% CI, 2.86e16.52), and septic shock (AOR, 5.77; 95% CI, 2.34e14.21).
5
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8
Table 4 e Multivariate analysis evaluating postoperative outcomes with preoperative platelet count of <150 per nL.
Table 5 e Multivariate analysis evaluating postoperative outcomes with preoperative platelet count of <100 per nL.
Outcome measure
Outcome measure
Mortality Overall morbidity Serious morbidity Wound complication Cardiopulmonary complication Postoperative transfusion Septic complication Return to the operating room Prolonged LOS Postoperative complication Pneumonia Unplanned reintubation Vent dependence >48 h DVT/PE Renal insufficiency/failure Cardiac arrest/MI Sepsis Septic shock
AOR (95% CI; platelet count <150 versus 150) 1.79 1.15 1.14 0.70 1.61 1.35 1.53 0.79 1.05
(1.08e2.97)* (0.89e1.49) (0.85e1.54) (0.47e1.10) (1.12e2.32)* (1.01e1.83)* (1.05e2.21)* (0.41e1.52) (0.80e1.38)
1.71 1.91 1.98 0.56 2.48 0.85 1.07 2.36
(1.05e2.80)* (1.23e2.94)* (1.25e3.14)* (0.26e1.20) (1.49e4.15)* (0.38e1.90) (0.66e1.74) (1.39e4.00)*
MI ¼ myocardial infarction. AOR obtained after controlling for age, gender, BMI, ASA, MELD score, and type of hepatectomy. * P 0.05.
4.
Discussion
The outcome of hepatic resection in cirrhotic patients has improved significantly over the last decade. With improvement in surgical techniques and perioperative care, recently published large series of liver resections for HCC have reported good results, with a significant reduction in postoperative morbidity, mortality, and liver decompensation [11,33]. These results have largely been possible due to the importance given to careful patient selection. The optimal treatment for HCC is determined after consideration of the extent of disease as well as the overall condition of the patient, with particular attention given to baseline hepatic function. There are a number of methods used to make a determination of the liver function, which include the ChildePugh score, the MELD score, and ICG clearance. None of these tests have emerged as the universal best assessment of preoperative liver function. The discriminatory ability of the Child score is limited by its use of subjective parameters such as ascites and encephalopathy. The ICG clearance, which is utilized mostly in Europe and Asia, is rarely performed in the United States. Although the MELD score has been shown to correlate with postoperative outcomes after liver resection for HCC [16,18], there are limited studies that evaluate risk factors associated with poor outcomes in patients who have low MELD scores and undergo major hepatic resections. Moreover, none of these scoring systems incorporate an assessment of PHT, which has been shown to be associated with postoperative liver insufficiency and increased morbidity [21,34,35]. Significant PHT is usually considered a contraindication to liver resection, as recommended by the EASL [22] and AASLD [23] guidelines for HCC treatment.
Mortality Overall morbidity Serious morbidity Wound complication Cardiopulmonary complication Postoperative transfusion Septic complication Return to the operating room Prolonged LOS Postoperative complication Pneumonia Unplanned reintubation Vent dependence >48 h DVT/PE Renal insufficiency/failure Cardiac arrest/MI Sepsis Septic shock
AOR (95% CI; platelet count <100 versus 100) 4.19 1.18 1.41 0.63 1.96 1.60 1.96 0.77 1.16
(2.18e8.04)* (0.77e1.83) (0.88e2.26) (0.31e1.28) (1.12e3.43)* (1.02e2.60)* (1.13e3.41)* (0.23e2.51) (0.73e1.82)
2.00 2.64 1.71 0.27 4.96 1.65 1.20 3.88
(0.95e4.20)y (1.39e4.99)* (0.82e3.58) (0.03e2.00) (2.60e9.45)* (0.48e5.66) (0.56e2.59) (1.94e7.75)*
MI ¼ myocardial infarction. AOR obtained after controlling for age, gender, BMI, ASA, MELD score, and type of hepatectomy. * P 0.05, yP ¼ 0.06.
The gold standard for measuring PHT is direct measurement of HVPG, and a pressure 10 mm Hg has been shown to be associated with postoperative liver decompensation [21]. However, direct measurement of HVPG has some limitations, such as cost, invasiveness, and the need for specific training to conduct and interpret the test, which may not be available in all centers [36]. Thus, the Barcelona Clinic Liver Cancer staging [37], as well as the EASL and AASLD guidelines, state clinically significant PHT as a contraindication for liver resection. However, PHT in the absence of other clinical signs, such as ascites, encephalopathy, and esophageal varices, could be easily underestimated during the selection criteria for hepatic resection, especially in the setting of a normal ChildePugh classification or MELD scores. With the goal to improve the selection criteria for liver resection, the focus of this study was to evaluate the association of preoperative platelet count, as a widely available and noninvasive indicator of PHT, with postoperative outcomes in patients undergoing hepatic resection for HCC using the NSQIP database. Although the EASL and AASLD guidelines clearly define indications for hepatic resection for HCC, patients with single HCC and completely preserved liver function without PHT, this recommendation has been subject to considerable debate [38,39]. Cucchetti et al. [39] retrospectively analyzed 241 cirrhotic patients who underwent resection for HCC and used propensity score matching to overcome the bias due to patient comorbidities. They concluded that on one-to-one matching, patients with and without PHT had the same preoperative characteristics and showed the same intraoperative course, postoperative occurrence of liver failure, morbidity, length of inhospital stay, and survival rate. In their study, the only predictors of postoperative liver failure were MELD score and
6
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8
extent of hepatectomy. Sugimachi et al. [38] evaluated 15 of 341 patients who had severe thrombocytopenia and underwent hepatic resection for HCC. They concluded that severe thrombocytopenia was not a contraindication for hepatectomy. On the contrary, numerous studies have supported the AASLD and EASL guidelines including the recent metaanalyses by Tang et al. [34] and Choi et al. [35], which have reported higher rates of morbidity and mortality in patients with PHT after hepatic resection for HCC. Maithel et al. [40] performed a multi-institutional analysis of 231 patients treated with liver resection for HCC and concluded that a platelet count of <150/nL was independently associated with an increased risk of 60-d mortality (odds ratio [OR], 4.6; P ¼ 0.009), postoperative liver insufficiency, as defined by a peak bilirubin >7 mg/dL or development of ascites (OR, 4.0; P ¼ 0.008), and increased risk of major complications (OR, 2.8; P ¼ 0.026). In the present study, we evaluated 2097 patients who underwent hepatic resection for HCC using the NSQIP database. To our knowledge, this is the largest series of patient population studied to evaluate the association of preoperative platelet count, which may act as a placeholder for PHT and postoperative outcomes. We performed a multivariate analysis to evaluate the independent association of thrombocytopenia and to overcome the bias due to patient characteristics (age, sex, BMI, and ASA score), MELD score, and the extent of hepatic resection. A low platelet count was significantly associated with an increased risk of 30d mortality, cardiopulmonary complications, need for blood transfusion, septic complications, pneumonia, unplanned reintubations, renal insufficiency/failure, and septic shock. The increase in the risk of unplanned reintubations and renal insufficiency and/or failure may suggest worsening of liver function in the form of hepatopulmonary and hepatorenal syndrome. This further supports the findings of the study by Maithel et al. that thrombocytopenia is associated with increased incidence of postoperative liver insufficiency. In addition, platelet count was also seen to be incrementally associated with poor postoperative outcomes as observed by the increases in the magnitude of the AOR’s between a preoperative platelet count of <150/nL and <100/nL with morbidity and mortality. Even in patients with low MELD scores (<10), platelet count was independently and incrementally associated with poor outcomes. The predominant mechanism of the association between low platelet count and the increase in the incidence of postoperative morbidity and mortality is likely secondary to worsening PHT after resection [41,42]. Other studies have suggested that thrombocytopenia may greatly compromise the critical role that platelets have in the initiation of liver regeneration, as demonstrated in experimental rodent models [43,44]. Recently, Starlinger et al. [45] provided the first evidence in vivo that platelet-derived serotonin is relevant to liver regeneration in humans. Our study had several limitations related to its retrospective design and its inherent biases. Data on the ChildePugh score were unable to be calculated based on the variables available in the NSQIP database and hence were unable to be controlled for, as a potential confounder in the multivariate analysis. Data on the MELD score nevertheless were
calculated using preoperative laboratory parameters and used to account for the baseline liver function in the patients. There were missing variables encountered in the database. However, any potential bias due to data not being missing completely at random was accounted for by using multiple imputation for all regression analyses that included variables with missing data. The NSQIP is a large national database, the integrity of individual data points is dependent on numerous data-entry sites, and opportunity exists for error at more than one site. Quality of data from NSQIP is maintained by continuity of indepth personnel training and study-site monitoring as discussed previously. The relationship between postoperative complications of reintubation, renal insufficiency, and septic complications with thrombocytopenia is not completely clear. We can speculate that these are surrogate for postoperative liver insufficiency. Because the NSQIP database does not capture postoperative liver insufficiency as a separate variable, future studies are warranted to evaluate this association further.
5.
Conclusions
A low preoperative platelet count is significantly and incrementally associated with adverse outcomes after hepatic resections for HCC, independent of the MELD score. Platelet count should be integrated to the selection criteria for hepatic resections for HCC. In patients with low platelet count, especially those with <100/nL, other curative therapies, such as ablation and transplantation, might be warranted instead of hepatectomy.
Acknowledgment Authors’ contributions: R.V. and F.B.M. contributed to the concept and design. R.V., J.R.H., and F.B.M. did the data acquisition and analysis. R.V., R.S.K., and F.B.M. did the drafting of the article and critical revisions.
Disclosure The authors declare that they have no conflicts of interests.
references
[1] El-Serag HB, Davila JA, Petersen NJ, McGlynn KA. The continuing increase in the incidence of hepatocellular carcinoma in the United States: an update. Ann Intern Med 2003;139:817. [2] Thompson Coon J, Rogers G, Hewson P, et al. Surveillance of cirrhosis for hepatocellular carcinoma: systematic review and economic analysis. Health Technol Assess 2007;11:1. [3] Nordenstedt H, White DL, El-Serag HB. The changing pattern of epidemiology in hepatocellular carcinoma. Dig Liver Dis 2010;42(Suppl 3):S206. [4] American Cancer Society. Cancer facts and figures 2014. Atlanta, GA: American Cancer Society; 2014.
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8
[5] Llovet JM, Di Bisceglie AM, Bruix J, et al. Design and endpoints of clinical trials in hepatocellular carcinoma. J Natl Cancer Inst 2008;100:698. [6] Forner A, Reig ME, de Lope CR, Bruix J. Current strategy for staging and treatment: the BCLC update and future prospects. Semin Liver Dis 2010;30:61. [7] Llovet JM, Bruix J, Gores GJ. Surgical resection versus transplantation for early hepatocellular carcinoma: clues for the best strategy. Hepatology 2000;31:1019. [8] Farinati F, Giacomin A, Vanin V, et al. Liver transplantation for hepatocellular carcinoma in clinical practice: the lesson from a 20-year multicentre experience in Italy. Eur J Gastroenterol Hepatol 2012;24:195. [9] Fu C, Liu N, Deng Q, Li X, Ma K, Bie P. Radiofrequency ablation vs. surgical resection on the treatment of patients with small hepatocellular carcinoma: a system review and meta-analysis of five randomized controlled trials. Hepatogastroenterology 2014;61:1722. [10] Jarnagin WR, Gonen M, Fong Y, et al. Improvement in perioperative outcome after hepatic resection: analysis of 1,803 consecutive cases over the past decade. Ann Surg 2002; 236:397. discussion 406e407. [11] Poon RT, Fan ST, Lo CM, et al. Improving perioperative outcome expands the role of hepatectomy in management of benign and malignant hepatobiliary diseases: analysis of 1222 consecutive patients from a prospective database. Ann Surg 2004;240:698. discussion 708e710. [12] Schiffman SC, Woodall CE, Kooby DA, et al. Factors associated with recurrence and survival following hepatectomy for large hepatocellular carcinoma: a multicenter analysis. J Surg Oncol 2010;101:105. [13] Lim KC, Wang VW, Siddiqui FJ, et al. Cost-effectiveness analysis of liver resection versus transplantation for early hepatocellular carcinoma within the Milan criteria. Hepatology 2015;61:227. [14] Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. The Br J Surg 1973;60:646. [15] Lau H, Man K, Fan ST, Yu WC, Lo CM, Wong J. Evaluation of preoperative hepatic function in patients with hepatocellular carcinoma undergoing hepatectomy. The Br J Surg 1997;84: 1255. [16] Cucchetti A, Ercolani G, Vivarelli M, et al. Impact of model for end-stage liver disease (MELD) score on prognosis after hepatectomy for hepatocellular carcinoma on cirrhosis. Liver Transpl 2006;12:966. [17] Teh SH, Sheppard BC, Schwartz J, Orloff SL. Model for endstage liver disease score fails to predict perioperative outcome after hepatic resection for hepatocellular carcinoma in patients without cirrhosis. Am J Surg 2008;195: 697. [18] Delis SG, Bakoyiannis A, Biliatis I, Athanassiou K, Tassopoulos N, Dervenis C. Model for end-stage liver disease (MELD) score, as a prognostic factor for post-operative morbidity and mortality in cirrhotic patients, undergoing hepatectomy for hepatocellular carcinoma. HPB 2009;11:351. [19] Hsu KY, Chau GY, Lui WY, Tsay SH, King KL, Wu CW. Predicting morbidity and mortality after hepatic resection in patients with hepatocellular carcinoma: the role of model for end-stage liver disease score. World J Surg 2009;33:2412. [20] Lam CM, Fan ST, Lo CM, Wong J. Major hepatectomy for hepatocellular carcinoma in patients with an unsatisfactory indocyanine green clearance test. The Br J Surg 1999;86:1012. [21] Bruix J, Castells A, Bosch J, et al. Surgical resection of hepatocellular carcinoma in cirrhotic patients: prognostic value of preoperative portal pressure. Gastroenterology 1996; 111:1018.
7
[22] Bruix J, Sherman M, Llovet JM, et al. Clinical management of hepatocellular carcinoma. Conclusions of the Barcelona2000 EASL conference. European Association for the Study of the Liver. J Hepatol 2001;35:421. [23] Bruix J, Sherman M. Practice guidelines committee AAftSoLD. Management of hepatocellular carcinoma. Hepatology 2005;42:1208. [24] Khuri SF, Daley J, Henderson W, et al. The Department of Veterans Affairs’ NSQIP: the first national, validated, outcome-based, risk-adjusted, and peer-controlled program for the measurement and enhancement of the quality of surgical care. National VA Surgical Quality Improvement Program. Ann Surg 1998;228:491. [25] Khuri SF, Henderson WG, Daley J, et al. Successful implementation of the department of Veterans Affairs’ national surgical quality improvement program in the private sector: the patient safety in surgery study. Ann Surg 2008;248:329. [26] Al-Refaie WB, Parsons HM, Henderson WG, et al. Major cancer surgery in the elderly: results from the American College of Surgeons National Surgical Quality Improvement Program. Ann Surg 2010;251:311. [27] Merkow RP, Bilimoria KY, McCarter MD, Bentrem DJ. Effect of body mass index on short-term outcomes after colectomy for cancer. J Am Coll Surg 2009;208:53. [28] Malinchoc M, Kamath PS, Gordon FD, Peine CJ, Rank J, ter Borg PC. A model to predict poor survival in patients undergoing transjugular intrahepatic portosystemic shunts. Hepatology 2000;31:864. [29] Venkat R, Valdivia PL, Guerrero MA. Resident participation and postoperative outcomes in adrenal surgery. The J Surg Res 2014;190:559. [30] Ingraham AM, Cohen ME, Bilimoria KY, Pritts TA, Ko CY, Esposito TJ. Comparison of outcomes after laparoscopic versus open appendectomy for acute appendicitis at 222 ACS NSQIP hospitals. Surgery 2010;148:625. discussion 635e637. [31] Cohen ME, Bilimoria KY, Ko CY, Richards K, Hall BL. Variability in length of stay after colorectal surgery: assessment of 182 hospitals in the national surgical quality improvement program. Ann Surg 2009;250:901. [32] Rubin DB, Schenker N. Multiple imputation in health-care databases: an overview and some applications. Stat Med 1991;10:585. [33] Grazi GL, Ercolani G, Pierangeli F, et al. Improved results of liver resection for hepatocellular carcinoma on cirrhosis give the procedure added value. Ann Surg 2001;234:71. [34] Tang YH, Zhu WJ, Wen TF. Influence of clinically significant portal hypertension on hepatectomy for hepatocellular carcinoma: a meta-analysis. Asian Pac J Cancer Prev 2014;15: 1649. [35] Choi SB, Kim HJ, Song TJ, Ahn HS, Choi SY. Influence of clinically significant portal hypertension on surgical outcomes and survival following hepatectomy for hepatocellular carcinoma: a systematic review and metaanalysis. J hepato-biliary-pancreatic Sci 2014;21:639. [36] Hernandez-Gea V, Turon F, Berzigotti A, Villanueva A. Management of small hepatocellular carcinoma in cirrhosis: focus on portal hypertension. World J Gastroenterol 2013;19: 1193. [37] Llovet JM, Bru C, Bruix J. Prognosis of hepatocellular carcinoma: the BCLC staging classification. Semin Liver Dis 1999;19:329. [38] Sugimachi K, Ikeda Y, Tomikawa M, et al. Appraisal of hepatic resection in the treatment of hepatocellular carcinoma with severe thrombocytopenia. World J Surg 2008; 32:1077.
8
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e8
[39] Cucchetti A, Ercolani G, Vivarelli M, et al. Is portal hypertension a contraindication to hepatic resection? Ann Surg 2009;250:922. [40] Maithel SK, Kneuertz PJ, Kooby DA, et al. Importance of low preoperative platelet count in selecting patients for resection of hepatocellular carcinoma: a multi-institutional analysis. J Am Coll Surg 2011;212:638. discussion 648e650. [41] Boleslawski E, Petrovai G, Truant S, et al. Hepatic venous pressure gradient in the assessment of portal hypertension before liver resection in patients with cirrhosis. The Br J Surg 2012;99:855. [42] Giannini EG, Savarino V, Farinati F. Influence of clinically significant portal hypertension on survival after hepatic
resection for hepatocellular carcinoma in cirrhotic patients. Liver Int 2013;33:1594. [43] Alkozai EM, Nijsten MW, de Jong KP, et al. Immediate postoperative low platelet count is associated with delayed liver function recovery after partial liver resection. Ann Surg 2010;251:300. [44] Murata S, Ohkohchi N, Matsuo R, Ikeda O, Myronovych A, Hoshi R. Platelets promote liver regeneration in early period after hepatectomy in mice. World J Surg 2007;31: 808. [45] Starlinger P, Assinger A, Haegele S, et al. Evidence for serotonin as a relevant inducer of liver regeneration after liver resection in humans. Hepatology 2014;60:257.