Impact of Abdominal Aortic Calcification Among Liver Transplantation Recipients

Transcription

1 ORIGINAL ARTICLE Impact of Abdominal Aortic Calcification Among Liver Transplantation Recipients Yuki Imaoka, 1 Masahiro Ohira, 1,2 Ryosuke Nakano, 1 Seiichi Shimizu, 1 Shintaro Kuroda, 1 Hiroyuki Tahara, 1 Kentaro Ide, 1 Tsuyoshi Kobayashi, 1 and Hideki Ohdan 1 1 Department of Gastroenterological and Transplant Surgery, Graduate School of Biomedical and Health Sciences, Hiroshima University, Hiroshima, Japan; and 2 Division of Regeneration and Medicine, Medical Center for Translational and Clinical Research, Hiroshima University Hospital, Hiroshima, Japan Abdominal aortic calcification (AAC) is known as a risk factor of coronary artery disease, stroke, hyperphosphatemia, chronic inflammation, diabetes, and decreased estimated glomerular filtration rate. However, the clinical implications of incidental AAC findings in liver transplantation (LT) have not been evaluated in terms of posttransplantation survival and complications. Therefore, we analyzed the relationships between the AAC level and the outcomes following LT. A total of 156 consecutive patients who underwent LT between January 2007 and December 2014 were divided into 2 groups according to their AAC level (<100 mm 3 or 100 mm 3 ), as calculated using the Agatston method. Even after propensity matching, the survival time was significantly longer in the low-aac group compared with that in the high-aac group (median survival time, 4.5 versus 3.0 years; P < 0.01). A multivariate analysis identified high AAC level (hazard ratio, 2.2) and old donor age (hazard ratio, 2.2) as prognostic factors for overall survival. In conclusion, high AAC is an independent unfavorable prognostic factor in LT. Liver Transplantation AASLD. Received March 2, 2018; accepted July 14, Aortic calcification is a risk marker for cardiovascular disease (CVD) because it has been associated with coronary artery disease and stroke in the general population. (1-3) Aortic calcification can be quantitatively measured on abdominal computed tomography (CT) as the aortic calcification index, a widely used measure that expresses calcification in 12 sectors as a percentage. In addition, abdominal aortic calcification (AAC) is easily evaluated on abdominal CT. (4) AAC prevalence, patterns, and severity in the living kidney donor Abbreviations: AAC, abdominal aortic calcification; BSI, bloodstream infection; CI, confidence interval; CIT, cold ischemia time; CKD, chronic kidney disease; CMV, cytomegalovirus; CRP, C-reactive protein; CT, computed tomography; CVD, cardiovascular disease; DM, diabetes mellitus; EDV, end-diastolic velocity; egfr, estimated glomerular filtration rate; GPS, Glasgow prognosis scale; GRWR, graft-to-recipient weight ratio; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; LT, liver transplantation; MELD, Model for End-Stage Liver Disease; mgps, modified Glasgow prognosis scale; OR, odds ratio; OS, overall survival; PSV, peak systolic velocity; PVF, portal vein flow; RI, resistance index. Address reprint requests to Masahiro Ohira, M.D., Ph.D., Department of Gastroenterological and Transplant Surgery, Graduate School of Biomedical and Health Sciences, Hiroshima University, Kasumi, Minami-ku, Hiroshima , Japan; Telephone: ; FAX: ; mohira@hiroshima-u.ac.jp population have been reported. (5) Living donors with a high AAC level require close observation because such patients have a higher probability of delayed renal function recovery after donation. (4) Furthermore, an initial report demonstrated that thoracic aortic calcification in renal transplantation recipients predicts both cardiovascular events and all-cause mortality. (6) However, the clinical implications of incidental AAC findings in liver transplantation (LT) have not been evaluated in terms of posttransplantation survival from complications. Therefore, the objective of the present study was to evaluate the impact of the AAC level on the outcomes after LT. Specifically, we assessed the relationship between the preoperative AAC level and overall survival (OS) after LT, as well as the impact of AAC on inflammatory markers and severe infectious diseases. Patients and Methods STUDY PATIENTS A total of 156 LTs were performed in our unit between January 2007 and December 2014, including 141 living donor LTs and 15 deceased donor LTs. Data concerning the recipient at the time of the transplant (age, sex, Model for End-Stage Liver Disease [MELD] score, ORIGINAL ARTICLE 79

2 Liver Transplantation, January 2019 Child-Pugh classification, graft-to-recipient weight ratio [GRWR], ABO incompatibilities, hepatitis B virus [HBV] surface antigen level, hepatitis C virus [HCV] antibody level, AAC level, presence of diabetes, presence of hepatocellular carcinoma [HCC], serum C-reactive protein [CRP], serum albumin, operative time, and bleeding volume), as well as the donor s age and cold ischemia time (CIT), were collected from electronic records. The rates of postoperative and longterm complications (including CVD, new incidence of diabetes, chronic renal failure, and de novo cancer) as well as OS were determined from posttransplantation electronic records. The study protocol was approved by the ethics committee of our hospital (E-1410). CHRONIC RENAL FAILURE Chronic renal failure was defined as a decreased estimated glomerular filtration rate (egfr < 15 ml/ minute/1.73 m 2 ) with the exception of acute renal dysfunction. BLOODSTREAM INFECTION Bloodstream infection (BSI) was defined according to the criteria proposed by the Centers for Disease Control and Prevention. (7) A blood culture test was performed in patients with an increased temperature of above 38 C. The isolation of bacteria and fungus (other than common skin contaminants) from 1 or more blood cultures in the presence of clinical symptoms was considered as proof of a BSI. The present study included cases in which infections developed within 30 days after surgery, and the first episode of BSI was documented. GLASGOW PROGNOSIS SCALE AND MODIFIED GLASGOW PROGNOSIS SCALE The Glasgow prognosis scale (GPS) was calculated as previously described (patients with a CRP of 1 mg/dl Copyright 2018 by the American Association for the Study of Liver Diseases. View this article online at wileyonlinelibrary.com. DOI /lt Potential conflict of interest: Nothing to report. 80 ORIGINAL ARTICLE were allocated a score of 0, those with a CRP > 1 mg/dl or albumin < 3.5 g/dl received a score of 1, and those with a CRP > 1 mg/l and albumin < 3.5 g/dl received a score of 2). (8) The modified Glasgow prognosis scale (mgps) was also calculated as previously described (patients with a CRP of 1 mg/dl were allocated a score of 0, those with a CRP > 1 mg/dl received a score of 1, and those with a CRP > 1 mg/l and albumin < 3.5 g/dl received a score of 2). (9) The GPS and mgps are well known and are the most extensively validated systemic inflammation-based prognostic scores. AORTIC ABDOMINAL CALCIFICATION CT angiographies were performed using a standardized examination protocol on a 320-detector row CT scanner (Aquilion ONE ViSION, Toshiba Medical Systems, Japan). The AAC score was calculated using AZE VirtualPlace Lexus64 Anatomia software (AZE Inc., Tokyo, Japan). Using the Agatston method, (9) the AAC volume was automatically calculated for calcifications located in the abdominal aorta (from the origin of the renal artery to the iliac bifurcation) with attenuation greater than the predefined 130 Hounsfield units level. Patients were divided into 2 groups according to the AAC level (<100 mm 3 or 100 mm 3 ). (4) DOPPLER ULTRASONOGRAPHY OF THE LIVER We evaluated the blood flow to the graft on postoperative day 1 using Doppler ultrasonography in 97 LT recipients between 2009 and The resistance index (RI), peak systolic velocity (PSV), and enddiastolic velocity (EDV) were investigated at the position of the hepatic artery anastomosis, and the portal vein flow (PVF) was checked at the level of the portal trunk. STATISTICAL ANALYSIS The longterm survival was estimated and compared using Kaplan-Meier and log-rank statistics. To adjust for differences in baseline characteristics, 1-to-1 propensity score models were constructed on the basis of each patient s estimated propensity score (based on the recipient s age, donor s age, sex, MELD score, Child- Pugh classification, GRWR, ABO incompatibilities, HBV surface antigen, HCV antibody level, AAC

3 Liver Transplantation, Vol. 25, No. 1, 2019 level, presence of diabetes, presence of HCC, operative time, bleeding volume, and CIT). Statistical analyses, including propensity score matching, were performed using JMP statistical software (JMP 13; SAS Institute Inc., Cary, NC). P values <0.05 were considered statistically significant. Results CHARACTERISTICS OF PATIENTS The cohort included 95 men and 61 women with an overall mean age at the time of transplant of 54.9 ± 10.4 years. There were 141 living donor LTs and 15 deceased donor LTs. Patient characteristics and surgical procedures prior to propensity score matching are summarized in Table 1. In the unmatched population, the high-aac group had significantly more men and a higher mean age compared with that in the low-aac group. For early mortality (within 1 year after LT, n = 27), the causes of death were as follows: infectious diseases, n = 20 (74.1%); liver failure, n = 3 (11.1%); CVD, n = 2 (7.4%); and other, n = 2 (7.4%). In contrast, for late mortality (over 1 year after LT, n = 21), the causes of death were as follows: infectious diseases, n = 9 (42.9%); HCC recurrences, n = 5 (23.8%); liver failure, n = 3 (14.3%); and CVD, n = 2 (9.5%). There was a significantly higher percentage of deaths related to infectious diseases in early mortality than in late mortality (74.1% versus 42.9%; P = 0.03). In late mortality, severe infectious diseases and CVD caused more than 50% of the deaths. PROPENSITY SCORE MATCHING TO EVALUATE THE INFLUENCE OF RECIPIENT A AC In the Kaplan-Meier survival curve analysis, the longterm survival was significantly worse in the high- AAC group than in the low-aac group (P = 0.004; Fig. 1). In the propensity score matched cohort (Fig. 2), the high-aac group still had significantly worse survival compared with that in the low-aac group. The 1-, 3-, and 5-year survival rates were 75.9%, 64.2%, and 57.5% in the high-aac group, respectively, and were 87.7%, 84.0%, and 79.2% in the low-aac group, respectively (Fig. 2). The postoperative complications are summarized in Table 3. TABLE 1. Patient Characteristics Prior to Propensity Score Matching Low-AAC Group (n = 90) High-AAC Group (n = 66) P Value Sex 0.11 Male Female Recipient age, years 51.6 ± ± 6.6 <0.01 Donor age, years 38.0 ± ± MELD points 17.8 ± ± Child-Pugh <0.01 A 14 1 B C GRWR, % 1.0 ± ± ABO incompatibilities Positive 12 7 Negative RISK FACTORS FOR A POOR PROGNOSIS 0.61 Virus 0.04 No virus HCV HBV 17 9 Diabetes 0.08 Positive Negative Hepatocellular carcinoma Positive Negative Operative time, minutes Bleeding volume, ml ± ± ± ± CIT, minutes ± ± Type of donor 0.85 Deceased 9 6 Living donor Univariate analyses revealed that a high donor age and high recipient AAC level were statistically significant predictors of a poor prognosis. In a multivariate logistic regression analysis of the factors associated with a poor prognosis in the univariate analyses, donor age 50 years (odds ratio [OR], 2.2; 95% confidence interval [CI], ; P = 0.01) and recipient AAC level 100 mm 3 (OR, 2.2; 95% CI, ; P < 0.01) remained as independent risk factors (Table 4). ORIGINAL ARTICLE 81

4 Liver Transplantation, January 2019 FIG. 1. Kaplan-Meier survival curves before propensity score matching. CORRELATIONS BETWEEN THE RECIPIENT A AC LEVEL AND LONGTERM COMPLICATIONS Correlations between the recipient AAC level and longterm complications are shown in Table 5. There were no significant group differences in the longterm complications, including CVD, diabetes, severe renal dysfunction, and de novo cancer among the 121 patients alive at 1 year after transplantation (high-aac group, n = 44; low-aac group, n = 77). CORRELATIONS BETWEEN THE A AC LEVEL AND INFLAMMATORY MARKERS Table 6 shows the correlations between the AAC level and inflammatory markers. The preoperative serum albumin level in the high-aac group was significantly lower than that in the low-aac group (2.9 ± 0.5 g/dl versus 3.2 ± 0.5 g/dl; P < 0.01). In the high-aac group, 9.1%, 65.2%, and 25.8% of patients had a GPS of 0, 1, and 2 points, respectively, whereas in the low-aac group, the distribution was 26.7%, 62.2%, and 11.1%, respectively (P < 0.01). In the high- AAC group, 74.2%, 0.0%, and 25.8% of patients had a mgps of 0, 1, and 2 points, respectively, whereas in the low-aac group, the distribution was 82.2%, 6.7%, and 11.1%, respectively (P < 0.01). AAC was strongly related to the systemic inflammation-based prognostic scores. After propensity score matching, there were no significant group differences in the GPS and mgps. FINDINGS ON ULTRASONOGRAPHY Figure 3 shows the findings on ultrasonography. The RI was significantly higher in the high-aac group than in the low-aac group (0.70 ± 0.09 versus 0.74 ± 0.09; P = 0.04). In addition, the EDV was also significantly lower in the high-aac group than in the low-aac group (10.4 ± 6.0 cm/seconds versus cm/seconds; P = 0.03). There were no significant group differences in the PSV and PVF. Discussion The present study revealed a high AAC level as an independent poor prognostic factor for LT recipients. Furthermore, a high AAC level was strongly associated with worse survival. As no previous publications have evaluated the impact of the degree of AAC on OS in LT, the present study is the first to 82 ORIGINAL ARTICLE

5 Liver Transplantation, Vol. 25, No. 1, 2019 TABLE 2. Patient Characteristics After Propensity Score Matching Low-AAC Group (n = 35) High-AAC Group (n = 35) P Value Sex 0.31 Male Female Recipient age, years 56.6 ± ± Donor age, years 37.5 ± ± MELD points 16.3 ± ± Child-Pugh 0.51 A 0 1 B C GRWR, % 1.0 ± ± ABO incompatibilities 0.23 Positive 5 2 Negative Virus 0.87 No virus HCV HBV 7 7 Diabetes 0.81 Positive Negative Hepatocellular carcinoma Positive 5 7 Negative Operative time, minutes ± ± Bleeding volume, ml ± ± CIT, minutes ± ± Type of donor 0.69 Deceased 4 3 Living donor demonstrate the influence of AAC on the risk of unfavorable longterm outcomes after LT. Aortic calcification is a well-known risk marker for CVD. (1-3) In addition, several studies have evaluated the clinical implications of AAC as a specific disease state and have shown that AAC is associated with hyperphosphatemia, chronic inflammation, diabetes, and decreased egfr among patients with chronic kidney disease (CKD). (10) Furthermore, a strong relationship between AAC and coronary artery calcification has been reported in patients with CKD. (11) Another group reported that elevated levels of several inflammatory biomarkers, including β2- microglobulin, fibroblast growth factor 23, interleukin 8, and interleukin 18, are associated with arterial calcification in patients with CKD. (10) Furthermore, aortic calcification was shown to be a strong predictor of stroke in a large previous study. (1) Thus, the present results add to the literature on the deleterious effects of high AAC by demonstrating that a high AAC level is a risk factor for a poor prognosis after LT. On the basis of the present results, the AAC level should be estimated preoperatively to plan the risk management. AAC can be detected using various complimentary methods, including plain lateral abdominal X-ray films (straightforward, low radiation dose, only semiquantitative at best) and CT-based examinations, with either manual or automatic algorithms to speedily and reproducibly detect and quantify calcified sections of the vessel walls. (12) Given recent advancements in imaging devices, the AAC level can be calculated and quantified accurately and automatically. (4) Diabetes has been associated with the presence of aortic calcification in patients on prevalent dialysis. (13) Diabetes is also associated with worse event-free survival and mortality. (6) However, in the present study, there was no difference in the new onset of diabetes between patients with low and high AAC. In addition, numerous reports have demonstrated that AAC is a risk factor of CVD. (1-3,6) However, in the present study, the high-aac group did not have a statistically higher risk of CVD or severe chronic renal dysfunction compared with that of the low-aac group, possibly due to the small sample size. Inflammatory cells present in arteriosclerotic lesions, such as macrophages and T lymphocytes, are known to be involved in the progression of vascular calcification. (14) In the present study, the AAC level had a close relationship with the GPS and mgps, and this level reflects the systemic inflammatory response. The systemic inflammatory response may be worse because of malnutrition and reduced preoperative immunocompetence, suggesting that compromised immunocompetence may influence infections. (15,16) The GPS was also reported as the better predictive system in patients with HCC after living donor LT in our population. (17) Therefore, the finding that a high AAC level leads to poorer survival and a high systemic inflammatory response is highly compelling. In elderly patients, frailty is a state of vulnerability to a poor resolution to homeostasis after stress and is a consequence of a cumulative decline in multiple physiological systems over the lifespan. (18) AAC is considered to be one of the most representative ORIGINAL ARTICLE 83

6 Liver Transplantation, January 2019 FIG. 2. Kaplan-Meier survival curves after propensity score matching. TABLE 3. Postoperative Complications Before and After Propensity Score Matching Before Matching After Matching Low AAC (n = 90) High AAC (n = 66) P Value Low AAC (n = 35) High AAC (n = 35) P Value Biliary complication 29 (32.2) 19 (28.8) (31.4) 9 (25.7) 0.60 CMV infection 39 (43.3) 31 (47.0) (34.3) 13 (37.1) 0.80 Portal vein stenosis 1 (1.1) 1 (1.5) (2.9) 1 (2.9) >0.99 Postoperative hemorrhage 17 (18.9) 16 (24.2) (8.6) 8 (22.9) 0.19 BSI within 30 days 16 (17.8) 22 (33.3) (14.3) 11 (31.4) 0.09 BSI within 1 year 23 (25.6) 28 (42.4) (22.9) 14 (40.0) 0.12 Hepatic artery thrombosis 0 (0) 0 (0) 0 (0) 0 (0) Clinical rejection 15 (16.7) 16 (24.2) (20.0) 7 (20.0) >0.99 Note: Data are given as n (%). factors of frailty. Consistent with this, the high- AAC group had a high prevalence of diabetes, CVD, severe renal dysfunction, inflammatory status, de novo cancer, and so on. Thus, the activities of daily living may decrease in recipients with a high AAC level. Furthermore, organ dysfunction is likely to occur in the long term. Surprisingly, recipient age was not a risk factor for OS in our department. In contrast, the present study showed that the recipient AAC level is an important risk factor of not only longterm but also short-term survival. In particular, high inflammatory status was increased in the high- AAC group. This finding suggests that AAC reflects one aspect of frailty in the recipients before LT. The present study results suggest that chronic inflammation led by arteriosclerosis caused partial organ dysfunction, potentially due to frailty. Although the present study provides new information on the impact of AAC on LT outcomes, it has several limitations because of the small sample size. In addition, many unknown issues involving aortic calcification, including the mechanisms 84 ORIGINAL ARTICLE

7 Liver Transplantation, Vol. 25, No. 1, 2019 TABLE 4. Risk Factors of Worse Longterm Survival Univariate Analysis Multivariate Analysis Factors n 3-Year Survival 5-Year Survival P Value HR 95% CI P Value Age years % 78.7% <60 years % 66.2% Sex 0.40 Male % 67.5% Female % 75.3% Donor age < years % 52.6% <50 years % 75.0% MELD % 64.9% < % 73.2% GRWR % % 68.8% <0.8% % 71.0% AAC < < % 57.5% < % 79.2% ABO incompatibilities 0.37 Yes % 63.2% No % 71.2% DM 0.33 Yes % 61.2% No % 72.5% Virus 0.62 Non % 68.2% HCV % 70.0% HBV % 75.7% HCC 0.62 Yes % 66.2% No % 73.8% TABLE 5. Longterm Complications in Patients Alive Beyond 1 Year After Transplantation Low-AAC Group (n = 77) High-AAC Group (n = 44) P Value CVD 1 (1.3) 3 (6.8) 0.14 New incidence of diabetes Severe renal dysfunction De novo cancer NOTE: Data are given as n (%). 23 (29.9) 14 (31.8) (2.6) 3 (6.8) (3.9) 4 (9.1) 0.26 TABLE 6. Correlations Between AAC and Inflammatory Markers Low-AAC Group (n = 90) High-AAC Group (n = 66) P Value CRP, mg/dl 0.7 ± ± Albumin, g/dl 3.2 ± ± 0.5 <0.01 GPS, n < mgps, n < ORIGINAL ARTICLE 85

8 Liver Transplantation, January 2019 FIG. 3. The findings on ultrasonography. involved, remain undetermined. Larger cohorts are necessary to further investigate the impact of AAC in LT. In conclusion, the present retrospective study demonstrated that recipient AAC is associated with OS after LT with a high AAC level as an independent unfavorable prognostic factor. REFERENCES 1) Hollander M, Hak AE, Koudstaal PJ, Bots ML, Grobbee DE, Hofman A, et al. Comparison between measures of atherosclerosis and risk of stroke: the Rotterdam Study. Stroke 2003;34: ) Wilson PW, Kauppila LI, O Donnell CJ, Kiel DP, Hannan M, Polak JM, Cupples LA. Abdominal aortic calcific deposits are an important predictor of vascular morbidity and mortality. Circulation 2001;103: ) van der Meer IM, Bots ML, Hofman A, del Sol AI, van der Kuip DA, Witteman JC. Predictive value of noninvasive measures of atherosclerosis for incident myocardial infarction: the Rotterdam Study. Circulation 2004;109: ) Yoon YE, Han WK, Lee HH, Chang MY, Huh KH, Jung DC, et al. Abdominal aortic calcification in living kidney donors. Transplant Proc 2016;48: ) Leckstroem DC, Bhuvanakrishna T, McGrath A, Goldsmith DJ. Prevalence and predictors of abdominal aortic calcification in healthy living kidney donors. Int Urol Nephrol 2014;46: ) DeLoach SS, Joffe MM, Mai X, Goral S, Rosas SE. Aortic calcification predicts cardiovascular events and all-cause mortality in renal transplantation. Nephrol Dial Transplant 2009;24: ) Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control 2008;36: ) McMillan DC, Forrest LM, O Gorman P, Angerson WJ, McArdle CS. Performance status of male and female advanced cancer patients is independently predicted by mid-upper arm circumference measurement. Nutr Cancer 2002;42: ORIGINAL ARTICLE

9 Liver Transplantation, Vol. 25, No. 1, ) McMillan DC. The systemic inflammation-based Glasgow Prognostic Score: a decade of experience in patients with cancer. Cancer Treat Rev 2013;39: ) Kiu Weber CI, Duchateau-Nguyen G, Solier C, Schell-Steven A, Hermosilla R, Nogoceke E, Block G. Cardiovascular risk markers associated with arterial calcification in patients with chronic kidney disease stages 3 and 4. Clin Kidney J 2014;7: ) Takayama Y, Yasuda Y, Suzuki S, Shibata Y, Tatami Y, Shibata K, et al. Relationship between abdominal aortic and coronary artery calcification as detected by computed tomography in chronic kidney disease patients. Heart Vessels 2016;31: ) Bellasi A, Raggi P. Vascular imaging in chronic kidney disease. Curr Opin Nephrol Hypertens 2012;21: ) Okuno S, Ishimura E, Kitatani K, Fujino Y, Kohno K, Maeno Y, et al. Presence of abdominal aortic calcification is significantly associated with all-cause and cardiovascular mortality in maintenance hemodialysis patients. Am J Kidney Dis 2007;49: ) Shioi A, Katagi M, Okuno Y, Mori K, Jono S, Koyama H, Nishizawa Y. Induction of bone-type alkaline phosphatase in human vascular smooth muscle cells: roles of tumor necrosis factor-alpha and oncostatin M derived from macrophages. Circ Res 2002;91: ) Sagawa M, Yoshimatsu K, Yokomizo H, Yano Y, Okayama S, Usui T, et al. Worse preoperative status based on inflammation and host immunity is a risk factor for surgical site infections in colorectal cancer surgery. J Nippon Med Sch 2017;84: ) Moyes LH, Leitch EF, McKee RF, Anderson JH, Horgan PG, McMillan DC. Preoperative systemic inflammation predicts postoperative infectious complications in patients undergoing curative resection for colorectal cancer. Br J Cancer 2009;100: ) Abe T, Tashiro H, Hattori M, Kuroda S, Tahara H, Ohira M, et al. Prediction of long-term survival by using the Glasgow Prognostic Score in patients with hepatocellular carcinoma after liver transplantation. Hepatol Res 2016;46: ) Stuck AE, Iliffe S. Comprehensive geriatric assessment for older adults. BMJ 2011;343:d6799. ORIGINAL ARTICLE 87