Effect of Testosterone Level on Mortality in Patients With Left Ventricular Assist Device

Effect of Testosterone Level on Mortality in Patients With Left Ventricular Assist Device Evrim Simseka,*, Salih Kilicb, Hatice Soner Kemalc, Sanem Nalbantgilb, Pelin Ozturkd, Ilgin Yildirime, Tahir Yagdid, Cagatay Engind, and Mustafa Ozbarand a Ege University Faculty of Medicine, Department of Cardiology, Izmir, Turkey; bDoctor Ersin Arslan Training and Research Hospital, Department of Cardiology, Gaziantep, Turkey; cNear East University Faculty of Medicine, Department of Cardiology, Cyprus; dEge University Faculty of Medicine, Department of Cardiovascular Surgery, Izmir, Turkey; and eEge University Faculty of Medicine, Department of Internal Medicine, Izmir, Turkey

ABSTRACT Background and objectives. Testosterone deficiency is associated with mortality in patients with heart failure; however, its effects on patients undergoing Left Ventricular Assist Device (LVAD) implantation are unclear. We investigated the role of total testosterone (TT) and free testosterone (FT) levels on mortality undergoing LVAD implantation. Methods. Between December 2010 and December 2014, 101 consecutive male patients who underwent LVAD implantation and had plasma testosterone measurement (TT and FT) in the last month before operation were included in the study. Demographics, followup, and mortality data were analyzed retrospectively. Results. The mean age of the patients was 51.7
11 years. TT and FT levels were in the below normal range of 31.6% (n ¼ 32) and 65.3% (n ¼ 66) of the patients, respectively. The mean follow-up time was 355
268 days, and 32 (31%) patients died during follow-up. Cumulative survival rates were significantly worse in patients with low TT and FT than patients in the normal range (P < .001 and P ¼ .029, respectively). Multivariate analysis after adjustment for clinical variables, age, albumin, C-reactive protein, total cholesterol, chronic kidney disease, diabetes mellitus (DM), and leukocytosis showed that low TT and FT were independently associated with poor survival (HR, 3.680; 95% CI, 1.6158.385 P ¼ .002 and HR, 3.816; 95% CI, 1.279-11.383, P ¼ .016, respectively). Conclusion. Low TT and FT levels were independent risk factors for mortality in patients with LVAD.

T

HE development of cachexia in heart failure (HF) is based on the fact that the metabolic balance is shifted in the catabolic direction by the activation of inflammatory and neuroendocrine systems [1,2]. This imbalance leads to structural changes in skeletal muscles and a reduction in exercise capacity [2]. Testosterone is the most important androgen in the human body and has anabolic effects that decrease protein degradation, increases lean body mass, and improves nitrogen storage and protein synthesis. It has been shown that total testosterone (TT) and free testosterone (FT) levels are lower in patients with heart failure compared to the healthy population; additionally, muscle strength and

ª 2019 Elsevier Inc. All rights reserved. 230 Park Avenue, New York, NY 10169

Transplantation Proceedings, XX, 1e6 (2019)

exercise capacity are lower in patients with decreased testosterone levels [3e7]. There are conflicting study results regarding mortality and testosterone levels in the HF population. Some studies showed that there is an independent relationship between mortality and testosterone level. However, some other

*Address correspondence to Dr Evrim Simsek, Ege Universitesi Tıp Fakultesi Hastanesi, Kardiyoloji A.D. Bornova, Izmir, Turkey. Tel: 0090 5058843974. E-mail: drevrimsimsek@ gmail.com 0041-1345/19 https://doi.org/10.1016/j.transproceed.2019.07.028

1

2

studies reported that after adjustments for clinical variables such as age and diabetes, this relation disappears [8,9]. Left Ventricular Assist Devices (LVAD) are commonly implanted with the aim of bridging to transplantation or as destination therapy in patients who are not eligible for heart transplant [10]. After LVAD implantation to improve outcomes, it is important to select patients who would have the most benefits of the LVAD. Physicians are focused on the preoperative factors that may affect morbidity and mortality after LVAD implantation. There is a lack of data about testosterone levels and mortality in the LVAD population. Association with testosterone levels and early- to midterm survival in male LVAD patients were evaluated in the present study METHODS Study Population One hundred fifty-two patients underwent LVAD implantation in the study center between December 2010 and December 2014 due to end-stage heart failure. The inclusion criteria for the study were to be male and have had blood testosterone levels tested a maximum of 1 month before LVAD implantation. The exclusion criteria were to be female or have a history of hormone replacement therapy or antiandrogen therapy (finasteride, doxazosin, or tamsulosin) due to any cause. Clinical and demographic characteristics were obtained from the patients’ hospital records retrospectively. Twenty-four female patients, 21 male patients who had no testosterone testing, and 6 patients who had tests over 1 month before LVAD implantation were not included in the study. Among the patients included in the study, there were no patients receiving hormone replacement therapy. A total number of 101 patients were included in the study. HVAD (HeartWare, Framingham, Mass, United States) was implanted in the majority of the patients, and 31 (30.7%) patients received HeartMate II (HMII; Thoratec Corporation, Pleasanton, Calif, United States). None of the patients were lost during the follow-up period. Chronic kidney disease (CKD) was defined as the Estimated Glomerular Filtration Rate of the patient being below 60/mL/min/m2. All patients were followed from the time of LVAD implantation. Patients who underwent heart transplantation were censored. The primary endpoint was defined as the occurrence of death on the waiting list under LVAD support.

Laboratory Measurements In all patients, venous blood samples were taken in the morning (before 10 AM) after 8 hours of fasting and a supine rest of at least 15 minutes. After centrifugation, the serum was collected and kept at -70 C until being analyzed. Serum levels of TT were measured with chemiluminescence immunoassay. Total testosterone levels were measured using 2 different kits including TESTOSTERONE (Beckman Coulter, Brea, Calif, United States) and Elecsys 2010 Testosterone II (Roche Diagnostics, Indianapolis, Ind, United States) during the study period. The interassay and intra-assay variability coefficients for Elecsys 2010 Testosterone II were 1.8% and 2.8% and for TESTOSTERONE as 5.09% and 2.52% respectively. Estimated FT levels were calculated using the Vermeulen method [11]. All measurements were performed in a single laboratory. The patients were grouped according to low, normal, and high TT and FT levels specified by age for both different kits.

SIMSEK, KILIC, KEMAL ET AL

Statistic Analysis Continuous data are presented as the means
standard deviation and median with maximum and minimum levels. The KolmogorovSmirnov test was used to evaluate whether the distribution of continuous variables was normal. Differences in continuous variables between the 2 groups were determined by Student t test or Mann-Whitney U Test. Categorical variables were summarized as percentages and compared with the c2 test or Fisher’s exact test. For FT (model 1) and TT (model 2), multivariable Cox regression models were established including variables with a P value of < .1 in univariate analysis. Age, DM, CKD, albumin, C-reactive protein (CRP), total cholesterol, and white blood cells > 103 were included in both models. A P value < .05 was considered to be statistically significant. A survival analysis between the 2 groups (free or total testosterone low versus normal) was performed using the Kaplan-Meier method with a log-rank analysis. Statistical analyses were conducted using the Statistical Package for the Social Sciences (SPSS 20.0) for Windows (IBM Corp, Armonk, NY, United States). The study was approved by the local ethics committee.

RESULTS

A total of 101 male patients (mean age 51.7
11 years) were included in the study. Baseline demographic and clinical characteristics of the patients are summarized in Table 1. The majority of the patients had ischemic cardiomyopathy (66.3%). The mean follow-up time was 355
268 days; 31.6% (n ¼ 32) of the patients had TT values below the normal range (lower total testosterone group: LTT). The demographic and clinical characteristics of the patients with normal TT and low TT levels are compared in Table 1. The median CRP level, number of patients with leukocytosis, and history of CKD were higher, and the mean hemoglobin level was lower in the LTT group than the normal TT group (Table 1). In total, 65.3% (n ¼ 66) of the patients were found to have FT values below the normal range (lower free testosterone group: LFT). The mean of total cholesterol and albumin levels were lower and the median CRP level was higher in the LTT group than the normal TT group. Also, the percentage of patients with CKD was higher in the LTT group than the normal TT group (Table 1). During the study period, 10 (9.9%) patients had heart transplantation (at mean 366
290 days), and 32 (31.7%) patients died (at mean 269
279 days). Among this last group, 21 (65.6%) died because of cardiovascular reasons and 11 (34.4%) because of other reasons (3 patients intracranial hemorrhage, 5 patients sepsis, and 3 for other reasons). The demographic and clinical outcomes of the survival and mortality groups are summarized in Table 2. The history of DM (50% vs 27.5%; P ¼ .027), CKD (50% vs 21.7%; P ¼ .004), and number of patients with leukocytosis (34.4% vs 14.5%; P ¼ .022) were significantly higher in the mortality group than the survival group. Besides, the median CRP level [(1.0 mg/dL (IQR 1.93) vs 0.54 mg/dL (IQR 0.86); P ¼ .006)] was significantly higher and the mean albumin level (3.8
0.5 g/dL vs 4.0
0.5 mg/dL; P ¼ .029) was significantly lower in the mortality group than the

TESTOSTERONE IN PATIENTS WITH LVAD

3

Table 1. Patient Characteristics and Laboratory Parameters Before LVAD Implantation Variables

All patients (n ¼ 101)

Low FT (n ¼ 66)

Normal FT (n ¼ 35)

P

Low TT (n ¼ 32)

Normal TT (n ¼ 69)

Age (years) BMI (kg/m2) DM, %, (n) HT, %, (n) HPL, %, (n) CKD, %, (n) ICMP, %, (n) COPD, %, (n) Smoking, %, (n) INTERMACS Profile £ 2 %, (n) Total protein (g/dL) Albumin (g/dL) Total cholesterol (mg/dL) LDL-C (mg/dL) CRP, (mg/dL) Median (IQR) WBC > 103, %, (n) Hb (g/dL) Plt x 103/mm3

51.
11.4 25.2
4.0 34.7 (35) 61.4 (62) 27.7 (28) 30.7 (31) 66.3 (67) 17.8 (18) 69.3 (70) 54.5 (55)

51.8
12 25
3.77 31.8 (21) 62.1 (41) 25.8 (17) 37.9 (25) 66.7 (44) 19.7 (13) 72.7 (48) 50 (33)

51.5
10 25.6
4.46 40 (14) 60 (21) 11 (31.4) 17.1 (6) 65.7 (23) 14.3 (5) 62.9 (22) 62.9 (22)

.904 .513 .411 .835 .545 .032 .923 .499 .306 .217

52.6
12.4 24.6
3.7 46.9 (15) 75 (24) 18.8 (6) 50 (16) 65.6 (21) 22.6 (7) 64.5 (20) 50 (16)

51
11.1 25.4
4.2 29 (20) 55.1 (38) 32.4 (22) 21.7 (15) 66.7 (46) 15.9 (11) 73.5 (50) 56.5 (39)

.528 .614 .079 .056 .158 .004 .918 .424 .361 .540

6.8
0.82 4.0
0.52 145
45 90
36 0.7 (1.2)

6.7
0.68 3.9
0.50 137
43 85
34 0.845 (1.63)

7.0
1.0 4.2
0.50 158
45 99
38 0.415 (0.57)

.680 .001 .033 .069 .026

6.7
0.75 3.9
0.6 134
43 84
36 1.545 (1.16)

6.9
0.85 4.0
0.5 151
46 93
37 0.44 (0.71)

.183 .097 .072 .161 < .001

20.8 (20) 12.7
1.8 244
77

22.7 (15) 12.5
1.74 245
77

17.1 (6) 13
1.95 242
78

.510 .139 .856

34.4 (119) 12.0
1.7 257
78

14.5 (10) 12.9
1.8 238
77

P

.022 .022 .299

Abbreviations: CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; CRP, C-reactive protein; DM, diabetes mellitus; FT, free testosterone; Hb, hemoglobin; HPL, hyperlipidemia; HT, hypertension; ICMP, ischemic cardiomyopathy; LDL-C, low density lipoprotein cholesterol; LFT, lower free testosterone; LTT, low total testosterone; Plt, platelet; TT, total testosterone; WBC, white blood cell.

survival group (Table 2). The patients were predominantly in INTERMACS profile 2 (32.7%, n ¼ 33) and 3 (43.6%, n ¼ 44) at the time of LVAD implantation. There was no significant difference between the groups in terms of the INTERMACS profile (P ¼ .610). The frequency of LFT and LTT levels was significantly higher in the mortality group (87.5% vs 55%; P ¼ .002 and 56.2% vs 20.3%; P < .001, Table 2. Laboratory Parameters and Clinical Characteristics of Survivor and Mortality Patients Before LVAD Implantation Variables

Age, years (mean ± SD) BMI (kg/m2) DM, %, n HT, %, n HPL, %, n ICMP, %, n COPD, %, n Smoking, %, n INTERMACS Profile £ 2 LFT, %, n LTT, %, n Total protein (g/dL) Albumin (g/dL) CRP (mg/dL), Median (IQR) CKD, %, n WBC > 103, %, n Hb (g/dL) Plt 3 103/mm3

Survivors (n ¼ 69) Mortality (n ¼ 32)

P

51.4
11.2 25.3
3.7 27.5 (19) 58 (40) 26.1 (18) 62.3 (43) 18.8 (13) 67.7 (46) 49.5(34) 55 (38) 20.3 (14) 6.8
0.8 4.0
0.5 0.54 (0.86)

52.2
12.0 24.8
4.6 50 (16) 68.8 (22) 32.3 (10) 75 (24) 16.1 (5) 77.4 (24) 37.5(12) 87.5 (28) 56.2 (18) 6.7
0.7 3.8
0.5 1.0 (1.93)

.742 .631 .027 .301 .631 .210 .744 .322 .269 .002 < .001 .402 .029 .006

21.7 14.5 12.8 243

50 (16) 34.4 (11) 12.3
1.8 246
73

.004 .022 .151 .861

(15) (10)
1.7
79

Abbreviations: BMI, body mass index; CKD, chronic kidney disease; COPD, Chronic Obstructive Pulmonary Disease; CRP, C-reactive protein; DM, diabetes mellitus; Hb, hemoglobin; HPL, hyperlipidemia; HT, hypertension; ICMP, ischemic cardiomyopathy; LFT, lower free testosterone; LTT; low total testosterone; Plt, platelet; WBC, white blood cell.

respectively). The Kaplan-Meier survival analysis showed that LTT or LFT patients were more likely to die under LVAD support compared with normal TT or FT patients (log rank, P < .001 and P ¼ .029, respectively; Figs 1 and 2). By the univariate analysis, albumin, CKD, DM, LFT, LTT, CRP and leukocytosis were significantly associated with mortality (Tables 3 and 4). Two different multivariable Cox regression analysis models were established to assess the effects of FT and TT on mortality. Both models showed that LFT (model 1) and LTT (model 2) were independently associated with mortality (HR, 3.314; %95 CI, 1.096-10.016; P ¼ .03, HR, 3.327 95% CI, 1.411-7.849; P ¼ .006, respectively). Other significant risk factors were a history of DM and elevated CRP in the both models (Tables 3 and 4). DISCUSSION

The effects of androgen deficiency on mortality in HF patients were previously investigated; however, it was not evaluated in the LVAD population. This study showed that low TT and FT levels were common in the LVAD-planned HF patients and they were independent predictors of mortality after LVAD implantation. Low TT and low FT levels are more common in patients with symptomatic heart failure, including the patients with reduced left ventricle ejection fraction (LVEF) and preserved LVEF, compared to the healthy population. Wu et al found low TT and FT percentages (21.7% and 27.4%, respectively) in patients with HF over 60 years of age [12]. However, this ratio depends on age of the patient population. Jankowska et al showed that low testosterone levels was more frequent in patients with HF under 45 years of age compared to older patients with HF (LTT levels were 39%

4

SIMSEK, KILIC, KEMAL ET AL

Fig 1. Survival curves of patients with low and normal free testosterone before LVAD implantation.

in patients under 45 years of age and 27% in patients over 66 years of age) [13]. The present study had a younger patient population (mean age 51.7) and had higher percentages of low TT and FT patients than Jankowska’s study (31.6% and 65.3%, respectively). Free testosterone is calculated over TT, albumin, and sex hormone binding globulin. It is known that sex hormone binding globulin levels are elevated in HF patients. That is the reason why low FT frequency is higher in the present study and some other studies. Guder et al showed that FT deficiency was 79.4% and TT deficiency was only 9.4% in symptomatic HF patients; because of that, it is stated that using FT or bioactive testosterone levels in HF population might give more consistent results [14]. The association between androgen deficiency and mortality among HF patients is not clearly shown in clinical studies. Some studies showed no association; however, some other studies showed benefits in univariate analysis but not in multivariate analysis, and some others showed androgen deficiency was an independent predictor of mortality. The reason for these controversial results is the heterogeneous patient population among studies. Guder et al investigated all symptomatic HF patients including the reduced and preserved ejection fraction. They showed that low FT levels was significantly higher in the mortality group at an average follow-up period of 2.3 years; however, this significance disappeared in the multivariable analysis [15]. Jankowska

et al evaluated 208 patients with HF at any age in whom the median age was 63 years and median LVEF was 33%; they showed that low FT and TT levels are independent risk factors of mortality at the end of 3 years follow-up as in our study (for TT HR: 0.84, 95% CI 0.72-0.97; P ¼ .02 and for FT HR: 0.94, 95% CI 0.89-0.99, P ¼ .01) [13]. There are a few studies handling younger HF patients. However, the present study includes younger and severe HF patients (45% of the patient population were in INTERMACS profile 2 and below), and these patients were never evaluated for androgen deficiency. The present study is also important in terms of handling this unique patient population. There are limited studies with small number of patients on testosterone deficiency and the benefits of replacement in heart failure patients [16,17]. Recent meta-analysis showed that androgen replacement therapies in HF patients remain optimally neutral about cardiovascular events [17]. These findings suggest that even though testosterone is regarded as an independent risk predictor of mortality in the HF population, it may only be a marker of poor clinical status instead of cause of mortality. LIMITATIONS

Our study has some important limitations. The most important limitations were the retrospective design and

TESTOSTERONE IN PATIENTS WITH LVAD

5

Fig 2. Survival curves of patients with low and normal total testosterone before LVAD implantation.

results of a single-center experience. The basal testosterone levels were measured once before listing, and no further measurements were performed after LVAD implantation. Because of that, the effect of changes in the testosterone values after LVAD implantation on mortality could not been evaluated. Moreover, testosterone measurements were performed with 2 different kits. Therefore, we were not able

Table 3. Cox Regression Analysis for Death Before Heart Transplantation (Model 1) Analysis Variables

Age Albumin Total cholesterol WBC > 103 CKD LFT CRP DM

Univariate P

HR [95% CI]

Multivariate

to give an exact cutoff value for TT or FT values for highrisk patients. CONCLUSION

Total and free testosterone deficiencies are common among LVAD-planned heart failure patients. Preoperation LTT

Table 4. Cox Regression Analysis for Death Before Heart Transplantation (Model 2) Analysis

P

HR [95% CI]

Variables

.799 1.004 [0.974-1.034] .014 0.381 [0.177-0.821] .155 0.993 [0.984-1.003]

-

-

.016 2.466 [1.184-5.134] .001 3.296 [1.627-6.679] .038 3.039 [1.063-8.689]

.016

Age Albumin Total cholesterol WBC > 103 CKD LTT CRP

3.816 [1.279-11.383] < .001 1.190 [1.098-1.290] < .001 1.209 [1.103-1.325] .007 2.690 [1.316-5.498] < .001 4.369 [1.980-9.7637]

Abbreviations: CKD, chronic kidney disease; CRP, C-reactive protein; DM, diabetes mellitus; LFT, lower free testosterone.

DM

Univariate Analysis P

Multivariable Analysis

HR [95% CI]

P

HR [95% CI]

.799 .014 .155

1.004 [0.974-1.034] 0.381 [0.177-0.821] 0.993 [0.984-1.003]

-

-

.016 .001 < .001

2.466 [1.184-5.134] 3.296 [1.627-6.679] 5.475 [2.491-12.035]

.002

< .001

1.190 [1.098-1.290]

.002

.007

2.690 [1.316-5.498]

.017

3.680 [1.615-8.385] 1.244 [1.081-1.432] 2.521 [1.180-5.383]

Abbreviations: CKD, chronic kidney disease; CRP, C-reactive protein; DM, diabetes mellitus; LTT, low total testosterone; WBC, white blood cell

6

and LFT values are independent predictors of high-risk mortality on patients under LVAD support. REFERENCES [1] Anker SD, Clark AL, Kemp M, Salsbury C, Teixeira MM, Hellewell PG, et al. Tumor necrosis factor and steroid metabolism in chronic heart failure: possible relation to muscle wasting. J Am Coll Cardiol 1997;30:997e1001. [2] Anker SD, Chua TP, Ponikowski P, Harrington D, Swan JW, Kox WJ, et al. Hormonal changes and catabolic/anabolic imbalance in chronic heart failure and their importance for cardiac cachexia. Circulation 1997;96:526e34. [3] Volterrani M, Rosano G, Iellamo F. Testosterone and heart failure. Endocrine 2012;42:272e7. [4] Kontoleon PE, Anastasiou-Nana MI, Papapetrou PD, Alexopoulos G, Ktenas V, Rapti AC, et al. Hormonal profile in patients with congestive heart failure. Int J Cardiol 2003;87:179e83. [5] Santos MRd, Sayegh ALC, Groehs RVR, Fonseca G, Trombetta IC, Barretto ACP, et al. Testosterone deficiency increases hospital readmission and mortality rates in male patients with heart failure. Arquivos Brasileiros de Cardiologia 2015;105:256e64. [6] Wehr E, Pilz S, Boehm BO, März W, Grammer T, Obermayer-Pietsch B. Low free testosterone is associated with heart failure mortality in older men referred for coronary angiography. Eur J Heart Fail 2011;13:482e8. [7] Oskui PM, French WJ, Herring MJ, Mayeda GS, Burstein S, Kloner RA. Testosterone and the cardiovascular system: a comprehensive review of the clinical literature. JAMA 2013;2: e000272.

SIMSEK, KILIC, KEMAL ET AL [8] Cunningham GR, Matsumoto AM, Swerdloff R. Low testosterone and men’s health. J Clin Endocrinol Metabol 2004;89. [9] Kalyani RR, Gavini S, Dobs AS. Male hypogonadism in systemic disease. Endocrin Metab Clin 2007;36:333e48. [10] Gustafsson F, Rogers JG. Left ventricular assist device therapy in advanced heart failure: patient selection and outcomes. Eur J Heart Fail 2017;19:595e602. [11] Vermeulen A, Verdonck L, Kaufman JM. A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metabol 1999;84:3666e72. [12] Wu H-Y, Wang X-F, Wang J-H, Li J-Y. Testosterone level and mortality in elderly men with systolic chronic heart failure. Asian J Androl 2011;13:759. [13] Jankowska EA, Biel B, Majda J, Szklarska A, Lopuszanska M, Medras M, et al. Anabolic deficiency in men with chronic heart failure: prevalence and detrimental impact on survival. Circulation 2006;114:1829e37. [14] Güder G, Frantz S, Bauersachs J, Allolio B, Ertl G, Angermann CE, et al. Low circulating androgens and mortality risk in heart failure. Heart 2010;96:504e9. [15] Gencer B, Mach F. Testosterone: a hormone preventing cardiovascular disease or a therapy increasing cardiovascular events? Eur Heart J 2015;37:3569e75. [16] Araujo AB, Dixon JM, Suarez EA, Murad MH, Guey LT, Wittert GA. Endogenous testosterone and mortality in men: a systematic review and meta-analysis. J Clin Endocrinol Metabol 2011;96:3007e19. [17] Toma M, McAlister FA, Coglianese EE, Vidi V, Vasaiwala S, Bakal JA, et al. Testosterone supplementation in heart failure clinical perspective: a meta-analysis. Circ Heart Fail 2012;5: 315e21.