#PAGE_PARAMS# #ADS_HEAD_SCRIPTS# #MICRODATA#

Clinically-defined preoperative serum phosphorus abnormalities and outcomes of coronary artery bypass grafting: Retrospective analysis using inverse probability weighting adjustment


Authors: Jungchan Park aff001;  Kwan Young Hong aff001;  Jeong Jin Min aff001;  Eunjin Kwon aff001;  Young Tak Lee aff002;  Wook Sung Kim aff002;  Hye Seung Kim aff003;  Kyunga Kim aff003;  Jong-Hwan Lee aff001
Authors place of work: Department of Anesthesiology and Pain Medicine, Samsung Medical Center, Sung Kyunkwan University School of Medicine, Seoul, Korea aff001;  Department of Thoracic and Cardiovascular Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea aff002;  Statistics and Data Center, Research Institute for Future Medicine, Samsung Medical Center, Seoul, Korea aff003;  Department of Digital Health, SAIHST, Sungkyunkwan University, Seoul, Korea aff004
Published in the journal: PLoS ONE 14(12)
Category: Research Article
doi: https://doi.org/10.1371/journal.pone.0225720

Summary

Background

Serum phosphorus is a well-known marker of vascular calcification, but the effects of serum phosphorus abnormalities defined by clinical criteria on the outcomes of coronary artery bypass grafting (CABG) remain unclear. We aimed to evaluate whether preoperative serum phosphorus abnormalities defined based on clinical criteria are associated with outcomes of CABG using a relatively new statistical technique, inverse probability weighting (IPW) adjustment.

Methods

From January 2001 to December 2014, 4,989 consecutive patients who underwent CABG were stratified into normal (2.5–4.5 mg/dl; n = 4,544), hypophosphatemia (<2.5 mg/dl; n = 238), or hyperphophatemia (>4.5 mg/dl; n = 207) groups depending on preoperative serum phosphorus level.

Results

The primary outcome was all-cause death during a median follow-up of 48 months. Secondary outcomes were cardiovascular death, graft failure, myocardial infarction, repeat revascularization, and stroke. In multivariate Cox analysis, preoperative hypophosphatemia was significantly associated with all-cause death (hazard ratio [HR] 1.76; 95% confidence interval [CI] 1.13–2.76; P = 0.01). However, this association varied depending on chronic kidney disease and emergent operation (p for interaction = 0.05 and 0.03, respectively). In addition, analysis after IPW adjustment demonstrated that preoperative serum phosphorus abnormalities were not significantly associated with all-cause death (P = 0.08) or any secondary outcomes except graft failure. Graft failure was significantly associated with preoperative hypophosphatemia (HR 2.51; 95% CI 1.37–4.61; P = 0.003).

Conclusion

Our study showed that preoperative serum phosphorus abnormalities in clinical criteria were not associated with outcomes after CABG except for graft failure. And, the association of hypophosphatemia with graft failure remains to be evaluated.

Keywords:

coronary heart disease – chronic kidney disease – Arteries – Coronary arteries – myocardial infarction – Ejection fraction – Calcification – Coronary artery bypass grafting

Introduction

Serum phosphorus plays a pivotal role in various biologic processes including energy production, membrane transport, and cellular signaling [1]. Serum phosphorus levels are tightly regulated via dietary absorption, bone formation, and renal excretion, and thus directly reflect renal dysfunction and vascular calcification [2,3]; therefore, patients with renal dysfunction and cardiovascular disease typically show a graded association between serum phosphorus and poor prognosis [47].

In addition to the well-known adverse effects of serum phosphorus abnormalities, an increase in morbidity after cardiovascular surgeries has also been reported [810]. However, in contrast to the results of percutaneous coronary intervention [11,12], previous studies have shown an inconsistent relationship between serum phosphorus abnormalities and coronary artery bypass grafting (CABG) outcomes [8,13]. In addition, to the best of our knowledge, no prior studies have used the clinical criteria definition of phosphorus abnormalities to perform a direct three-group comparison of the outcomes of CABG among normal, hypo- and hyperphosphatemia groups.

Considering that serum phosphorus is a well-known marker of vascular calcification and an assay is readily available, the effect of preoperative serum phosphorus abnormalities defined based on clinical criteria on the outcomes of CABG need to be re-evaluated. Therefore, in the present study, we evaluated whether preoperative serum phosphorus abnormalities are related to outcomes of CABG using the definition of serum phosphorus abnormalities from clinical criteria and a relatively new statistical technique, inverse probability weighting (IPW) adjustment.

Materials and methods

Study population and data collection

This study was approved by the Institutional Review Board of Samsung Medical Center (IRB No. 2017-10-109-003) and conducted in accordance with the principles of the Declaration of Helsinki. Considering the nature of this retrospective study and the minimal risk to patients, the need for informed consent was waived by the institutional review board. From January 2001 to December 2014, a consecutive 5,456 adult patients who underwent CABG at our institution were enrolled initially. After excluding 467 patients without preoperative phosphorus data, a total of 4,989 patients were analyzed. Serum phosphorus levels were measured during preoperative evaluation (Roche/Hitachi cobas c 701/702, Mannheim, Germany). The normal range for phosphorus was provided by the manufacturer (2.5 to 4.5 mg/dl). The patients were stratified into three groups according to serum phosphorus levels (i.e., normal, hypophosphatemia, and hyperphosphatemia).

Perioperative medical data were collected using a standardized form and protocol. Perioperative medications, laboratory findings, and echocardiographic data were extracted automatically from electronic medical records with the aid of the Institutional Medical Information Department. Death records were collected from the national database. Other postoperative clinical outcomes and causes of death were collected though manual review of each case by independent researchers who were blinded to perioperative medical data including preoperative phosphorus level.

Study outcomes and definitions

Clinical outcomes were defined according to a report on cardiovascular events in clinical trials from the American College of Cardiology Foundation/American Heart Association task force [14]. The primary outcome was all-cause death within a median follow-up of 48 months. Secondary outcomes were cardiovascular death, graft failure, and a composite of major adverse cardiovascular and cerebral events (MACCE), including all-cause death, myocardial infarction (MI), repeat revascularization, and stroke. Cardiovascular death was defined as death related to MI, cardiac arrhythmia, heart failure, stroke, or any other vascular causes. Graft failure was confirmed by coronary angiogram and defined as stenosis more than 50%, complete occlusion of the graft and/or an anastomosed coronary artery, or by string sign. Recurrent MI was defined as recurrent symptoms with new electrocardiographic changes compatible with MI or cardiac marker elevation according to the Fourth Universal Definition of MI [15]. Repeat revascularization was defined as revascularization on either target or non-target vessel. Stroke was defined as a new ischemic or hemorrhagic lesion with a neurological deficit lasting longer than 24 hours.

Statistical analysis

We used ANOVA or Kruskal-Wallis test to compare differences in baseline characteristics. Covariates with a univariable effect of P < 0.1 or clinically relevant coavarates were retained in multivariable analysis. The following were covariates for adjustment: male sex, age, diabetes, ejection fraction <40%, chronic kidney disease, old MI, acute coronary syndrome, left main artery disease, emergent operation, off-pump technique, valve-combined operation, and use of the right internal thoracic artery and the right gastroepiploic artery. To further reduce selection bias and maximize the study power while maintaining the balance in confounding factors between the three groups, we conducted rigorous adjustment for differences in baseline characteristics of patients using weighted Cox proportional hazards regression models with IPW [16]. According to this technique, weights for patients with hypo- and hyperphosphatemia were the inverse of the propensity score and weights for patients with normal phosphorus level were the inverse of 1 ± the propensity score. The reduction in the risk of outcome was compared using a stratified Cox regression model. Adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) were reported. In addition, we performed subgroup analysis to reveal hidden interactions with male sex, chronic kidney disease, ejection fraction <40%, valve-combined operation, emergent operation, and the off-pump technique. Statistical analyses were performed with SAS version 9.4 (SAS Institute, Cary, NC) and R 3.5.1 (Vienna, Austria; http://www.R-project.org/). All tests were 2-tailed and P < 0.05 was considered statistically significant.

Results

Patient characteristics

The baseline characteristics of the three groups are summarized in Table 1. Of the enrolled 4,989 patients, 4,544 (91.1%) 238 (4.8%), and 207 (4.1%) were stratified into the normal, hypophosphatemia, and hyperphosphatemia groups, respectively (Table 1). Mean serum phosphorus levels were 3.5 (±0.5) mg/dl, 2.1 (±0.3) mg/dl and 5.1 (±0.6) mg/dl in each group. The three groups had significantly different baseline characteristics including male sex, age, diabetes, ejection fraction < 40%, chronic kidney disease, old MI, emergent operation, off-pump technique, valve-combined operation, and the use of the right gastroepiploic artery. After IPW adjustment, the balance between the three groups was improved and presented as the change in the average standardized absolute mean difference in S1 Fig.

Tab. 1. Baseline characteristics.
Baseline characteristics.

Serum phosphorus abnormalities and clinical outcomes

The overall incidence of all-cause death was 4.1% (206/4989), with a 3.7% incidence in the normal group, 9.7% in the hypophosphatemia group, and 6.3% in the hyperphosphatemia group (P < 0.6). In multivariate Cox regression analysis only hypophosphatemia was significantly associated with the incidence of all-cause death (HR 1.76; 95% CI 1.13–2.76; P = 0.01) (Table 2). However, on further analysis using IPW adjustment, the association between hypophosphatemia and all-cause death was not significant (HR 1.52; 95% CI 0.89–2.61; P = 0.12). The incidence of graft failure was significantly higher in the hypophosphatemia group after adjustment using multivariate Cox regression analysis and IPW analysis (HR 2.14; 95% CI 1.22–3.75; P = 0.01, HR 2.51; 95% CI 1.37–4.61; P = 0.003, respectively).

Tab. 2. Clinical outcomes.
Clinical outcomes.

To determine whether the incidence of all-cause death and graft failure varied based on other covariates, we calculated the odds ratios (ORs) of various complex subgroups (Figs 1 and 2). Preoperative hypophosphatemia was associated with an increased incidence of all-cause death only in patients with chronic kidney disease or those undergoing non-emergent operation (P for interaction = 0.05 and 0.03, respectively). Other variables showed no interaction with all-cause death or graft failure.

Subgroup analysis of all-cause death according to the following variables: male sex, chronic kidney disease, ejection fraction<40%, valve combined operation, emergent operation, and the off-pump technique.
Fig. 1. Subgroup analysis of all-cause death according to the following variables: male sex, chronic kidney disease, ejection fraction<40%, valve combined operation, emergent operation, and the off-pump technique.
Subgroup analysis of graft failure according to the following variables: male sex, chronic kidney disease, ejection fraction<40%, valve combined operation, emergent operation, and the off-pump technique.
Fig. 2. Subgroup analysis of graft failure according to the following variables: male sex, chronic kidney disease, ejection fraction<40%, valve combined operation, emergent operation, and the off-pump technique.

Discussion

In the present study, preoperative hypophosphatemia defined based on clinical criteria appeared to be associated with all-cause death and graft failure on univariate analysis. However, further analyses using IPW adjustment revealed that the association with all-cause death was not significant, and subgroup analysis indicated that this association varied based on chronic kidney disease and emergent operation. Interestingly, preoperative hypophosphatemia was independently associated with graft failure after CABG.

The relationship between serum phosphorus level and vascular calcification has been established based on a large body of evidence [2], and serum phosphorus abnormalities have been implicated in cardiovascular mortality and calcification of coronary arteries [4,5]. However, previous studies of patients undergoing coronary revascularization have shown inconsistent results [1113]. Interestingly, all of these studies stratified patients into multiple groups in a graded manner by applying discrete criteria within normal range. Such methods might be convenient for showing statistical relationships but, in our opinion, does not reflect the clinical significance of phosphorus abnormalities. Moreover, the accuracy of the statistics applied in those studies to adjust covariates for comparing more than two groups is also questionable. Therefore, we stratified patients into normal, hypo-, and hyperphosphatemia groups using phosphorus levels defined based on clinical criteria and analyzed data using IPW adjustment.

Our study demonstrated that hyperphosphatemia is not associated with all-cause mortality. This result corresponds with the findings of a recent study on CABG [13]. The suggested explanation was that diseased arteries are fully replaced by endothelialized arteries or vein grafts in CABG, in contrast to other cardiovascular disease or coronary interventions, and so vascular calcification biomarkers are relatively useless in predicting outcomes [13]. On the other hand, in this study, hypophosphatemia appeared to be associated with all-cause death in multivariate analysis. Although this relationship has been previously suggested in open heart surgey [8], our subgroup analysis revealed that the association varied significantly based on other covariates and was limited to those with chronic kidney disease or those undergoing non-emergent operation. Considering that phosphorus-relevant biomarkers such as bone minerals, vitamin D, and parathyroid hormone are closely related to increases in cardiovascular risk, especially in the patients with kidney disease [7,17,18], these results might be related to the three main regulation mechanisms (i.e., diet intake, bone formation, and renal excretion), which interact dynamically. Poor oral intake is frequently observed in patients undergoing emergent procedures [19].

For adjustments among the three groups, we used a relatively new statistical method in this study. The propensity scores of relevant covariates are generally used to allow an observational study to mimic some of the particular characteristics of randomized trials [20] and there are four broad ways to use calculated propensity scores: matching, IPW, stratification, and covariate adjustment [21]. Although matching is the most frequently used method, using a logistic regression model to estimate the propensity score may lead to biased results in misspecified models [22]. Moreover, adequate adjustment with these methods could be challenging in comparisons among more than two groups. Therefore, using a generalized boosted regression has been proposed as an alternative approach for estimating propensity scores in IPW, and has also been suggested to be a suitable substitute matching method in some models [20]. For this study, institutional statistical experts participated in the design, and chose IPW as the most suitable statistical method to directly compare both hypo- and hyperphosphatemia to the normal group.

Interestingly, even after adjustment, hypophosphatemia was independently associated with graft failure. We suspect that this was related to endothelial dysfunction in hypophosphatemia [23]. In endothelial cells, nitric oxide is produced from L-arginine and molecular oxygen, and acts as a potent vasodilator that determines vascular remodeling [24]. However, in an experimental study, simulated hypophosphatemia decrease endothelial nitric oxide synthase activity and nitric oxide production via reduced intracellular calcium and increased protein kinase C β2 [23]. In addition, previous studies suggested that hypophosphatemia can cause a wide variety of cardiovascular diseases such as heart failure after cardiac surgery [8,25]. However, considering that graft failure after CABG can be caused not only by stenosis or thrombosis but also by native vessel flow or subclavian steal [26], further investigation of the relationship between hypophosphatemia and graft failure in CABG is needed.

There were several limitations to this study. First, since our study had a retrospective observational design, the possibility that hidden bias or unmeasured confounding factors might have influenced the results even after adjustment cannot be completely ruled out. Second, serum phosphorus abnormalities were determined based on a single preoperative measurement, and the duration of hypo- or hyperphosphatemia could not be considered in our analysis. Rapid turnover and tight regulation might have led to variation according to the timing of measurement. So, whether the correction of serum phosphorus level would improve clinical outcomes remains uncertain in this study and needs further investigation. Third, our study could not explain the underlying pathophysiologic mechanisms or determine whether phosphorus supplementation could improve outcomes. Finally, since we analyzed 14 years of data, our institutional protocols for surgical and perioperative management would have changed during the study period.

Conclusions

In conclusion, preoperative serum phosphorus abnormalities defined based on clinical criteria were not associated with clinical outcomes of CABG. However, further studies will be needed to reveal the relationship between hypophosphatemia and graft failure.

Supporting information

S1 Fig [tif]
The change in the average standardized absolute mean difference before and after inverse probability weighting adjustment.

S1 File [xlsx]
Dataset.


Zdroje

1. Komaba H, Fukagawa M. Phosphate-a poison for humans? Kidney Int. 2016; 90:753–63. doi: 10.1016/j.kint.2016.03.039 27282935

2. Podd D. Hyperphosphatemia: understanding the role of phosphate metabolism. JAAPA. 2010; 23:32–7. doi: 10.1097/01720610-201007000-00008 20690439

3. O'Neill WC. Pyrophosphate, alkaline phosphatase, and vascular calcification. Circ Res. 2006; 99:e2. doi: 10.1161/01.RES.0000234909.24367.a9 16857967

4. Tonelli M, Sacks F, Pfeffer M, Gao Z, Curhan G, Cholesterol, et al. Relation between serum phosphate level and cardiovascular event rate in people with coronary disease. Circulation. 2005; 112:2627–33. doi: 10.1161/CIRCULATIONAHA.105.553198 16246962

5. Tonelli M, Curhan G, Pfeffer M, Sacks F, Thadhani R, Melamed ML, et al. Relation between alkaline phosphatase, serum phosphate, and all-cause or cardiovascular mortality. Circulation. 2009; 120:1784–92. doi: 10.1161/CIRCULATIONAHA.109.851873 19841303

6. Adeney KL, Siscovick DS, Ix JH, Seliger SL, Shlipak MG, Jenny NS, et al. Association of serum phosphate with vascular and valvular calcification in moderate CKD. J Am Soc Nephrol. 2009; 20:381–7. doi: 10.1681/ASN.2008040349 19073826

7. Kestenbaum B, Sampson JN, Rudser KD, Patterson DJ, Seliger SL, Young B, et al. Serum phosphate levels and mortality risk among people with chronic kidney disease. J Am Soc Nephrol. 2005; 16:520–8. doi: 10.1681/ASN.2004070602 15615819

8. Cohen J, Kogan A, Sahar G, Lev S, Vidne B, Singer P. Hypophosphatemia following open heart surgery: incidence and consequences. Eur J Cardiothorac Surg. 2004; 26:306–10. doi: 10.1016/j.ejcts.2004.03.004 15296888

9. Andersen PT, Nielsen LK, Moller-Petersen J, Henneberg EW, Egeblad K. Severe hypophosphatemia following elective abdominal aortic bypass grafting. Acta Chir Scand. 1987; 153:641–6. 3434107

10. van Kuijk JP, Flu WJ, Chonchol M, Valentijn TM, Verhagen HJ, Bax JJ, et al. Elevated preoperative phosphorus levels are an independent risk factor for cardiovascular mortality. Am J Nephrol. 2010; 32:163–8. doi: 10.1159/000315856 20606420

11. Oh PC, Lee K, Kim TH, Moon J, Park HW, Jang HJ, et al. Prognostic impact of alkaline phosphatase measured at time of presentation in patients undergoing primary percutaneous coronary intervention for ST-segment elevation myocardial infarction. PLoS One. 2017; 12:e0171914. doi: 10.1371/journal.pone.0171914 28182682

12. Park JB, Kang DY, Yang HM, Cho HJ, Park KW, Lee HY, et al. Serum alkaline phosphatase is a predictor of mortality, myocardial infarction, or stent thrombosis after implantation of coronary drug-eluting stent. Eur Heart J. 2013; 34:920–31. doi: 10.1093/eurheartj/ehs419 23242189

13. Kim TK, Hong DM, Choi YH, Koo CH, Cho YJ, Park JB, et al. Preoperative Serum Alkaline Phosphatase and Clinical Outcome of Off-Pump Coronary Artery Bypass Surgery. Circ J. 2017; 81:799–805. doi: 10.1253/circj.CJ-16-1288 28250357

14. Hicks KA, Tcheng JE, Bozkurt B, Chaitman BR, Cutlip DE, Farb A, et al. 2014 ACC/AHA Key Data Elements and Definitions for Cardiovascular Endpoint Events in Clinical Trials: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Data Standards (Writing Committee to Develop Cardiovascular Endpoints Data Standards). Circulation. 2015; 132:302–61. doi: 10.1161/CIR.0000000000000156 25547519

15. Thygesen K, Alpert JS, Jaffe AS, Chaitman BR, Bax JJ, Morrow DA, et al. Fourth Universal Definition of Myocardial Infarction (2018). J Am Coll Cardiol. 2018; 72:2231–64. doi: 10.1016/j.jacc.2018.08.1038 30153967

16. Austin PC, Stuart EA. Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies. Stat Med. 2015; 34:3661–79. doi: 10.1002/sim.6607 26238958

17. Block GA, Hulbert-Shearon TE, Levin NW, Port FK. Association of serum phosphorus and calcium x phosphate product with mortality risk in chronic hemodialysis patients: a national study. Am J Kidney Dis. 1998; 31:607–17. doi: 10.1053/ajkd.1998.v31.pm9531176 9531176

18. Bellasi A, Mandreoli M, Baldrati L, Corradini M, Di Nicolo P, Malmusi G, et al. Chronic kidney disease progression and outcome according to serum phosphorus in mild-to-moderate kidney dysfunction. Clin J Am Soc Nephrol. 2011; 6:883–91. doi: 10.2215/CJN.07810910 21393493

19. Uribarri J. Phosphorus homeostasis in normal health and in chronic kidney disease patients with special emphasis on dietary phosphorus intake. Semin Dial. 2007; 20:295–301. doi: 10.1111/j.1525-139X.2007.00309.x 17635818

20. Austin PC. An Introduction to Propensity Score Methods for Reducing the Effects of Confounding in Observational Studies. Multivariate Behav Res. 2011; 46:399–424. doi: 10.1080/00273171.2011.568786 21818162

21. Austin PC, Stuart EA. The performance of inverse probability of treatment weighting and full matching on the propensity score in the presence of model misspecification when estimating the effect of treatment on survival outcomes. Stat Methods Med Res. 2017; 26:1654–70. doi: 10.1177/0962280215584401 25934643

22. McCaffrey DF, Ridgeway G, Morral AR. Propensity score estimation with boosted regression for evaluating causal effects in observational studies. Psychol Methods. 2004; 9:403–25. doi: 10.1037/1082-989X.9.4.403 15598095

23. Peng A, Wu T, Zeng C, Rakheja D, Zhu J, Ye T, et al. Adverse effects of simulated hyper- and hypo-phosphatemia on endothelial cell function and viability. PLoS One. 2011; 6:e23268. doi: 10.1371/journal.pone.0023268 21858050

24. Dimmeler S, Fleming I, Fisslthaler B, Hermann C, Busse R, Zeiher AM. Activation of nitric oxide synthase in endothelial cells by Akt-dependent phosphorylation. Nature. 1999; 399:601–5. doi: 10.1038/21224 10376603

25. Heames RM, Cope RA. Hypophosphataemia causing profound cardiac failure after cardiac surgery. Anaesthesia. 2006; 61:1211–3. doi: 10.1111/j.1365-2044.2006.04839.x 17090245

26. Rasmussen C, Thiis JJ, Clemmensen P, Efsen F, Arendrup HC, Saunamaki K, et al. Significance and management of early graft failure after coronary artery bypass grafting: feasibility and results of acute angiography and re-re-vascularization. Eur J Cardiothorac Surg. 1997; 12:847–52. doi: 10.1016/s1010-7940(97)00268-6 9489868


Článek vyšel v časopise

PLOS One


2019 Číslo 12
Nejčtenější tento týden
Nejčtenější v tomto čísle
Kurzy Podcasty Doporučená témata Časopisy
Přihlášení
Zapomenuté heslo

Zadejte e-mailovou adresu, se kterou jste vytvářel(a) účet, budou Vám na ni zaslány informace k nastavení nového hesla.

Přihlášení

Nemáte účet?  Registrujte se

#ADS_BOTTOM_SCRIPTS#