OUP user menu

Impact of radiocontrast use during left ventricular pacemaker lead implantation for cardiac resynchronization therapy

Gregory A. Tester, Amit Noheria, Heather L. Carrico, Jennifer A. Mears, Yong-Mei Cha, Brian D. Powell, Paul A. Friedman, Robert F. Rea, David L. Hayes, Samuel J. Asirvatham
DOI: http://dx.doi.org/10.1093/europace/eur282 243-248 First published online: 24 October 2011

Abstract

Aims The risk of contrast-induced nephropathy (CIN) with radiocontrast use during left ventricular (LV) lead placement for cardiac resynchronization therapy (CRT) is unknown. It is unclear as to whether minimizing contrast use impacts adequacy of LV lead placement.

Methods and results A retrospective analysis was performed of all LV leads placed for CRT at Mayo Clinic, Rochester, MN from 16 March 2001 to 1 April 2009. The primary goal was to assess risk of CIN and adequacy of lead placement depending on the amount of contrast administered during CRT placement. Contrast-induced nephropathy was defined as a ≥25% increase in serum creatinine ≥48 h post-procedurally. Adequacy of lead placement was assessed in a blinded fashion by review of procedural fluoroscopic and post-procedural radiographic images. Eight hundred and twenty-two subjects were divided based on the amount of procedural contrast used into tertile 1 (<55 mL, 257 patients), tertile 2 (55–94 mL, 261 patients), and tertile 3 (≥95 mL, 304 patients). Contrast-induced nephropathy occurred in 5.4% of patients in tertile 1, 5.4% in tertile 2 and 11.8% in tertile 3 (P = 0.004). Among the tertiles, lead positioning was optimal in 95, 80 and 66%, respectively (P < 0.0001). Fluoroscopic time was 34 ± 23, 42 ± 26, and 48 ± 30 min in tertiles 1, 2, and 3 (P < 0.0001).

Conclusion Risk of CIN with CRT implantations was substantial. Increased volume of radiocontrast used for LV lead placement was associated with substantially increased risk of CIN. Minimal contrast use was associated with decreased procedural times without adverse impact on adequacy of lead placement.

  • Contrast-induced nephropathy
  • Cardiac resynchronization therapy (CRT)
  • LV lead
  • Device implantation
  • Radiocontrast

Introduction

Heart failure is a leading cause of worldwide morbidity and mortality affecting over 23 million individuals, and a cause of 1 in every 8.6 deaths in the USA.1 Management of heart failure is largely based on pharmacotherapy to improve haemodynamics and cardiac remodelling. Within the last decade, cardiac resynchronization therapy (CRT) has emerged as a novel device-based therapy for individuals with symptomatic New York Heart Association (NYHA) class III/ambulatory class IV heart failure, severe left ventricular (LV) systolic dysfunction (left ventricular ejection fraction, LVEF ≤35%), and a QRS duration ≥120 ms despite optimal pharmacotherapy. In addition, the 2010 European Society of Cardiology focused update on device therapy in heart failure added a class I indication to CRT for NYHA function class II patients with LVEF ≤35% and QRS duration ≥150 ms.2,3 In this population of patients, CRT has been shown to improve NYHA functional class, quality of life, 6-min-walk distance, QRS duration, and LVEF.4 Cardiac resynchronization therapy's proposed mechanism of benefit is achieved by improving mechanical synchrony, which is associated with reverse cardiac remodelling and improved LV function.4

Cardiac resynchronization therapy improves mechanical synchrony by pacing both the septal and LV free wall. Pacing of the LV is achieved by cannulation of the coronary sinus and delivering a pacemaker lead into the coronary veins on the epicardial surface. Final location and site of pacing via the LV lead has been shown to affect clinical outcomes. Multiple studies have shown varying data on the site of optimal placement, but the majority supports an optimal site on the lateral LV.59 Left ventricular lead placement can be technically challenging.1013 Typically, radiocontrast is used to visualize the tributaries of the coronary venous system to assist selective cannulation of veins draining the free wall of the LV.1416 Contrast use is well known to produce nephropathy in at-risk patients undergoing other invasive procedures.17,18 Patients requiring biventricular pacing usually have advanced heart failure and could potentially be at an increased risk for development of CIN. However, contrast is not universally used for LV lead implantation, and there is no evidence to suggest whether or not contrast use affects LV lead placement. Specifically, the effect of contrast use on patient outcomes, development of CIN or improving the adequacy of lead placement, is unknown. The purpose of this study, therefore, was to determine if contrast use for LV lead placement, first, increases the risk of CIN, and secondly, improves the rate of optimal lead placement.

Methods

The patient population for the analyses comprised all patients who underwent LV lead placement for CRT at the Mayo Clinic, Rochester, MN between 16 March 2001 and 1 April 2009. They were identified from the prospectively maintained implantable cardioverter–defibrillator (ICD) database maintained at the Mayo Clinic. Data on demographic and clinical variables including NYHA functional class19 and vital status were collected by retrospective review of the electronic medical record. Data on serum creatinine levels were obtained from the laboratory reports on each patient. Serum creatinine at Mayo Clinic is measured by an institutional standardized assay. The estimated glomerular filtration rate (eGFR) was estimated using the 4-variable Modification of Diet in Renal Diseases (MDRD) equation.20,21 Contrast-induced nephropathy was defined as a ≥25% increase in the ≥48 h post-procedure serum creatinine (obtained within a week of the procedure) compared to pre-procedural value.17 Data on the details of the procedure and on the device and leads implanted were obtained from the ICD database and confirmed by review of the procedure notes. This included information on volume of contrast used, whether coronary venoplasty was performed, and fluoroscopic exposure time. Adequacy of LV lead position was assessed by review of the procedural fluoroscopic images and the post-procedural chest radiographs by an electrophysiologist who was blinded to other information about the patients (Figure 1). Lead position was classified as optimal, suboptimal, and poor. Mid-lateral, distal-lateral, basal-lateral, mid-anterolateral, and mid-posterolateral positions were considered optimal; basal anterolateral, distal anterolateral, basal posterolateral, and distal posterolateral were considered suboptimal; and anterior, anteroseptal, and posterior were considered poor. Last available echocardiographic variables from the echocardiogram reports prior to the procedure were obtained as baseline, and similarly, those at 0–6 and 6–12 months from the date of the procedure, if available, were recorded as follow-up. Mortality was assessed from the clinical chart as death within 1 month of the procedure and last available vital status (assessed in February 2010).

Figure 1

Example of fluoroscopic assessment of position of left ventricular lead—the right anterior oblique (RAO) projections were used to evaluate anterior vs. posterior and basal vs. mid vs. distal positioning of the lead. In this example, the lead is positioned in the anterior distal wall (arrow). The left anterior oblique (LAO) projections were used to determine the septal vs. lateral-free wall location of the lead. In this example, the lead is located on the lateral-free wall of the left ventricle (arrow). Thus, the lead position here was determined to be distal-anterolateral, a suboptimal position.

Statistical analyses

Statistical analysis was done using SAS version 9.1.3 for Windows, Cary, NC. Values of baseline covariates if missing were imputed using the Markov-chain Monte Carlo multiple imputation method. The study population was divided into tertiles based on increasing volume of contrast used during CRT implantation. Chi-squared test for comparison of proportions and one-way ANOVA (analysis of variance) for comparison of means were used to compare the distributions of baseline characteristics and outcomes between the tertiles. The primary analysis was the 3-tertile comparison of the proportion of patients developing CIN. Secondary analyses comprised 3-tertile comparisons of (i) proportion of patients with optimal lead position and (ii) mean fluoroscopic times. Logistic and linear regressions were subsequently used to assess the association of tertile 2 and tertile 3 as compared with tertile 1 with outcomes. The regression analyses were performed sequentially (i) unadjusted for covariates, (ii) adjusted for age, sex, and baseline eGFR, and (iii) additionally adjusted for other variables [date of procedure, radiocontrast use in the week prior to the procedure, venoplasty done or not, contrast used or not for (i) cannulation of coronary sinus, (ii) venogram, or (iii) both, history of diabetes, hypertension, hyperlipidemia, stroke or transient ischaemic attack, peripheral vascular disease, tobacco use, baseline LVEF, LV end-diastolic dimension, and NYHA functional class]. Any effect of additional adjustment for operators and optimal lead positioning was assessed. Interaction terms were added to evaluate any effect modification by age, diabetes, sex, and eGFR. Sensitivity analyses were performed by removing the imputed missing values, and by using volume of contrast and percent change in creatinine as continuous variables instead of being categorized as tertiles and occurrence of CIN, respectively. Operator-adjusted time trends in volume of contrast use, fluoroscopy time, and adequate lead placement were assessed by regressing on date of the procedure. All statistical tests were two tailed and assumed a significance threshold for type 1 error of 0.05. Proportions are reported as number (percentage), continuous distributions as mean ± standard deviation, and regression results as estimate (95% confidence intervals, P-value). The study was approved by the Mayo Clinic Institutional Review Board.

Results

We analysed 823 consecutive patients who had CRT system placed at our institution. Among these, one did not have pre-procedural creatinine available and was excluded for all analyses. The remaining 822 patients were divided into tertiles 1, 2, and 3 comprising 257, 261, and 304 patients who received <55 cc, ≥55 to <95 cc, and ≥95 cc of contrast, respectively. There were 108/822 (13%) patients who did not receive any contrast for the CRT lead implantation. Another four patients received <10 cc contrast, and four received 100 cc. Among the other 714/822 (87%) patients who received >0 cc of contrast, 210/822 (26%) received contrast only for venogram. Values were missing in 67 patients for baseline LVEF, 113 patients for baseline LV end-diastolic dimension, 55 patients for baseline NYHA functional class, and 27 patients for procedural fluoroscopy exposure time, and were imputed. Overall, the mean volume of contrast was 77 ± 54 mL, age 67 ± 12 years, baseline eGFR 57 ± 21 mL/min, LVEF 24 ± 9%, males 623 (76%), diabetics 289 (35%), hypertensives 451 (55%), history of smoking in 175 (21%), and 721 (88%) had NYHA functional class III or IV status. The three tertiles were similar with regard to distributions of the baseline variables including the prevalence of diabetes, hypertension, history of stroke, creatinine/eGFR, LVEF, and NYHA functional class, although there was a trend across the respective tertiles of increasing mean age and prevalence of contrast use in the prior week (Table 1).

View this table:
Table 1

Distribution of variables by tertiles of contrast volume

VariableTertile 1Tertile 2Tertile 3P value
n = 257n = 261n = 304
Age66 ± 1367 ± 1269 ± 110.02
Male187 (73)208 (80)228 (75)0.2
Diabetes92 (36)96 (37)101 (33)0.7
Hypertension129 (50)148 (57)174 (57)0.2
Hyperlipidemia149 (58)179 (69)184 (61)0.03
Stroke39 (15)36 (14)53 (17)0.5
Tobacco use50 (19)59 (23)66 (22)0.7
Peripheral vascular disease15 (6)20 (8)27 (9)0.4
Creatinine (mg/dL)1.4 ± 0.61.4 ± 0.51.4 ± 0.50.6
eGFR (mL/min/1.73 m2)57 ± 2357 ± 2056 ± 190.8
Ejection fraction (%)25 ± 924 ± 824 ± 90.2
LVEDD (mm)65 ± 966 ± 967 ± 110.3
NYHA class0.7
 I3 (1)5 (2)6 (2)
 II33 (13)23 (9)30 (10)
 III179 (70)187 (72)214 (70)
 IV41 (16)46 (18)54 (18)
Contrast use in prior week25 (10)33 (13)52 (17)0.03
Venoplasty performed12 (5)12 (5)13 (4)1.0
Volume of contrast (mL)18 ± 1874 ± 9129 ± 42<0.0001
  • eGFR, estimated glomerular filtration rate; LVEDD, left ventricular end-diastolic dimension; NYHA, New York Heart Association.

Overall, CIN developed in 64 (8%) patients. The mean fluoroscopy exposure time was 42 ± 28 min. The lead position was anteroseptal/anterior in 46 (6%), anterolateral in 251 (31%), lateral in 371 (45%), posterolateral in 128 (16%), and posterior in 26 (3%). The leads were in the basal region in 75 (9%), mid-segments 661 (80%), and distal location in 86 (10%). In total, lead placement was adjudicated to be optimal in 657 (80%), sub-optimal in 93 (11%), and poor in 72 (9%) (Figure 2). The fluoroscopy exposure time in the three tertiles, respectively, was 34 ± 23, 42 ± 26, and 48 ± 30 min (P < 0.0001). There were differences between the tertiles in the distribution of poor, sub-optimal or optimal lead position (P < 0.0001) with optimal lead placement decreasing from 245 (95%) to 210 (80%) to 202 (66%) with increasing volume of contrast. Contrast-induced nephropathy occurred in 14 (5.5%), 14 (5.4%), and 36 (11.8%), respectively (P= 0.004) (Table 2).

View this table:
Table 2

Distribution of main outcomes by tertiles of contrast volume

OutcomeTertile 1Tertile 2Tertile 3P value
n = 257n = 261n = 304
Contrast-induced nephropathy14 (5)14 (5)36 (12)0.004
Lead-positiona<0.0001
Poor4 (2)24 (9)44 (14)
Sub-optimal8 (3)27 (10)58 (19)
Optimal245 (95)210 (80)202 (66)
Radiation exposure time (min)34 ± 2342 ± 2648 ± 30<0.0001
  • aLeft ventricular epicardial lead position was categorized as optimal if it was reckoned to be located at the basal, mid or distal lateral, the mid-anterolateral, or mid-posterolateral segments. It was suboptimal for the basal and distal anterolateral and posterolateral segments. Other positions were considered poor.

Figure 2

Depiction of the distribution the lead position as assessed from fluoroscopy and radiographs for the 822 patients. The dots represent individual patients and are arranged per the estimated lead location. The area with the dark shade was considered the optimal lead position and the light shade sub-optimal, the rest being poor lead positions.

Compared with tertile 1, mean fluoroscopy exposure time in tertile 2 increased by 8.7 min (4.1–13.3, P= 0.0002) and in tertile 3 by 14.5 min (10.0–18.9, P< 0.0001). However, compared with tertile 1, non-optimal lead positioning was more common in tertile 2, odds ration (OR) = 5.0 (2.7–10.0, P< 0.0001), and tertile 3, OR = 10.3 (5.7–20.3, P< 0.0001). These associations remained statistically significant on multivariate adjustments and on sensitivity analyses. Similarly, poor lead placement was more likely in tertiles 2 and 3 compared with tertile 1 [ORs 6.4 (2.4–22.0, P< 0.0001) and 10.7 (4.3–35.9, P< 0.0001)].

Compared with tertile 1, the odds of developing CIN was similar in tertile 2, OR 0.98 (0.46–2.11, P= 0.97), and higher in tertile 3, OR 2.33 (1.23–4.43, P= 0.01). No substantial changes were noted on multivariate analysis. Adjusting for age, sex, and baseline eGFR, ORtertile-2 0.96 (0.45–2.07; P= 0.92) and ORtertile-3 2.32 (1.22–4.42, P= 0.01). On additional multivariate adjustments for baseline factors, ORtertile-2 1.05 (0.46–2.40, P= 0.91) and ORtertile-3 2.54 (1.25–5.13, P= 0.01). On further adjustment for lead positioning and operator performing the lead placement, ORtertile-2 1.59 (0.53–4.80, P= 0.41) and ORtertile-3 3.98 (1.44–11.02; P= 0.008). There was no significant effect modification by age, diabetes, sex, and eGFR, and the analyses were robust to sensitivity analyses. None of the other covariates in the multivariate model predicted risk of CIN.

Other adverse effects were not statistically different across the three tertiles. Bleeding complications including procedural bleeding or pocket haematoma occurred in 7%, infections within a month of implantation occurred in 2%, pneumothorax occurred in 0.7%, and readmission to hospital for any cause within a week occurred in 4%. Within a month of implantation, 2 (1%), 3 (1%), and 5 (2%) patients died in the three tertiles, respectively (P = 0.4). Mortality until the last follow-up occurred in 76 (30%), 75 (29%), and 90 (30%) patients, respectively (P= 1.0). For those patients where follow-up information up to a year following the procedure was available, there was no association of tertile of contrast volume with change in LVEF (n= 504, P= 0.4), LV end-diastolic dimension (n= 489, P= 0.1) or NYHA functional class (n= 577, P= 0.2) (Table 3). There was a trend towards higher long-term mortality associated with development of CIN, OR 1.6 (0.94–2.7, P= 0.08).

View this table:
Table 3

Distributions of adverse events, echocardiographic parameters, functional status and mortality on follow-up by tertiles of contrast volume

OutcomeTertile 1Tertile 2Tertile 3P value
n = 257n = 261n = 304
Bleeding19 (7)14 (5)24 (8)0.5
Infection within a month5 (2)4 (2)6 (2)0.9
Pneumothorax2 (1)04 (1)0.2
Readmission within a week12 (5)11 (4)12 (4)0.9
Change in ejection fraction (%)a,b5 ± 127 ± 127 ± 110.4
Change in LVEDD (mm)a−1 ± 7−3 ± 7−3 ± 80.1
Change in NYHA classa0.2
 Improvement79 (44)93 (50)113 (53)
 No change82 (46)73 (39)82 (39)
 Worsening17 (10)21 (11)17 (8)
Mortality
 Within a month5 (2)2 (1)3 (1)0.4
 Long term76 (30)75 (29)90 (30)1.0
  • LVEDD, left ventricular end-diastolic dimension; NYHA, New York Heart Association.

  • aLimited follow-up information available—ejection fraction (n = 504), LVEDD (n = 489), NYHA (n = 577).

  • bChange in ejection fraction reflects absolute percentage point change.

Assessing for time trends adjusted for the operator performing the procedure, the volume of contrast use decreased with time, mean estimated annual decrease 3.8 mL (2.2–5.4, P< 0.0001). Similarly, adjusting for operator, the proportion of LV leads placed in a non-optimal position or in the poor position decreased with time, OR per additional year 0.88 (0.80–0.98, P= 0.02) and 0.72 (0.61–0.83, P< 0.0001), respectively. The estimated annual decrease in radiation time was 1.9 min (0.9–2.9, P= 0.0002).

Discussion

In this study on 822 patients undergoing CRT implantation, we found the development of CIN to be significantly associated with the volume of contrast used with an overall risk of 7.9%. The odds of developing CIN were over two-fold higher in the highest tertile of contrast used during the procedure compared with the lowest tertile even on adjusting for other covariates. The highest tertile compared with the lowest further had on average 14 min of extra radiation exposure time and substantially lower optimal location of lead placement (66 vs. 95%). We did not observe any differences in adverse events or mortality across the tertiles.

Cardiac resynchronization therapy using biventricular pacing has emerged as a device-based therapeutic option for selected patients with systolic heart failure that is refractory to medical management. With the increasing incidence of heart failure, more biventricular pacing devices are likely to be implanted. As heart failure is itself a risk factor for renal dysfunction, placement of the LV CRT lead through the epicardial venous system using radiocontrast in this population places the patient at an increased risk of CIN.22,23 The risk of 7.8% for development of CIN in this study is higher than the 3.3% reported in a study on percutaneous coronary artery interventions.18 This increased risk of CIN might be indicative of the relatively higher susceptibility of the population undergoing CRT implantation for renal dysfunction. Moreover, in context of other radiology and cardiovascular procedures, CIN has been shown to be associated with increased morbidity and mortality.17,18,24

To our knowledge, this is the first study evaluating CIN associated with biventricular pacemaker implantation. Our findings correlate with reports from other interventional procedures showing a higher risk of CIN with contrast volumes ≥100 mL.25 We observed a greater risk for CIN in those receiving ≥95 mL of contrast. This elevated risk persisted, and became more pronounced, on adjusting for various potential confounders. Further, this association was independent of the operator performing the procedure and whether the lead location was optimal, suboptimal, or poor. We, however, did not find significant differences in other adverse events or mortality associated with higher contrast use. There was a trend towards higher overall mortality associated with CIN.

In this retrospective study, minimal contrast use was not associated with inadequacy of lead placement. In fact, surprisingly, better lead position was significantly associated with less contrast use. Potential explanations for this include the possibility that more contrast was used when lead positioning was difficult, and perhaps without contrast use, even less adequate lead placement or failed implant may have resulted. However, we noted that simultaneously less contrast use and better lead positioning occurred with increasing operator experience and certain operators preferentially performed the procedure without using contrast in their patients. These results suggest that factors beyond difficulty with lead placement alone results in increased contrast use. Given the retrospective nature of this study, we could not clearly quantify difficulty of lead implant with these procedures, and we could also not exclude the possibility that contrast use itself (dissection and selective cannulation) may have contributed to lead positioning difficulty.26 A significant number of successful and adequate LV lead positioning was done without any contrast use at all, suggesting that contrast is not routinely required for LV lead implantation. Finally, as noted above, although primarily difficult venous anatomy may have resulted in increased contrast requirement, using additional contrast did not appear in our study to eventually result in optimal lead position.

Implications

  1. Excessive radiocontrast use appears to increase the risk of CIN, although it is unclear if this complication impacts long-term outcomes.

  2. In some cases, LV lead implantation can be performed successfully without the use of radiocontrast.

  3. The adequacy of LV lead implantation might not be adversely affected when contrast use is minimized.

Conclusion

The population of patients undergoing LV lead implantation are at significant risk of CIN. An increased amount of contrast is associated with further increased risk of CIN and may not lead to substantial improvement in the adequacy of LV lead placement. While we could not exclude the possibility in this retrospective study that complex LV lead implantation led to simultaneously increased contrast use and decreased adequacy of LV lead placement, operators should be cognizant for the potential of CIN with excessive contrast use while appreciating that radiocontrast may not be required routinely in all patients for adequate lead placement.

Conflict of interest: none declared.

References

View Abstract