Ambulatory monitoring unmasks hypertension among kidney transplant patients: single center experience and review of the literature

In kidney transplant recipients, hypertension (HTN) is an independent risk factor for cardiovascular diseases (CVD) [1, 2], death [3] and graft malfunction independent of death [3‐7], and its prevalence is 50–90% [2, 8, 9]. However, there are no randomized control trials to define target BP levels in renal transplant recipients [10]. The Kidney Disease Improving Global Outcomes (KDIGO) [11] and the American College of Cardiology (ACC) and American Heart Association (AHA) [12] recommendations are to target clinic BP of kidney transplant recipients to levels lower than 130/80 mmHg (a 2C-grade recommendation according to KDIGO). Yet, the recommendations of the National Kidney Foundation are to adjust individually BP levels due to lack in evidence based recommendations [13], while the recent European Society of Hypertension (ESH) and the European Society of Cardiology (ESC) guidelines [14] suggest that the target BP range for patients with chronic kidney disease (CKD) should be 130–139 mmHg, and the Eighth Joint National Committee (JNC 8) recommended values of less than 140/90 mmHg in all CKD patients, not excluding transplant recipients [15].

In the general population, clinic BP measurement often differs from ambulatory BP monitoring (ABPM) [16]. Ambulatory BP is monitored at the patient’s ordinary environment, and the measurements are recorded every 20–30 min during a period of 24 h and capture sleep. Hence, ABPM is considered a more reliable measure of BP [17]. Importantly, in the general population ABPM correlates better than clinic BP with end organ damage and cardiovascular outcomes [18‐20].

In addition to measuring mean BP, ABPM reveals important data on other values, including awake period BP, asleep period BP, BP variability, morning surge and the sleep-related dipping pattern [21]. Furthermore, it can frame differences between clinic BP measurements and home BP monitoring, by establishing masked hypertension and white-coat hypertension [21, 22]. While white coat hypertension leads to overtreatment [23], masked hypertension represents under-diagnosis and therefore under-treatment [24]. Hence, there is a need in parameters in the daily clinical practice that will help in identifying patients with masked hypertension or a prominent masking effect [25].

Only a few studies have examined the concordance between clinic BP and ABPM in kidney transplant recipients. Though some found a high rate of white-coat hypertension [26‐29], others found higher rates of masked hypertension [30‐32] in this population. Differences in hypertension definitions can account for some of the discrepancy between the studies. Significantly, ABPM better predicts renal graft function [7].

A non-dipping BP pattern (defined as a < 10% reduction in asleep period BP relative to awake period BP) was reported in up to 79% of the patients with kidney allografts [7, 28, 29]. In the general population, a non-dipping pattern is associated with more complications, including higher rates of stroke, dementia, left ventricular hypertrophy, microalbuminuria and increased carotid intima-media thickening [9, 33]. In kidney transplantation, a non-dipping pattern links with poor allograft function and high Doppler resistive index [34].

The lack of well-established data with regards to diagnosis, treatment and monitoring of hypertension in the renal transplant population gave rise to a contemporary “Call to Action” published by several European scientific societies, encouraging performance of surveys aimed at assessing the prevalence of hypertension (including white-coat and masked hypertension) [35]. Accordingly, in this study, our objectives were to describe clinic and ambulatory BP, dipping patterns and clinic-awake differences in kidney transplant patients and identify associated clinical parameters.

We enrolled patients post kidney transplantation followed-up in our renal transplant clinic. Inclusion criteria were age 18 years and above and a period of at least 3 months from the transplantation. Pregnant women were not included. Renal transplant physicians recruited potential participants during regular clinic visits. The Helsinki Committee of the Hadassah Medical Organization approved the protocol for this study. All volunteers provided written informed consent prior to participation.

We collected baseline data including age, weight, transplantation details, underlying kidney disease, medications and BP measurements from medical files, and height and missing information directly from the patients. We calculated the aggregate dose of antihypertensive medications as the sum of percentages of each drug’s dose from its maximal dose. We quantified urinary protein excretion and creatinine clearance from a 24-h urine collected during a period before or after the monitoring (up to 3 months for creatinine clearance; for urinary protein, we used data from up to 4 months for renal transplant recipients who were transplanted more than 1 year before the monitoring and 2 months for recipients who were transplanted up to 1 year before the monitoring). Hemoglobin and serum creatinine were measured in the day of monitoring in most patients; in 10 patients, they were measured up to 1 month before or after the monitoring. In a single patient, we collected data from a visit that occurred 4 months before the ABPM. We estimated glomerular filtration rate (eGFR) using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation. Trough levels of tacrolimus, everolimus, sirolimus and cyclosporine were measured in the morning of monitoring, before taking the morning dose. In five patients, plasma levels were measured within 8 days after the monitoring; in ten patients, they were measured up to 14 days before the monitoring; in a single patient it was measured 4 months before the monitoring, and in two patients, plasma levels are missing. We make use of information from our institution’s ABPM dataset of referred patients for perspective and comparison with this study’s patients (registry control) [36]. This registry includes patients (47% women) referred predominantly by the primary physicians for accepted clinical indications. Three percent had diabetes and 65% had hypertension (as defined by their ambulatory awake BP).

BP was measured either by a nurse (observed) using an oscillometric device (Welch Allyn, 52000 series) before entering the physician’s office (n = 16), or by the physician, using aneroid auscultatory sphygmomanometry (Accucare BP wall mount sphygmomanometer), at least 5 min after arrival at the clinic (n = 60). Typically, single measurements were performed. Our study protocol did not address the particulars of clinic measurements, as our aim was to compare real-life clinic measurements with standardized ambulatory monitoring. Measurements were done on an arm without an arterio-venous fistula, while it rested on a table. Clinic BP was calculated as an average of 2–3 visits during the 3 months before conducting ABPM. If only one measurement was performed in this period, an additional measurement taken 3–12 months before the ABPM was included in the average clinic BP. Hypertension medications were in general not changed during this period. However, adherence and as needed use were not systematically assessed.

Ambulatory BP was monitored with Oscar2 devices (SunTech Medical). Three adult cuff sizes were available. Measurements were recorded every 20 min during the day and every 30 min during nighttime sleep (according to the participant’s projection) over a 24-h monitoring period. In the ABPM analysis, awake period limits and asleep period limits were further defined according to the participant’s log. Participants with at least 20 awake period measurements and 8 sleep measurements were considered to have a valid ABPM recording. One participant discontinued the monitoring before sleep; his 20 awake period records were included in the study. ABPMs were processed with AccuWin Pro v3 (SunTech Medical).

Clinic HTN was defined as mean clinic SBP ≥ 140 mmHg and/or mean clinic DBP ≥ 90 mmHg, according to JNC 8 report [15]. Awake period HTN (ABPM) was defined as mean SBP ≥ 135 mmHg and/or mean DBP ≥ 85 mmHg; asleep period HTN was defined as SBP ≥ 120 mmHg and/or DBP ≥ 70 mmHg; and 24 h HTN was defined as SBP ≥ 130 mmHg and/or DBP ≥ 80 mmHg, according to the ESC recommendations [37].

The primary outcomes of this study were the differences between clinic SBP/DBP and ambulatory awake period SBP/DBP. The secondary outcomes were prevalence of clinic HTN, awake period ambulatory HTN, asleep period ambulatory HTN; white-coat HTN (defined as clinic HTN without awake period ambulatory HTN); masked HTN (defined as awake period ambulatory HTN without clinic HTN); isolated nocturnal HTN; non-dipping pattern (defined as < 10% reduction in SBP from awake period to sleep time) and a reverse-dipping pattern (defined as higher sleep SBP than awake SBP); as well as assessment of awake and sleep periods’ SBP and DBP variability (standard deviation, SD), and average awake period and sleep period heart rates.

Previous studies have also shown negative clinic-ambulatory BP differences in renal transplant populations, albeit of lesser extent. However, in several other studies opposing results have been noted. These previous studies differ one from another in patients’ age, clinic BP and ABPM methods and in exclusion criteria (Table 6).

In a study of 98 patients from New-Zealand [30], the mean differences between clinic SBP and DBP and average 24 h SBP and DBP were − 3.5 mmHg and − 7.2 mmHg, respectively. These values are also negative, but they are less negative than the current study's values. There are several explanations for this difference: (1) In this study, clinic SBP was compared to 24 h BP, which was lower than awake BP. (2) The clinic BP was measured as an average of all the measurements in the study period, apparently including measurements that were done after (and may have been affected by) the ABPM. (3) The study was conducted at a research center, with which the patients were not familiar. (4) Patients with unstable BP levels (not defined) were not included. In a 113-patient Turkish study [32], the mean differences between clinic SBP and DBP and average awake SBP and DBP were − 9.0 mmHg and − 6.4 mmHg, respectively. Patients with uncontrolled HTN (not defined) were excluded. Lastly, in 172 Italian patients, the clinic – awake SBP and DBP differences were − 1 mmHg [42]; clinic BP was measured only in the monitoring day.

On the other hand, in a 260-patient Italian study, the clinic – awake SBP and DBP differences were + 6 and 0 mmHg, respectively [43]. The clinic measurements in the Italian study were done after the ABPM, a fact that could has influenced the results, and the clinic BP was measured as the mean of BP measurements within a mean period of 3.7 years. In another 87-patient Turkish study, the clinic – awake SBP and DBP differences were + 6 and + 4.5 mmHg, respectively [29]; however, several exclusion criteria have been used in this Turkish study (history of diabetes mellitus, heart failure, ischemic heart disease, cardiomyopathy, or significant valvular heart disease; hemoglobin level < 10 g/dL; and serum creatinine level > 1.5 mg/dL). Among 244 Canadian patients, the clinic – awake SBP and DBP differences were + 3.6 and + 2.5 mmHg, respectively [28]. In this Canadian study, one clinic measurement was recorded, obtained within 5 days of ABPM (not mentioned if before or after the ambulatory monitoring). More importantly, the average clinic BP in the Canadian study was 137/79 mmHg, higher than the value in our study. Higher clinic BP levels are associated with less negative clinic – awake BP differences (Fig. 1b).

In a 49-patient Indian study, the clinic – awake SBP and DBP differences were positive at 2, 4, 6, and 9 months after transplantation [26]; however, the mean age of the participants in the Indian study was 35 years, and the methodology of clinic BP measurement was different (it was calculated as a mean of a clinic measurement, the first ABPM measurement and the last ABPM measurements). In a Spanish multicenter study (868 patients), the clinic – awake SBP and DBP differences were + 7 and + 1 mmHg, respectively [27]; however, kidney transplant recipients with Scr > 2.5 mg/dL or with eGFR < 30 ml/min/1.73m² were not included, and description of clinic and ABPM measurement methodology is lacking. In a 94-patient Italian study, the clinic – awake SBP and DBP differences were + 4 and − 1 mmHg, respectively [7]. A possible explanation for this difference is that in this Italian study the clinic BP was 150/86 mmHg on average, and in our study higher clinic BP measurements associated with less negative delta values, as mentioned previously (Fig. 1b).

As in our study, the tendency for a negative clinic–awake BP difference has been observed in other (non-transplant) CKD populations. In an US cross-sectional study on 1492 CKD patients, the differences between clinic and ambulatory awake SBP and DBP were − 5.9 and − 6.4 mmHg, respectively [44]. In a Japanese study [45], masked HTN was more prevalent than white coat HTN among stage 3–5 CKD patients.

In addition to the clinic-awake discrepancy, nocturnal measurements have revealed that 83% of our patients have sleep HTN. This finding is in line with previous results (69% [42], 71% [30]). Also, only 23% of the patients in our study were normal dippers. This too is similar to previous findings (21% [7], 32% [29], 14% [27], 27% [46]). High rates of non-dipping have also been shown in liver transplant patients, treated with cyclosporine or tacrolimus [47], and in cyclosporine treated heart transplant recipients [48]. In our study, a non-dipping pattern was associated with tacrolimus use, even after adjustment for diabetes, time since transplantation and age. Low hemoglobin levels were linked to lower SBP and DBP dipping ratio; this finding is consistent with data from the general population [49]. As compared with BP dipping, heart rate dipping during sleep, an independent predictor of mortality [50], was relatively preserved in our patients.

In the absence of significant relationships with age, diabetes and BMI, the elevated magnitude of BP masking and non-dipping observed in our patients is possibly a consequence of disturbed volume status, as has been shown for tacrolimus treatment [51], and derangements in autonomic nervous system activity. Also, in as much as anxiety and stress may be involved in the white-coat BP response [52], it is tempting to speculate that for a kidney transplant patient in a routine visit, the kidney transplant clinic constitutes a comforting environment, with familiar fellow patients and more importantly staff, with whom not rarely a patient may have more than a decade long familiarity (being that kidney transplant physicians and nurses also care for dialysis and pre-dialysis patients). Thus, hypertension may be masked in the kidney transplant clinic much more often than a white coat response is elicited.

Recruitment of participants was performed predominantly by their transplant nephrologists, possibly leading to referral biases toward patients who their physicians thought them to have indications for ABPM. Also, clinic BP measurements were not done in a uniform manner: 5 of 7 physicians used aneroid sphygmomanometry, while two relied on oscillometric measurements taken by the clinic nurse. Only one measurement was typically recorded in each visit. Our study has no dedicated control group, although we did use parallel information from our institution’s general ABPM dataset for perspective. Our study’s relatively small size (76 patients who take different antihypertensive medications) is a limitation too, and caution is advised in interpreting some results also due to their post-hoc nature.

On the other hand, our study’s strength is in examining the clinic-awake BP difference as a continuous variable, and therefore independent of hypertension definitions. To our knowledge, it is the first study that describes diurnal-based BP differences between immunosuppressant regimens, thus generating hypothesis for further studies.