Sleep Breath DOI 10.1007/s11325-014-0991-z
REVIEW
Effect of renal sympathetic denervation on apnea-hypopnea index in patients with obstructive sleep apnea: a systematic review and meta-analysis Ghanshyam Palamaner Subash Shantha & Samir Bipin Pancholy
Received: 7 February 2014 / Revised: 17 March 2014 / Accepted: 23 April 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Purpose Recent evidence associates sympathetic tone with severity of obstructive sleep apnea (OSA). Renal sympathetic denervation (RDN), by decreasing sympathetic tone, has the potential to decrease OSA severity. Small observational studies that assessed this hypothesis lacked precision. Hence, in this meta-analysis, we have attempted to pool available data from studies that have assessed the effect of RDN on OSA severity in patients with OSA. Methods Medline, Embase, Cochrane central, Ovid, Cinahl, web of science, and conference abstracts were searched for eligible citations by two independent reviewers using key words “renal denervation,” “hypertension,” and “obstructive sleep apnea.” From a total of 2,863 identified citations, using metaanalysis of observational studies in epidemiology method, five studies were assessed eligible and included in the meta-analysis. Results All five studies followed an observational study design, involved patients with OSA and HTN, and reported an apnea-hypopnea index (AHI) 6 months post-RDN. Four were “before and after” studies and one compared continuous positive airway pressure with RDN. In the pooled analysis, involving 49 patients, RDN was associated with a significant reduction in mean AHI [weighted mean difference −9.61 (95 % CI −15.43 to −3.79, P=0.001)] 6 months post-RDN. One study also reported improvement in oxygen desaturation index and Epworth sleepiness scale score 6 months post-RDN. Conclusions RDN is associated with significant improvement in OSA severity. However, our results need validation in G. P. S. Shantha : S. B. Pancholy (*) The Wright Center for Graduate Medical Education, 501 Madison Avenue, Scranton, PA 18010, USA e-mail: [email protected] S. B. Pancholy The Commonwealth Medical College, 501 Madison Avenue, Scranton, PA 18510, USA
RCTs that assess effect of RDN in patients with OSA, which can potentially broaden the clinical applicability of RDN. Keywords Renal denervation . Obstructive sleep apnea . Apnea-hypopnea index . Sympathetic tone
Introduction Obstructive sleep apnea (OSA) is the most common form of sleep-disordered breathing affecting 3–7 % of the general population [1, 2]. Given the obesity epidemic with nearly 35 % of adult Americans being obese [3], the public health burden of OSA will increase in the near future [4]. OSA significantly increases the risk for cardiovascular events, especially stroke [5]. Association between OSA and hypertension (HTN) is clearly established. Patients with OSA have 1.42 times higher odds of incident HTN compared to those without OSA [6], and even small increases in apnea-hypopnea index (AHI) (a common measure of OSA severity) significantly increase HTN risk [6]. Patients with AHI between 0– 4.9 events/h, 5–14.9 events/h, and >15 events/h had 1.42, 2.03, and 2.89 times higher odds of HTN compared to those with 0 events/h [6]. Also, OSA is reported to be prevalent in >80 % of patients with drug-resistant HTN [7, 8]. Since it is established that effective treatment of OSA results in significant improvements in blood pressure both short and long term, expert panels have recommended OSA management as an important element in the management of resistant HTN [9–12]. Hence, it is possible that OSA and HTN potentially share a common pathogenic mechanism that may be a cause or consequence of both disease states. Excess sympathetic tone, evident in patients with OSA, is a possible mechanistic pathway that links OSA to HTN and the development of resistance to pharmacotherapy of HTN.
Sleep Breath
However, it is unclear if excess sympathetic tone is a cause of or consequence to OSA, as there is evidence in support of both mechanisms [11, 12]. Excess sympathetic tone, by increasing pharyngeal wall thickness and peri-pharyngeal fluid accumulation, may lead to OSA [11], whereas OSA, due to recurrent episodes of hypoxia, sleep fragmentation, and airway obstruction, may lead to increased sympathetic tone [11, 12]. Renal sympathetic denervation (RDN), by administering discrete, low-dose radiofrequency energy to the renal artery endothelial surface via a percutaneous catheter-based procedure, is shown to reduce renal sympathetic outflow by severing the sympathetic nerve supply to the kidney, which in turn leads to decreased systemic sympathetic tone [13–15]. Hence, irrespective of excess sympathetic tone being a cause of or consequence to OSA, RDN mechanistically has the potential to decrease this excess sympathetic tone associated with OSA. Small observational studies did assess this hypothesis [16, 17]. These studies, though showing a trend toward significant reduction in AHI following RDN, are limited by small sample size and hence have lacked precision [16, 17]. The objective of this pooled analysis is to evaluate, with better precision, the association between RDN and change in OSA severity in patients with OSA. If the association between RDN and improvement in OSA severity is established, not only will RDN be identified as a potential treatment modality for OSA but will also give mechanistic insights to the link between OSA and autonomic nervous system.
Methods Data sources We followed the guidelines reported by the meta-analysis of observational studies in epidemiology (MOOSE) for planning, conducting, data abstraction, and reporting of this systematic review and meta-analysis [18]. We searched Medline, Embase, CINAHL, OVID, Cochrane library database, and Google scholar for studies assessing the association between RDN and OSA in patients with OSA. We did not use any language or time restriction for the search. Fifth of January 2014 was the date of last search. Though we planned to include randomized controlled trials (RCTs) and observational studies in the review, we could not identify any RCTs that reported data relevant to our research question. The search terms included sleep-disordered breathing, obstructive sleep apnea, sleep apnea, renal denervation, renal sympathetic denervation, resistant hypertension, drug-resistant hypertension, treatment-resistant hypertension, apnea-hypopnea index, prospective study, cohort study, longitudinal study, and follow-up study. Also, we reviewed the reference lists of retrieved articles for additional studies. We further searched conference abstracts, hand searched, and gray literature searched, and contacted authors if needed.
Study selection We initially reviewed the title and abstracts of retrieved citations. Then, full texts of those citations considered relevant at this stage were assessed for eligibility for inclusion into the review. Inclusion criteria for studies were as follows: (1) studies that followed a prospective study design (before and after intervention study design or studies comparing another intervention or no intervention to RDN); (2) studies where patients with OSA formed part or the complete study cohort; (3) studies involving adult patients (age >18 years); (4) exposure assessed was RDN; (5) the outcome reported was mean change (with standard deviation) in AHI with follow-up times; if instead of mean change in AHI, median, or another summary method was reported, individual patient data should be available to calculate mean and standard deviation of AHI change, or if authors responded to our query on mean change in AHI though the manuscript reported another summary measure; and (6) study used standard polysomnography to assess OSA severity. We included conference abstracts that reported data relevant to our research question. We excluded case reports. Two reviewers independently assessed studies for eligibility. Discrepancy was resolved by consensus. Data extraction Two reviewers independently extracted data from the included full text citations and entered into an electronic datasheet using standardized protocol. Discrepancy was resolved by consensus. The following information was abstracted: the last name of the first author; publication year; country where the study was performed; study design (observational or randomized trials); age and gender distribution of the study participants with OSA; mean baseline BMI specific to patients with OSA; exposure (RDN); baseline AHI in patients with OSA pre-RDN; followup times in months; outcomes, namely, AHI change in patients with OSA during these follow-up times; other measures denoting OSA severity, namely, oxygen saturation index; and Epworth sleepiness scale score if reported in the included studies. Further, if studies reported number of responders (patients who demonstrated blood pressure improvement following RDN), we abstracted that data. We agreed not to assess quality of included studies because some of the included studies were conference abstracts, containing data specific to our research question, although not containing sufficient information to make a definitive assessment of study quality. Statistical analysis From the included studies, pre-RDN mean AHI and post-RDN AHI and the number of patients with OSA in each study were abstracted and entered into RevMan 5.1 software (Nordic Cochrane Center, Kobenhavn, Denmark) statistical software program [19]. Data were summarized as weighted mean
Sleep Breath
difference (WMD) of continuous variables with 95 % confidence intervals (CI), and combined using DerSimonian and Laird random effects model with inverse variance weighting. Heterogeneity across studies was assessed with the Cochran’s Q statistic (χ2) with a P20
Germany, Europe 10
Thakur et al. [23]
Observational Total 21 patients, 6 had NK (abstract) OSA
NK
Witkowski et al. [17]
Observational Total 10 patients, all 10 had OSA, 8 only OSA, and 2 mixed sleep apnea Observational Total 31 patients, 15 (abstract) OSA, 16 CPAP
Warsaw, Poland
8
China
NK
Zhao et al. [24]
Not known if treated for OSA AHI15, n=5
Untreated OSA and HTN
OSA obstructive sleep apnea, HTN hypertension, AHI apnea-hypopnea index, CPAP continuous positive airway pressure, NK not known
contacting authors of both studies [22, 23], as this detail was not mentioned in the abstracts. Period of follow-up was unclear in these two studies; however, we did obtain AHI measures pre-RDN and 6 months post-RDN. Schmiedel et al. [22] concluded that RDN did not improve blood pressure in patients with untreated OSA, while Thakur reported that RDN improved blood pressure in their total sample but did not mention blood pressure change specific to patients with OSA [23]. Witkowski et al. [17] primarily aimed at assessing the role of RDN on AHI, blood pressure, and glycemic control. They assessed AHI using standard polysomnography pre-RDN and 6 months post-RDN, and their patients were identified to have OSA if their AHI was ≥5. Eight of the 10 patients in their study had OSA and were untreated. Two had mixed sleep apnea. Five patients had AHI5. They had mentioned median AHI change 6 months post-RDN. We contacted the authors to get mean AHI measures pre-RDN and 6 months post-RDN. They concluded that though RDN improved OSA severity, their results did not reach statistical significance due to decreased precision. Further, they did show that RDN improved blood pressure and glycemic control (glycosylated hemoglobin) in their study patients [17]. The study by Zhao et al. [24], another conference abstract, reported on patients with OSA and HTN that did not qualify to be defined as drug-resistant HTN and compared RDN with CPAP for its efficiency in reducing OSA severity. Of the total 31 patients with OSA and HTN in this study, 16 were treated with CPAP, and 15 underwent RDN, and the treatment allocation details were not mentioned. OSA definition and AHI assessment method in this study are not known. The authors concluded that though both CPAP and RDN improved OSA severity, CPAP performed better (10 decrease in AHI at 6 months of CPAP use) with AHI improvement than RDN (5 decrease in AHI 6 months post-RDN; P=
0.029). The common features of the included studies that are of relevance to our systematic review is that RDN was the common intervention, all included patients had OSA and HTN, and AHI was the OSA severity measure in all the included studies, and AHI change at 6 months post-RDN was obtainable for all five included studies. Blood pressure change with RDN Of the five included studies, three reported blood pressure change at 6 months postRDN [17, 22, 24]. There was a significant decrease in mean office systolic blood pressure (−14.72 mmHg, 95 % CI −25.06 to −4.38 mmHg) and a non-significant decrease in mean office diastolic blood pressure (−6.96 mmHg, 95 % CI −15.21 to 1.29 mmHg) 6 months post-RDN. Change in OSA severity with RDN In the pooled analysis of five studies involving 49 patients, weighted mean difference in AHI 6 months post-RDN was −9.61 (95 % CI −15.43 to −3.79, P=0.001; Fig. 1). I2 for AHI change was 33 % (Fig. 2). When the analysis was restricted to studies with patients with OSA and resistant HTN (n=42) [16, 17, 23, 24], the weighted mean difference in AHI 6 months post-RDN was −12.23 (95 % CI −20.62 to −3.85, P=0.004). Also, Witkowski et al. [17] reported non-significant (P=0.1) improvement in oxygen saturation index (median 13.0 events/h, interquartile range 7.1 to 43.8 events per hour versus median 8.7 events per hour, interquartile range 1.9 to 26.2 events/h) and significant (P
REVIEW
Effect of renal sympathetic denervation on apnea-hypopnea index in patients with obstructive sleep apnea: a systematic review and meta-analysis Ghanshyam Palamaner Subash Shantha & Samir Bipin Pancholy
Received: 7 February 2014 / Revised: 17 March 2014 / Accepted: 23 April 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Purpose Recent evidence associates sympathetic tone with severity of obstructive sleep apnea (OSA). Renal sympathetic denervation (RDN), by decreasing sympathetic tone, has the potential to decrease OSA severity. Small observational studies that assessed this hypothesis lacked precision. Hence, in this meta-analysis, we have attempted to pool available data from studies that have assessed the effect of RDN on OSA severity in patients with OSA. Methods Medline, Embase, Cochrane central, Ovid, Cinahl, web of science, and conference abstracts were searched for eligible citations by two independent reviewers using key words “renal denervation,” “hypertension,” and “obstructive sleep apnea.” From a total of 2,863 identified citations, using metaanalysis of observational studies in epidemiology method, five studies were assessed eligible and included in the meta-analysis. Results All five studies followed an observational study design, involved patients with OSA and HTN, and reported an apnea-hypopnea index (AHI) 6 months post-RDN. Four were “before and after” studies and one compared continuous positive airway pressure with RDN. In the pooled analysis, involving 49 patients, RDN was associated with a significant reduction in mean AHI [weighted mean difference −9.61 (95 % CI −15.43 to −3.79, P=0.001)] 6 months post-RDN. One study also reported improvement in oxygen desaturation index and Epworth sleepiness scale score 6 months post-RDN. Conclusions RDN is associated with significant improvement in OSA severity. However, our results need validation in G. P. S. Shantha : S. B. Pancholy (*) The Wright Center for Graduate Medical Education, 501 Madison Avenue, Scranton, PA 18010, USA e-mail: [email protected] S. B. Pancholy The Commonwealth Medical College, 501 Madison Avenue, Scranton, PA 18510, USA
RCTs that assess effect of RDN in patients with OSA, which can potentially broaden the clinical applicability of RDN. Keywords Renal denervation . Obstructive sleep apnea . Apnea-hypopnea index . Sympathetic tone
Introduction Obstructive sleep apnea (OSA) is the most common form of sleep-disordered breathing affecting 3–7 % of the general population [1, 2]. Given the obesity epidemic with nearly 35 % of adult Americans being obese [3], the public health burden of OSA will increase in the near future [4]. OSA significantly increases the risk for cardiovascular events, especially stroke [5]. Association between OSA and hypertension (HTN) is clearly established. Patients with OSA have 1.42 times higher odds of incident HTN compared to those without OSA [6], and even small increases in apnea-hypopnea index (AHI) (a common measure of OSA severity) significantly increase HTN risk [6]. Patients with AHI between 0– 4.9 events/h, 5–14.9 events/h, and >15 events/h had 1.42, 2.03, and 2.89 times higher odds of HTN compared to those with 0 events/h [6]. Also, OSA is reported to be prevalent in >80 % of patients with drug-resistant HTN [7, 8]. Since it is established that effective treatment of OSA results in significant improvements in blood pressure both short and long term, expert panels have recommended OSA management as an important element in the management of resistant HTN [9–12]. Hence, it is possible that OSA and HTN potentially share a common pathogenic mechanism that may be a cause or consequence of both disease states. Excess sympathetic tone, evident in patients with OSA, is a possible mechanistic pathway that links OSA to HTN and the development of resistance to pharmacotherapy of HTN.
Sleep Breath
However, it is unclear if excess sympathetic tone is a cause of or consequence to OSA, as there is evidence in support of both mechanisms [11, 12]. Excess sympathetic tone, by increasing pharyngeal wall thickness and peri-pharyngeal fluid accumulation, may lead to OSA [11], whereas OSA, due to recurrent episodes of hypoxia, sleep fragmentation, and airway obstruction, may lead to increased sympathetic tone [11, 12]. Renal sympathetic denervation (RDN), by administering discrete, low-dose radiofrequency energy to the renal artery endothelial surface via a percutaneous catheter-based procedure, is shown to reduce renal sympathetic outflow by severing the sympathetic nerve supply to the kidney, which in turn leads to decreased systemic sympathetic tone [13–15]. Hence, irrespective of excess sympathetic tone being a cause of or consequence to OSA, RDN mechanistically has the potential to decrease this excess sympathetic tone associated with OSA. Small observational studies did assess this hypothesis [16, 17]. These studies, though showing a trend toward significant reduction in AHI following RDN, are limited by small sample size and hence have lacked precision [16, 17]. The objective of this pooled analysis is to evaluate, with better precision, the association between RDN and change in OSA severity in patients with OSA. If the association between RDN and improvement in OSA severity is established, not only will RDN be identified as a potential treatment modality for OSA but will also give mechanistic insights to the link between OSA and autonomic nervous system.
Methods Data sources We followed the guidelines reported by the meta-analysis of observational studies in epidemiology (MOOSE) for planning, conducting, data abstraction, and reporting of this systematic review and meta-analysis [18]. We searched Medline, Embase, CINAHL, OVID, Cochrane library database, and Google scholar for studies assessing the association between RDN and OSA in patients with OSA. We did not use any language or time restriction for the search. Fifth of January 2014 was the date of last search. Though we planned to include randomized controlled trials (RCTs) and observational studies in the review, we could not identify any RCTs that reported data relevant to our research question. The search terms included sleep-disordered breathing, obstructive sleep apnea, sleep apnea, renal denervation, renal sympathetic denervation, resistant hypertension, drug-resistant hypertension, treatment-resistant hypertension, apnea-hypopnea index, prospective study, cohort study, longitudinal study, and follow-up study. Also, we reviewed the reference lists of retrieved articles for additional studies. We further searched conference abstracts, hand searched, and gray literature searched, and contacted authors if needed.
Study selection We initially reviewed the title and abstracts of retrieved citations. Then, full texts of those citations considered relevant at this stage were assessed for eligibility for inclusion into the review. Inclusion criteria for studies were as follows: (1) studies that followed a prospective study design (before and after intervention study design or studies comparing another intervention or no intervention to RDN); (2) studies where patients with OSA formed part or the complete study cohort; (3) studies involving adult patients (age >18 years); (4) exposure assessed was RDN; (5) the outcome reported was mean change (with standard deviation) in AHI with follow-up times; if instead of mean change in AHI, median, or another summary method was reported, individual patient data should be available to calculate mean and standard deviation of AHI change, or if authors responded to our query on mean change in AHI though the manuscript reported another summary measure; and (6) study used standard polysomnography to assess OSA severity. We included conference abstracts that reported data relevant to our research question. We excluded case reports. Two reviewers independently assessed studies for eligibility. Discrepancy was resolved by consensus. Data extraction Two reviewers independently extracted data from the included full text citations and entered into an electronic datasheet using standardized protocol. Discrepancy was resolved by consensus. The following information was abstracted: the last name of the first author; publication year; country where the study was performed; study design (observational or randomized trials); age and gender distribution of the study participants with OSA; mean baseline BMI specific to patients with OSA; exposure (RDN); baseline AHI in patients with OSA pre-RDN; followup times in months; outcomes, namely, AHI change in patients with OSA during these follow-up times; other measures denoting OSA severity, namely, oxygen saturation index; and Epworth sleepiness scale score if reported in the included studies. Further, if studies reported number of responders (patients who demonstrated blood pressure improvement following RDN), we abstracted that data. We agreed not to assess quality of included studies because some of the included studies were conference abstracts, containing data specific to our research question, although not containing sufficient information to make a definitive assessment of study quality. Statistical analysis From the included studies, pre-RDN mean AHI and post-RDN AHI and the number of patients with OSA in each study were abstracted and entered into RevMan 5.1 software (Nordic Cochrane Center, Kobenhavn, Denmark) statistical software program [19]. Data were summarized as weighted mean
Sleep Breath
difference (WMD) of continuous variables with 95 % confidence intervals (CI), and combined using DerSimonian and Laird random effects model with inverse variance weighting. Heterogeneity across studies was assessed with the Cochran’s Q statistic (χ2) with a P20
Germany, Europe 10
Thakur et al. [23]
Observational Total 21 patients, 6 had NK (abstract) OSA
NK
Witkowski et al. [17]
Observational Total 10 patients, all 10 had OSA, 8 only OSA, and 2 mixed sleep apnea Observational Total 31 patients, 15 (abstract) OSA, 16 CPAP
Warsaw, Poland
8
China
NK
Zhao et al. [24]
Not known if treated for OSA AHI15, n=5
Untreated OSA and HTN
OSA obstructive sleep apnea, HTN hypertension, AHI apnea-hypopnea index, CPAP continuous positive airway pressure, NK not known
contacting authors of both studies [22, 23], as this detail was not mentioned in the abstracts. Period of follow-up was unclear in these two studies; however, we did obtain AHI measures pre-RDN and 6 months post-RDN. Schmiedel et al. [22] concluded that RDN did not improve blood pressure in patients with untreated OSA, while Thakur reported that RDN improved blood pressure in their total sample but did not mention blood pressure change specific to patients with OSA [23]. Witkowski et al. [17] primarily aimed at assessing the role of RDN on AHI, blood pressure, and glycemic control. They assessed AHI using standard polysomnography pre-RDN and 6 months post-RDN, and their patients were identified to have OSA if their AHI was ≥5. Eight of the 10 patients in their study had OSA and were untreated. Two had mixed sleep apnea. Five patients had AHI5. They had mentioned median AHI change 6 months post-RDN. We contacted the authors to get mean AHI measures pre-RDN and 6 months post-RDN. They concluded that though RDN improved OSA severity, their results did not reach statistical significance due to decreased precision. Further, they did show that RDN improved blood pressure and glycemic control (glycosylated hemoglobin) in their study patients [17]. The study by Zhao et al. [24], another conference abstract, reported on patients with OSA and HTN that did not qualify to be defined as drug-resistant HTN and compared RDN with CPAP for its efficiency in reducing OSA severity. Of the total 31 patients with OSA and HTN in this study, 16 were treated with CPAP, and 15 underwent RDN, and the treatment allocation details were not mentioned. OSA definition and AHI assessment method in this study are not known. The authors concluded that though both CPAP and RDN improved OSA severity, CPAP performed better (10 decrease in AHI at 6 months of CPAP use) with AHI improvement than RDN (5 decrease in AHI 6 months post-RDN; P=
0.029). The common features of the included studies that are of relevance to our systematic review is that RDN was the common intervention, all included patients had OSA and HTN, and AHI was the OSA severity measure in all the included studies, and AHI change at 6 months post-RDN was obtainable for all five included studies. Blood pressure change with RDN Of the five included studies, three reported blood pressure change at 6 months postRDN [17, 22, 24]. There was a significant decrease in mean office systolic blood pressure (−14.72 mmHg, 95 % CI −25.06 to −4.38 mmHg) and a non-significant decrease in mean office diastolic blood pressure (−6.96 mmHg, 95 % CI −15.21 to 1.29 mmHg) 6 months post-RDN. Change in OSA severity with RDN In the pooled analysis of five studies involving 49 patients, weighted mean difference in AHI 6 months post-RDN was −9.61 (95 % CI −15.43 to −3.79, P=0.001; Fig. 1). I2 for AHI change was 33 % (Fig. 2). When the analysis was restricted to studies with patients with OSA and resistant HTN (n=42) [16, 17, 23, 24], the weighted mean difference in AHI 6 months post-RDN was −12.23 (95 % CI −20.62 to −3.85, P=0.004). Also, Witkowski et al. [17] reported non-significant (P=0.1) improvement in oxygen saturation index (median 13.0 events/h, interquartile range 7.1 to 43.8 events per hour versus median 8.7 events per hour, interquartile range 1.9 to 26.2 events/h) and significant (P
