Insights into the Treatment of SBD With Imaging Richard J. Schwab, - - PDF document

1

Insights into the Treatment of SBD With Imaging

Richard J. Schwab, M.D. Professor of Medicine Interim Chief, Division of Sleep Medicine Medical Director, Penn Sleep Centers University of Pennsylvania Perelman School of Medicine

Insights into the Treatment of SBD With Imaging - Disclosures

• NIH grants - PPG (phenotyping and OSA)
• ResMed Grant/Registry to study OSA/CSA

and CPAP in hospitalized patients

• Jazz clinical trial for daytime sleepiness in

OSA

• Inspire CT study to examine upper airway

anatomy with hypoglossal nerve stimulation

2

Insights into the Treatment of SBD with Upper Airway Imaging

• Treatment of sleep apnea

– Weight loss – CPAP – Oral appliances

• Mandibular repositioning devices
• Upper airway surgery

– Transoral robotic surgery (TORS) – Hypoglossal nerve stimulation

Weight Loss and Sleep Apnea

• How much weight loss results in clinical

improvement?

– Weight loss of 5 - 10% may be effective

• Does size of parapharyngeal fat pads and

tongue (tongue fat) decrease with weight loss?

• Does size of lateral pharyngeal walls, soft

palate decrease with weight loss?

– Weight loss is associated with reductions in both fat (75%) and fat-free mass (25%)

3

Demographic Comparison between OSA and Control Patients after 6 Months of Weight Loss

Variable All Participants Weight Stable* Weight Loss* Control OSA p† Control OSA p† Control OSA p† N 18 51 – 3 8 – 14 38 – Age (years) 40.4 ± 13.6 49.6 ± 10.6 0.0151 30.3 ± 10.1 51.5 ± 10.1 0.0244 42.3 ± 14.1 48.6 ± 11.2 0.1467 Male 16.67% 41.18% 0.0605 66.67% 50.00% >0.999 7.14% 39.47% 0.0251 White 47.06% 46.00% 0.9397 100.00% 37.50% 0.1818 38.46% 48.65% 0.5261 Height (inches) 65.9 ± 2.8 67.0 ± 4.2 0.2436 67.0 ± 5.3 68.0 ± 4.8 0.6082 66.0 ± 2.2 66.9 ± 4.1 0.2877 Weight (pounds) Baseline 226 ± 50.4 271 ± 54.0 0.0033 232 ± 28.6 303 ± 64.3 0.0412 228 ± 55.0 266 ± 52.5 0.0340 Follow-up 206 ± 36.5 241 ± 50.8 0.0029 230 ± 28.9 301 ± 62.2 0.0412 202 ± 37.8 225 ± 38.1 0.0690 Change

• 20.41 ± 26.85 -29.94 ± 32.76 0.2307 -2.60 ± 3.65 -1.42 ± 4.72 0.5403 -26.05 ± 27.98 -41.82 ± 29.08 0.0871

Percent Change

• 7.98 ± 8.33 -10.50 ± 10.59 0.3115 -1.14 ± 1.64 -0.40 ± 1.37 0.4142 -10.22 ± 8.06 -14.78 ± 8.52 0.0874

*Weight Stable defined as <2.5% weight change, and weight loss as >2.5% decrease in weight; †p-value from T-test (parametric) or Kruskal-Wallis (non-parametric) test for continuous measures and chi-square or Fisher’s exact test

for categorical measures.

26 pound weight loss (10%) in controls and a 42 pound weight loss (15%) in apneics

Demographic Comparison between OSA and Control Patients after 6 Months of Weight Loss

Variable All Participants Weight Stable* Weight Loss* Control OSA p† Control OSA p† Control OSA p† BMI (kg/m2) Baseline 36.6 ± 7.8 42.4 ± 7.5 0.0095 37.3 ± 10.6 46.1 ± 8.2 0.2207 36.7 ± 7.8 41.8 ± 7.5 0.0447 Follow-up 33.3 ± 5.9 37.8 ± 7.3 0.0128 36.8 ± 10.3 45.9 ± 8.0 0.1530 32.5 ± 5.0 35.3 ± 5.4 0.0925 Change

• 3.26 ± 4.20
• 4.64 ± 5.13

0.2675 -0.45 ± 0.58 -0.21 ± 0.64 0.5403 -4.17 ± 4.35

• 6.53 ± 4.49

0.0989 Percent Change

• 7.98 ± 8.33 -10.50 ± 10.59 0.3115 -1.14 ± 1.64 -0.40 ± 1.37 0.4142 -10.22 ± 8.06 -14.78 ± 8.52 0.0874

AHI (events/hour) Baseline 4.7 ± 3.2 39.5 ± 27.7 <0.0001 7.1 ± 1.8 43.9 ± 34.3 0.0143 3.9 ± 3.1 39.5 ± 27.4 <0.0001 Follow-up 6.4 ± 7.1 21.8 ± 20.8 <0.0001 12.1 ± 6.7 37.4 ± 20.8 0.0244 5.2 ± 7.1 16.3 ± 19.1 0.0039 Change 1.66 ± 5.83 -18.20 ± 23.41 <0.0001 5.00 ± 4.96 -6.47 ± 28.22 0.3074 1.33 ± 5.97 -23.69 ± 20.28 <0.0001 Percent Change 99.38 ± 221.2 -40.39 ± 60.43 0.0165 61.32 ± 54.61 8.41 ± 55.41 0.2207 117.6 ± 247.6 -63.44 ± 29.31 0.0171

*Weight Stable defined as <2.5% weight change, and weight loss as >2.5% decrease in weight; †p-value from T-test (parametric) or Kruskal-Wallis (non-parametric) test for continuous measures and chi-square or Fisher’s exact test

for categorical measures. No significant changes in AHI in controls or apneics in the weight stable groups. In the

weight loss groups no significant change in AHI in controls but in apniecs AHI reduced significantly by 24 events/hour

4

Percent Changes in Airway Size within and between Controls and Apneics after 6 Months of Weight Loss

Airway Measures All Participants Controls OSA Cases Controls

• vs. OSA

N mean ± SD p† N mean ± SD p† N mean ± SD p† p‡

RP Airway Volume (mm3)

51 15.35 ± 45.21 0.0190 14 17.38 ± 46.07 0.1815 37 14.58 ± 45.49 0.0590 0.8459

RP Cross Sectional Area (mm2)

51 21.64 ± 45.12 0.0012 14 22.42 ± 46.39 0.0937 37 21.35 ± 45.28 0.0069 0.9404

RP Minimum Area (mm)

47 49.12 ± 110.3 0.0038 12 40.12 ± 76.7 0.0972 35 52.20 ± 120.6 0.0150 0.7473

RP Minimum AP Distance (mm) 46

6.34 ± 37.72 0.2605 12 9.69 ± 31.22 0.3053 34 5.15 ± 40.12 0.4591 0.7245

RP Minimum Lateral Distance

46 13.93 ± 33.26 0.0067 12 16.33 ± 33.14 0.1159 34 13.09 ± 33.76 0.0305 0.7754

RG Airway Volume (mm3)

51 6.65 ± 33.65 0.1643 14 14.73 ± 31.96 0.1082 37 3.59 ± 34.19 0.5268 0.2961

RG Cross Sectional Area (mm2)

51 4.91 ± 26.05 0.1846 14 11.36 ± 22.12 0.0769 37 2.47 ± 27.27 0.5856 0.2812

RG Minimum Area (mm)

46 6.79 ± 41.07 0.2680 12 24.07 ± 59.05 0.1856 34 0.69 ± 31.48 0.8984 0.2133

RG Minimum AP Distance (mm) 45 -1.39 ± 28.15 0.7421 12 -3.17 ± 26.64 0.6883 33 -0.74 ± 29.05

0.8842 0.8016

RG Minimum Lateral Distance

45 7.82 ± 40.29 0.1994 12 21.05 ± 51.02 0.1808 33 3.01 ± 35.32 0.6272 0.1873

†p-value from paired T-test examining significance of within group change; ‡p-value from T-test comparing changes between OSA and controls

Weight loss resulted in significant increases in the percent changes in Retropalatal cross sectional area and minimum area in apneics but not controls. No change in RG region

Percent Changes in Soft Tissues within and between Controls and Apneics after 6 Months of Weight Loss

Soft Tissue Measures All Participants Controls OSA Cases Controls

• vs. OSA

N mean ± SD p† N mean ± SD p† N mean ± SD p† p‡ Total Soft Tissue (mm3) 51

• 6.90 ± 7.39

<0.0001 14

• 3.48 ± 5.05

0.0230 37

• 8.20 ± 7.77

<0.0001 0.0404 Soft Palate (mm3) 51 -1.62 ± 16.67 0.4898 14 -1.85 ± 10.86 0.5346 37 -1.54 ± 18.53 0.6168 0.9409 Genioglossus (mm3) 51

• 4.65 ± 9.22

0.0007 14

• 1.96 ± 8.57

0.4071 37

• 5.66 ± 9.37

0.0008 0.2044 Tongue Fat (mm3) 51 -19.71 ± 16.35 <0.0001 14 -13.17 ± 18.73 0.0207 37 -22.19 ± 14.89 <0.0001 0.0788 Tongue Volume (mm3) 51

• 4.83 ± 8.68

0.0002 14

• 1.43 ± 6.29

0.4102 37

• 6.11 ± 9.17

0.0003 0.0855 Fat Pads (mm3) 51 -29.17 ± 14.85 <0.0001 14 -27.48 ± 16.45 <0.0001 37 -29.81 ± 14.39 <0.0001 0.6226 Pterygoid (mm3) 51 -11.77 ± 13.30 <0.0001 14 -10.15 ± 12.06 0.0077 37 -12.38 ± 13.84 <0.0001 0.5973 RP Lateral Walls (mm3) 51 -11.61 ± 16.06 <0.0001 14 -6.83 ± 18.41 0.1882 37 -13.41 ± 14.95 <0.0001 0.1946 RG Lateral Walls (mm3) 51 -1.85 ± 18.76 0.4858 14 4.46 ± 19.19 0.4007 37 -4.23 ± 18.30 0.1683 0.1418 Total Lateral Walls (mm3) 51 -8.24 ± 11.86 <0.0001 14 -2.61 ± 14.30 0.5071 37 -10.37 ± 10.23 <0.0001 0.0357

†p-value from paired T-test examining significance of within group change; ‡p-value from T-test comparing changes between OSA and controls

Significant reductions in the volumes of the tongue, fat pads, pterygoids, lateral walls,

total soft tissue and tongue fat (but not soft palate) in apneics with weight loss

5 Apneic Pre Weight-Loss Apneic Post Weight-Loss Normal Pre Weight-Loss Normal Post Weight-Loss

Correlations between Percent Change in AHI and Percent Change in Soft Tissues among OSA Cases

Airway Measure Unadjusted Adjusted Model† N rho p N rho p Weight (pounds) 50 0.68 <.0001 49 0.67 <.0001 Total Soft Tissue (mm3) 49 0.25 0.0799 48 0.21 0.1565 Soft Palate (mm3) 49 -0.10 0.4942 48

• 0.12

0.4517 Genioglossus (mm3) 49 0.17 0.2373 48 0.13 0.4022 Tongue Fat (mm3) 49 0.63 <.0001 48 0.62 <.0001 Total Tongue Volume (mm3) 49 0.18 0.2257 48 0.17 0.2523 Fat Pads (mm3) 49 0.28 0.0531 48 0.20 0.1973 RP Lateral Walls (mm3) 49 0.41 0.0038 48 0.39 0.0083 RG Lateral Walls (mm3) 49 0.06 0.6864 48

• 0.01

0.9633 Total Lateral Walls (mm3) 49 0.35 0.0147 48 0.29 0.0501

†Partial correlation adjusted for age, gender, and race; ‡Partial correlation adjusted for age, gender, race, baseline weight and percent

change in weight.

Significant correlations between reductions in tongue fat and the RP lateral walls and reductions in AHI in apneics with weight loss

6

0 cm H20 5 cm H20 10 cm H20 15 cm H20

CPAP - Airway 3D Volumes

RP RG RP RG

Schwab et al, AJRCCM 154:1106-1116, 1996

CPAP - 0 cm H20 CPAP - 15 cm H20

Schwab et al, AJRCCM 154:1106-1116, 1996

7

0 cm H20 5 cm H20 10 cm H20 15 cm H20

CPAP Settings Schwab et al, AJRCCM 154:1106-1116, 1996

0 cm H20 15 cm H20

Mid-sagittal MRI with and without CPAP in a Normal Subject

(Schwab et al, AJRCCM 154:1106-1116, 1996)

8

How do Oral Appliances Change Upper Airway Geometry?

• Do oral appliances simply pull mandible and

tongue forward?

• How important is vertical bite opening?
• Studies indicate airway caliber increases in lateral

dimension with oral appliances

– Gao et al Am J Orthod Dentofacial Orthop 125, 191-199, 2004 – Isono et al, J Appl Physiol, 79:2132-8, 1995

• Structures lateral to airway may be important in

understanding how oral devices maintain upper airway patency

Custom-Made Two-Piece Mandibular Advancement Splint (MAS) (a Modification of the Somnodent MAS) Chan et al. Thorax 2010;65:726-732

9

Representative Axial Images From a Responder and Non-Responder

Chan et al. Thorax 2010;65:726-732. MAS = mandibular advancement split.

10 Volumetric Reconstructions of the Upper Airway in a Responder Showing the Increase in Caliber of the Upper Airway With Mandibular Advancement Splint (MAS) Chan et al. Thorax 2010;65:726-732 Chan et al. Thorax 2010;65:726-732

11 Movement of Centroids of Soft Tissue Structures With Mandibular Advancement Spint Chan et al. Thorax 2010;65:726-732

Lingual Tonsils Can Narrow the Retroglossal Airway

Apneic with Lingual Tonsils Apneic

12 Radiologic and clinical preoperative view of severe tongue base

• hypertrophy. Pre and postoperative result using TORS

Vicini C, et al. Head Neck 2012; 34:15-22. TORS = Transoral Robotic Surgery

S/P TORS Pre-OP

Upper Airway MRI before and after TORS

• To quantitatively measure volumetric changes in upper

airway soft tissue structures using MRI pre- and post transoral robotic surgery for OSA

• Apneics undergoing TORS, which included bilateral

posterior hemiglossectomy with limited pharyngectomy and uvulopalatopharyngoplasty, had upper airway MRIs pre- and postoperatively

• Changes (percent and absolute values) in upper airway

and surrounding soft tissue volumes were calculated

• Correlations between upper airway soft tissue volumetric

changes and AHI changes were assessed

Chiffer et al. Volumetric MRI Analysis Pre- and Post Transoral Robotic Surgery for Obstructive Sleep Apnea (OSA-TORS). Laryngoscope 2015

13

Upper Airway MRI before and after TORS

Magnetic resonance imaging pre- and postoperatively showing increase in airway volume at the retropalatal airway level

Chiffer et al. Volumetric MRI Analysis Pre- and Post Transoral Robotic Surgery for Obstructive Sleep Apnea (OSA-TORS). Laryngoscope 2015

Upper Airway MRI before and after TORS

Magnetic resonance imaging showing an increase in airway volume at the retroglossal airway level

Chiffer et al. Volumetric MRI Analysis Pre- and Post Transoral Robotic Surgery for Obstructive Sleep Apnea (OSA-TORS). Laryngoscope 2015

14

Upper Airway MRI before and after TORS

Magnetic resonance imaging showing an increase in airway volume at the retro-epiglottal airway level

Chiffer et al. Volumetric MRI Analysis Pre- and Post Transoral Robotic Surgery for Obstructive Sleep Apnea (OSA-TORS). Laryngoscope 2015 Chiffer et al. Volumetric MRI Analysis Pre- and Post Transoral Robotic Surgery for Obstructive Sleep Apnea (OSA-TORS). Laryngoscope 2015

Percent Changes in MRI Structures with TORS

Total airway volume increased by 19.4% (P = 0.030). Soft palate and tongue volumes decreased by 18.3% (P = 0.002) and 5.8% (P = 0.013). Retropalatal and total lateral wall volumes decreased by 49.8% (P = 0.0001) and 17.9% (P = 0.008)

15

Chiffer et al. Volumetric MRI Analysis Pre- and Post Transoral Robotic Surgery for Obstructive Sleep Apnea (OSA-TORS). Laryngoscope 2015

Correlations Between Changes in MRI Structures and Changes in AHI

Change in AHI and both RP (rho = 0.53, P = 0.023) and total (rho = 0.56, P = 0.028) lateral walls with TORS. In both instances, greater decreases in lateral wall volume correlated with greater decreases in AHI

Conclusions

• Total airway volume increased postoperatively,

whereas the volumes of the soft palate, tongue, and total and retropalatal lateral pharyngeal walls decreased

• Changes in lateral wall volume also correlated

with changes in apnea/hypopnea index (AHI)

• Volumetric upper airway MRI is an important

tool to better understand postsurgical anatomical changes following all types of sleep surgery and the reasons for surgical success

Chiffer et al. Volumetric MRI Analysis Pre- and Post Transoral Robotic Surgery for Obstructive Sleep Apnea (OSA-TORS). Laryngoscope 2015

16

Strollo et al, N Engl J Med, 2014 Patient Remote Physician Programmer Implanted System

Implantable Hypoglossal Nerve Stimulator: Components

Anatomic Predictors of Response to Upper Airway Stimulation Therapy for Obstructive Sleep Apnea

• RJ Schwab1, SHC Wang1, J Verbraecken2, OM

Vanderveken2, P Van de Heyning2, WG Vos3, JW DeBacker3, BT Keenan1, Q Ni4, W DeBacker2 –1University of Pennsylvania, Philadelphia, PA;

2Antwerp University Hospital and University of

Antwerp, Antwerp, Belgium; 3FluidDA, Kontich, Belgium; 4Inspire Medical Systems, Maple Grove, MN

• Accepted to SLEEP

17

Methods (Schwab et al, Accepted to Sleep)

• Subjects with mild-to-moderate
• bstructive sleep apnea (AHI > 15

events/hour) were recruited from clinical trials of upper airway stimulation

• Subjects were grouped into one of two

categories:

− Responders (AHI reduction of 50% or more and residual AHI < 20 events/hour) − Non-responders (AHI improvement of less than 50% or residual AHI > 20 events/hour)

CT Scanning with HGNS (Schwab et al, Accepted to Sleep)

• During a post-operative visit, CT scans of the upper

airway were performed under two conditions:

• Without stimulation
• During a 5-second period of stimulation during

inspiration

• Subjects were instructed to inhale slowly during the

scans while supine in the CT scanner

• Stimulation was delivered at the voltage level that

was titrated and set for home use during an in-lab sleep study

• Among subjects, the amplitude ranged from 1.4 to

2.6 volts

18

Results (Schwab et al, Accepted to Sleep)

All Subjects Responders Non‐Responders P* Characteristic (n=13) (n=7) (n=6) Age (years) 53.00 ± 9.15 50.32 ± 8.24 56.13 ± 9.89 0.252 BMI (kg/m2) 27.77 ± 1.57 27.74 ± 1.93 27.81 ± 1.20 0.775 Baseline AHI (events/hour) 33.47 ± 10.42 36.71 ± 9.17 29.68 ± 11.31 0.087 AHI with Upper Airway Stimulation Therapy (events/hour) 21.36 ± 17.47 14.09 ± 4.86 29.85 ± 23.31 0.032 AHI Change (events/hour) ‐12.11 ± 15.60 ‐22.63 ± 6.54 0.17 ± 14.04 0.003 *P‐value from non‐parametric Wilcoxon rank sum tests comparing change from baseline between responders and non‐responders. Values presented as Mean ± SD

Table 1A – Demographic Data in Responders and Non-Responders

AHI was significantly reduced in responders with no change in the non-responders

Results (Schwab et al, Accepted to Sleep)

All Subjects Responders Non‐Responders P* Upper Airway Measure (n=13) (n=7) (n=6) RP Airway Volume (mm3) 8642.98 ± 4986.41 6838.85 ± 3256.40 10747.79 ± 6089.03 0.199 Average RP Airway Area Per Slice (mm2) 214.95 ± 93.51 178.79 ± 54.04 257.13 ± 116.26 0.199 RG Airway Volume (mm3) 12429.94 ± 5760.36 12711.03 ± 5643.46 12102.00 ± 6417.00 0.668 Average RG Airway Area Per Slice (mm2) 309.34 ± 123.67 306.72 ± 122.33 312.4 ± 136.86 0.775 Pharyngeal length (mm) 80.92 ± 7.63 80.35 ± 8.29 81.57 ± 7.50 0.568 Soft Palate Volume (mm3) 9991.63 ± 2348.54 8789.33 ± 1811.37 11394.33 ± 2217.07 0.032 Total Tongue Volume (mm3) 156558.7 ± 16222.3 157986.9 ± 11300.2 154892.5 ± 21729.5 0.568 Mandible‐Hyoid Distance (mm) 41.00 ± 5.01 42.30 ± 5.55 39.49 ± 4.27 0.317 Mandibular Plane‐Hyoid Distance (mm) 27.24 ± 3.53 26.66 ± 4.00 27.92 ± 3.09 0.775 Hyoid‐Spine Distance (mm) 40.17 ± 3.40 41.23 ± 4.11 38.93 ± 2.03 0.391 *P‐value from non‐parametric Wilcoxon rank sum tests comparing change from baseline between responders and non‐responders. Values presented as Mean ± SD

Table 1B - Baseline Anatomic Data Comparing Responders and Non- Responders

Soft palate volume larger in non-responders than responders. No other differences in the baseline size of the upper airway or soft tissue structures

19

CT Image Analysis: Non-responder

Figure 4A - Midsagittal reconstruction of a representative non-responder before and during stimulation. The stimulation caused tongue tip protrusion and inferior displacement but also resulted in tongue base retrusion causing narrowing of the retroglossal airway.

CT Image Analysis: Responder

Figure 4B - Midsagittal reconstruction of a representative responder before and during stimulation. The stimulation caused anterior and inferior displacement of the tongue and hyoid bone with a resultant increase in the retropalatal and retroglossal airway

20

Results (Schwab et al, Accepted to Sleep)

Baseline vs. Stimulation Difference P* All Subjects Responders Non‐Responders Measure (n=13) (n=7) (n=6) Pharyngeal Length (mm) 0.71 ± 5.34 ‐0.89 ± 5.84 2.57 ± 4.44 0.391 RP Airway Volume (mm3) 712.57 ± 3403.86 1992.07 ± 3483.33 ‐780.18 ± 2877.64 0.116 Average RP Area Per Slice (mm2) 36.31 ± 119.80 69.31 ± 129.27 ‐2.19 ± 105.30 0.317 Mid‐RP AP Distance (mm) 2.38 ± 6.10 2.64 ± 4.23 2.07 ± 8.23 0.391 Mid‐RP Lateral Distance (mm) ‐1.53 ± 3.66 ‐1.16 ± 2.74 ‐1.97 ± 4.77 0.617 Mid‐RP Cross‐Sectional Area (mm2) 57.15 ± 184.83 101.47 ± 180.56† 5.44 ± 192.12 0.199 Min‐RP AP Distance (mm) 1.86 ± 4.05 2.99 ± 3.98 0.54 ± 4.05 0.317 Min‐RP Lateral Distance (mm) ‐2.12 ± 7.20 ‐0.01 ± 4.35 ‐4.59 ± 9.40 0.253 Min‐RP Cross‐Sectional Area (mm2) 9.65 ± 75.92 38.60 ± 49.58† ‐24.12 ± 91.33 0.199 Max‐RP AP Distance (mm) 1.19 ± 5.21 0.72 ± 2.07 1.75 ± 7.71 0.886 Max‐RP Lateral Distance (mm) 0.40 ± 2.14 ‐0.30 ± 1.50 1.22 ± 2.61 0.153 Max‐RP Cross‐Sectional Area (mm2) 17.37 ± 159.90 43.46 ± 209.25 ‐13.07 ± 82.18 0.775 *P‐value from non‐parametric Wilcoxon rank sum tests comparing change from baseline between responders and non‐responders; †Within group change significantly different from zero (p<0.05 in signed rank‐test); Values presented as Mean ± SD

Table 2A - Retropalatal Airway Absolute Changes with Upper Airway Stimulation between Responders and Non-Responders

Changes in RP airway volume with hypoglossal nerve stimulation were larger in responders than non-responders but these differences were not statistically significant. Lateral and AP changes were not different between responders than non-responders

Results (Schwab et al, Accepted to Sleep)

Baseline vs. Stimulation Difference P* All Subjects Responders Non‐Responders Measure (n=13) (n=7) (n=6) RG Airway Volume (mm3) 1490.05 ± 4402.30 3481.20 ± 3450.68† ‐832.95 ± 4494.20 0.032 Average RG Area Per Slice (mm2) 38.60 ± 127.49 96.65 ± 78.02† ‐29.12 ± 146.57 0.032 Mid‐RG AP Distance (mm) 2.10 ± 5.39 2.66 ± 5.03 1.44 ± 6.20 0.668 Mid‐RG Lateral Distance (mm) 2.67 ± 4.41† 2.84 ± 5.00 2.47 ± 4.07 0.886 Mid‐RG Cross‐Sectional Area (mm2) 11.16 ± 146.24 46.67 ± 89.65 ‐30.26 ± 194.57 0.568 Min‐RG AP Distance (mm) 1.47 ± 5.70 1.47 ± 6.01 1.48 ± 5.90 >0.999 Min‐RG Lateral Distance (mm) 4.15 ± 4.50† 3.26 ± 4.93 5.18 ± 4.13† 0.568 Min‐RG Cross‐Sectional Area (mm2) ‐4.65 ± 145.74 12.27 ± 130.40 ‐24.38 ± 172.34 0.391 Max‐RG AP Distance (mm) 3.85 ± 5.14† 5.54 ± 4.01† 1.89 ± 5.96 0.199 Max‐RG Lateral Distance (mm) 4.04 ± 5.45† 3.63 ± 5.20 4.52 ± 6.19 >0.999 Max‐RG Cross‐Sectional Area (mm2) 84.28 ± 160.90 155.96 ± 106.33† 0.65 ± 181.59 0.087 *P‐value from non‐parametric Wilcoxon rank sum tests comparing change from baseline between responders and non‐responders; †Within group change significantly different from zero (p<0.05 in signed‐rank test); Values presented as Mean ± SD

Table 2B - Retroglossal Airway Absolute Changes with Upper Airway Stimulation between Responders and Non-Responders

Changes in retroglossal airway volume with hypoglossal nerve stimulation were significantly larger in responders than non-responders. Lateral and anterior- posterior changes were not different between responders than non-responders

21

Results (Schwab et al, Accepted to Sleep)

Baseline vs. Stimulation Difference P* All Subjects Responders Non‐Responders Measure (n=13) (n=7) (n=6) Mandible‐Hyoid Distance (mm) ‐6.93 ± 7.92† ‐11.41 ± 7.44† ‐1.71 ± 4.82 0.010 Mandibular Plane‐Hyoid Distance (mm) ‐5.03 ± 5.36† ‐6.86 ± 5.69† ‐2.91 ± 4.48 0.199 Hyoid‐Spine Distance (mm) 5.20 ± 5.87† 8.09 ± 5.40† 1.82 ± 4.72 0.063 *P‐value from non‐parametric Wilcoxon rank sum tests comparing change from baseline between responders and non‐responders; †Within group change significantly different from zero (p<0.05 in signed‐ rank test); Values presented as Mean ± SD

Table 3 - Absolute Craniofacial Measure Changes with Upper Airway Stimulation between Responders and Non-Responders Mandible to hyoid distance with hypoglossal nerve stimulation was significantly larger in responders than non-responders. Hyoid to spine distance was larger in responders than non-responders

CT Image Analysis - Tongue Centroid

Figure 5: 3D reconstruction

• f a responder showing the

location of the total tongue centroid (circled)

• f

the tongue at baseline and with stimulation. The centroid tended to shift anteriorly and inferiorly with stimulation, with responders showing greater anterior deviation towards the left.

Schwab et al, Accepted to Sleep

22

Results (Schwab et al, Accepted to Sleep)

Baseline vs. Stimulation Difference P* All Subjects Responders Non‐Responders Measure (n=13) (n=7) (n=6) Total Tongue Anterior Displacement (mm) 1.37 ± 2.56 2.54 ± 2.01† ‐0.01 ± 2.59 0.063 Total Tongue Inferior Displacement (mm) 3.22 ± 2.80† 2.78 ± 2.39† 3.73 ± 3.36† 0.668 Total Tongue Lateral Displacement (mm) 0.40 ± 1.53 to the left 0.52 ± 1.53 to the left 0.26 ± 1.68 to the left 0.568 *P‐value from non‐parametric Wilcoxon rank sum tests comparing change from baseline between responders and non‐responders; †Within group change significantly different from zero (p<0.05 in signed‐ rank test); Values presented as Mean ± SD

Table 4A - Absolute Total Tongue Centroid Changes with Upper Airway Stimulation between Responders and Non-Responders Tongue anterior displacement was larger in responders than non- responders with hypoglossal nerve stimulation

Conclusions (Schwab et al, Accepted to Sleep)

• Baseline soft palate volumes were greater in

non-responders compared to responders

• Hyoid movement was greater in responders,

due to increased tongue base recruitment

• These differences were reflected as greater

increases in RP and RG airway volume and average cross-sectional area

• These results suggest that larger soft palate

volumes at baseline and impaired tongue movement anteriorly may hinder response to upper airway stimulation

23

Insights in the Management of SBD with Upper Airway Imaging - "Take Home Messages"

• Weight loss decreases tongue fat and fat pad size
• CPAP increases the upper airway primarily in the

lateral dimension by decreasing the lateral walls

• Examining the centroids of the soft palate and

tongue leads to understanding the mechanism of action of oral appliances

• Volumetric MRI is an important tool to understand

the postsurgical anatomic changes and reasons for success in all sleep surgeries (including TORS)

• We can understand the mechanism of action of

hypoglossal nerve stimulation with CT scanning

Thank you for your attention! Any Questions?

rschwab@pennmedicine.upenn.edu

Insights into the Treatment of SBD With Imaging