Kobe University Repository : Thesis

学位論文題目Tit le

Noninvasive and Simple Assessment of Cardiac Output and PulmonaryVascular Resistance With Whole-Body Impedance Cardiography IsUseful for Monitoring Pat ients With Pulmonary Hypertension(肺高血圧症患者の診療における全身インピーダンス心電計を用いた非侵襲的な心拍出量および肺血管抵抗測定法の有用性)

氏名Author Taniguchi, Yu

専攻分野Degree 博士（医学）

学位授与の日付Date of Degree 2014-03-25

公開日Date of Publicat ion 2015-03-01

資源タイプResource Type Thesis or Dissertat ion / 学位論文

報告番号Report Number 甲第6054号

権利Rights

JaLCDOI

URL http://www.lib.kobe-u.ac.jp/handle_kernel/D1006054※当コンテンツは神戸大学の学術成果です。無断複製・不正使用等を禁じます。著作権法で認められている範囲内で、適切にご利用ください。

PDF issue: 2020-08-06

Circulation Journal OR",cial Journal of lheJapanese Circulation Society http: //www.j-circ.or .jp

ORIGINAL ARTICLE Pulmonary Circulation

Noninvasive and Simple Assessment of Cardiac Output and Pulmonary Vascular Resistance With Whole-Body

Impedance Cardiography Is Useful for Monitoring Patients With Pulmonary Hypertension

Yu Taniguchi, MD; oriak i Emoto. MD. PhD: Kazu a Mi ag:n n. MD. PhD: Kazuhiko akayama. MD, PhD; Hi roto Kinutani . MD: Hidekazu Tanaka. MD. PhD:

Toshir Shinke, MD, PhD; K n-ichi Hirata. MD. PhD

Background: Right heart catheterization (RHC) is the gold standard for the diagnosis of pulmonary hypertension

(PH) and a useful tool for monitoring PH. However, there are some disadvantages in the regular use of RHC because

it is invasive. Noninvasive methods for monitoring hemodynamics are needed to manage patients with PH. In this

study, we aimed to evaluate the reliability of noninvasive hemodynamic assessment with whole-body impedance cardiography (Non-Invasive Cardiac System [NICaS]) for PH.

Methods and Results: We investigated 55 consecutive patients undergoing RHC. Two- thirds of them had pulmo­

nary arterial hypertension and one-third had chronic thromboembol ic PH; 25% of the patients were receiving medi­cal therapy. Cardiac output (CO) was estimated by NICaS (NI-CO), thermodilution (TO-CO), and the Fick method

(Fick-CO). There was a strong correlation between NI-CO and TO-CO (r=O.715. P<O.0001 ) and Fick-CO (r=O.553, P<O.0001). Noninvasive pulmonary vascular resistance (PVR) was estimated using a conventional invasive equation

with NI-CO, mean pulmonary arterial pressure was calculated by echocardiographic measurement, and pulmonary capillary wedge pressure was estimated at 10 mmHg in all cases. NICaS-derived PVR was very strongly correlated

with invasive PVR (TO-PVR : r=0.704 , P<0.0001 ; Fick-PVR: r=0.702, P<O.OOO1 ).

Conclusions: Noninvasive measurement of CO and PVR using NICaS and echocardiography is a useful tool for

the assessment of PH . (Cire J 2013; 77: 2383- 2389)

Key Words : Cardiac output; Noninvasive assessment ; Pulmonary hypertension; Pulmonary vascular resistance; Whole-body impedance cardiography

Pulmonary aneri al h perren. ion (P H) is a progre .. ive

disea.se chamclerized by ele aled pulmonary vascular reo i. tan e (PVR ) because of pu lmonary va. cular re­

modeling. This lead. to a decl-ea. e in cardi ac output (CO) and ultimately ueath . Recently. turgeted medical therap for PAH paLients with endothelin-receptor antagonists. phosphodies­tera! e-5 inh ibitors. and pro. tacyclin analog. h~ been estab­l i~hed. 1 and the prognosi of PAH hn improved.1 However. there is no uni ersally accepted consensus on the treatment goals or follow-up strategy for PA H patients. Right he.u1 cath­eterization (RH ) i. not onl y the gold standard for the diagno­sis of PAH. but i ' a.lso a useful tool for monitoring PA H. and is recommended 3-6 months after ne\ tre~llments and in the ~ e of c1ini al wOI.'ening.1 Hemod namie monitoring with

RHC i. predicti ve of . urvi al and effecti ve in a goal-oriented treatment stmtegy . .I ·~ and has been recommended by a recent guide line: I however. there are . orne di. advantage. in the reg­ular use of RH as a follow-up procedure. spe iall wi th regard to invasivene . Non in va~ive and Ie).). comp li ca ted methods for monitoring hemooynami s are needed to mana!!e patient! wi th pulmonary hypenen. ion (pH ). Le .. inv~ ive hemod namic monitori ng ha recen tly been sll gge~ted a f~l­sible in some situation ~.s The on- Invasi e Cardiac System ( ICaS: I M edical. Hod-Ha ·haran . Israel) is a device for calculating CO noninvajvel wi th whole-bod impedance ardiogmph (lCGwB . The NI a -deri ved CO I-CO) ha '

been !>hown to be as rel iable a ' the RHC-deri ed CO and i ' appl icable for the non in asi e asse .. ment of cardiac function

. 2U1 3; rc i cd manu, rill received May 1. 20 13: :Iccerlcd May 2. 20 13: rclea cd onl ine June 12. 20 13 Time 1'01'

primary re, iew: 42 day~

Di,·i,ion orCardiova~culllr 1edi inc. Departll1cnl ur Inlemal Medicine. Kobe ni\'crsi ly Gr.Jdualc hool or Medicinc. Kobe ( .T .. .E .. K. 1.. K. .. H.K .. H.T .. T .S .. K.H .): C linical Phanna y. K be Pharmaceulical ni verloi lY. Kobe ( '.E .. K .. ). Japdl

1ail ing addreS!: 'oriaki 111 010. MD. PhD. Oi vi~ion o r Cardi ov~L, -ul ar Medi ·inc. Deparlmcnl of Inlcmal Medic inc. Kobe nivcrsily Gradua ll.! School or Medicine. 7-5-1 KU~lInok i. Chllo-kll. Kohe 650-0017. Japan. "-mail : el11111o @l11cd.l-obe-1I .a ·.jp

ISSI -1346 .. 9 43 doi : IO. I25 leir j . J- 13-0172

All right>. are n'_erved 10 the J. pane e Circulation Societ . . For permbsion . . p lease e-mail : cj .iir .or.jp

Circulation Journal Vol 77. Seplember 201'3

2384

Table 1. Clinical Characteristics of All Patients at Initial Hospitalization

Age (years) 62±14

Female(%) 39 (65)

Diagnosis (%)

PAH 38 (63)

IPAH 12 (20)

CTO-PAH 24 (40)

Po-PAH 2 (3)

PH associated with respiratory disorders 3 (5)

CTPH 20 (33)

WHO-Ic (%)

1 1 (1 .7)

2 22 (37.3)

3 32 (54.2)

4 4 (6.8)

Treatment (%) 25 (24)

Bosentan 13 (22)

Sildenafil 14 (23)

Beraprost 10 (17)

Hemodynamic variables

sPAP (mmHg) 53.9±21 .3

mPAP (mmHg) 31 .7±t2.0

RAP (mmHg) 3.7±4.2

PCWP (mmHg) 7.0±4.3

CO (TO) (Umin) 4.90±1.62

CO (Fick) (Umin) 3.92±2.08

PVR (TO) (dyn · S- 1 . cm..s) 433±244

PVR (Fick) (dyn · S-I . cm-5) 581±344

HR (beats/min) 73±11

CO. cardiac output; CTO-PAH, collagen tissue disease associ­ated PAH; CTEPH, chronic thromboembolic pulmonary hyperten­sion; HR, heart rate; IPAH, idiopathic PAH; mPAP, mean PAP; PAH , pulmonary arterial hypertension ; Po-PAH, portal PAH; PCWP, pulmonary capillary wedge pressure; PVR, pulmonary vascular resistance; RAP, right atrial pressure; sPAP, systolic pulmonary arterial pressure; TO, thermodilution; WHO-fc, World Health Organization functional class.

in pati e n ~ with left-sided chron ic hean failure .6•7 but it. fea. i­bility in patients with PH has not been evaluated. The purpose of th i .. tud \Va. to evaluate the reliabili ty of nonin asi\ .. e mea. urement of CO and P R \ ith ICGwB in patien~ with PH .

Editorial p 2251

Methods Thi ' 'lUdy was approved by Kobe niversi ty Huspi talln ·titu ­tional Re iew Board and the pali en~ pro ided wri tten infonlled consent to participate.

Patients We enrolled 6S concec uti ve pati ent · with known or . uspecled pulm nary hype l1en ion h . pitalized in Kobe ni el ity H -pital fro m April 20 10 to ugu. t 2011. All pati enu who were . cheduled for RHC without fulfil ling one of the exclu:ion criteria were eligible for th e study . The e elusion criteria in­cl uded restlessness and/or uru·table patient condition. 'evere aortic £lIve regurgi tation and/or aorti c ·tenosis. aonic an u­r m. heart rate > 130 beats/min. intra- and ex tra ' ardiac shunt. . severe peripheral vascular dise" 'e. severe pitting edema. -ep-

TANIG CHI Y et al.

Table 2. Hemodynamic Measurements

Parameter

TO-CO (Umin)

Fick-CO (Umin)

Echo-CO (Umin)

NI·CO (Umin)

TO-PVR (dyn · S-" cm-5)

Fick-PVR (dyn · s-' . cm-5)

Echo·PVR (dyn · s-1 ·cm-5)

NI-PVR (dyn ,s-l ·cm-5)

Value

4.92±1 .56

3.87±1.24

4.34±1 .11

4.40±1 .32

446±249

583±362

660±363

544±316

Echo-CO, echocardiography-derived cardiac output: Echo-PVR. echocardiography-derived PVR: Fick·CO, cardiac output derived by the modified Fick method; Fick-PVR, PVR derived by modified Fick method; NI-CO. NICaS-derived cardiac output: NI·PVR. NICaS with echocardiography·derived PVR: PVR, pulmonary vascular resistance; TO· CO, thermodilution-<ierived cardiac output; TO-PVR, thermodilution-derived PVR.

~is . and lial y. i ~. all of' which interfere ith the a curate mea­'urement of iml edance-derived 0 wi th I as. a~ previ­ou. ly described . ~ Patient · with elevated pulmonary capillary wedge pre~~ure (PC'll P > IS mIllHg) on RH were al 0 e '­<.: Iuded: 5 pali nl! were e eluded becau 'e of the I re~ence I' an intra-cardiac shunt : 29 pa tient ~ were reevaluated 3- 6 months after nelV treatments or at lini al worsening.

Hemodynamics HemOdynamic data were derived fro m slandard RH in all patients using a 6Fr Swan-Gan7. catheter (Baxter I-Iealthcare. Irvine, CA. SA). The calheter ~ as illt.rodu ed into the pul ­monary artery under flu oro 'w pic guidan<.:c. 1 ' an pulmonary arterial pres 'ure (mPAP). s stolic and end-diaslOli pulmomu'_ art.erial pre. sure (. PAP and dPAP). mean right atri al pre . . ure. and PCWP were mea ured. 0 IVa. measured u ~i ng the fol­lowing te<.:hnillue .

Thermodilution-Derived CO (TO-CO) A -1111 bolus of iced SCll' gluco. e . olution Wa! injected 5 times at the . ame rare. The re~ult:-. of _ in.iec tion ~ wi thin 150/1' I' their e treme di~parity were a eraged to derive the TD-CO value.

Modified Fick method (Fick-CO) Blood samples were ob­tained from . y. temic and pulmonal aneries. II . ample. were measured for oxygen saturation with the same dev ice (Radi ­ometer ABL 715. Copenhagen, Denmark).

NI-CO The. e measurements were pelformed . imultane­ously ~ ith the mea. urement of TD-CO and Fick-CO during RHC. The melnlrement of 'I-CO followed the method ,,_ previously repoI1ed : ~ an altemating eleclJical CUITent of l .4 m with a 3D-kH z. frequen y i~ pa~!,ecl through the patient ia _ pair ' of t Lrapolar elecLrode ' - I pair placed on the wrist above the radial pulse. and the other pair placed on the contralater.11 ank le above the posterior tibiali . al1erial pul se . If the anerial pulse ' in the legs are ei ther absent or of poor quality. the sec­ond pair 01' !ectrooe ' is placed on the contralateral wrist.

The NI a apparatus calculat e. the . troke volume (S ) by FrineIl11an· . formula :8

SV =dRlR x px L2IRi x (cx +f)/f X KW x HF

where dR is the impedan e hange: R i the ba, al re ' istance; p i' the blood electrical re~i . ti vi ty: L i. the patient' height : Ri is the corrected basal resistance according to 'ex and age: KW i ' a correction faclOr for weight according to ideal value: HF i ule hydration factor. whi ch take into acco unt the bod water composition. a+f3 i Ctlualto ule ECG R-R wave interval

C"culatJon Journal Vol.77, September 2013

Noninvasive Assessment of PH 2385

A ll B 12

11 .- 1:' . p<tt .OttIJ I • 1I.6 'i.l. p' II .IIIIUI

:: OJ :: OJ

~ ~ • 6

:::. (,

0 0 v I... 6 "" '- J " J ~

0 0 0 J 6 9 12 0 3 6 9 12

"I-CO ( Illllill) 'II -CO (Umill)

C 11

r = 0.795, 11<0.000 I

c: 'J

E • ::J

/' 0 6

• • 0

oX • • .= J

• 0

0 6 9 12

TO-CO (I./mi II)

D E 12 12

r ~ tt _" 12 . 1,<11 .11011 I • O..l61 · IY tt.ttttttl

:: 9 C 9

~ E ::J :::. (, 6

0 0 U V 6 "" ... I- 3 ~ 3

II 0 0 3 (. 9 12 0 3 6 9 12

Echo-CO ( lIli II) Echo-CO (Llmi II)

Figure 1. Linear corre lation analysis between ill-CO and NI -CO (A), Fick-CO and NI-CO (6 ). ill-CO and Fick-CO (e ), ill-CO and Echo-CO (D), and Flck-CO and Echo-CO (E). TO-CO, thermodilution-derived cardiac output; NI-CO, NICaS-derived cardiac output; Fick-CO. cardiac output derived by the modified Fick method; Echo-CO, echocardlog raphy-derived cardiac output.

and f3 is the diastolic time interval. To ca lculate the CO_ SY is multiplied by the heart rate. Because the I-CO value: are cil lcul ated evelY 20 . . the average or 3 mea. urements obtai ned con~ecuLi ve l y during 60 . uf monitoring i ' considered to be the

I-CO value for ench individual case.

Echocardiography I:.chocardi ography \ a~ performed lI 5ing a ivid 5 . ~le lll and il 3.5- 1Hz triln. ducer (G ingmed it msollnd S, HOrLen.

orway). Two-dimen. ion ill Doppler examinati on. were per­formed in th e uSLI al manner. 0 was mea ured by tracing the len venlli cular ejection now (Echu-CO). Echo- 'PAP was c -timated frOIll the peak veloc ity or the lIi cuspid regurgitation jet plus estimated ri ght alrial pres. lire (Echo-RAP).9

Measurement of PVR PYR (dyn ' S-I 'CIll-5) was ca lculated u 'ing RHC l'rolllthe equa­tion:

PYR =80x (mPAP- PCWP)/COro. Fock (TD-PVR. Fick-PVR).

P R wa. also estimated nonin vasi ely u. ing a oll1bination of ICaS and echocardiography. and b echoca rdi ography alone. In cho-IllPAP wa~ alcul ated , Echo- P PXO.6 1+ 2 IllmHg. as previoU! Iy de. ribed_ 1I and non in va~ive P R WilS calculated as 80x(Ech -mPAP-PCWP)! O Echo. NI (Ech -PYR.

I-PYR ). PCWP for the ca lculati on of n nin va~ i ve P R wa eSlimated at 10 mmHg in all cases. III

Statistical Analysis Quantitative data are presenled as mean±SD. The correlation '

CirculaOon Journal Vol.77, Seplember 2013

2386

A 6,-------------------------,

~ E 4 ::J tn t-lI 11 -4 2 ' U ... • • ------------ -_. ------- --- --.-.-.--- -- --- . -- ----o 111 011 11 ••••• • '" ..!. 0 F=-~rwtE_FlI!--=------; z , -) o -

V 6-4 ~ ~+_----r_--_,----_r----_r----4

C 6

:: E 4

o , -I

o

o 6 10

-eragr of TO-CO and ;\ I-CO ( miu)

• •

~ ~+_---,-----r----._--_,----~ o 6

.\ >erage ofTl>-CO and Echo- '0 ( 10

mi II)

TAN IG CH I Y ct a l.

B 6,-------------------------,

• • :5 4

o ~ O~~--or~~~~~L-~.·~~ o -2 V ~-4 .~ L<.

• ~+_----_r----~------~----~

o 4 6 8

.\'·~ rag~ ofFick-CO and _'I-CO ( lII i n)

D = 6,---------------------------,

• o

=- 0 nU' 11 11 ~ l..!!!.!..!!!!._ """ w.; ~ .. o -2 !'!!_ ..... ;~_~ fJ. ____ .,.----.!-- ________ • ________ __ , -I ~

" ~ ~+_----r_--_,----_r----,_----; o 4 6 8

A>r""l!e ofFick-CO and Ecbo- 0 ( 10

min)

Figure 2. Bland-Altman plots with mean difference (solid line) ±2S0 (dotted line) comparing TO-CO and NI-CO (A). Rck-CO and NI-CO (B), TO-CO and Echo-CO (e ), and Fick-CO and Echo-CO (D). TO-CO, thermodilution-derived cardiac output; NI-CO, NICaS-derived cardiac output; Rck-CO, cardiac ou put derived by the modified Fick method; Echo-CO. echocardiography-derived cardiac output; SO, standard deviation

among TO-CO, Fick-CO. Echo-CO. and I-CO and between Echo-mPAP and mPAP measured by RHC (RHC-mPAP) were determined b calc ulating the Spearman-. rank corre­lation coerficienl. P<O.OS wa ' con ~idered to be -ignificant. Agreemelll between method ' was analyzed by the B land-Alt ­man method. l ! nle limit; of the agreemelll we re e pre .. cd a. the mean±SO. The 9S% confidence interva l ' (Ci s) or the bia ' were also calcu lated. Receiver-operating characteri stic (ROC) curve. were generated for the detection of elevated I VR de­fi ned a. > 240dyn · S-I. cm-5 (3 Wood units I W U I). The area under the curve (AUC). cut -orr value. sen ·itivity. and : peci lic­i ty we re e. timated by the R curve .. A ll st::uist ical analy:es were performed using GraphPad Pri rn ve l' ion 5 (Graph Pad Software. La Jolla. CA. SA).

Results The baseli.ne haracteristics of al l patients at initial hospital­iz.ation are . ummarized in T able 1. Approximately two-third. of the patients had PAH ( World Health OrganiL.ati on IWHOI cia . . ification of PH group I ) and the other one-t hird of the patient. had chronic thromboembol ic PH (CT EPH : WHO group 4): SIk of the pati ents were c ia .. ified as WHO group 1. At enrollment. 24 ~ o f the patients were receiving medical thempy.

Relationships Among Parameters The mean CO alues from all mea uremenl in these sub-

jects for TO-CO. Fick -CO. Echo-CO. and ' I-CO were 4.9 __ 1.56 Llmin . 3.87± 1.24 Llmin, 4.34± 1. 11 U min. and 4.40± 1.12 U min. re, pecli ve l (Table 2). . ignifi ant and very 'Iron2 correlation was ob 'en 'ed between TO-CO and I-CO (1'=0.7 I S. 1'<0.0(0 1) and between TO-CO and Fick-CO (r= O.79S. P<O.OOO I ) by 2-lailed Spearman'_ rank correlati on test (Figure I). There was a strong correlation between Fick-CO and I-CO (1'=0.653. P<O.OOO I ). Howe er. the correlation between Echo-CO and TO-CO or Fick- 0 \Va: . ignificalll but not stron g (1'=0.5 12 or 0.461. P<O.Ooo I. re pecti ely). The di fference bet ween 2 measurement. were plotteu 'lccording to the Bland-A ltman method (Figure 2). The mean bia. and l imit ofagl emem between TO-CO and I-CO. Fick-CO and II-CO, and TO-CO and Fick -CO were O.SO±I.0 (-1.6 1 to

2.61) Llmin. and --O.S4±I .04 (- 2.S7 to 1,49) Llmin. and I .O_± 0.86 (--0.68 La 2.7 1) Umin_ respecti ve l . The limits of agree­ment between TO-CO anu Echo-CO. and rick- 0 and Echo­CO were 0.64±1.'3 (- 1.97 10 3._6) Umin and -O.4:L I. 18 (- 2.73 to 1.88) Umin. respective ly. T here \Va. no c lear differ­ence in the mea. urement. of CO among the patients w ith id­iopathic PA H, collagen Li ~ ue di ' ea e % 0 iatecl PAH or CTEPH (Figure , 1).

Comparison of Invasive and Noninvasive Measurement of mPAP and PVR The mean values of all measurements of invasive mPAP and Echo-mPAP were 32.9± 1.28m mHg and 43.0± I. -9 mmHg. re­specti ve ly. There wa: a very ·trong correlation between inva-

Circulation Journal Vol.77, September 2013

Noninvasive Assessment of PH

~ i\'e mP P and Echo-mPAP (r=O.70:. P<O.()OO I: Figur 3A ). The limit~ of agreement between in vasive mPAP and 8:ho­mP P were - 9.6. _ 10.2 (- 29.6 to 10.-1 ) mmHg (Figure 38 ).

The mean va lue ' of all measurement!> of TD-P R. Fick­P R. Echo-P R. and I-P R wen: 446-2-l9dyn· S- I. em-5.

SR. ±362dyn · . - I . m-5. 660±. 63 dyn· . l· cm-5 . and 644±J 16 d, n · - I . m-5. reo pecti ely (Tahle 2). There were ~ignifica nt and el ~lrOng correlatiol1 \ betwee n TD-P R anti I-P R (r=O.7n-l. P<O.OOQ I). berween Fi k-P R and I-P R (r=O.702. P<OJlOO I). and bel\ een TD-PVR aJ1d Fi k-P R (1'=00.942. P<O.OOO I) (Figu" ~A.I» . Howe\ r. the orrelatiol1 betwee n P R meas ured by in vasi \ e method ' and Echo-P R was not as trong (r=0.602 or 0.603. P<O.OOO I. resp ctive ly: Figure ~G.J ) a~ that between invas ive method~ and I-PVR . Figure 4B and d n· S- I. cnr~ 'hows the Bland-Allman plol! of the difkren 'es between TD-P R. Fiek-PVR. and NI-P R. The limits of agreement between TD-P R nnd NI -P R. FicJ..-P R and I-P R. and TD-P R and Fi k-PVR were - 195±....6 (- 71-to 3_6) d n ·s-I·cm-5. - 35_3_5 ( 73 to 603) dy n ·s-l ·cm-5. nnd - 135_ lo-l ( .-7 to I 7) dyn ·S-I · cm-'. respe ti vely. 111e l i ll1it ~ of agreement between TD-P R and Echo-P R. and Fick-P R and E ho-PVR ~ ere - 19 1±166 (- 7 13 to 330) dyn· .-I· cnr . and - 33±34 1 (- 703 to 635) dyn · s-I. 111- 5• respecti vel (figur ~II . K ) . The A for NI ­p R tociete<:l i nc rea~d P R >240 dyn · ~- '· cnr (3 )

again: t TD and Fick-P R ",ere 0.84 (95% CI. 0.72 .96) and 0.92 (95'7r C I. O.R4-0.99). reo pecti vely (Figures 3 .F ). nnd optimal cut-off alue. were 4 11 dyn · . - I . cm- (. en. itivit y: 8 LV/c. speci fi cit y: 75 %-) and 40U dyn · . - I . cm-5 (sen iti it y: O.Yif-. spcc ifici ty: IOOC,'f,). respe tively. The C for Eeho-

I' R nga in. t TD and Fiek-P R were lower: 0.75 (95 Yc CI. 0.57-0.92) and O. ' 3 (95 '7. CI. 0.6(r..().99) (Figur ~J.L ) com­pared wi th that for NI-PVR against TD and Fick-PVR.

Discussion We repon on lhe reli ability of a nonin vasive and si mple meth­od of assessing CO and P R usi ng ICOwn in patient!'> with PH. Prev iou. repons have inui ·ated the rea ibility or hemody­namic as e menl lI ~ ing arious method~ in compalison \ ith RHC in a nlilge of cl inical sell i ngs:'-' however, a reliable meth­od for the as. e. sment of PH h,~ no! yet been e. tabl i. hed. In panicular. there are few . tudie. that have addre. , ed the nonin­vasive a ' se~~ment of hemodynamic ' in PH. Hemod namic a se .. melll u. i ng cardi a magnetic re onance or echocardiog­raph have been . ho\\~l 10 be re li able, 111.14-17 but the. e method. requi re e ' pensive equi pment and trd ined opera tor~ . Th raci impedance cardiography has been used for the mea urement or o noninvasive l in PAH . ' ~ and its reliability w~ . hown 10 be

compromised in cardiac patienl in a meta-analys is. ' 9

We demonstrated strong correlations among the ICaS. TO. and the Fick methods for the measurement of CO. Although the limit s of agreement between I-CO and TD- 0 or Fick­CO estimated by Ihe Bland-Altman approach were not small. they were acceptable when 'ompared with pre ious repons. ' ~ Therefore. e believe that I as can be a re liabl e tool for the nonin va ive as,e " ment of 0 in PH. H wever. Illpared with N I-CO. the orrelat ion between Echo-CO and TD-CO or Fi k- 0 was weaker and the limits of agreemenl s were larger. The relati ve inaccuracy of 0 men. ured by echocardiography was con~is t enl wi th a pre iOll. reporl. :!U and Illay be a onse­quence of using the Doppler method. 'evere !ricu 'pid regurgi­tation. and operator-dependency.

We also demonstrated Ihe feasibilit y of nonin a .. ive and simple measurement of P R using a combination of NICaS

A 100~----------------------------,

~ 80

E E

Q..

<:

60

Q.. -10 E o v 10 W

0

B '2 ::t: 60 E E -Q..

40

-< Q..

E 20

b 0 -£ ..: , -20 ::.. .... Q.. -40 E 0 -60 ::t: ~

r = 0.703. p<O.OOO I

0 10 -10 60 80 RH -IIlI'AJ' (IIImHj!)

0 20 40 60 80 100

2387

Awrage ofRJl C -mPAP ami Echo-mPAP (mmlig)

Figure 3. linear correlation analySIS between RHC-mPAP and Echo-mPAP (A) Bland-Altman plots With mean diffe ence (solid line) ±2S0 (dotted line) companng RHC-mPAP and Echo-mPAP (8 ). RHC. nght heart calhe erization . mPAP. mean pulmonary arterial pressure; Edlo-mPAP. mPAP calcu­laled by echocardiography; SO. standard devialion.

and echocardiograph_ . Kouzu et "I showed that tricu, pid re­gurgitant pressure gradie nt (TRPO )/right ventri ular time­veloc ity integral (T I) is reliable for the e. timati on of P R.21 Alth ugh TRPOrr I h::t~ been confi rmed n. a reliable method for estimating PVR.16 accurate measurement of TVI need: a . killed operator. Lindqvi. t et nl reponed the nccurac fa . imple Doppler-derived me~ urement of I' R Wilh the con­ventional in va ive equ ation in pat ienls \ ith PH :lo h wever. their stud y excluded patients wi th se cre tri u 'pid regurgita­tion. \ hich cau. e. innccuraey in 0 measure ment using ceh -cardiography. In the present IUdy. we estimated P R usi ng the conventional invasive equation with the ombination or NI-CO and Echo-mPAP. There \Va ' very trong correlation nOI onl y between in asive mPAP and Echo-mPAP. but nl. 0

between inva ive PVR and NI-P R. Furthennore. a 't ronger 'olTclalion of P R was found for the use of I-CO compared wilh cho-CO. The lim iLs of agret;men~ estimated by the B land- Itman anal . i. were large. but comparnble with previ ­ous reports that sho\ ed the fea ibility of PVR derived by e hocardiography again. t inv~ ive PVR.lUl Furthermore. the hi gh C. , en iti ity. and l-pec ifici ty for I I-PVR to detect inc rea~ed PVR >240dyn · . I. cm-5 (3 W ) 31..0 indicate::. the reliability of noninvasive P R asse 'sment using ICaS.

In our ·lUdy. the va lu c for TD-CO wa 'igni ficantly hi gher than the CO va lue ' wi th \ ther meth ds, incl uding Fi k- O. and therefore. the value of TD-P R was underestimated. This

CirculatJon Journal VoL 77. September 2013

2388 T ANIG UCHI Y et aJ.

A B C lOtiO I~OU 1011

r 11'11 , 1"(1 111101

! .r lOon 80 ~ I ~OO

~ ~OO ' . ... n ... ! ... u ,:. C ::: :. 60

It f II ~ • :e 1000 ;,. • • _ h lll Q'''I,ih 1_'-• :z '" :: JO (,,(Irtfl" ~~ ...

E: z -,00 ., ~ UO 'J'J

0 ~ - IOUO 20

:: -1,00 n 300 1000 I~OO 1000 "- 0 !OO 1000 I~OO 0 20 JII .0 80 100

~I · P\ R hi,' " · ~/t:. III ~ \H' ra~ of lD-P\ H ami 'I - I'VI{ ( U) rl ' !HCfl1 ~) 100·". - ~r.ermrh~ ·~

0 E F l OOO .r I~OO 100

11 .... 01. P ".HIIHI E

.r • ~ 1000 • = • • • • • 80 l!ill{i III ~~" :;:-:.u. • v • §' • t , • • ,00 • • •• • (,0

~ • • • > III II.Y; € IfllJO <>- . .... , .. , '<"n\I',\I', ~II_\·. a: • c: 40 r>< lifo 11\ .. InO-. 7 -~OO 11.. .... ~~~ .... 1_, _\ .... _ ......... u 0-

~QO .1Jl.~I!O .-;~

'" t ; . 1000 10 ~ "-

~ · I!\UO U !'oo 100U I ~OO lOOO ..: u ~oo IUOO I~OO 0 20 JU 60 8U IOU

\1 -1'1 I{ ( d~ n ·s lrlll 'J \ rru~r of ~irk - I'\ 'n ond \I -f'l n (d) n'>lr m' ) 100~. - 1)(' c'iriC'i l ~ • .•

G H .. I~OO ~ I~OO 100

I 11 .• 01 . P 0111/111 = .- • ~. 1000 :e •

~ 10UO , ::: ~ou '? .... } •. '! : ;~~) C ~ 9- ';: = "- It ( O .-~ ...

:e ~ . - '"'4. """\11\ x.~· . v-a: <:; • :: 'f·tI(hit~ C:O"II

~ ,00 ." -~OO • # ..

• • 'J'J Q '" mrs ll ·2~U 10 ~ 't -1000

~ -t,OO U 500 10110 1500 lOOO Il 500 1000 ISOO 0 ,0 100 150

~J:h o-I'I I{ ( d~' n · s/r ln ', .'1\Croge of l1)- PVI{ nnd Echo- I'I I{ ( d~ n·s/rm ' , 100% - pt'cifid r~ -/ (1

J K L

woo ~ 1500 100 r O.b(IJ . 1' lIU~UI . - ~ . 1000 •

E • :e 80

~ 1,00

" " • 500 ;,. Ai: 60 ... \t ( 0 .\ :e 1000 ~ .- ~n\lh\11\ iI . I', a: ~ •• • •• '" J O • :: 1' .. :dOdh IUJ ... -500 • "" . ~

~UO , •. •. .... _ •..• .... .. !I! ._ ... .... .. .... !I.I !']I.'!. .·k~I). _ '" ~ ~ -1000 • 20

..: "-t ·1500 0

, 00 1000 1500 2000 to: 0 500 1000 1500 2000 0 20 J O 611 80 100

Echo- I'\ R ( dyu . "/cll1 ~) Hr3M.t' uf Fid. · .. ' R and "::':IH~I'\' R ( LI~ " O) /CII1 5-) 100% - prtlnd l) '10

Figure 4. Diagnostic reliability of NI-PVR compared with invasive PVR. Linear correlation analySIS between TD-PVR and NI-PVR (A), Fick-PVR and NI-PVR (0 ), TO-PVR and Echo-PVR (G). and Rck-PVR and Echo-PVR (J). Bland-Altman plot with mean differ­ence (solid line) ±2SD (dotted line) comparing TO-PVR and NI-PVR (B), Rck-PVR and NI-PVR (E), TO-PVR and NI-PVR (H), and Fick-PVR and NI-PVR (K). Area under the receiver-operating characteristics curve with 95% confidence interval or NI-PVR and Echo-PVR to detect increased PVR (>240dyn · s I . cm-5 [3 WU]) against TD-PVR (C,I) and Rck-PVR (F,L) PVR, pulmonary vascu­lar resistance; NI-PVR, NICaS w ith echocardiography-derived PVR; TD-PVR, thermodlluYlon-derived PVR; Fick-PVR. PVR defived by the modified Fick method; Echo-PVR, echocardiography-derived PVR.

could be cau. ed by overestimation of the va lue of TO-CO in Ihe presence of low CO. con iSlen l Wi lh pre iOlls rep rt , .B

Study Limitations The main limitation of thi .. rud wa. Ihe need to mea. ure the Doppler parameter for esti mating I-PVR . Proper alignment of the 1I1t r.tsound beam i . crucial for the Doppler parameter to

be determined aprr printely. Thi. may have reo ull ed in bia. in the mea. uremeJ1l of I-P R. In our study, the Doppler p,u'<lI1l ­eler needed in order to e. timme ' I-P R wa nl TRPG. and there was no pntient in whom I e were unable to obtain lhat vallie. Second. we used the conventional invasive equation for e, timating I-PVR. 'Ii e had to e limate PCWP at IOmmHg in aU case - as previously reported. 1O whi 'h m3 also have re-

Circulation Journal Vol 77. September 2013

Noninvasivc ASSeSS I11cnt of P H

su ited in the measurement l r NI-PVR : howl! er. in generJ I. 11

wide variation in PCWP is not usually obser ed among pa­lient. wilh PH. Third. becau. e 0 measurel1lenl u. ing '1CaS in pa l i e n L~ wilh cardiac ~hul1ls i. known to be unreliable.2-1 we excluded case of PAH ;L~ 'ociated with cardiac shunts. Founh. nonin"a. ive estimalion of 0 and PVR with NICaS wa. fea­. ible: however. there ~ ere ~ome paliems who had large diver­gen 'e bl!tween I-CO or '1-PVR and in ~ ive CO or P R. FUllher . Iudie. are needed 10 clarif Ihe factor. l.hal lead to inaccurale me:u urement. of 0 and P R. Fifth . in our . rudy. the:: number of patient wilh WHO functi ona l clas ' ..! was small . MO:I palients were WHO functi onal cla~ ' 2 or 3. The reliabili ty o f ICa ill palicnts wi th se ere PH i!> to be exam­ined in future ~1Udi es. r ina lly. the slud. sample ~ile \ a re i a­lively 'mall llnd originated from a single center. We believe Ihal a lar!!er. mu ll icenter stud i ~ needed to approplialely COnfill1l the re liahilit. of the melhod.

Ithough recen l advan e. in treatment option ' and manage­ment have improved the olltcom . ror patient.'; with PH. treat­menl goa ls and fol low- up slralegy are stil l not we ll defi ned. Hemod. namic moniloring wi th RH C L recommended in a goal-oriented Irealmenl ~ tr:ltegy for PH I and i~ Ihe I!o ld ~ t a ll­

dard for Ihe as!>" 'ment of PA H: however. t.he invasivencs ' of RHC i~ a nilic:J1 factorin~ ils regular use a a follow- up pro­cedure. A noni n va~ive. accu rme. and .imple method i. re­quired for the management of patien ts \ ith PH. We have demon .. lrated noninvlU ive l1le~ urement of CO and PVR u. ing onl : imple parameter. . Echo-sPAP i. needed to e. timate P R. but Echo-sPAP ha been estab li . hed as a . imple. relj ­ab le screening parall1eler for PH .25 Thi. noninvasi e. reliable, and . imple ~ . e .. menl an be a usefu l tool for monilOring and managing patients with PH.

Conclusion onin a~ive mea uremenl or c o and PVR using NICa i, as

reliable a invasive RH .. Thi ' simple as. e sment ould help physicians to manage lhei r pat i enL~ with PH.

Disclosures 'one.

References I . Galie N. Hoeper MM. Humbcn 1, Torbicki A. Vachiery JL. Barbera

JA. el aI . Guidelines for the diagnosis and treaUllenl of pulmonary hypcnen. ion: The Tn! k Force for the Diagnosis and Treatment of Pulmonary Hypellension o f the European Soc' iely of Cardiology (ESC) and Ihe European Rcspir.1tOI)' ocicry (ERS), endorsed by the Imer­nalional SocielY of Heart and LUlig Trun. plantalion ( I HLT). Ellr Hean 1 2009; .'10: 2~93 -2537.

2. Benza RL. 1iller DP. BarS! RJ . Bade: h DB . Fro. I AE. McGoon MD. An e a lu:u ion o f long-leml urvival from timc of diagno i in pu lm nary allcrial hypenen. ion from the REVEA L Rcgistry . Che.SI 20 12; 142: 44 - 456.

3. Fukumoto Y. Shimokawa H. Recent prog.rc in the manag.ement of pulmonary hyperten ·ion. Cire J 201 1: 75: 180 1- 18 10.

~ . Hoeper M 'I. 1arke ych I. Spieke rkoener E, Welle T . ' iedcrmeyer J . Goal-oriented trealment and combinulion lhemp. for pulmo nary allcrial hypencn i 11 . Ellr Re 'pir J 2005 : 26 : 858 - 863.

5. Alhashemi JA, Cecconi M. Holer CK. Cardiac outpul moniloring: An integrative pe pective. Cri/ Care 20 II ; .15: 2 14 .

6. Paredc OL. hile J. hinke T. Walanabe S. Olake H. Ma umoto D. el aI. Impedan(:e cardiograph, for cardiac o utpul estimation: Reliabil­ily of wriq-t ·ankle e lectrode nfiguraliol1. Cire 1 2006: 70 : 11 64 -116 .

7. Tanino Y, Shile J, Panxl OL, Shinke T , Ogasawara D, awada T, el aI . Whole body bioimpedance moniloring for outpatienl hronic hean fai lure follow up. Cire 1 _009: 73: 1074- 1079.

2389

Cohen AJ . Amaudov D. Z heed a D. chullheis L. Lashinger J. Schachner A. o n-invasive mca~uremel1l of card ia outpul during coronary an er}' bypa grafting . Etlr J Cardio/lliI/'ac S"rg 199 : 14: 64 - 69.

9. Yeo T . Dujardin KS. Tei C. Maho ney D\ . McGooo MD. eward JB. Value o f :I Doppler-derivcd index combining y~ tol i and dia­lolie time inlervals in predicting oUlcome in primary pulmonary

hypellen ion. Am J Cardial 1998: 8 1: 1157 - I 161 . 10. Lindqvisl P. oderberg S. Gon7..alcz 1 . To. savainen E. Hencin MY.

Echocardiography based C! tim31iOIl of pulmonary vascutar rcsi tancc in patienl Wilh pulmonary hypencn ion: A imultaneou D ppler echocardiography and cardia cathelerization study. Ellr J Ed,ocar· diogr_OI I: 12: 96 1- 966.

II. Chemla O. Caslelain . Humbert M. Heben JL. imonneau G. Lecarpentier Y. el al. c\ fonnula for predi ling me.~n pul monary arlery pressure u 'ing sy tolic pulmonary anery pre'MlTe. ClIl'S/ _004: 126: 1313-1317.

12. Bland JM . Allman DG. Sialislical mclhods or ru se,si ng agreement bel ween lWO methods of clinical measuremem. Lallce/ 19 6: I : 07-310.

13. Manlha S. Roi£en MF. Fleisher LA. 'lbi sled R. Fo:,s J. CompllTing methods of clinica l mea~urcmclll: Reponing sl311dard for bland and alLman analy is. AliI'. III Allal 2000: 90: 593-602.

14. Inaba T. Yao A. akao T. Hman M. Maki H. Imamura T. Cl al. o lumetric and funcl ional 3.! sessmcnt of venuic lcs in pulmonary

hypelle nsion on 3-di mell ional cchocardi ography. Cire J 20 12; 77: 19 -206.

15. Garcia-AlvafL"l A, Femandez-Friera L. ircli JG. awi l . air . Kallman J. ct al. on-invasive eSlimation of pulmonary va: ular re ' i '­lance with cardiac magnetic resonance. Ellr Nean J 20 II : 32: 243 -2445.

16. Abba., AE. Fonuin FD, Schiller B, Appleton P. 1orcno . Lestcr SJ. A imple method for noninvasive' tim31io l1 of pulm nary v. -cu lar resistance. J Alii Coli CardinI2003: 41: 1021 - 1027.

17. Kang KW. Chang HJ . Kim YJ , Choi BW. Lee H .. Yang WI. et a l. Cardiac magncl ic re nan e imaging-deri ved pulmo nary artery di . tensibilily index correl ales w i ~l pulmonary artery slifTnCl> and pre­dic fu ncti ona l capaci ty in patienl wilh pulmonary an erial hyper­lension. Circ 1 20 11 : 75: 2244- _25 1.

18. Yung GL. Fedull o PF, Kinningcr K, Johnson \ , Channi k R . Com­parison of impedance cardiography lo di reci Fick and th~nnodi lulion caroiac OUlPUI delermination in pulmonary ancriaJ h. penen ion. COli '

ges/ Helin FlIil_OO4: 10: 7- 10. 19. Raa ij make E. Fa' TJ. Scho lten RJ , Goovacns HG , Heethaar R 1.

mela-analysis of publi hed lUdiCl> c ncemin' the a lidity o f lho­racicimpedancecardiography. Alln ' Acod Sci 1999: 873: 12t - 127.

:W. Fi,her 1R. Fortia PRo Challlcra E. Huu,len-Hams T. hal11pion H . Girgis RE. el al. Accuracy of Doppler echocardiography in the he­modynamic as ssmenl of pulmonary hypellen: ion. Alii J Rupir Cri/ Care Med2009: 179: 6 15-6_1.

21. Kouzu H. akalani S. Kyollmi S. Kan zaki H. akanishi , Kitakaze M. Noninva~ive estimation of pulmonary vascular resislance by Dop· pie r cchocardiography in patiellls with pulmonary anerial hyperten ­io n. Am.! CardinI2009: 103: 872 - R76.

22. Selimovic . Rundq vi l B, Bergh CH. A nder son B, Pelersson S. Johansson L. el al . A. melll of pulmonary vascu lar resi. tan e b. Doppler echocardiography in patienls with pulmonary arlerial hyper­lcn. ion. J Hean Lung Trallsl'lalll2007 ; 26: 927- 934.

23. Toumadre JP. Chassard O. Muchada R. O ver limation o f low ar­diac OUIPUII11ea.;ured by thermodiJ uti n. Br J Allaes/h 1997: 79: 514-516.

24. Kauppinen PK. Koobi T. Hyn inen J. Malmi\'Uo J . gmental com-posilion of whole-body impedance ca rdi ogram eHi malcd by com­pUler imulalions and c linical experiments. Ciill Physio/ 2000: 20: 106- 113.

25. Taleb M. Khuder S. Tinkel J. Khouri SJ . The diagno tic a cura yof Doppler echocardiography in assessment of pu lmonary anery y lo li pres ure: A meta-analysis. Echocardiography :!01 3: 30: 258-265.

Supplementary Files upplcmenlary File 1

figure S1. Linear correlation belween I-CO and TO-CO (A ) or Fick-CO ( ll ) in IPA H (e. red line). CTD-PA H (e. blue line) and CTEPH (.&. green line ).

PIl'<I.>C fi nd ,upplementary file(,): http://dx.doi.orgl I0 .1253/circj.o-13-0 172

Circulation Journal Vol. 77 , September 2013