Ventricular Diastolic Functions of Ankylosing Spondylitis Patients by Using Conventional Pulsed?Wave Doppler, Myocardial Performance Index, and Tissue Doppler Imaging

C© 2007, the AuthorsJournal compilation C© 2007, Blackwell Publishing, Inc.DOI: 10.1111/j.1540-8175.2007.00541.x

Ventricular Diastolic Function of AnkylosingSpondylitis Patients Using Conventional PulsedWave Doppler, Myocardial Performance Indexand Tissue Doppler ImagingTaha Okan, M.D.,∗ Ismail Sari, M.D.,† Servet Akar, M.D.,† Hasan Cece, M.D.,‡Ozhan Goldeli, M.D.,∗ Sema Guneri, M.D.,∗ and Nurullah Akkoc, M.D.‡∗Department of Cardiology, †Department of Internal Medicine Division of Rheumatology, and‡Department of Radiology, Dokuz Eylul University School of Medicine, Izmir, Turkey

Objective: The aim of this study was to evaluate ventricular functions by using standard Dopplerechocardiography (SDE), myocardial performance index (MPI), and pulsed wave tissue Doppler imag-ing (PW-TDI) in patients with ankylosing spondylitis (AS) and healthy controls. Methods: Forty-nineAS patients (38 ± 11 years, 25 M/24 F) and 33 controls (36 ± 9 years, 17 M/16 F) were studied.Two-dimensional, M-Mode, SDE, PW-TDI echocardiography examinations were performed. Spinalmobility was assessed by the Bath ankylosing spondylitis metrology index (BASMI) measurement.Patients were also evaluated using the Bath ankylosing spondylitis functional index (BASFI) andthe Bath ankylosing spondylitis disease activity index (BASDAI). Results: Four control subjects andsix AS patients met the left ventricular (LV) diastolic dysfunction (DD) criteria by using conventionalDoppler echocardiography (p > 0.05). However, using PW-TDI method 22 patients in the AS groupand six subjects in the control group were diagnosed to have LV DD (Em/Am < 1). Pseudonormal-ized pattern was present in 16 AS patients and two control subjects. Correlation analysis revealedsignificant moderate negative correlations between Em/Am and BASMI, age and body mass index(p < 0.05; r = −0.3, −0.6, and −0.4, respectively). No correlation was observed between Em/Amand disease duration, BASFI, BASDAI, CRP, and ESR. We could not detect any right ventricularfunction involvement either by conventional or by recently introduced echocardiography methods. Therisk of developing LV DD was found to be 3.7 times higher in AS patients. Conclusion: When sensitiveechocardiographic Doppler techniques such as MPI, TDI-derived MPI, and PW-TDI are utilized, DDcan be detected in a significant proportion of patients with AS without cardiovascular (CV) diseasewhich may contribute CV mortality in these patients. (ECHOCARDIOGRAPHY, Volume 25, January2008)

ventricular dysfunction, spondylitis, ankylosing, echocardiography, Doppler, echocardiography,Doppler, pulsed

Ankylosing spondylitis is a chronic inflamma-tory rheumatic disease of the spine that affects0.2–0.9% of the population.1 There are severalcharacteristic extra-articular manifestations ofankylosing spondylitis involving organs such asthe eye, gastrointestinal system, kidneys, lung,and heart.2

Address for correspondence and reprint requests: Nurul-lah Akkoc, M.D., Dokuz Eylul Universitesi Tip Fakul-tesi, Ic hastaliklari ABD Immunoloji-Romatoloji BD 35340Inciralti, Izmir, Turkey. Fax: 00-90-232-2792739; E-mail:[email protected]

Cardiac abnormalities may be a result of scle-rosing inflammatory process which primarilyinvolves the aortic root and the aortic valvecusps. This inflammation may cause aortic an-nular dilatation, cusp retraction, and aortic re-gurgitation. Furthermore, disease process mayalso extend into the ventricular septum andthe atrioventricular nodal blood supply, leadingto conduction disorders.3–5 Although the exactprevalence is not well defined, involvement ofmyocardium and pericardium has also been re-ported.3,5–7

Diastolic dysfunction (DD) is a relativelycommon problem that may be mild and

Vol. 25, No. 1, 2008 ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech. 47

OKAN, ET AL.

asymptomatic. When present in advancedforms, DD presents as diastolic heart failurewith symptoms similar to systolic heart fail-ure.8,9 Clinical significance of asymptomatic DDin the general population is not yet entirelyclear; however, some studies suggest it may beassociated with increased all cause and cardio-vascular (CV) mortality.9–11 DD has been re-ported to occur frequently in AS patients vary-ing from 20% to 49% in studies using standardDoppler echocardiography (SDE).6,7,12–14

In this study, we assessed diastolic functionof both ventricles in ankylosing spondylitis pa-tients and healthy controls with new echocar-diographic technologies such as pulsed wavetissue Doppler imaging (PW-TDI), myocardialperformance index (MPI), and TDI-derived MPIin addition to SDE.

Patients and Methods

Fifty-nine consecutive patients with anky-losing spondylitis who attended the rheuma-tology outpatient clinic of our hospital fromMarch 2004 to June 2004 were screened. Forty-nine patients who denied a previous history ofmyocardial infarction, diabetes mellitus (DM),and hypertension (HT) and who accepted toparticipate in the study were included. Allselected patients fulfilled the modified NewYork criteria for ankylosing spondylitis.15 Theyall had normal blood pressure (BP) and nor-mal oral glucose tolerance test (OGTT) at theinitial examination. All subjects were in si-nus rhythm. Echocardiography was performedon 49 patients (38 ± 11 years, range: 21–66;25 males/24 females) and 33 healthy controls(36 ± 9 years, range: 24–53; 17 males/16 fe-males) with a comparable age and sex distribu-tion who were recruited from volunteers fromhospital staff. Same exclusion criteria were ap-plied to controls as to patients.

Interview and Questionnaires

The following demographic and clinical datawere collected: sex, age, cigarette smoking, andhistory of previous and current disease. Cur-rent medications for all subjects and diseaseduration for patients were also recorded. Par-ticipants who reported smoking at least onecigarette per day during the year before the ex-amination were classified as smokers.

Laboratory Evaluation

In the morning, after an overnight fast,venous blood was sampled for the measure-ment of serum concentrations of glucose, totalcholesterol, high density lipoprotein, low den-sity lipoprotein, triglycerides, erythrocyte sed-imentation rate (ESR), and C-reactive protein(CRP) levels. Thereafter, OGTT was performedusing a glucose load containing 75 g anhydrousglucose dissolved in water.

Echocardiographic Procedures

In all participants; two-dimensional, M-Mode, SDE, and PW-TDI echocardiography ex-aminations were performed by the same physi-cian. A HP model Sonos 4500 (Hewlett Packard,Andover, MA) imaging system with S4 adult2–4 MHz phased array cardiac probe was usedfor examination. Cardiac dimensions were mea-sured according to the recommendations ofthe American Society of Echocardiography.16

Valvular regurgitation was assessed qualita-tively and graded none, trace, mild, moderate,or severe as described by Singh et al.17 Thetracings were recorded over five cardiac cy-cles at a sweep speed of 100 mm/s and storedon VHS videotapes for later playback anal-yses. Prerecorded data were evaluated afterthe echocardiographic procedure was complete.Prerecorded tapes were read by two experi-enced echocardiographers who were blindedto any clinical data of patients and controls.Among the recorded echocardiographic data theinvestigators averaged the three best consecu-tive cardiac cycles for all variables.

Doppler Measurement

The sample volume (size 2 mm) of the pulsedwave Doppler was placed between the tips ofthe mitral leaflets in the apical four-chamberview. The mitral inflow velocity was traced andthe following variables derived: peak velocityof early (E) and late (A) filling and decelera-tion time (DT) of the E wave velocity. The ratioof early to late peak velocities (E/A) was calcu-lated. Same parameters were obtained for rightventricle (RV) in the apical four-chamber view,placing the sample volume at the tips of the tri-cuspid valve.

Myocardial Performance Index

Doppler time intervals which are used todefine MPI (Tei index) and TDI-MPI are

48 ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech. Vol. 25, No. 1, 2008

VENTRICULAR DIASTOLIC FUNCTIONS OF ANKYLOSING SPONDYLITIS PATIENTS

ECG

LV/RV outflow

c

d

b

ET

IVCT IVRT

Mitral/Tricuspid inflow

a

EA

EA

MPI =(a-b)/b

Figure 1. Schema of Doppler flow velocity spectra repre-senting the time intervals used for calculation of the Tei in-dex. Interval a extends from the cessation to the onset ofmitral or tricuspid inflow. It includes the IVCT, ET, andIVRT. Interval b is the duration of left or right ventric-ular outflow (ET). The Tei index (MPI) is equal to (a −b)/b. IVRT is measured by subtracting interval d (betweenR-wave and cessation of LV or RV outflow) from intervalc (between R-wave and onset of mitral or tricuspid in-flow). IVCT is obtained by subtracting IVRT from (a − b).ET = ejection time; IVCT = isovolumic contraction time;IVRT = isovolumic relaxation time; MPI = myocardial per-formance index.

schematized in Figures 1 and 2, respectively.MPI was assessed by using both spectral andTDI methods for right and left ventricles.18,19

Tei index was calculated as isovolumetric relax-ation time (IVRT) plus isovolumetric contrac-tion time (IVCT) divided by the ejection time(ET): (IVCT + IVRT)/ET = (a − b)/b.

Tissue Doppler

The PW-TDI was performed by activating thetissue Doppler function in the same echocar-diography machine. In the apical four-chamberview, a 5 mm pulsed Doppler sample volumewas placed at the level of the lateral mitraland tricuspid valves. TDI velocity time inter-vals were measured from the sites at the mi-tral and tricuspid annulus as demonstrated inFigure 2.

ECG

PW-TDI

Am EmAm

Sm

IVCTm IVRTm

a'

b'

c'

d'

TDI-MPI=a'-b'

b'

Figure 2. Tissue Doppler echocardiography time intervalsmeasured from the lateral mitral and tricuspid valves. Nor-mal TDI patterns during sinus rhythm consist of threemajor signals: a single systolic signal (Sm) and two dis-tinct signals in early (Em) and late (Am) diastole. Myocar-dial early (Em) and atrial contraction (Am) peak velocities,Em/Am ratio, and myocardial relaxation time (IVRTm)(time interval between the end of Sm and the onset ofEm) were determined as diastolic measurements. MPI wascalculated from the TDI technique using the same for-mula: TDI − MPI = (a − b)/b = (IVCTm + IVRTm)/ET.IVCT = isovolumic contraction time; IVRT = isovolumic re-laxation time; MPI = myocardial performance index.

Definition of Diastolic Dysfunction

The following parameters were used to defineventricular diastolic function:

1. LV DD (conventional Doppler): E/A<50 years< 1, or E/A>50 years < 0.5, or IVRT<30 years> 92 ms, or IVRT30–50 years > 100 ms, orIVRT>50 years > 105.20

2. LV DD (PW-TDI): Em/Am < 1.21

3. RV DD (conventional Doppler): E/A < 1.4. RV DD (TDI): Em/Am < 1.

Various stages of DD are schematized inFigure 3).

Definition of Hypertension

BP was measured by using mercury sphyg-momanometer after 5 minutes of rest, in thesitting position. Two readings were taken halfan hour apart and averaged. HT was definedas a mean systolic BP ≥ 140 mmHg or mean


OKAN, ET AL.

EA E

A EA

EA

Em

Am Em

Am

Em

Am

EmAm

Normal Pseudo-normalization (Grade 2)

Restrictive pattern (Grade 3-4)

E>A

Em>Am

E<A

Em<Am

E>A

Em<Am

E>>A

Em>Am

Abnormal relaxation (Grade 1)

MVF

TDI

Figure 3. Schematic diagrams of thestages of the diastolic dysfunction ob-tained by using standard Dopplerechocardiography and tissue Dopplerimaging. The diagrams show typicalmitral valve flow (MVF) and tissueDoppler mitral annulus velocity (TDI)of the various stages of the diastolicdysfunction. A = mitral A-wave veloc-ity; Am = tissue Doppler atrial con-traction wave; E = mitral E-wavevelocity; Em = tissue Doppler early fill-ing wave.

diastolic BP ≥ 90 mmHg or if the subject wason antihypertensive medication.

Definition of Diabetes Mellitus

Current use of medications prescribed totreat DM or fasting serum glucose levels ≥126 mg/dL or second hour glucose duringOGTT ≥ 200 mg/dL.

Other Measurements

Body-mass index (BMI) was calculated.Spinal mobility was assessed by the Bath anky-losing spondylitis metrology index (BASMI)measurements.22 Patients were also evaluatedusing the Turkish version of the Bath anky-losing spondylitis functional index (BASFI)23,24

and the Bath ankylosing spondylitis disease ac-tivity index (BASDAI).25,26

Statistical Analysis

Results were presented as mean ± standarddeviation. In comparison to the continuous vari-ables student t-test was used. Differences be-tween categorical variables were analyzed bychi-square test. The relationships between leftventricle (LV) diastolic function and differentvariables were analyzed by the Pearson corre-lation test. Binary logistic regression was ap-plied to assess the risk of developing LV DD.The statistical analysis was carried out by usingStatistical Package of Social Science, version11.0 (SPSS Inc., Chicago, IL, USA). A p-valueof < 0.05 was considered as statistically signifi-cant.

Results

Clinical characteristics of the patient andcontrol groups are presented in Table I. Age,sex, BMI, smoking status, and serum lipids

TABLE I

Clinical and Laboratory Characteristics of the Subjects

Patients Controls(n = 49) (n = 33) P-value

Age (years) 38 ± 11 36 ± 9 0.4Male/female 25/24 17/16 1Height (cm) 163.6 ± 10 166 ± 9 0.3Weight (kg) 66.2 ± 13.4 69.3 ± 13.4 0.3BMI (kg/m2) 24.83 ± 5.1 25.14 ± 4.2 0.8Smoking (%) 31 27 0.1Systolic BP (mm

Hg)120.6 ± 12.8 116.5 ± 13.4 0.2

Diastolic BP (mmHg)

75 ± 7.9 74.3 ± 10 0.4

Fasting glucose(mg/dL)

89.1 ± 12.5 86.7 ± 11.4 0.4

Second hourglucose (mg/dL)

96 ± 31.8 96.4 ± 29.7 0.9

Total cholesterol(mg/dL)

174.9 ± 39.3 176.9 ± 37.9 0.8

LDL cholesterol(mg/dL)

103.2 ± 33.8 101.7 ± 37.3 0.8

HDL cholesterol(mg/dL)

50.9 ± 12.6 52.3 ± 10.9 0.6

Triglyceride(mg/dl)

103.4 ± 51 114.7 ± 52.1 0.3

ESR (mm/h) 22.6 ± 18.3 6.9 ± 3.9 <0.001CRP (mg/L) 21 ± 13.9 1.9 ± 1.6 <0.001

Data are means ± SD.BP = blood pressure; BMI = body-mass index; ESR = ery-throcyte sedimentation rate; CRP = c-reactive protein.



were comparable between the two groups (p >0.05). However, acute phase reactants (ESR andCRP) were significantly higher in the patientgroup than those of controls (22.6 ± 18.3 vs6.9 ± 3.9 mm/h and 21 ± 13.9 vs 1.9 ± 1.6 mg/L,respectively; p < 0.05). The mean disease dura-tion for patients was 12.4 ± 9.2 years. BASFIand BASDAI were 2.5 ± 2.5 and 3.5 ± 2.0 re-spectively. BASMI was 1.6 ± 2.3.

Cardiac dimensions and systolic functionparameters measured by M-mode echocardio-graphic examination in patients and controlsrevealed no difference (Table II). Valvular ab-normalities detected by color flow Dopplerechocardiography were as follows: mild mitralregurgitation (MR) in two patients and in onecontrol subject; mild MR with mitral valve pro-lepse in one patient; mild aortic regurgitationin one patient and one healthy control.

SDE data for LV revealed prolonged DT,shortened ET, and greater MPI, whereas for RVonly prolonged DT was significantly differentfrom controls (Table III). By SDE, four controlsubjects and six patients met the LV DD crite-ria by using conventional Doppler but the dif-ference was not significant (p = 0.98). However,using PW-TDI method, 22 patients and six con-trols had Em/Am ratio < 1 (p = 0.02). When onlythose patients and controls under the age of 50were analyzed, Em/Am ratio < 1 was present in16 out of 41 patients and three out of 28 controlsrespectively (p = 0.02). On the other hand E/Em

TABLE II

Cardiac Dimensions and Systolic Function ParametersMeasured by M-mode Echocardiographic Examination in

Patients and Controls


Aortic root (cm) 3.3 ± 0.3 3.2 ± 0.3 0.1Left atrial

dimension (cm) 3.7 ± 0.4 3.6 ± 0.4 0.2LVEDD (cm) 4.7 ± 0.4 4.6 ± 0.4 0.2LVESD (cm) 3.0 ± 0.4 2.9 ± 0.3 0.4RVD (cm) 2.4 ± 0.3 2.4 ± 0.2 0.6IVS (cm) 0.99 ± 0.1 0.97 ± 0.1 0.2EF (%) 65.0 ± 4.4 65.0 ± 3.9 0.9Number of valvular

abnormalities 4 2 0.9

Data are means ± SD.LVEDD = left ventricular end-diastolic dimension; LVESD= left ventricular end-systolic dimension; RVD = right ven-tricular dimension; IVS = interventricular septum; EF =ejection fraction.

ratio was similar between the two groups andthere were no patients with E/Em ratio greaterthan 15. Although in seven patients the E/Emratio was greater than 8, this value was lowerthan 8 in all healthy subjects (p = 0.03). TDIvelocity data in the LV and RV are summarizedin Table IV. The ones that were found to haveDD by SDE and PW-TDI combination were cat-egorized as grade I DD (6 patients vs 4 controls)and as grade II DD (pseudonormalized pattern)(16 patients vs 2 controls).

LV MPI was greater in patients with ankylos-ing spondylitis than in controls (0.47 ± 0.07 vs0.39 ± 0.04; p < 0.001). Similarly TDI-MPI forLV was also significantly increased in patientsthan in controls (0.47 ± 0.07 vs 0.39 ± 0.04;p < 0.001) (Table III and Table IV).

RV diastolic indices including E/A andEm/Am ratios, RV MPI and RV TDI MPI werenot different in patients and controls (p > 0.05).In addition, the number of subjects who hadE/A < 1 and Em/Am < 1 for RV were also com-parable in both groups (p > 0.05).

Correlation analysis revealed significantmoderate negative correlations betweenEm/Am and BASMI, age, and BMI (p <0.05; r = −0.3, −0.6, and −0.4, respectively)(Fig. 4). No correlation was observed betweenEm/Am and disease duration, BASFI, BASDAI,CRP, and ESR (p,>,0.05; r = −0.10, −0.11,0.15, −0.09, −0.04, respectively). The risk ofdeveloping LV DD was 3.7 times higher in

TABLE III

Standard Doppler Echocardiographic Data of the LeftVentricle


Mitral inflowPeak E velocity

(cm/s) 77.31 ± 15.26 79.17 ± 15.08 0.6Peak A velocity

(cm/s) 61.95 ± 17.12 59.68 ± 13.24 0.5Peak E/A ratio 1.30 ± 0.35 1.36 ± 0.27 0.5DT (ms) 212.76 ± 42.24 183.42 ± 22.48 <0.001IVCT (ms) 44.8 ± 8.71 41.6 ± 8.24 0.1IVRT (ms) 86.02 ± 11.77 78.30 ± 12.08 0.005ET (ms) 267.96 ± 24.68 292.06 ± 38.36 0.001MPI 0.49± 0.06 0.41 ± 0.04 <0.001

Data are means ± SD.ET = ejection time; DT = deceleration time; IVCT = isovo-lumic contraction time; IVRT = isovolumic relaxation time;MPI = myocardial performance index.


OKAN, ET AL.

TABLE IV

TDI Analysis of Left Ventricular Lateral Mitral Annulus


Mitral lateral annulusEm peak

velocity(cm/s)

15.91 ± 4.29 16.5 ± 3.8 0.5

Am peakvelocity(cm/s)

13.91 ± 3.38 12.2 ± 2.5 0.01∗

Peak Em /Amratio

1.21 ± 0.45 1.36 ± 0.27 0.05∗

Sm peak (m/s) 11.69 ± 2.32 11.76 ± 2.64 0.9IVRTm (ms) 89.92 ± 14.27 77.7 ± 9.93 <0.001∗IVCTm (ms) 47.55 ± 8.73 43.48 ± 7.34 0.03∗ETm (ms) 287.14 ± 19.98 306.15 ± 20.36 <0.001∗Em /Am <1 22 6 0.02∗E/Em 5.16 ± 1.5 4.96 ± 1.2 0.5MPI TDI 0.47 ± 0.07 0.39 ± 0.04 <0.001∗

Data are means ± SD.DTm = deceleration time; IVRTm = isovolumic relaxationtime; IVCTm = isovolumic contraction time; ETm =ejection time; MPI = myocardial performance index.∗All the significant p-values remained significant whenonly those patients and controls under the age of 50 wereanalyzed.

ankylosing spondylitis patients (OR = 3.7; 95%CI = 1.3–10.5; p = 0.01).

Discussion

We have found an increased rate of DD inankylosing spondylitis patients compared tohealthy controls with the application of recentlyintroduced echocardiographic techniques, butnot with conventional Doppler echocardiogra-phy.

Earlier studies have also shown impaireddiastolic function in ankylosing spondylitispatients varying from 20% to 49%.6,7,12–14,27

Disparities between studies may be due tothe methodological differences and limitationsof old echocardiographic methods to diagnoseDD. The first study using M-mode and two-dimensional echocardiography reported earlydiastolic abnormalities in 16 of 30 (49%) pa-tients indicated by increased mitral valve open-ing in milliseconds and increased interval fromaortic closure to mitral valve opening in mil-liseconds which were significantly differentfrom controls.6 These results were confirmedand extended by a subsequent study using SDEwhich found significantly shorter diastolic fill-

ing period, reduced velocity of early mitral in-flow, and lower ratios of early/late inflow ve-locities in ankylosing spondylitis patients con-sistent with impaired ventricular relaxationin some patients.12,14 In another study usingnuclear angiography, together with standardtwo-dimensional and M-mode echocardiogra-phy, global nuclide LV function was found tobe more significantly affected in ankylosingspondylitis patients compared to controls; how-ever, no differences in echocardiographic find-ings could be detected in the same study.27 Twoother studies using two-dimensional, M-mode,and Doppler echocardiography have concludedthat LV DD occurs frequently in patients withankylosing spondylitis. These two studies mea-sured very similar parameters, but DD seems tobe defined differently.7,13 A recent study whichassessed the rate of selected cardiac pathologiesincluding DD found a higher rate of DD in pa-tients with long-standing ankylosing spondyli-tis compared to controls, and speculated thatthe higher occurrence of DD in ankylosingspondylitis patients might be caused by thepresence of other CV risk factors such as ageand HT.28 Our patient group consisted of con-siderably younger patients (38 ± 11 years) withshorter disease duration (12 ± 9. years) com-pared to that study (55 ± 10 and 33 ± 10 years,respectively). Moreover, patients with possiblecontributing factors such as HT, DM, and priorhistory of CV diseases were excluded from ourstudy.29,30 Despite the younger age of the studypopulation and exclusion of patients with HT,we could still detect asymptomatic DD in almosthalf of the patients, significantly more commonthan the controls (OR: 3.7), which suggests thatdisease related factors may be responsible forDD.

Several mechanisms may be responsible forDD in ankylosing spondylitis. One very rea-sonable explanation is a diffuse increase inmyocardial interstitial connective tissue whichhas been shown in a small number of autopsycases.6 Amyloidosis, aortic insufficiency, con-duction disturbances, mitral valve disease, car-diomyopathy, and pericarditis,5,6,31,32 may alsocontribute to the impairment of LV diastolicfunction in different ankylosing spondylitis pa-tients.

Although invasive methods remain the goldstandard,20 echocardiography and Dopplertechniques as used in the above-mentionedstudies are widely used for assessing ventric-ular functions. However, these methods haveseveral limitations. One major limitation of



Figure 4. Correlations between Em/Am, age, body mass index and BASMI. The figure shows negative correlations betweenEm/Am, age, body mass index, and BASMI (p < 0.05; r = −0.6, −0.4, and −0.3, respectively).

conventional Doppler in assessing diastolic ven-tricular functions is its dependence of loadingconditions and heart rate.33 In case of worsen-ing LV diastolic function, there is a compen-satory increase in left atrial pressure result-ing in an increase in E-wave velocity of the mi-tral inflow and therefore pseudonormalizationof the filling pattern (normal E/A ratio and de-celeration time).34

TDI is a new technique which offers usefulinformation about ventricular diastolic func-tions.35,36 Myocardial or annular velocities areeasy to obtain, are load independent, and of-fer a rapid way to differentiate between nor-mal and pseudo-normal patterns and constric-tive and restrictive physiology.34,36–38 By usingSDE, we detected no significant difference be-tween AS patients and controls with regard tothe measurements made to evaluate ventricu-lar DD. However, by PW-TDI technique, 47% ofankylosing spondylitis patients had isolated LV

DD according to the criteria Em/Am < 1. In re-cent years another indices of diastolic function,E/Em ratio, is increasingly used.36 The majorstrength is that ratios below 8 and above 15 canconfidently separate normal from elevated fill-ing pressures. The most important limitationof this index is seen in patients with a valuebetween 8 and 15. In the present study we donot have any value higher than 15. However,in seven subjects the E/Em ratio was at the in-definite zone of 8 to 15 and strikingly all thosesubjects were ankylosing spondylitis patients.

In this study, LV MPI was also significantlyprolonged in ankylosing spondylitis patients byusing both spectral and PW-TDI methods. MPI,which is a relatively new Doppler-derived in-dex, was found to be a sensitive parameterof global cardiac function and related to mor-bidity and mortality in many CV diseases.18,39

This index combines systolic and diastolic in-tervals and has been used to assess global


OKAN, ET AL.

ventricular function in a noninvasive way.18

MPI can be combined with TDI and this combi-nation eliminates the disadvantages of conven-tional MPI by not being affected from loadingconditions and heart rate fluctuations.40 Thus,TDI-derived MPI serves more reliable globalcardiac performance.19,40

Conventional Doppler studies conducted forRV DD have the same limitations for LV, likeits dependence on preload, afterload, and heartrate.33 Furthermore, RV Doppler-derived di-astolic parameters are also affected from therespiration.41 Information concerning RV func-tions in ankylosing spondylitis is limited. Onestudy, which used conventional techniques, hasdemonstrated RV involvement.12 In this study,we could not detect any RV function involve-ment neither by conventional nor by relativelynew echocardiographic methods.

We found a negative correlation of Em/Amwith age and BMI in agreement with earlierstudies.9,42 Em/Am also showed a negative cor-relation with BASMI which is a composite indexreflecting spinal mobility and disease activity,but not with any other disease activity parame-ters. Ankylosing spondylitis patients are proneto develop progressive kyphotic deformity andit is known that severe kyphosis may cause pul-monary restriction.43 In this respect, the corre-lation between Em/Am which is an indicator ofLV DD and BASMI is of noteworthy. However,most of our study group had no or only mildspinal deformities as reflected by the low meanBASMI score of the study group.

The most commonly encountered valve dis-ease in ankylosing spondylitis patients are theaortic and mitral insufficiency.31,44 Valvular dis-ease, especially aortic stenosis, may contributeto the impairment of diastolic functions.45 Inour study mild valvular abnormalities were ob-served in only few subjects, both in the patientand control groups.

Clinical significance of asymptomatic (iso-lated) LV DD in ankylosing spondylitis popula-tion is unclear.5 However, it has been suggestedthat the presence of LV DD in asymptomaticpersons is a risk factor for the future develop-ment of congestive heart failure.29 Moreover, ithas been shown that asymptomatic LV DD isassociated with increased mortality.9,10 Someauthors have suggested an increased CV mor-tality in ankylosing spondylitis.46,47 Diastolicabnormalities in ankylosing spondylitis pa-tients may contribute to the reported in-creased CV mortality. There is growing evi-dence suggesting that accelerated atherosclero-

sis occurs in various inflammatory rheumaticdiseases including ankylosing spondylitis.32,48

Current evidence indicates that nonsteroidalanti-inflammatory drug (NSAID) treatmentwhich is the mainstay of therapy in ankylos-ing spondylitis may also contribute to the CVrisk. An increased CV risk due to usage of cox-ibs has now been demonstrated for several ofthese compounds. The risk of CV events re-lated to non-selective NSAIDs remains contro-versial,49 although a recent study has reportedincreased non-fatal myocardial infarction withsome conventional NSAIDs.50 Thus, it remainsto be shown how DD will add to the picture ofCV risk in ankylosing spondylitis patients.

One limitation of this study was that we didnot do any inter- and intraobserver variabil-ity analysis. Another important limitation isthat although we used controls with a similarmean age and gender distribution to patients,two groups were not completely matched for ageand sex and the patient group included subjectsolder by as much as 10 years than the controlgroup. However, when we analyzed only thosepatients and controls below the age of 50, all thedifferences between the two groups remainedstill significant (Table IV).

In conclusion, when sensitive echocardio-graphic Doppler techniques such as MPI, TDI-derived MPI, and PW-TDI are used, DD canbe detected in nearly half of the ankylosingspondylitis patients. This finding obtained in agroup of relatively young patients suggests thatDD may contribute CV mortality in ankylosingspondylitis.

References

1. Akkoc N, Khan MA: Epidemiology of ankylosingspondylitis and related spondyloarthropathies. InWeisman MH, Reveille JD, van der Heijde D (eds):Ankylosing Spondylitis and the Spondyloarthropathies.St. Louis, MO Mosby, 2006, pp. 117–131.

2. Khan MA: Update on spondyloarthropathies. Ann In-tern Med 2002;136(12):896–907.

3. Bergfeldt L: HLA-B27-associated cardiac disease. AnnIntern Med 1997;127(8 Pt 1):621–629.

4. Huffer LL, Furgerson JL: Aortic root dilatation withsinus of valsalva and coronary artery aneurysms as-sociated with ankylosing spondylitis. Tex Heart Inst J2006;33(1):70–73.

5. Lautermann D, Braun J: Ankylosing spondylitis–cardiac manifestations. Clin Exp Rheumatol 2002;20(6Suppl 28):S11–S15.

6. Brewerton DA, Gibson DG, Goddard DH, et al: Themyocardium in ankylosing spondylitis. A clinical,echocardiographic, and histopathological study. Lancet1987;1(8540):995–998.

7. Yildirir A, Aksoyek S, Calguneri M, et al: Echocardio-graphic evidence of cardiac involvement in ankylosingspondylitis. Clin Rheumatol 2002;21(2):129–134.



8. Gaasch WH: Diastolic dysfunction of the left ven-tricle: importance to the clinician. Adv Intern Med1990;35:311–340.

9. Redfield MM, Jacobsen SJ, Burnett JC Jr, et al: Burdenof systolic and diastolic ventricular dysfunction in thecommunity: Appreciating the scope of the heart failureepidemic. J Am Med Assoc 2003;289(2):194–202.

10. Bella JN, Palmieri V, Roman MJ, et al: Mitral ratio ofpeak early to late diastolic filling velocity as a predic-tor of mortality in middle-aged and elderly adults: theStrong Heart Study. Circulation 2002;105(16):1928–33.

11. Schillaci G, Pasqualini L, Verdecchia P, et al: Prog-nostic significance of left ventricular diastolic dys-function in essential hypertension. J Am Coll Cardiol2002;39(12):2005–2011.

12. Sun JP, Khan MA, Farhat AZ, et al: Alterations incardiac diastolic function in patients with ankylosingspondylitis. Int J Cardiol 1992;37(1):65–72.

13. Crowley JJ, Donnelly SM, Tobin M, et al: Dopplerechocardiographic evidence of left ventricular diastolicdysfunction in ankylosing spondylitis. Am J Cardiol1993;71(15):1337–1340.

14. Khan MA, Sun JP, Balher RC: Echocardiographic evi-dence of altered cardiac diastolic function in ankylosingspondylitis. Lancet 1987;1(8548):1501.

15. van der Linden S, Valkenburg HA, Cats A: Evaluationof diagnostic criteria for ankylosing spondylitis. A pro-posal for modification of the New York criteria. ArthritisRheum 1984;27(4):361–368.

16. Quinones MA, Otto CM, Stoddard M, et al: Recom-mendations for quantification of Doppler echocardiog-raphy: A report from the Doppler Quantification TaskForce of the Nomenclature and Standards Committeeof the American Society of Echocardiography. J Am SocEchocardiogr 2002;15(2):167–184.

17. Singh JP, Evans JC, Levy D, et al: Prevalence and clin-ical determinants of mitral, tricuspid, and aortic regur-gitation (the Framingham Heart Study). Am J Cardiol1999;83(6):897–902.

18. Tei C, Ling LH, Hodge DO, et al: New index of com-bined systolic and diastolic myocardial performance: Asimple and reproducible measure of cardiac function—astudy in normals and dilated cardiomyopathy. J Cardiol1995;26(6):357–366.

19. Tekten T, Onbasili AO, Ceyhan C, et al: Novel approachto measure myocardial performance index: Pulsed-wave tissue Doppler echocardiography. Echocardiogra-phy 2003;20(6):503–510.

20. European Study Group on Diastolic Heart Failure.How to diagnose diastolic heart failure. Eur Heart J1998;19(7):990–1003.

21. De Boeck BW, Cramer MJ, Oh JK, et al: Spectral pulsedtissue Doppler imaging in diastole: A tool to increaseour insight in and assessment of diastolic relaxation ofthe left ventricle. Am Heart J 2003;146(3):411–419.

22. Jenkinson TR, Mallorie PA, Whitelock HC, et al:Defining spinal mobility in ankylosing spondylitis(AS). The Bath AS metrology index. J Rheumatol1994;21(9):1694–1698.

23. Calin A, Garrett S, Whitelock H, et al: A new approachto defining functional ability in ankylosing spondyli-tis: The development of the Bath ankylosing spondylitisfunctional index. J Rheumatol 1994;21(12):2281–2285.

24. Karatepe AG, Akkoc Y, Akar S, et al: The Turkish ver-sions of the Bath ankylosing spondylitis and Dougadosfunctional indices: Reliability and validity. RheumatolInt 2005;25(8):612–618.

25. Garrett S, Jenkinson T, Kennedy LG, et al: Anew approach to defining disease status in ankylos-ing spondylitis: The Bath ankylosing spondylitis dis-ease activity index. J Rheumatol 1994;21(12):2286–2291.

26. Akkoc Y, Karatepe AG, Akar S, et al: A Turkish ver-sion of the Bath ankylosing spondylitis disease ac-tivity index: Reliability and validity. Rheumatol Int2005;25(4):280–284.

27. Gould BA, Turner J, Keeling DH, et al: Myocardialdysfunction in ankylosing spondylitis. Ann Rheum Dis1992;51(2):227–232.

28. Brunner F, Kunz A, Weber U, et al: Ankylosingspondylitis and heart abnormalities: Do cardiac con-duction disorders, valve regurgitation and diastolic dys-function occur more often in male patients with diag-nosed ankylosing spondylitis for over 15 years than inthe normal population? Clin Rheumatol 2006;25(1):24–29.

29. Gaasch WH, Zile MR: Left ventricular diastolic dys-function and diastolic heart failure. Annu Rev Med2004;55:373–394.

30. Poirier P, Bogaty P, Garneau C, et al: Diastolic dysfunc-tion in normotensive men with well-controlled type 2 di-abetes: Importance of maneuvers in echocardiographicscreening for preclinical diabetic cardiomyopathy. Dia-betes Care 2001;24(1):5–10.

31. O’Neill TW, Bresnihan B: The heart in ankylosingspondylitis. Ann Rheum Dis 1992;51(6):705–6.

32. Peters MJ, van der Horst-Bruinsma IE, DijkmansBA, et al: Cardiovascular risk profile of patientswith spondylarthropathies, particularly ankylosingspondylitis and psoriatic arthritis. Semin ArthritisRheum 2004;34(3):585–592.

33. Choong CY, Herrmann HC, Weyman AE, et al: Preloaddependence of Doppler-derived indexes of left ventric-ular diastolic function in humans. J Am Coll Cardiol1987;10(4):800–808.

34. Bruch C, Schmermund A, Bartel T, et al: Tissue Dopplerimaging: a new technique for assessment of pseudonor-malization of the mitral inflow pattern. Echocardiogra-phy 2000;17(6 Pt 1):539–546.

35. Sutherland GR, Bijnens B, McDicken WN: Tis-sue Doppler echocardiography: Historical perspectiveand technological considerations. Echocardiography1999;16(5):445–453.

36. Ommen SR, Nishimura RA, Appleton CP, et al: Clin-ical utility of Doppler echocardiography and tissueDoppler imaging in the estimation of left ventricularfilling pressures: A comparative simultaneous Doppler-catheterization study. Circulation 2000;102(15):1788–1794.

37. Wierzbowska-Drabik K, Drozdz J, Plewka M, et al: Theutility of pulsed tissue Doppler parameters for the diag-nosis of advanced left ventricular diastolic dysfunction.Echocardiography 2006;23(3):189–196.

38. Appleton CP, Hatle LK, Popp RL: Relation of transmi-tral flow velocity patterns to left ventricular diastolicfunction: New insights from a combined hemodynamicand Doppler echocardiographic study. J Am Coll Car-diol 1988;12(2):426–440.

39. Doin FL, Borges Mda R, Campos O, et al: Effectof central hypothyroidism on Doppler-derived my-ocardial performance index. J Am Soc Echocardiogr2004;17(6):622–629.

40. Su HM, Lin TH, Voon WC, et al: Differentiationof left ventricular diastolic dysfunction, identificationof pseudonormal/restrictive mitral inflow pattern and


OKAN, ET AL.

determination of left ventricular filling pressure by Teiindex obtained from tissue Doppler echocardiography.Echocardiography 2006;23(4):287–294.

41. Yu CM, Sanderson JE: Right and left ventricular di-astolic function in patients with and without heartfailure: Effect of age, sex, heart rate, and respira-tion on Doppler-derived measurements. Am Heart J1997;134(3):426–434.

42. Pascual M, Pascual DA, Soria F, et al: Effects of iso-lated obesity on systolic and diastolic left ventricularfunction. Heart 2003;89(10):1152–1156.

43. Morgan WC, Hodge HL: Diagnostic evaluation of dysp-nea. Am Fam Physician 1998;57(4):711–716.

44. Gupta BK, Panwar RB, Kabra PK, et al: Ankylos-ing spondylitis in association with mitral stenosis, mi-tral regurgitation, and aortic regurgitation: A case re-port and review of the literature. Echocardiography2003;20(3):275–277.

45. Mandinov L, Eberli FR, Seiler C, et al: Diastolic heartfailure. Cardiovasc Res 2000;45(4):813–825.

46. Radford EP, Doll R, Smith PG: Mortality among pa-tients with ankylosing spondylitis not given X-ray ther-apy. N Engl J Med 1977;297(11):572–576.

47. Lehtinen K: Mortality and causes of death in 398 pa-tients admitted to hospital with ankylosing spondylitis.Ann Rheum Dis 1993;52(3):174–176.

48. Hall FC, Dalbeth N: Disease modification and cardio-vascular risk reduction: two sides of the same coin?Rheumatology 2005;44(12):1473–1482.

49. Maillard M, Burnier M: Comparative cardiovascularsafety of traditional nonsteroidal anti-inflammatorydrugs. Expert Opin Drug Saf 2006;5(1):83–94.

50. Garcia Rodriguez LA, Gonzalez-Perez A: Long-term useof non-steroidal anti-inflammatory drugs and the risk ofmyocardial infarction in the general population. BMCMed 2005;3:17.


Ventricular Diastolic Functions of Ankylosing Spondylitis Patients by Using Conventional Pulsed?Wave Doppler, Myocardial Performance Index, and Tissue Doppler Imaging

Documents