Influence of intrauterine growth restriction on cardiac time intervals evaluated by fetal magnetocardiography

Early Human Development 74 (2003) 1 – 11 www.elsevier.com/locate/earlhumdev

Influence of intrauterine growth restriction on cardiac time intervals evaluated by fetal magnetocardiography Barbara Grimm a, Christiane Kaehler a,*, Ekkehard Schleussner a, Uwe Schneider a, Jens Haueisen b, Hans J. Seewald a a

Department of Obstetrics, University Hospital, Friedrich-Schiller-University, Bachstrasse 18 D-07743 Jena, Germany b Biomagnetic Centre, Department of Neurology, University Hospital, Friedrich-Schiller-University of Jena, Jena, Germany Accepted 14 April 2003

Abstract Objective: Differences in the cardiac excitation cycle between normotrophic and intrauterine growth-restricted fetuses were to be investigated by fetal magnetocardiography (fMCG). Study design: In this study, the time intervals of the fMCG signal in dependence on gestational age were compared between a group of 30 growth-restricted fetuses and 60 normotrophic fetuses by using Spearman’s correlation coefficient and two-way analyses of variance. Results: A significantly increasing duration of the P wave and the QRS complex could be observed with advancing gestational age in the normotrophic collective. This prolongation was not evident in the group of growthrestricted fetuses. The QRS complex showed a significant difference between both groups. In regard to the duration of the PR and the QT intervals, neither a distinct increase nor a clear difference between both groups was observable. Conclusion: In contrast to the observations in the normally grown fetuses, none of the cardiac time intervals in the group of the growth-restricted fetuses were significantly correlated with gestational age. More especially, the results of the QRS complex could be an indicator of the altered conditions when intrauterine life is complicated by intrauterine growth restriction (IUGR). D 2003 Elsevier Ireland Ltd. All rights reserved. Keywords: Fetal magnetocardiography; Intrauterine growth restriction; Cardiac excitation cycle

* Corresponding author. Tel.: +49-3641-933-409; fax: +49-3641-933-986. E-mail address: [email protected] (C. Kaehler). 0378-3782/$ - see front matter D 2003 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/S0378-3782(03)00079-3

2

B. Grimm et al. / Early Human Development 74 (2003) 1–11

1. Introduction Intrauterine growth restriction (IUGR) occurs in approximately 3 – 10% of all pregnancies and is associated with an increased incidence of perinatal morbidity and mortality [1,2]. Genetic factors, infectious diseases or metabolic disturbances of the mothers make for about 20 –30% of those cases in which fetuses appear to be growth restricted. The majority of 70 –80% of IUGR is caused by primary placental insufficiency, often associated with chronic intrauterine hypoxaemia [3]. The effect of IUGR on cardiac time intervals was mainly investigated in the fetal ECG. Pardi et al. [4] reported a reduced duration of the QRS complex in growthrestricted fetuses in comparison to normally grown fetuses. The influence of hypoxaemia during labour was associated with alterations of the PQ interval, the ST segment, and the T wave [5,6]. These measurements were performed invasively with a fetal scalp electrode after the rupture of the membranes. Recording the fetal electrocardiogram (fECG) from the maternal abdominal wall is burdened with considerable technical problems, e.g., a low signal-to-noise ratio or interference of the much larger field caused by maternal cardiac activity. Furthermore, the isolating effect of the vernix caseosa is disturbing the registration of the fECG between the 28th and 32nd week of gestation. The resulting waveforms may be considerably distorted [7,8]. Therefore, Pardi et al. [4] only demonstrate the development of the QRS complex in growth-restricted fetuses, but not in the whole excitation cycle. Fetal magnetocardiography (fMCG) is an alternative method used to investigate the fetal cardiac excitation cycle. The fMCG enables to precisely record both the physiological cardiac conduction pattern and the beat-to-beat intervals [9]. The fMCG signal has successfully been observed from the 13th week of gestation onwards [10]. The magnetic fields can be recorded noninvasively without direct skin contact and are less affected by the conductive characteristics of the tissues, including those of the vernix caseosa [7,8]. The latter is the major reason for its superior signal quality in comparison to the fECG. In the meantime, the fMCG is well established in the diagnosis of fetal arrhythmia [11 – 13]. Furthermore, the determination of the cardiac time intervals of healthy fetuses in dependence on gestational age has been described by several authors [14 – 18]. In this study, we used the advantages of the fMCG to investigate the effect of intrauterine growth restriction (IUGR) on cardiac time intervals (P wave, PR interval, QRS complex, and QT time) of the cardiac conductive cycle in comparison to normally grown fetuses. The cardiac time intervals are the expression of the duration of the spread of excitation within the cardiac muscle. Unless hypoxic injury or congenital abnormality distorts this spread of excitation, the cycle is supposed to be shortened in a heart that appears to be of smaller size [4– 6,13]. Hence, we hypothesize that IUGR is associated with reduced durations of the cardiac time intervals, in particular, those thought to represent the growth dynamics of myocardial mass.

B. Grimm et al. / Early Human Development 74 (2003) 1–11

3

2. Materials and methods Before starting the measurements, the study was approved by the ethics commission of the Friedrich-Schiller-University of Jena. All patients gave their written consent to measurement and the storage and analysis of the associated data. The study group consisted of 30 women pregnant with singletons showing intrauterine growth restriction. These subjects were compared to a control group of 60 women with fetuses showing adequate growth dynamics according to gestational age. Detailed information about the study and control groups is given in Table 1. Pregnancies with multiplets, pregnancies complicated by severe fetal or maternal pathology, under acute risk of preterm labour or other obstetric emergencies were primarily excluded from recruitment (for details, see Table 2). The IUGR fetuses were recruited on the basis of ultrasound findings (Toshiba SSA 270, 3.75 MHz abdominal transducer). The diagnosis of IUGR was assumed after estimation of the fetal weight according to Hadlock based on measurements of the head and abdominal circumferences and the femur length [19]. IUGR was diagnosed by an estimated weight below the fifth percentile with reference to Voigt et al. [20]. Only if the actual birth weight confirmed the diagnosis of IUGR, the data was included into further analysis. Gestational age (GA) at the time of the study ranged from the 27th to the 41st completed week postmenstruation (Table 3). This distribution was normalised in both groups by the casual recruitment of two normal subjects per case of IUGR according to GA. Each acquisition represents a different fetus. The fMCG was recorded at the Biomagnetic Centre of the Friedrich-Schiller-University of Jena inside a magnetically shielded room (AK3b, Vakuumschmelze Hanau). A 31channel SQUID biomagnetometer (Philips) consisting of first-order gradiometers was used. The measurements were taken either in supine or left side posture. Before starting the session, the fetal heart was located by ultrasound. For a better distinction of the maternal and the fetal heart signals, a single lead ECG of the mother was simultaneously recorded. The biomagnetometer was placed as close as possible over the mother’s abdomen without actually touching it directly over the fetal heart. The duration of the measurement was at least 2 min with a sampling rate of 1000 Hz. The hardware filters were set with a band pass from 0.3 to 500 Hz. The raw data was processed in order to remove overlying maternal activity and to obtain a high-quality averaged excitation cycle of the fetal heart. Methodological details and extensive tests of the performance of the underlying algorithms have been described

Table 1 Age, number of pregnancies, and parity (mean/S.D./minimum/maximum)

Age (years) Number of pregnancies Parity

Study group (IUGR), N = 30

Controls, N = 60

27.5/6.8/18/43 2.1/0.96/1/4 0.93/0.91/0/3

27.0/5.3/18/44 1.77/0.83/1/4 0.71/0.77/0/3

4

B. Grimm et al. / Early Human Development 74 (2003) 1–11

Table 2 Maternal and fetal exclusion criteria of the study Maternal

Fetal

Age < 18 years Consent denied or withdrawn Diabetes mellitus, gestational diabetes Medication with effect on the fetal heart (e.g., h-mimetics, corticosteroids)

fetal arrhythmias congenital heart defects somatic malformations chromosomal abberations fetal conditions, requiring immediate delivery or uninterrupted surveillance (e.g., abnormalities in the CTG)

Gemini, multiplets Unknown gestational age Premature contractions Any diagnoses requiring immediate delivery, bed rest or contraindicating transportation away from obstetric facilities (e.g., premature rupture of the membranes, placental complications/haemorrhage, pre-eclampsia)

elsewhere [21]. With reference to Schneider et al. [21], only a brief explanation of the processing steps will be given here. A secondary, namely, auxiliary, data set was received after smoothing the raw data by a Savitzky – Goolay filter in order to improve the signal-to-noise ratio prior to the application of a maximum coherence matching (MCM) algorithm on the basis of a representative QRS complex. MCM was then performed in the maternal ECG first to determine the time instants of all maternal heartbeats to average the maternal excitation cycles in each of the magnetic channels of the raw data set and, eventually, to subtract the maternal MCG from the raw data. MCM of the fetal signals was consecutively performed in the magnetic channel with the highest signal-to-noise ratio. Beat-to-beat variability was used for plausibility control. As described by Schneider et al. [21], an initial high-quality run of the algorithm, using a correlation threshold of 0.9 between template and matches, is predictive of stable QRS Table 3 Distribution of the gestational age of the growth retarded and normotrophic subjects Week of gestation

Growth-retarded fetuses, N = 30

Normotrophic fetuses, N = 60

27 29 30 34 35 36 37 38 39 40 41

1 1 1 3 5 2 8 4 3 1 1

2 2 2 6 10 4 16 8 6 2 2

B. Grimm et al. / Early Human Development 74 (2003) 1–11

5

Fig. 1. Fetal MCG: butterfly plot of all 31 channels. The cardiac time intervals: P wave, QRS complex, PR interval, and QT time were determined by the onset (first visible deflection from baseline) and offset (return to baseline) in the indicated way.

6

B. Grimm et al. / Early Human Development 74 (2003) 1–11

complexes throughout the measurement. To increase the number of matches further, allowances on the criteria are possible as long as beat-to-beat variability does not indicate false-positive matching of artefacts, usually within inter-beat intervals [21]. Furthermore, the choice for a baseline correction between an isomagnetic interval prior to the onset of the P wave or within the PQ interval does not make a measurable difference to the time intervals of the averaged fMCG [21]. Hence, both alternatives were used according to appropriateness [22]. The averaged fetal heart signals were visualised in a multichannel plot (so-called butterfly plot) of all 31 channels and the baseline corrected (Fig. 1). From the multichannel plot, the durations of the P wave and the QRS complex, the PQ interval, and the QT time were extracted (Fig. 1). To avoid biased observation, the data was blinded to the observer by encoding the numbers.

Fig. 2. Duration of the cardiac time intervals in dependence on gestational age of both groups: IUGR fetuses: —, normotrophic fetuses: 4 - - - (a) P wave, (b) QRS complex, (c) PR interval, (d) QT time. In normotrophic fetuses, the P wave and the QRS complex both showed a significant positive correlation between gestational age and duration. There was a significant difference between the durations of the QRS complexes between normotrophic and IUGR fetuses determined by two-way analyses of variance ( p = 0.009). The separated regression lines for each group are used for the purpose of illustration and for elucidating the development of the parameters in the course of pregnancy.

B. Grimm et al. / Early Human Development 74 (2003) 1–11

7

Two statistical analyses were performed by the use of SPSS 10.0 for Windows. Statistical significance was assumed for p < 0.05. 1. Each interval obtained was separately correlated to the gestational age for both the IUGR fetuses and the control group by the calculation of the Spearman’s correlation coefficient. 2. The differences of the cardiac time intervals in dependence on the gestational age between IUGR fetuses and normotrophic fetuses were examined by means of two-ways analyses of variance. Basic requirement to the application of this method is an initial testing for equal variances on the basis of a nonsignificant result in the Levene test (refusal of the assumption of differences on a p-value >0.05).

3. Results Depending on the duration of the measurements and the gestational age, at least 150 fetal heart signals were averaged. The detection rates of the P wave and the QRS complex reached 100%, whereas the T wave was recognizable without doubt in 93% of the hypotrophic and 96% of the normotrophic fetuses. The development of the cardiac time intervals in dependence on gestational age in both groups is shown in Fig. 2. The healthy fetuses showed a significant correlation between gestational age and the increase in the length of the P wave as well as that of the QRS complex (Fig. 2a –b). In contrast to that, the growth-restricted fetuses showed neither a significant correlation of the P wave nor of the QRS duration in dependence on gestational age (Table 4). Nevertheless, there was an insignificant increase observable regarding the QRS duration and the P wave (Fig. 2a– b). In contrast to the P wave, the difference in the QRS durations between normotrophic and hypotrophic fetuses was statistically significant ( p = 0.009). The PQ interval and the QT time in both groups did not show a clear prolongation, and there was no significant correlation with gestational age (Fig. 2c– d). Concerning these two intervals, no differences between normotrophic fetuses and IUGR fetuses could be observed. Table 4 Correlation between cardiac time intervals and gestation age Parameter

Growth-restricted fetuses, N = 30

Normotrophic fetuses N = 60

Correlation coefficient

Significance

Correlation coefficient

Significance

P wave PQ interval QRS complex QT time

0.078 0.102 0.164 0.111

0.340 0.141 0.193 0.286

0.222 0.041 0.318 0.065

0.044* 0.266 0.007* 0.314

Correlation coefficients (Spearman) and level of significance (*p < 0.05) of the P wave, PQ interval, QRS complex, and QT time in growth-restricted and normotrophic fetuses.

8

B. Grimm et al. / Early Human Development 74 (2003) 1–11

4. Discussion This study was performed to investigate the influence of intrauterine growth restriction in contrast to normal growth dynamics regarding the cardiac time intervals in fMCG. The earliest registration of the fMCG of an IUGR fetus was performed during the 27th week of gestation. Most cases were diagnosed from the 34th week onwards. The very high detection rates of the P wave and the T wave are likely to be caused by the better quality of the magnetic signal in the latter stages of pregnancy. Other authors reported increasing detection rates in dependence on gestational age [16,17]. In agreement with other investigations, we found a positive correlation between QRS duration and gestational age in the normotrophic collective [14 – 18]. Brambati and Pardi [23] found similar results by using fECG. The prolongation of the QRS complex is considered to reflect the growth and maturation of the fetal heart during pregnancy as well as the increasing number of muscle fibres [14]. In addition to the QRS duration, the P wave of the normotrophic fetuses showed a statistically significant correlation with gestational age. This prolongation reflects the growth of the muscle mass of the atria. Here, our data confirmed the results of other authors [4,14,18]. Whereas several authors investigated the development of the cardiac time intervals in uncomplicated pregnancies by fetal magnetocardiography, only a few studies have been performed in growth-restricted fetuses. Our study has determined some differences in comparison to normotrophic fetuses. In the case of IUGR, we found a low but not significant ascent of the P wave and QRS duration. Van Leeuwen et al. [24] showed a significantly shorter duration of the P wave, QRS complex, PQ and PQRS interval in IUGR fetuses in comparison to normotrophic fetuses in a smaller number of subjects. Our data confirm these results only with regard to the QRS duration. In agreement with this observation, Pardi et al. [4] reported a reduced QRS duration in growth-restricted fetuses in comparison to normotrophic fetuses by the use of fECG. In the case of intrauterine growth restriction, the weight of the fetal heart is significantly reduced [4]. The reduced myocardial mass is likely to be the reason for the lower duration of the QRS complex [24]. By utilising the fECG, Brambati and Pardi [23] observed a significant relation between birth weight and QRS duration and confirmed the particular influence of fetal weight on QRS duration. The small prolongation of the PR interval in the course of pregnancy was reported to be caused by the increased duration of the P wave [11]. In contrast to earlier investigations, in our study, the prolongation of the PR interval turned out to be not significant [15,17]. The PR interval is assumed to be influenced by various factors such as the fetal heart rate or hypoxaemia [6]. Our data suggest that these factors are to be taken into consideration in the interpretation of the relations between the absolute PR interval and the gestational age. The exact determination of the length rather than the pure detection of the T wave remains a particular source of doubt concerning fMCG studies, which may explain some of the inconsistencies in the results of the QT interval [8,17]. Nevertheless, we found an

B. Grimm et al. / Early Human Development 74 (2003) 1–11

9

insignificant increase of the QT duration during pregnancy which is likely to be caused by the rising QRS duration. The PR interval is the expression of the atrioventricular conduction time. The QT interval reflects the time interval of the de- and repolarization of the ventricles. Regarding these two time intervals, we failed to find significant difference between both groups. Apart from a lower muscular mass, changes of the atrial or ventricular conduction system are not necessarily associated with IUGR [6]. Summarizing the results, we could demonstrate the correlation between the cardiac time intervals to be associated with intramuscular conduction and, hence, muscular mass of the growing heart and gestational age in normotrophic fetuses. Although the QRS complex of the hypotrophic fetuses was significantly different from that of the normotrophic fetuses, we failed to find a statistically significant correlation between any of the cardiac time intervals and gestational age. This is largely due to the fact that IUGR fetuses are a very heterogeneous group in terms of aetiology and severity of the growth restriction. Furthermore, the individual compensatory ability of the fetal organism to the interaction of these factors (e.g., hypoxaemia) is influencing the development of the different organs. In the case of hypoxaemia, the fetus initially shows a compensating redistribution of blood flow (brain-sparing effect). A vasodilatation at the level of the cerebral vessels, the coronary arteries, adrenal arteries, and the splenic artery is accompanied by somatic peripheral vasoconstriction. The redistribution results in an increased preload and after load of the developing heart followed by myocardial hypertrophy [25]. The imminent severity of hypoxaemia is obviously difficult to determine in a relatively crude study design as there are no clinically apparent cutoff levels available in the individual case [26]. Hence, the study group has to be considered inconsistent in terms of necessity and ability to compensate the effects of chronic hypoxaemia, and this is likely to be mirrored in the widespread distribution of the observed parameters. Secondly, it has to be borne in mind that only a small number of fetuses presenting with IUGR before 30 weeks GA are available in a study beyond the boundaries of the maternity unit. Single results in this subgroup as well as the different numbers of fetuses in both groups may have severely influenced the overall results. In normally growing fetuses, the cardiac time intervals can be considered as reasonable markers for the dynamics of cardiac growth. In fetuses presenting with low estimated weight on ultrasound examination, growth dynamics, developmental restriction, compensating necessity, and capabilities overlay each other and need to be dismantled in future studies in order to characterise the confounding parameters of development and maturation of the fetal heart.

Acknowledgements The authors would like to acknowledge Marcus Huck, MSc for programming the software. Wolfgang Michels, PhD, Department of Obstetrics, FSU Jena and Heike Hoyer, PhD, Institute of Medical Statistics assisted in performing statistical analyses. Uwe Schulze, Biomagnetic Centre, Deparment of Neurology, FSU Jena, and Sylvia

10

B. Grimm et al. / Early Human Development 74 (2003) 1–11

Bausch, Department of Obstetrics, FSU Jena participated in the performance of the measurements.

References [1] Vandenbosche RC, Kirchner JT. Intrauterine growth retardation. Am Fam Physician 1998;58:1384 – 90. [2] Severi FM, Rizzo G, Bocchi C, et al. Intrauterine growth retardation and fetal cardiac function. Fetal Diagn Ther 2000;15:8 – 19. [3] Doubilet PM, Benson CB. Sonographic evaluation of intrauterine growth retardation. AJR 1995;164: 709 – 17. [4] Pardi G, Marconi A, Ferrazzi E. The intraventricular conduction time of fetal heart in pregnancies with suspected fetal growth retardation. Br J Obstet Gynaecol 1986;93:250 – 4. [5] Jenkins HML, Symonds EM, Kirk DL, Smith PR. Can fetal electrocardiography improve the prediction of intrapartum fetal acidosis? BJOG 1986;93:6 – 12. [6] Luzetti R, Erkkola R, Hasbargen U, Mattson LA, Thoulon JM, Rosen KG. European community multicentre trial ‘‘Fetal ECG Analysis During Labour’’: the P-R interval. J Perinat Med 1997;25:27 – 35. [7] Oostendorp TF, van Oosterom A. Modelling the fetal magnetocardiogram. Clin Phys Physiol Meas 1991;12(Suppl. A):15 – 8. [8] Stinstra JG, Peters MJ. The volume conductor may act as a temporal filter on the ECG and EEG. Med Biol Eng Comp 1998;36:711 – 6. [9] Van Leeuwen P, Schuessler M, Bettermann H, Lange S, Hatzmann W. Magnetokardiographie zur Erfassung fetaler Herzaktionen [Magnetocardiography for assessment of fetal heart actions]. Geburtsh Frauenheilk 1995;55:642 – 6. [10] Seppenwolde Y, Broeck SP, Peters MJ. Volume conduction effects: the fetal magnetocardiogram. Biomed Tech 1997;42(Suppl. 1):261 – 4. [11] Achenbach S, Moshage W, Menendez T, Beinder E, Bittl A, Hofbeck M, et al. Investigation of fetal cardiac arrhythmias by magnetocardiography. Biomed Tech 1997;42(Suppl. 1):253 – 5. [12] Van Leeuwen P, Hailer B, Bader W, Geissler J, Trowitsch E, Groenemeyer D. Magnetocardiography in the diagnosis of fetal arrhythmia. Br J Obstet Gynecol 1999;106:1200 – 8. [13] Kaehler C, Grimm B, Schleussner E, Schneider A, Schneider U, Nowak H, et al. The application of fetal magnetocardiography (fMCG) to investigate fetal arrhythmias and congenital heart defects (CHD). Prenat Diagn 2001;21:176 – 82. [14] Horigome H, Takahashi MI, Asaka M, Shigemitsu S, Kandori A, Tsukada K. Magnetocardiographic determination of the developmental changes in PQ, QRS and QT intervals in the foetus. Acta Paediatr 2000;89: 64 – 7. [15] Leuthold A, Wakai RT, Martin CB. Noninvasive in utero assessment of PR and QRS intervals from the fetal magnetocardiogram. Early Hum Dev 1999;54:235 – 43. [16] Menendez T, Achenbach S, Moshage W, Flug M, Beinder E, Kollert A, et al. Praenatale Registrierung fetaler Herzaktionen mit Magnetokardiographie [Prenatal recording of fetal heart action with magnetocardiography]. Z Kardiol 1998;87:111 – 8. [17] Quinn A, Weir A, Shahani U, Rhoderick B, Maas P, Donaldson G. Antenatal fetal magnetocardiography: a new method for fetal surveillance? Br J Obstet Gynecol 1994;101:866 – 70. [18] Kaehler C, Schleussner E, Grimm B, Schneider A, Schneider U, Seewald HJ. Fetal magnetocardiography: development of the fetal cardiac time intervals. Prenat Diagn 2002;22:408 – 14. [19] Hadlock FP. Sonographic estimation of fetal age and weight. Radiol Clin North Am 1990;28:39 – 50. [20] Voigt M, Schneider KT, Jhrig K. Analyse des Geburtengutes des Jahrgangs 1992 der Bundesrepublik Deutschland. Teil 1: Neue Perzentilenwerte fu¨r die Ko¨rpermasse von Neugeborenen [Analysis of a 1992 birth sample in Germany: 1. New percentile values of the body weight of newborn infants]. Geburtsh Frauenheilk 1996;56:550 – 8. [21] Schneider U, Schleussner E, Haueisen J, Nowak H, Seewald HJ. Signal analysis of auditory evoked cortical fields in fetal magnetencephalography. Brain Topogr 2001;14:64 – 80. [22] Leder U, Unger R, Baier V, Haueisen J, Nowak H, Figulla HR. Einfluß der Wahl des Basislinien-Korrek-

B. Grimm et al. / Early Human Development 74 (2003) 1–11

[23] [24]

[25] [26]

11

turintervalls auf die Lokalisation der elektrischen Herzaktivita¨t [Influence of the Selection of the Baseline Correction Interval on the Localisation of the Electrical Heart Activity]. Biomed Tech 2000;45:114 – 8. Brambati B, Pardi G. The intraventricular conduction time of fetal heart in uncomplicated pregnancies. Br J Obstet Gynocol 1980;87:941 – 8. Van Leeuwen P, Lange S, Hackmann J, Klein A, Hatzmann W, Gro¨nemeyer D. Assessment of intra-uterine growth retardation by fetal magnetocardiography. In: Nenonen J, Ilmoniemi RJ, Katila T, editors. BIOMAG 2000. Proc 12th int conf on biomagnetism 2001, Helsinki Univ Technology, Espoo, Finland; 2001. p. 603. Hecher K. Dopper velocimetry in high-risk pregnancy: fetal hypoxemia. Textbook of perinatal medicine. London: The Parthenon Publishing Group; 1998. p. 441 – 46. Carrera JM, Deves R, Torrents M, Munoz A, Comas C, Mortera C, et al. Natural history of fetal compromise in intrauterine growth retardation. In: Kurjak A, editor. Textbook of perinatal medicine. London: The Partenon Publishing Group; 1998. p. 1226 – 43.