MAGNETIC RESONANCE

Improvement of MR cholangiopancreatography (MRCP)images after black tea consumption

Hossein Ghanaati & Hadi Rokni-Yazdi &Amir Hossein Jalali & Firouze Abahashemi &Madjid Shakiba & Kavous Firouznia

Received: 26 April 2011 /Revised: 11 June 2011 /Accepted: 2 July 2011 /Published online: 5 August 2011# European Society of Radiology 2011

AbstractObjective Evaluation of the efficacy of black tea as thenegative oral contrast agent in MRCP.Materials and methods MRCP was performed before and5 and 15 min after tea consumption for 35 patients.Depiction of the gall bladder (GB), cystic duct (CD),proximal and distal parts of the common bile duct (CBD),intrahepatic ducts (IHD), ampulla of vater (AV), mainpancreatic duct (MPD) and signal loss of stomach andthree different segments of the duodenum were investigatedaccording to VAS and Likert scores.Results Twenty-one of the patients (60%) were female(mean age, 50.319.2 years). Regarding visibility ofdifferent anatomical parts of the pancreatobiliary tree, thepost procedure images were better visualized in the distalpart of CBD, AV and MPD in Likert and VAS scoring (allP0.001). Regarding obliteration of high signal in thestomach and three different parts of the duodenum, all postprocedure images showed significant disappearance of highsignal in Likert and VAS scoring systems (all Ps0.001).Conclusion Black tea is a simple and safe negative oralcontrast agent which reduces the signal intensity ofgastrointestinal tract fluid and provides improved depictionof the MPD, the distal CBD and the ampulla during MRCP.

Key Points Tea is an effective negative oral contrast agent forgastrointestinal MRI

Ingestion of black tea improves conspicuity of the distalCBD in MRCP

Keywords Cholangiopancreatography .Magneticresonance . Contrast media . Biliary tract . Gastrointestinaltract . Tea

Introduction

Magnetic resonance cholangiopancreaticography (MRCP)is a non-invasive imaging investigation for the evaluationof the pancreaticobiliary system [14]. The magneticresonance techniques applied for the pancreaticobiliarytract are mainly based on heavily T2-weighed sequences[5]. Fluid in the stomach and the duodenum may obscurethe pancreatic and biliary ducts and may interfere with theirdepiction or may mimic certain lesions [5, 6].

We can overcome this problem by asking the patients to fastbefore the examination or by using multiple acquisitions ofsequence in different planes [68]. Unfortunately, sometimesdespite using these methods, unwanted signal may not beeliminated.

Oral negative contrast agents which have a highamount of high molecular metal ion such as manganeseand iron are used to shorten the T2 relaxation time andalso to decrease the T2 signal from gastric and intestinefluid [9, 10].

Among the large number of negative oral contrast agentswhich have been used such as blueberry and pineapplejuice, some are expensive and not available in all countries.

H. Ghanaati :H. Rokni-Yazdi :K. Firouznia (*)Radiology, Advanced Diagnostic and Interventional RadiologyResearch Center (ADIR), Imam Khomeini Hospital,Tehran University of Medical Sciences,Tehran, Irane-mail: k_firouznia@yahoo.com

A. H. Jalali : F. Abahashemi :M. ShakibaAdvanced Diagnostic and Interventional Radiology ResearchCenter (ADIR), Imam Khomeini Hospital,Tehran University of Medical Sciences,Tehran, Iran

Eur Radiol (2011) 21:25512557DOI 10.1007/s00330-011-2217-0

Indeed, ingestion of a large amount of some of theseagents such as blueberry is difficult.

The first publication on the use of tea as a negative oralcontrast agent was by Varavithya et al. who used Rosellaflower tea [11].

Tea, whether green or black, made from camellia sinensisleaves is the commonest drink after water in the world [12].

Because of the natural quality of this drink, the highamount of mineral content, especially manganese [13], andalso its low price we propose black tea as a negative oralcontrast agent for MRCP.

The purpose of this study was to investigate the use ofblack tea for the first time as a negative oral contrast agentfor MRCP studies.

Materials and methods

Because of the medium level of manganese in black tea, 3tea-bags of non-flavored tea were soaked in 300 mL ofboiled water for l0 min without further heating to make thetea infusion. We made different tea infusion samples withnon-flavored tea-bags of different trade marks to assesswhether tea brand has an impact on black tea signalintensity. Then we added 40 g of sugar to each sample tomake them more pleasant and tolerable. Before volunteerstudy we examined 15 mL of each tea infusion using 1.5Tesla MR (Signa, echo-speed, General Electric, Milwaukee,USA) with routine MRCP sequences: body coil, SingleShot FSE, TR/TE: 3,000/800, 3,500/900,4,000/800.

The signal intensity of samples were measured by a circularROI (Area: 1 cm) put on the area of each sample. Signalintensity of the air was measured similarly and assumed asnoise (loss of signal) and the samples with a signal intensityequal to noise level were assumed as signal void or loss ofsignal.

Volunteer examination

Three healthy volunteers were evaluated according to theabove mentioned protocol.

Initially, Axial T2-weighted slices through the liver andpancreas were performed to survey the liver and to find theCBD location. Then the MR examination was continuedwith routine MRCP protocols before and after 5 and 15 miningestion of 300 mL of tea infusion using the previouslymentioned MR system. MRCP protocols included:

Twelve radial slabs of 40 and 20 mm thickness with10 inter-slab angle.

coronal oblique 20 mm thickness with 50% over-lapping slabs at the best angle to see the CBD andampulla of vater.

The study parameters were as follows: Torso coil,

AX.T2: frFSE; TR/TE: Automatic selection/90, FoV:matched with patients size, matrix: 256160

12 Radial 40 mm/10 slabs, SSFSE; TR/TE: 3,500/900,matrix: 256256, FoV: 34 cm

12 radial 20 mm/10 slabs, SSFSE; TR/TE: 3,500/900,matrix: 224224, FoV: 34 cm

Coronal oblique 20 mm overlapped slabs: SSFSE; TR/TE: 3,500/900, matrix: 224224 FoV: 34 cm

The volunteers were questioned about the tolerability oftaste and the volume of tea infusion they had no problemabout. No side-effects were observed during and afterexamination.

Image quality investigation for volunteer examinations

The images were checked by an expert radiologist. Thesignal intensity in the stomach and duodenum wereinvestigated in the images obtained before and 5 and15 min after tea consumption. Signal loss was observed inthese areas after tea ingestion.

The signal intensity of samples were measured by e-filmworkstation 2.1 program and with circular ROI (Area:1 cm) put on the area of each sample.

ROI diameter for large organs such as liver, CBD, etc.and for air (as noise) was 1 cm but for thin and smallorgans such as pancreatic duct,cystic duct, etc. the ROI sizewas defined according to the size of organ to include onlythe organ, not the tissue around it.

Patient examination

Patients who were referred to our university affiliatedhospital for MRCP underwent examination after receivingthe informed consent and were ordered to keep a non-fattydiet 14 h before the examination and to fast on the day ofexamination.

Patients with ascites and diabetes mellitus were excludedfrom this study. The permission for the study and teaconsumption by the patients was confirmed by the EthicsCommittee of our university affiliated hospital. The MRsystem and the protocols were the same for volunteers and200300 mL tea infusion was consumed by the patientsaccording to their tolerability. Additional coronal oblique45 mm slices with the same angle of coronal oblique20 mm overlapped slices were performed after 15 minprotocols for better evaluation of CBD with the followingparameters: SSFSE, TR/TE: 3,500/900, Matrix: 224224,FoV: 34 cm.

We ordered 300 mL of tea infusion for all patients ofwhich some did not tolerate this volume, so we asked thesepatients to use at least 200 mL of tea and to use extra

2552 Eur Radiol (2011) 21:25512557

volumes of tea up to the mentioned 300 mL. None of themhad a problem with 200 mL tea mixture.

Image quality investigation for patient examinations

Images were checked and scored by an expert radiolo-gist. The visibility of gall bladder (GB), cystic duct(CD), proximal and distal parts of the common bile duct(CBD), intra hepatic ducts (IHD), ampulla of Vater (AV),main pancreatic duct (MPD) and signal loss of thestomach and three different segments of the duodenumhad been investigated before and 5 and 15 min after teaingestion.

Two scales: Likert [14] and visual analogue scale (VAS)[15] were used to score the value of visibility.

The Likert scale is defined as follows:

A) Visibility and detectability of pancreato-biliary system:

1 Poor: completely not visible; 2- Moderate: difficultto detect the anatomy; 3- Good: the anatomy isvisible but with some difficulty; 4- Excellent:completely visible.

Table 1 Visibility of different biliary tree parts according to Likertand VAS scores for before and 5 and 15 min after tea consumption

Different imaging sessions regardingtea ingestion in terms ofanatomical parts

Assessmentmethod

Mean SD

GB before. Likert 3.60.5

VAS 85.913.7

GB 5 min after Likert 3.60.6

VAS 86.515.8

GB 15 min after Likert 3.60.5

VAS 85.913.7

CD before Likert 2.70.9

VAS 61.718.3

CD 5 min after Likert 2.70.8

VAS 62.515.9

CD 15 min after Likert 2.80.8

VAS 66.517.5

Prox. CBD before Likert 3.80.5

VAS 89.113.4

Prox. CBD 5 min after Likert 3.80.5

VAS 89.412.8

Prox. CBD 15 min after Likert 3.80.5

VAS 89.413.0

CHD before Likert 3.60.8

VAS 87.420.3

CHD 5 min after Likert 3.70.8

VAS 88.319.5

CHD 15 min after Likert 3.60.9

VAS 86.820.7

IHD before Likert 2.90.7

VAS 69.414.1

IHD 5 min after Likert 2.90.7

VAS 69.714.6

IHD 15 min after Likert 30.7

VAS 70.314.5

Dis. CBD & AMP before Likert 2.10.9

VAS 48.321.9

Dis. CBD & AMP 5 min after Likert 2.31

VAS 53.422.7

Dis. CBD & AMP 15 min after Likert 2.41

VAS 54.424.0

MPD before Likert 2.31.1

VAS 51.426.0

MPD 5 min after Likert 2.61.1

VAS 57.425.1

MPD 15 min after Likert 2.71.1

VAS 59.726.6

Prox proximal, Dis distal

Table 2 Obliteration of bright signal of different gastrointestinal partsaccording to Likert and VAS scores for before and 5 and 15 min aftertea consumption

Different imaging sessions regardingtea ingestion in terms ofanatomical parts

Assessmentmethod

Mean SD

Stomach before Likert 1.70.6

VAS 38.318.7

Stomach 5 min after Likert 3.70.5

VAS 86.613.7

Stomach 15 min after Likert 3.80.4

VAS 90.612

First part of duodenum before Likert 1.40.7

VAS 3019.6

First part of duodenum 5 min after Likert 3.60.8

VAS 85.420.5

First part of duodenum 15 min after Likert 3.80.7

VAS 90.316.6

Second part of duodenum before Likert 1.30.7

VAS 26.320

Second part of duodenum 5 min after Likert 3.10.9

VAS 73.421.1

Second part of duodenum 15 min after Likert 3.10.9

VAS 72.619.4

Third part of duodenum before Likert 2.10.7

VAS 49.723.1

Third part of duodenum 5 min after Likert 3.60.7

VAS 86.619.1

Third part of duodenum 15 min after Likert 3.70.7

VAS 89.118.9

Eur Radiol (2011) 21:25512557 2553

B) Visibility and detectability of the GI system :

1 Poor: completely visible; 2-Moderate: visible but thesignal is low; 3-Good: little visibility; 4-Excellent:completely not visible

All SNR measurements were acquired using SSFSE;TR/TE: 3,500/900 sequences, 12 radial 20 mm/10 slabs.Although we know the liver parenchyma in thesesequences is almost signal free, we still acquired themeasuremens of signal for all organs in the samesequence. In the duodenum the signal of its contentwas measured, as also done for measurement of thesignal in stomach.

The formula for calculating SNR was: [mean of SI ineach region]/[standard deviation of air SI].

The formula for calculating CNR was [mean of SI ineach region- mean of SI in adjacent tissue]/[standarddeviation of air SI].

SPSS program version 11.5 was used for statisticalanalyses. Difference of all three Likert measurementsbefore and after tea consumption was assessed by theFriedman test. For VAS measurements, at first we evaluatedthe normality of the data; if the data had normaldistribution, we used the repeated measure ANOVA;otherwise, the data were analyzed by the Friedman test.

A P-value less than 0.05 was considered statisticallysignificant.

Results

Thirty five patients of which 21 were female (%60) and 14were male (%40) were enrolled into the study. The meanage of the patients was 50.319.2 years [women, 54.816.9 years; men 43.522.4 years. (P=0.13)].

Black tea was tolerated well by all patients and there wasno sign of nausea, vomiting, diarrhea or abdominal pain.

All tea infusions made by different trade mark tea-bagsmade the same result and were signal void.

Comparison of the visibility of the mentioned parts ofthe billiary tree by Likert method showed a statisticallysignificant difference of scores before and after teaconsumption in the distal part of the CBD and the ampullaof Vater [P=0.001], and the main pancreatic duct [P

the fact that none of the variables had normal distribution,all pairwise comparisons (one pre and two post interventiongroups) were performed by the Wilcoxon signed rank testwith Bonferroni correction.

As demonstrated, the statistically significant variableswere visibility of the distal part of the CBD and ampulla ofVater and the main pancreatic duct and also signalobliteration in the stomach and three parts of the duode-num. Pairwise comparisons of the three pairs (before versus5 min after ingestion, before versus 15 min after ingestion,and 5 min after versus 15 min after ingestion) in all the sixmentioned variables showed the pre-consumption imageswere statistically different compared to each of the twopost-consumption images, but there was no statisticallysignificant difference between the two post-consumptionimages. This pattern was similar for all the assessmentswith Likert and VAS methods. (all P-values

Discussion

MRCP is a useful noninvasive imaging technique intro-duced in 1991 for morphological evaluation of the biliarytract and pancreatic duct [14].

This technique is based on heavily T2-Weighted sequencesto enhance signal from fluid.

High signal intensity from the stomach and intestinalfluid may obscure the MRCP images because it super-imposes on the biliary tract. The signal from the GI tract isespecially problematic when single thickness slice imagesare obtained without a thin multislice data set [16]. Fastingbefore MRCP is not sufficient for elimination of thesesignals in the gastro-intestinal tract [16].

Oral negative contrast agents depend on high amounts ofhigh molecular metal ions such as iron or manganese whichhave paramagnetic and superparamagnetic properties.These qualities will increase magnetic susceptibility causingmarked shortening in T2 relaxation time [17, 18] due to rapidT2 decay. There are some studies that require negative oralcontrast agents - eg MRCP.

Although many agents may significantly obscure signalintensity of the GI tract, their effects in depiction of someparts of the biliary tract such as IHD may be limited. (11)

Several negative oral contrast agent products includingblueberry [19], pineapple juice [16] and Roselle [11] havebeen used as negative oral contrast agents. All these agentsare characterized by a high manganese concentration.

One of the most usual drinks among Iranian people is blacktea which has 3502,200 g/gr manganese in dry leaf [13].

In this study, we have proposed that black tea is a goodalternative for signal suppression from the GI tract structures.We used 300 mL of black tea as a negative oral contrast agent.

Because Iranians like to drink sweet tea, we added 40 grsugar in every 300 mL tea, and there was no change in thenegative contrast property of black tea.

Quantitative analysis using VAS and Likert showedsignificant improvement in MPD, the distal part of theCBD and ampulla.

Indeed, we found that black tea effectively reducedsignal intensity of the stomach and the duodenum.

Depiction of the MPD, the distal part of the CBD andampulla significantly improved statistically 5 and 15 minfollowing black tea ingestion. This suggests that follow-ing black tea ingestion, the optimum time for MRCP is5 min.

We noticed that visualization of the distal part of theCBD significantly improved following black tea ingestion.It might be due to the fact that duodenal signals only affectthe distal part of the CBD and there is no overlap betweenGI signals and the proximal part of the CBD.

Similar to this study, in a study published byVaravithya et al., the authors found that Rosella flower

tea can effectively reduce signal intensity of the stomachand duodenum, and they found slight improvement ofampulla and main pancreatic duct depiction in theirpatients [11]. Chan and his colleagues used dilutedgadopentetate dimeglumine in their study in 23 patientsand found that gadopentetate dimeglumine with a concen-tration of 1:15 is significantly effective for depiction of theCBD and MPD in MRCP. For GB and CD, a slight tomoderate improvement was seen after oral gadopentetatedimeglumine and they did not evaluate CHD, IHD andampulla in their study [20].

Papanikolaou et al. observed that there was a statisticallysignificant improvement in the depiction of CBD, CHD,ampulla and MPD after using 430 mL of blueberry juice in37 patients who suffered obstructive jaundice [19].

Riordan and his colleagues in 2004 published their studyabout using pineapple juice (PJ) as a negative oral contrastagent in MRCP. They demonstrated that PJ may be used asa suitable negative oral contrast agent in MRCP [16].

It should be kept in mind that, although the use ofnegative oral contrast agents is beneficial in suppressing thesignal in the stomach and intestine, visualization of someparts of the biliary tract may be limited. Furtehrmore, whena negative oral contrast agent is to be used, the patientsclinical condition should be carefully evaluated. Particularly,when a patient has a history of endoscopic sphincterotomy,negative oral contrast agent should not be given at firstbecause of the bile counterflow.

One limitation of this study was that only one radiologistreported the images.

Another limitation was the exact volume of tea based onthe patients tolerance which was not equal in all cases. Forthis reason we will design another study with use of lowervolumes of tea.

We conclude that black tea is an affordable, available,safe and efficient oral negative contrast agent for MRCPthat reduces the signal intensity of fluids in the gastrointes-tinal tract and also better depicts the MPD, the distal part ofthe CBD and ampulla.

References

1. Pilleul F, Courbire M, Henry L et al (2004) La cholangio-IRMdans le diagnostic tiologique des stenosis biliaires: corrlationanatomopathologique. J Radiol 85:2530

2. Hoeffel C, Azizi L, Lewin M et al (2006) Normal andpathologic features of the postoperative biliary tract at 3DMR cholangiopancreatography and MR imaging. Radio-graphics 26:16031620

3. Yu J, Turner MA, Fulcher AS et al (2006) Congenital anomaliesand normal variants of the pancreaticobiliary tract and thepancreas in adults: part 1, biliary tract. AJR Am J Roentgenol187:15361543

2556 Eur Radiol (2011) 21:25512557

4. Yu J, Turner MA, Fulcher AS et al (2006) Congenital anomaliesand normal variants of the pancreaticobiliary tract and thepancreas in adults: part 2, pancreatic duct and pancreas. AJRAm J Roentgenol 187:15441553

5. Arriv L, Coudray C, Azizi L et al (2007) Pineapple juice as anegative oral contrast agent in magnetic resonance cholangiopan-creatography. J Radiol 88(11 Pt 1):16891694

6. Coppens E, Metens T, Winant C et al (2005) Pineapple juicelabeled with gadolinium: a convenient oral contrast for magneticresonance cholangiopancreatography. Eur Radiol 15:21222129

7. Matos C, Metens T, Devire J et al (1997) Pancreatic duct:morphologic and functional evaluation with dynamic MRpancreatography after secretin stimulation. Radiology203:435441

8. Hirohashi S, Hirohashi R, Uchida H et al (1997) MR cholangio-pancreatography and MR urography: improved enhancement witha negative oral contrast agent. Radiology 203:281285

9. Bernardino MR, Weinreb JC, Mitchell DG et al (1994) Safety andoptimum concen- tration of a manganese chloride based oral MRcontrast agent. J Magn Res Imag 4:872876

10. Schreiber WE (1989) Iron, porphyrin and bilirubin metabolism.In: Kaplan LA, Pesce AJ (eds) Clinical chemistry: theory,analysis. Mosby, St. Louis, pp 496511

11. Varavithya V, Phongkitkarun S, Jatchavala J et al (2005) Theefficacy of roselle (Hibicus sabdariffa Linn.) flower tea as oralnegative contrast agent for MRCP study. J Med Assoc Thai 88(Suppl 1):S35S41

12. Weisburger JH (1997) Tea and health: a historical perspective.Cancer Lett 114:315317

13. Wrbel K, Wrbel K, Urbina EM (2000) Determination of totalaluminum, chromium, copper, iron, manganese, and nickel andtheir fractions leached to the infusions of black tea, green tea,Hibiscus sabdariffa, and Ilex paraguariensis (mate) by ETA-AAS.Biol Trace Elem Res 78:271280

14. Kato J, Kawamura Y, Watanabe T et al (2001) Examination ofintra-gastrointestinal tract signal elimination in MRCP: combineduse of T(1)-shortening positive contrast agent and single-shot fastinversion recovery. J Magn Reson Imaging 13:738743

15. Wewers ME, Lowe NK (1990) A critical review of visualanalogue scales in the measurement of clinical phenomena. ResNurs Heal 13:227236

16. Riordan RD, Khonsari M, Jeffries J et al (2004) Pineapple juice asa negative oral contrast agent in magnetic resonance cholangio-pancreatography: a preliminary evaluation. Br J Radiol 77:991999

17. Kim YK, Kim CS, Lee JM et al (2006) Value of adding T1-weighted image to MR cholangiopancreatography for detect-ing intrahepatic biliary stones. AJR Am J Roentgenol 187:W267W274

18. Broglia L, Tortora A, Maccioni F et al (1999) Optimization ofdosage and exam technique in the use of oral contrast media inmagneticresonance. Radiol Med 97:365370

19. Papanikolaou N, Karantanas A, Maris T et al (2000) MRcholangiopancreatography before and after oral blueberry juiceadministration. J Comput Assist Tomogr 24:229234

20. Chan JH, Tsui EY, Yuen MK et al (2000) Gadopentetatedimeglumine as an oral negative gastrointestinal contrast agentfor MRCP. Abdom Imaging 5:0508

Eur Radiol (2011) 21:25512557 2557

Copyright of European Radiology is the property of Springer Science & Business Media B.V. and its contentmay not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's expresswritten permission. However, users may print, download, or email articles for individual use.