Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Incidental Extra-cardiac Findings on Clinical CMR; A Comparison of 3 HASTE Techniques. Bruce Irwin1*, Tom Newton2* Charles Peebles3, Alexander Borg1, David Clark4, Chris Miller1, Nick Abidin1, Melanie Greaves2, Matthias Schmitt1,5 1 Department of Cardiology, University Hospital of South Manchester (UHSM) Department of Radiology, UHSM 3 Department of Radiology, Southampton General Hospital 4 Alliance Medical, Wythenshawe CMR unit 5 Biomedical Imaging Institute, The University of Manchester 2 BI and TN contributed equally to this work Word count; Abstract: 393 Text (including Tables/Figures): 4163 References: 416 Address for Correspondence Dr Matthias Schmitt Department of Cardiology Office 4, North West Heart Centre University Hospital of South Manchester Southmoor Road Manchester M23 9LT Tel./Fax 01612914940 e-mail; [email protected] ABSTRACT Objectives: First, we sought to assess the frequency of incidental extra-cardiac findings (IEF) found in a consecutive series of clinical Cardiac MR scans. Second, we compared the 3 clinically used HASTE acquisition protocols in this context. Third, we determined the impact of the three different HASTE protocols on acquisition time and image quality. Methods: Three consecutive groups of 238 patients (total 714), referred for clinically indicated CMR, were scanned with either breath-hold HASTE (BH, Group 1), free breathing HASTE (FB, Group 2) or diaphragmatic navigated HASTE (NAV, Group 3) in addition to multi-slice, single shot steady state sequences in 3 orthogonal planes. All 714 clinical reports were reviewed regarding the presence of incidental extracardiac findings and the recommendations on the need for further investigation, follow up, and/or clinical correlation. Finally, to determine the impact of each HASTE protocol on acquisition time and image quality an additional cohort of 15 patients underwent all 3 protocols back to back in a random fashion. The length of each acquisition was timed and image quality was reviewed and scored externally. Results: A total of 180 IEF were found in 162 (22.7%) out of 714 patients. There was no significant difference in frequency of IEF between the 3 HASTE groups. Out of 180 IEF 88 were considered minor and 92 major findings. Of the latter, 8 (1.1%) were considered highly significant including one bronchoalveolar carcinoma stage 1B requiring lobectomy, 2 cases of florid sarcoidosis in patients presenting with VT and “structurally normal hearts” on Echo, one case of pulmonary aspergillosis, 2 cases of advanced pulmonary fibrosis, one ascending thoracic aortic aneurysm and a case of iatrogenic liver haemorrhage following placement of a pericardial drain. FB HASTE acquisition (69±2.5s) was significantly faster than BH (105±3.8s) and NAV (121±2.7s), p<0.001 but also produced the lowest image quality on a 5 point scale; 3.5 (FB) versus 3.9 (BH) versus 3.8 (NAV), p=0.08. Conclusion: Overall, IEF are common and lead to follow on investigations in a substantial minority of cases. However, the overall incidence of highly significant findings in the current study was low (~1%). There was no difference in the frequency of incidental extra-cardiac findings between the 3 HASTE protocols. Whilst the free breathing HASTE technique is statistically significantly faster than breath hold and navigated HASTE the absolute time saving is small and probably out-weight by the resulting lesser image quality. INTRODUCTION: Cardiac magnetic resonance (CMR) is now an established imaging modality with well described clinical indications1 and appropriateness criteria2. In addition to the heart, a typical CMR examination will also image adjacent thoracic and abdominal structures. Indeed most, but not all, centres initially perform gradient echo “scout”-imaging of several slices in coronal, saggital and axial views followed by axial imaging of the entire chest, conventionally using half fourier turbo spin echo imaging (e.g. HASTE). Consequently, findings incidental to the cardiac examination may be encountered some of which may be clinically relevant. Although well described in the computed tomography (CT) literature3-7 there is paucity of data regarding the frequency and clinical impact of incidental extra-cardiac findings (IEF) discovered during routine CMR. The few studies5, 8, 9 and preliminary reports10, 11 available to date demonstrate considerable variability, with IEF rates ranging from 7.6% to 81%, and indicate unsurprisingly, that the prevalence of IEF is dependent on multiple factors, not least the population studied, the image sequences applied and the definitions utilised to classify and categorise IEF. Furthermore, it is unclear if the various methods of performing axial HASTE sequences (i.e breath-hold (BH), free breathing (FB), or diaphragmatic navigated (NAV)) impact on the frequency of detecting IEF. In the current study we sought to assess the frequency of IEF in consecutive patients referred for clinically indicated CMR, in an NHS tertiary care setting. We also aimed to determine the impact of the 3 most commonly applied HASTE acquisition protocols on the frequency of IEF, as well as on image acquisition time and image quality. METHODS: The study was planned in May 2008 and carried out from 12/08/2008 to 14/09/2009 in a dedicated CMR unit run collaboratively as a Managed Care Service by Alliance Medical at the University Hospital of South Manchester (UHSM), in Wythenshawe. Patients: Prospectively, three subsequent groups of 238 adult patients each (i.e 714 patients in total), all referred for clinically indicated CMR, were scanned with either BH HASTE (Group 1), FB HASTE (Group 2), or NAV HASTE (Group 3) sequences. All 714 clinical studies were acquired by 4 CMR trained radiographers, employed by Alliance Medical, and reported by one of 3 Physicians (2 Cardiologists (MS, NA) + 1 Radiologist (MG)) employed by UHSM Foundation Trust. All reports were subsequently reviewed and audited by either a CMR-Level II trained Cardiology (BI) or CMR-Level II trained Radiology Fellow (TN) regarding the presence of IEF and the recommendations on the need for further investigation, follow up and/or clinical correlation. Both Fellows closely collaborated and assured that they applied identical criteria for classification/categorisation of IEF in line with the following definition. Definition: We prospectively decided to classify IEF in minor and major findings whereby minor findings were considered benign or of no clinical importance not requiring clinical correlation or follow up. Major findings were those IEF potentially or definitely considered to be of clinical significance and/or requiring clinical correlation or further follow and/or work-up. Whilst there was variability in case load amongst the reporting physicians, their relative reporting contributions did not change between the HASTE groups. Furthermore, pleural effusions and/or ascites co-existing with significant ventricular impairment were in contrast to some previous studies not classified as IEF. CMR: All clinical CMR studies were vetted and coded by a single Consultant (MS) and for auditing purpose grouped in to one of 8 clinical indication groups (UK national MRI codes MCORV = Cardiac Ventriculogram and MCVVS = LV volume study were grouped together). Also, as we did not perform coronary artery imaging in isolation, we decided for audit purposes to classify MACOA (= MRA coronary arteries) as either a ventricular function study if no contrast was administered, or as a MCVIA (viability study) if prior contrast administration took place (in which case “late enhancement” imaging was always performed additionally). All scans were performed according to locally agreed scanning protocols, tailored where necessary to the individual patient. These were largely based on standardised protocols recommended by the Society of Cardiac Magnetic Resonance (SCMR). As a minimum all studies included 3 localising single shot steady state sequences in 3 orthogonal planes followed by one of three HASTE protocols (see below), assessment of left and right ventricular function by means of breath hold cine steady state free precession (SSFP) sequences, predominantly retrospectively gated but triggered in the presence off significant arrhythmia. Additional sequences, were performed as clinically indicated. Eighty nine percent of patients received 0.1 to 0.2mmol/kg of intravenous gadolinium either as dimeglumine gadopenetetate (Magnevist®, Bayer Schering Pharma), Gadobutrol (Gadovist®, Bayer Schering Pharma), or Gadobenate® dimeglumine (MultiHance, Bracco). Group 1; Axial breath hold HASTE (BH) Axial multi-slice HASTE sequence was acquired from the top of the aortic arch to the diaphragm. The field of view (FOV) chosen was dependant on both patient size and shape and typically ranged from 340mm x 233mm to 390mm x 344mm. Base resolution and phase resolution were 256 and 59% respectively. Slice thickness and slice gap were 8mm and 2mm respectively. This resulted in a spatial resolution typically ranging from 2.3mm x 1.3mm x 8mm to 2.5mm x 1.5mm x 8mm. Repetition time (TR) varied with patient heart rate. Interleaved slices were therefore acquired in 3, 4 or 5 concatenations (and therefore 3, 4 or 5 breath holds) in order to keep scan time to approximately 10 to 15 seconds per breath hold acquisition. Group 2; Axial free breathing HASTE (FB) Axial slices were acquired in a single concatenation and with the patient breathing shallowly.The total acquisition time was heart rate dependent and typically varied between 60-80 seconds. Group 3; Axial diaphragmatic navigated HASTE (NAV) As with the breath hold HASTE sequence, concatenations were set to maintain a breath hold time of 10 to 15 seconds. A single navigator echo was placed over the right hemi-diaphragm in order to monitor its position at each breath hold. If the position of the diaphragm varied then the slice position for each concatenation was altered in order to maintain contiguous slices. In the current study therefore, spatial resolution, was predominantly affected by the patientssize and shape rather than a result of the HASTE modality chosen. Acquisition time (AT) and Image quality (IQ); To determine the impact of each HASTE protocol on acquisition time and image quality an additional cohort of 15 patients underwent all 3 protocols back to back in a random fashion. The length of each acquisition was timed with a stopwatch and IQ was reviewed and scored separately in a blinded fashion by an external investigator (CP) using a 5 point scoring system whereby 1 marked unsatisfactory IQ, 2 = satisfactory IQ, 3=good IQ, 4= very good IQ, 5=excellent IQ. Statistics: The prevalence of IEF is presented in absolute and per-terms. Data with respect to image quality and acquisition speed are presented as mean ± SEM. Comparisons for image acquisition speed and image quality were performed by repeated measures ANOVA with post-hoc Bonferoni for pair-wise testing where appropriate. Proportionate comparison between HASTE protocols were analysed by Pearson Chi Square tests. A two-sided p value of <0.05 was considered significant. RESULTS A) A total of 714 studies were included in the analysis, split evenly between each cohort. Demographic characteristics are shown in table 1. Group 3 (NAV) had a somewhat greater proportion of male subjects, although age distribution and BSA remained consistent. Table 1. Demographic data Mean Age Group 1 Group 2 Group 3 p 53.9 (17-85) 54.8 (16-85) 54.8 (15-85) ns 1.93±0.016 1.90±0.014 1.92±0.015 ns 1.42 1.69 2.2 0.07 (Range) Body Surface Area (BSA) Male: Female ratio An analysis of the clinical indications for the study cohort is displayed in figure 1 and tables 2. This represents the typical workload of our centre, with little variation seen in the 2 most prevalent clinical indications, i.e. stress perfusion and viability imaging, over the time period of the study. However, overall there was a statistical difference between clinical indications between the 3 groups (p=0.009) which was predominantly driven by the proportionally larger number of Stress function (i.e. Dobutamine stress) and “pure” LV volume studies (i.e non-contrast studies) in Group 1. There was no correlation of IEF with the CMR indication. Figure 1. Indications for CMR scanning across all groups Because of the small number of first pass rest perfusion this indication was pooled with stress perfusion for the purpose of statistical analysis. Table 2. Indication for CMR study by group Indication Group 1 Group 2 Group 3 No. Patients (%) No. Patients (%) No. Patients (%) Myocardial stress perfusion 92 (38.7) 89 (37.4) 105 (44.1) Stress function study* 10 (4.2) 3 (1.3) 3 (1.3) Myocardial viability study† 74 (31.1) 92 (38.6) 88 (37.0) Congenital anomaly study 20 (8.4) 20 (8.4) 11 (4.6) MRA thoracic aorta 15 (6.3) 15 (6.3) 16 (6.7) LV volume study ‡ 23 (9.7) 10 (4.2) 11 (4.6) Valvular function study 1 (0.4) 8 (3.4) 3 (1.3) Rest perfusion study 3 (1.3) 1 (0.4) 1 (0.4) * Dobutamine stress with wall motion analysis Includes investigation of arrhythmogenic right ventricular cardiomyopathy ‡ No gadolinium administered. † From the 714 patients scanned, a total of 180 IEFs were discovered in 162 patients. This gives an overall prevalence of IEFs in the study of 25.2%, affecting 22.7% of patients scanned. As outlined above, these findings were further characterised into those perceived as clinically significant or ‘major’, and those considered insignificant or ‘minor’. Eightyeight extra-cardiac findings of minor and ninety-two of major significance were reported (see tables 3 + 4). Seven patients (1.0%) were diagnosed with more than one IEF. Highly significant Findings: Amongst the major IEF 8 (1.1% ) findings were considered highly significant including one bronchio-alveolar carcinoma stage 1B, 2 cases of florid pulmonary sarcoidosis in patients presenting with VT and “structurally normal hearts” on Echo, one case of pulmonary aspergillosis, 2 cases of advanced pulmonary fibrosis, one ascending thoracic aortic aneurysm (6.6cm) requiring surgery, in a patients with atypical chest pains, and a case of iatrogenic liver haemorrhage following placement of a pericardial drain. No significant difference was found between each of the groups when analysed for the total extra-cardiac findings identified, both minor and major. Table 3. Findings of minor significance Abnormality Group 1 Group 2 Group 3 Total all groups Atelectasis 3 0 3 6 Large Axillary lymph node 0 1 0 1 Azygos lobe 2 0 0 2 Bronchogenic cyst 1 0 0 1 Diaphragmatic Lipoma 0 1 0 1 Hiatus Hernia 3 5 2 10 High take-off of Bronchus 1 0 0 1 Hepatic cyst (solitary) 8 4 3 15 Hepatic haemangioma 0 1 0 1 Kyphoscoliosis 0 0 1 1 Lipomatosis (epicardial/ 2 3 1 6 5 1 10 16 1 0 0 1 Small pleural effusions 1 0 2 3 Simple Renal cysts ≤Bosniak II 0 2 0 2 Thyroid goitre 0 3 2 5 Vascular anomaly 3 2 7 12 Vertebral body haemangioma 0 4 0 4 Total minor abnormalities 30 27 31 88 (solitary) mediastinal) Mediastinal lymph nodes (<1cm in short axis) Pleural thickening (postpleurodesis) P value for inter-group difference Table 4. Findings of major clinical significance Abnormality Group 1 Group 2 Group 3 Total all groups Dilated aorta* 2 3 2 7 Dilated bile ducts 0 1 0 1 Dilated oesophagus (achalasia) 1 1 0 2 Dilated main pulmonary artery† 7 5 4 16 Scarring of/around defect breast 1 0 0 1 Empyema 0 1 0 1 Hepatic cysts (multiple) 6 3 4 13 Hepatic mass awaiting diagnosis 0 0 1 1 (Iatrogenic) liver haemorrhage 1 0 0 1 Lung nodule 0 0 1 1 Lung mass 3 2 3 8 Thyroid mass 0 0 1 1 Parenchymal lung abnormality 6 11 6 23 Pleural effusion, cause unknown 0 0 2 2 Pleural thickening 2 6 1 9 Renal cysts (multiple/septated) 1 2 1 4 0 1 0 1 30 36 26 92 implant ≥Bosniak IIF Vertebral body lesion, aetiology unknown Total major abnormalities P value for inter-group difference * Defined as > 4.0cm at pulmonary artery bifurcation level †Defined as > 3.0cm or larger then ascending aorta Lung nodule defined as <or equal 3cm, Lung mass >3cm B) Image Acquisition Time and Quality Data for the 15 patients who underwent all three methods of HASTE acquisition is summarised in table 5. FB HASTE took the shortest time, with average scan duration of 69 seconds. BH and NAV methods took significantly longer, on average 105 and 121 seconds, respectively. The individual image quality scores were averaged for each of the three methods. FB resulted in the lowest image quality, with an average quality score of 3.5, where 5 is the theoretical maximum. BH and NAV methods displayed superior quality, with average scores of 3.9 and 3.8 respectively. The overall difference in image quality between the acquisitions showed a trend towards significance. Table 5. Acquisition Time (in seconds) and Image Quality (5 point scoring system) Patient No. Group 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Average SD SEM Total sequence time (s) BH FB NAV G1 G2 G3 107 81 130 112 79 135 134 60 134 123 70 115 95 85 111 102 69 125 98 62 103 105 80 128 112 61 131 75 48 102 89 65 127 92 67 116 103 62 116 103 74 121 123 68 115 105 69 121 14.8 9.8 10.5 3.8 2.5 2.7 p<0.001 Image quality score (1-5, 5 = best) BH FB NAV G1 G2 G3 5 4 4 4 4 4 3 3 2 4 3 4 4 4 5 5 5 5 4 4 4 4 3 4 4 3 3 3 3 3 4 3 4 4 3 4 4 4 4 3 4 4 4 3 3 3.9 3.5 3.8 0.59 0.64 0.78 0.15 0.17 0.2 BH= Breath hold; FB= Free breathing; NAV= diaphragmatic navigation Image Comparison (best 1st) BH>NAV>FB BH>NAV>FB FB>BH>NAV BH>NAV>FB NAV>BH>FB FB>BH>NAV BH>NAV>FB NAV>BH>FB BH>NAV>FB FB>BH>NAV BH>NAV>FB NAV>BH>FB BH>NAV>FB FB>NAV>BH BH>NAV>FB p=0.08 DISCUSSION: The literature describing the prevalence of IEF during cross-sectional cardiac imaging is predominated by publications based on cardiac CT. In contrast there is a paucity of data characterising the prevalence and clinical significance of IEF on CMR, especially in a pure clinical setting. What the current study adds in the context of previous publications on IEF: The present study is to our knowledge the second largest consecutive series reporting the prevalence of (previously unknown) IEF in clinically indicated CMR studies. In keeping with the cardiac CT literature (~21% prevalence) our study found that the overall detection rate of IEF on clinically indicated CMR is relatively high (22.7%). Interestingly, our pick up rate was almost identical to a preliminary report of identical size also performed within the UK-NHS setting 11. The group from the London Chest Hospital found, in a slightly older population (61.3 ± 15years, range 16-86 years, 67% male), IEF in 218 (24%) out of 714 patients. These pathologies included renal cysts (9%), aortic pathology (3%), lung collapse/consolidation (2%), liver cysts (2%), lung mass (1%). They also found 2 cases of mediastinal masses, 1 mediastinal lymphadenopathy, 6 increased signal from the hila, a single renal mass, and one large hiatus hernia. In contrast with these 2 UK based CMR studies, both reporting comparable detection rates to that of cardiac CT literature, the so far largest published US based study by Chan et al 8 concluded that non-cardiac pathology is uncommonly reported. In their cohort (1534 patients, 62%male, age 50±15years) 116 studies (7.6%) had at least one non-cardiac finding of which 48 (3.1%) reports were deemed to demonstrate major and 70 (4.6%) minor findings. Major findings were; lymphadenopathy (n=22 / 40% of major findings), Lung abnormalities (encompassing nodules, masses and infiltrates; n=19 / 35%), Mediastinal masses (n=6 / 11%), Breast lesion (n=4 / 7%), Ascites (n=3 / 5%), soft tissue mass (n=1 / 2%). Minor findings were: Pleural effusion (n=30 /40%), benign liver cyst (n=15 / 20%), renal cysts (n=14 / 19%) ), Hiatus Hernia (n=7 / 9%), diaphragmatic abnormality (n=2 / 3%), Splenic abnormality (n=2 / 3%), paraspinal lipoma (n=2 /3%), anomalous vasculature non-cardiac (n=2 / 3%). Importantly only 8 findings in 6 reports (0.4%) of 1534 reports were ultimately deemed to be new and clinically significant. Of note, not unexpected, the age of those with clinically significant non-cardiac pathology was greater (54±16). Why these (apparent) discrepancies ? The few studies5, 8, 9 and preliminary reports10 available demonstrate considerable variability in IEF diagnosis with rates ranging from 7.6% (0.4% for major findings) to 81% (17%) and indicate that, not surprisingly, the prevalence of IEF is dependent on multiple factors including the cohort studied, the “clinical setting”, the sequences applied, and the reading session format as well as, possibly most importantly the definitions utilised to categorise a IEF. With respect to minor IEF it is possibly not unreasonable to assume, especially given that most (if not all) of the published literature is based on retrospective review of reports rather than images, that the bulk of variation in the literature reflects different attitudes and weighting of the most common abnormalities (such as simple renal and hepatic cysts and vascular abnormalities) by the reporting clinicians. This assumption is possibly supported by the significantly smaller variation in “major” IEF in the same cohorts. What constitutes a “major” IEF, is unfortunately not standardised and confused in the literature with definitions “ranging from one that requires follow-up or clinical correlation to one that needs immediate evaluation or treatment.” 12 With respect to referral bias it is apparent that the population in the study by Chan clearly differs, not only in age but equally important in scan indication, significantly from that in both UK based studies. Whilst the assessment of IHD (i.e. MCORP and MCVIA) was the predominant clinical indication in both UK studies the 2 largest referral cohorts in the US based study were assessment of ventricular function (34%) and interestingly assessment of pulmonary vein anatomy (25%). The difference, in particular with respect to the latter indication may be more representative of local practise variation and the academic nature of the US institution rather than being representative of differing technology application between healthcare systems. However, whilst partly coming to different conclusions with respect to the frequency of IEF, the overriding commonality between the current study and both above detailed studies, and in fact most of the published literature, is that highly significant findings (i.e. a finding that in line with the 2010 Expert Consensus requires immediate evaluation or treatment) occur with a prevalence varying between 0.4%-2%. Nevertheless, even taken together this literature remains insufficient to answer the important question if the incidental finding of an extra-cardiac pathology during cross-sectional cardiac imaging has a significant impact on patient health and long term outcome. Equally, little is known with respect to the cost-effectiveness and potential increased anxiety of follow on investigations triggered by IEF on crosssectional cardiac imaging. On the other hand in the increasingly litigious healthcare environment of western societies it appears unwise not to have policies in place that report and deal with potentially significant IEF. What is the impact of the sequences used ? A recent meta-analysis investigating the prevalence of incidental findings on magnetic resonance imaging (MRI) of the brain found unsurprisingly that sequences applied and sequence resolution were important determinant of the frequency with which IEF where identified.13 Also Chan et al pointed out that: “a better understanding of how various sequences impact on the ability to detect non-cardiac pathology would help the development of guidelines in CMR training”.8 The current study is the first to compare three different clinically utilised HASTE protocols with respect to the frequency of IEF, image quality and acquisition time and found no significant differences with respect to the IEF detection rate. Whilst the FB HASTE (Group 2) was acquired significantly faster than the two other acquisition protocols the overall time saving (of 36 or 51 seconds respectively) was small (on the background of an average study length of 43 minutes) and occurred against a significant trade off in image quality. Should one alter protocols during scanning to address IEF ? This important question off course only arises if the IEF is actually identified during the acquisition of the study which will be affected by the local set-up with respect to physician supervised or unsupervised study acquisition. However, even if detected during the acquisition one must consider to what degree additional sequences performed to evaluate the IEF further would impact on the workflow within a unit. Equally important, one must not ignore the possibility that the deviation (in particular if lengthy) from the original protocol may detrimentally effect the likelihood of answering the original clinical question posed. In this light whilst minor changes to cover the IEF with further slices/sequences may be appropriate more significant changes to the protocol may not be desirable as dedicated specialised sequences, and or contrast and expertise may be required, and /or alternative imaging modality may be more appropriate and / or cost effective.14 The true significance of IEF? Beyond the medico-legal aspects, the true clinical relevance probably more likely relates to the fact that a number of IEF may actually account for patient symptoms and thereby at least partly negate the originally entertained differential diagnosis. Common examples in support of this are lung pathology such as atelectasis, Emphysema, pleural effusions, pleural and interstitial lung diseases causing shortness of breath. Equally important IEF are not infrequently a clue to either a symptom relevant co-morbidity or again intimately related to the patients presenting problems. Examples in the current study are the patients presenting with VT in the setting of pulmonary sarcoidosis. Furthermore, IEF can provide indirect explanation for cardiac pathology such as abnormal pulmonary venous drainage contributing to SOB or coarctation, causing increased afterload and driving hypertension, leading to LVH, diastolic dysfunction and breathlessness. Also, a range of IEF, whilst unrelated to a subjects symptoms and at the time of diagnosis of no immediate clinical significance may become highly relevant during the patients future diagnostic workup and potential subsequent treatment and therefore should be documented and communicated in a way that the benefit of this information is not subsequently lost. Examples of such cases are the finding of vascular abnormalities such as interrupted IVC or retro-oesophageal course of a right subclavian artery, both highly relevant for potential future cardiac catheterisation. Finally, it is intuitive to assume that in individual circumstances action taken based on the detection of an IEF, e.g. such as the lung malignancies found in the current study, will improve outcome. Study limitations: Overall study and subgroup size was only moderate. However, this work is still the second largest published series and even our sub-groups are larger than most currently published work in this field. Despite prospective study design, report analysis for IEF reports (not images!) and review of follow on investigations and therapy was retrospective and limited to medical and electronic records of in-house patients. Furthermore, patients were not randomised to one of the HASTE groups but for organisational reasons scanned in subsequent cohorts with the inherent risk of temporal bias. With respect to this care was given that avoid protocol changes and unchanged relative reporting contribution amongst the 3 Consultant during the study period. Most importantly, like with all previous work in this field the definition of minor and major criteria, also pre-specified and in line with the subsequently published 210 Expert Consensus on Coronary computed Tomographic angiography 12, remains a subjective one and at least minor variations in weighting of IEF, e.g. if simple hepatic, even if multiple, and / or renal cysts are mentioned in a report can not be excluded. Conclusions Overall, IEF are common and require follow or additional investigations in a substantial minority of cases. However, the overall incidence of highly significant findings in the current study was low (~1%) and very similar to that found in the CT literature. Furthermore, there appears to be no substantial difference in the frequency of incidental extra-cardiac findings between the 3 clinically applied HASTE protocols. Whilst the FB HASTE technique is statistically significantly faster than BH and NAV HASTE the absolute time saving is small and probably out-weight by the resulting lesser image quality. Figure 2: A) B) C) D) Figure 3: A) C) B) D) Figure 4: A) B) C) Figure 5: A) C) D) B) E) Figure 6: A) B) C) Figure Legend: Figure 1: Clinical indication for CMR. Studies could have more than clinical indication / question but for coding and auditing purposes were allocated to a single indication as displayed in the pie chart. Data are presented as percentages. Figure 2: Fifty-six year old male patient, referred for stress perfusion imaging to evaluate cardiac sounding chest pain, is found to have 2 IEF on CMR. Images A-D; A: Depicts a well demarcated lung lesion in the left lower zone (yellow arrow) with low central and higher peripheral signal on coronal scout image. Additionally, (red arrow) a well rounded liver lesion with lower than surrounding liver signal was seen. The liver lesion was further investigated with Ultrasound and judged to reflect a haemangioma. No change in size has occurred over 24 months. The presence of the lesion was confirmed by lateral chest x-ray (B) and PA (C) chest radiographs and further evaluated with contrast enhanced CT (D). The nodule demonstrated a benign pattern of dense calcification and as such was followed with plain chest radiographs. It has remained unchanged in size for 2 years. Figure 3: 43 year old female non-smoker referred for evaluation of atypical chest pains. Coronal Localiser (A) and axial HASTE imaging (B) depict an ill defined region of increased signal in the apical segment of the left lower lobe. Axial CT “lung windows” (C) confirms the presence of a 2.7cm ill defined soft tissue nodule. Image D) illustrates the tumour (red arrow) axial CT “bone window” guided needle (yellow arrow) biopsy. The patient underwent curative resection of a broncho-alveolar carcinoma T2 M0 N0 stage 1B. Follow up CT 18/12 post surgery demonstrates freedom from local recurrence. Figure 4: Images of 3 Patients with Sarcoidosis. Image A: HASTE images in a 35 year old man being investigated for sustained ventricular tachycardia. The three axial cuts from the upper thorax demonstrate right supraclavicular and bilateral upper paratracheal lymphadenopathy (yellow arrows). The diagnosis of sarcoidosis was subsequently confirmed with transbronchial biopsy. There was no evidence of localised myocardial oedema on T2 weighted imaging nor localised hyperenhancement on myocardial delayed gadolinium imaging. Image B: Axial HASTE image at the level of the upper arch demonstrating bilateral pulmonary parenchymal high signal within the posterior upper lobes (yellow arrows) in a patient presenting with non-sustained VT and subsequently diagnosed with sarcoidosis. Again the patient had no evidence of localised hyperenhancement on myocardial delayed gadolinium imaging. Image C: Phase sensitive inversion recovery (PSIR) delayed Gadolinium image of a patient investigated for heart failure symptoms. The image shows multiple areas of localised hyperenhancement (yellow arrows) affecting both ventricles. The patient also had mild mediastinal lymphadenopathy and subsequent mediastinal lymph node biopsy confirmed the diagnosis of sarcoidosis. Figure 5: Examples of additional incidental findings: Image A; Coronal Scout image demonstrating a large hiatus hernia (arrow). Image B: Axial HASTE image at the level of the pulmonary bifurcation in an 80 year old patient with known extrinsic allergic alveolitis demonstrating a rather patulous air filled oesophagus. Image C: Lower thoracic HASTE demonstrating isolated rounded high signal vertebral body lesion judged to reflect a vertebral body haemangioma. Image D: Axial HASTE image though the upper thorax reveals an azygos fissure which is a normal variant of no clinical significance. Image E: Axial HASTE through the lung apex demonstrates generalised enlargement of the thyroid with a 25mmx25mm nodule, with mildly increased signal as compared to surrounding thyroid tissue, within the left lobe. Further investigation with ultrasound demonstrated the nodule to be hypoechogenic with a hyperechogenic (calcific) rim. Further smaller nodules of the same echogenicity but no lymphadenopathy was seen. Features were judged to be in keeping with a multinodular goitre. Figure 6: 55 year old woman referred for bilateral neck swelling and vague history of an abnormal aorta. Image A: Still frame of a coronal SSFP cine demonstrating an elongated right sided cervical aortic arch (yellow arrow) as well as a large left sided septated cystic thyroid mass (red arrow). Images B, C: Volume rendered contrast enhanced Aortogram (anterior view - image B, posterior view - image C) highlight the abnormally arising arch vessels whereby the first arch vessel is the left common carotid, followed by the right vertebral artery (normally arising from the right subclavian), the right common carotid, the right subclavian (arising from the posterior aspect of the arch), and the left subclavian arising from an aortic diverticulum (Kommeral). Reference List (1) Pennell DJ, Sechtem UP, Higgins CB et al. Clinical indications for cardiovascular magnetic resonance (CMR): Consensus Panel report. Eur Heart J 2004 November;25(21):1940-65. (2) Hendel RC, Patel MR, Kramer CM et al. ACCF/ACR/SCCT/SCMR/ASNC/NASCI/SCAI/SIR 2006 appropriateness criteria for cardiac computed tomography and cardiac magnetic resonance imaging: a report of the American College of Cardiology Foundation Quality Strategic Directions Committee Appropriateness Criteria Working Group, American College of Radiology, Society of Cardiovascular Computed Tomography, Society for Cardiovascular Magnetic Resonance, American Society of Nuclear Cardiology, North American Society for Cardiac Imaging, Society for Cardiovascular Angiography and Interventions, and Society of Interventional Radiology. J Am Coll Cardiol 2006 October 3;48(7):1475-97. (3) Haller S, Kaiser C, Buser P, Bongartz G, Bremerich J. Coronary artery imaging with contrast-enhanced MDCT: extracardiac findings. AJR Am J Roentgenol 2006 July;187(1):105-10. (4) Onuma Y, Tanabe K, Nakazawa G et al. Noncardiac findings in cardiac imaging with multidetector computed tomography. J Am Coll Cardiol 2006 July 18;48(2):402-6. (5) Dewey M, Schnapauff D, Teige F, Hamm B. Non-cardiac findings on coronary computed tomography and magnetic resonance imaging. Eur Radiol 2007 August;17(8):2038-43. (6) Hunold P, Schmermund A, Seibel RM, Gronemeyer DH, Erbel R. Prevalence and clinical significance of accidental findings in electron-beam tomographic scans for coronary artery calcification. Eur Heart J 2001 September;22(18):1748-58. (7) Horton KM, Post WS, Blumenthal RS, Fishman EK. Prevalence of significant noncardiac findings on electron-beam computed tomography coronary artery calcium screening examinations. Circulation 2002 July 30;106(5):532-4. (8) Chan PG, Smith MP, Hauser TH et al. Noncardiac pathology on clinical cardiac magnetic resonance imaging. JACC Cardiovasc Imaging 2009 August;2(8):980-6. (9) McKenna DA, Laxpati M, Colletti PM. The prevalence of incidental findings at cardiac MRI. Open Cardiovasc Med J 2008;2:20-5. (10) Romney BP, Khosa F, Costa DN, et al. Non-cardiac findings on cardiovascular magnetic resonance imaging are common: impact of imaging sequence and reading session format. Circulation 2008;18.S784. (11) Burchel Thomas LD, Mathur Anthony, Davies Ceric, Westwood Mark. Prevalence of non-cardiac incidental findings during routine clinical CMR assessment. Journal of Cardiovascular Magnetic Resonance 11 (Suppl 1):P48. 2009. (12) Mark DB, Berman DS, Budoff MJ et al. ACCF/ACR/AHA/NASCI/SAIP/SCAI/SCCT 2010 expert consensus document on coronary computed tomographic angiography:areport of the American College of Cardiology Foundation Task Force on Expert Consensus Documents. Circulation 121[22], 2509-2543. 8-6-2010. (13) Morris Z, Whiteley WN, Longstreth WT, Jr. et al. Incidental findings on brain magnetic resonance imaging: systematic review and meta-analysis. BMJ 2009;339:b3016. (14) Charles Peebles. My approach to extra-cardiac findings on CMR. 2010. Ref Type: Online Source