World Journal of Cardiovascular Diseases, 2013, 3, 389-393 WJCD
http://dx.doi.org/10.4236/wjcd.2013.35060 Published Online August 2013 (http://www.scirp.org/journal/wjcd/)
3-D multidetector computed tomography in
reoperative cardiac surgery
Tsuyoshi Kaneko1, Micheal L. Steigner2, Igor Gosev1, Sary F. Aranki1*
1Department of Cardiac Surgery, Brigham and Women’s Hospital, Boston, USA
2Department of Radiology, Brigham and Women’s Hospital, Boston, USA
Received 19 May 2013; revised 27 June 2013; accepted 14 July 2013
Copyright © 2013 Tsuyoshi Kaneko et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ABSTRACT
Reoperative cardiac surgery is becoming more com-
mon surgery although it carries significant risk due to
possible injury to vital structures under the sternum.
Three-dimensional multidetector computed tomogra-
phic angiography (3-D MDCTA) allows identifying
the relationship between the sternum and the medias-
tinal structures. Evidence shows that 3-D MDCTA
guides surgical strategy and enables to perform this
challenging surgery safely.
Keywords: 3-D MDCTA; Reoperative Cardiac Surgery
1. INTRODUCTION
Reoperative cardiac surgery carries a significantly higher
risk compared to first time surgery. Several reasons exist;
technical difficulties include reoperative sternotomy, me-
diastinal dissection and higher risk in the reoperative
population. However, the cardiac surgeons are facing
more and more reoperations, given the longer expectancy
in the modern era. The recently published mortality rate
for reoperative cardiac surgery in previous coronary ar-
tery bypass graft surgery (CABG) has ranged from 2.5%
to 8.3% [1-7]. The major concern during sternal reentry
in these patients is the potential injury to the cardiac
structures (right ventricle), previous grafts, innominate
vein and Aorta. Surgical strategies, such as interposition
of the lungs in the left internal mammary artery (LIMA)
[8] or no dissection technique [9] have been described in
the literature, but do not eliminate the chances of injury
to these structures.
Contrast-enhanced multidetector computed tomographic
angiography (MDCTA) has emerged as a useful method
to evaluate heart and mediatinal structures [10-11]. In
addition, three-dimensional (3-D) volume rendering al-
lows detailed identification of the previous grafts as well
as the right ventricle and Aorta.
The aim of this paper is to support the usage of 3-D
MDCTA for preoperative evaluation for reoperative car-
diac surgery.
2. METHODS
2.1. 3-D MDCTA
State-of-the-art MDCTA for redo sternotomy typically is
performed with 64 - 320 detector row CT scanners.
Helical acquisitions are performed from above the clavi-
cular heads to the base of the heart with thin slice colli-
mation (0.5 - 1.0 mm) at either 100 - 120 kV and at least
350 mA. Current scanners employ z-axis tube current
modulation as an option for the mA adjustment. Since
small vessel CT angiography requires high contrast to
noise ratios, the highest available concentration of con-
trast (370 mg/dL) is injected at high flow rates (4.5 - 6
ml/sec). Volume of contrast is determined by the scan
time varying from 80 - 100 ml. B-blockade for heart rate
control and nitroglycerine for vasodilation are typically
not required for the assessment of vessel patency and
course; however, these techniques are favored if stenosis
evaluation of the prior bypass grafts is paramount. Since
bypass grafts don’t move with the same velocity as na-
tive coronary arteries, acquisitions can typically be per-
formed with prospective gating with a single phase of the
cardiac cycle. Dynamic assessment of retrosternal struc-
tures has recently been described as an adjunct to the
proximity assessment <ref Malguria, et al. “Static and
Cine CT Imaging to Identify and Characterize Mediasti-
nal Adhesions as a Potential Complication Undergoing
Redo Sternotomy”. AJR ACCEPTED FOR PUBLICA-
TION NOT YET IN PRESS. This technique requires
retrospective gating at the expense of a higher radiation
exposure; however, additional studies are required to
determine its value.
*Corresponding author.

T. Kaneko et al. / World Journal of Cardiovascular Diseases 3 (2013) 389-393 390
2.2. Interpretation of the 3-D MDCTA
Preoperative evaluation of 3-D MDCTA allows the fol-
lowing assessment.
1) Proximity of LIMA and Other Grafts in Relation to
the Sternum.
2) Adhesion of the Right Ventricle (RV) to the Sternal
Bed (Especially in Light of Pulmonary Hyper Tension).
3) Location of the Innominate Vein and Aorta.
Distance of <1 cm from the sternum to the Aorta, RV
and CABG graft crossing the midline within 1 cm of
posterior to the sternumwere typically used for high risk
upon sternal reentry [12-14].
A survey of 2046 significant bleeding during sternal
entry showed that the most commonly injured structures
were RV (39%), Saphenous vein graft (SVG: 20%),
Aorta (15%), IMA (12%) and innominate vein (6%) [15].
Radiologist and Surgeon must keep this in mind to find
the relationship between the sternum and these structures
when 3-D MDCTA is assessed preoperatively.
2.3. Operative Strategy Based on 3-D MDCTA
After assessing these structures, surgeons will consider
the following for operative planning.
2.3.1. Cannul ation Str ategy
1) Direct central cannulation.
2) Guidewire placement for emergency.
3) Peripheral cannulation.
a) Femoral artery cannulation.
b) Axillary cannulation.
First step of operative decision making is cannulation
strategy. Cannulation strategy must be decided prior to
incision with a backup plan. If there is good distance
from the sternum, sternotomy can be performed and di-
rect central cannulation can be performed following me-
diastinal dissection. Mediastinal dissection should be
performed around the Aorta and right atrium first so that
cardiopulmonary bypass (CPB) can be established in
case of emergency.
It is our common practice to place guidewire in femo-
ral artery and femoral vein to prepare for emergency.
Guidewire is inserted under transesophageal echocardio-
gram (TEE) guidance. If any structure is injured during
sternal reentry, sternum is approximated and CPB is es-
tablished following Seldinger technique placement of
arterial and venous cannula.
If adhesions present, peripheral cannulation should be
considered. Femoral artery/vein can be used. Axillary
artery can be used when femoral artery use is concern
(peripheral vascular disease, obesity etc). Our practice is
to place 8 - 10 mm dacron graft to preserve distal flow.
2.3.2. Cardio pulmonary By p ass
1) Go on cardiopulmonary bypass prior to sternotomy.
2) Deep Hypothermic circulatory arrest.
3) No cardiopulmonary bypass.
In severe adhesions, consideration of establishing CPB
prior to sternotomy should be made. This will allow de-
compression of RV and decreases the chances of injury
during sternal reentry. Luciani, et al. reported their ex-
perience with 158 patients who underwent cardiopul-
monary bypass prior to creating sternotomy [16]. This
decision was made on the CT evidence of dense retros-
ternal adhesions, depressed ejection fraction or previous
mediastinitis. Use of CPB prior to sternotomy was asso-
ciated with less re-entry injuries, reduced operative time,
less postoperative bleeding and shorter ICU stay.
Deep hypothermic circulatory arrest can be used for
extreme cases with severe adhesion or injury to the aorta
occurred. For low risk cases, no CPB will allow blood
less dissection since no heparin is used prior to dissec-
tion.
2.3.3. Sternotomy vs Non-Sternotomy
1) Reoperative sternotomy-grafts in relation to the sternal
wires.
2) Anterior thoracotomy.
3) Transcatheter treatment.
Access strategy can be altered depending on the op-
eration being performed. If prior CABG patient with
significant graft adhesion is undergoing mitral valve sur-
gery, literature supports use of right thoracotomy for ac-
cess [17]. Left thoracotomy can be used for reoperative
CABG to circumflex artery and avoiding sternal entry
[18].
If the patient has high risk comorbidity and high risk
findings on 3-D MDCTA, consideration of transcatheter
treatment (percutaneous coronary intervention for coro-
nary disease, transcatheter valve replacements for valve
disease) if indicated.
2.3.4. C anc ellation
When the findings show extremely high risk, surgeon
can decide to cancel the case. 3-D-CTMDA will aid this
decision making. Report from Cleveland Clinic de-
scribed 4% cancellation in their 167 adult reoperative
cardiac surgery series [14]. Patent RIMA with additional
jump graft supplying a left dominant system crossing the
sternal midline at an unsafe distance in a severe valvular
disease patient was one of the examples used in this se-
ries. Other series report 6% - 12% cancellation rate in
their series [13,19].
2.4. Patient Images
Figures 1(a) and (b) show the 3-D MDCTA on 65 years
old male who underwent CABG (SVG to posterior de-
scending artery) and came back for aortic valve replace-
ment (AVR). Preoperative MDCTA showed SVG graft

T. Kaneko et al. / World Journal of Cardiovascular Diseases 3 (2013) 389-393 391
(a)
(b)
Figure 1. 65 years old male after CABG. Previous graft is seen
below the sternum. Proximity of the graft and sternum was
considered high risk. This patient underwent femoral artery
cannulation prior to sternotomy. (Arrow shows the location of
the SVG).
lying behind the sternum <1 cm and was crossing mid-
line. This was considered high risk and operative strategy
was discussed preoperatively. Femoral artery cannulation
was performed prior to sternal reentry. Careful dissection
with anticipation of SVG behind the sternum allowed
safe dissection of the SVG. No injury was encountered
and AVR was performed successfully.
Figure 2 shows the 3-D MDCTA on 70-year-old male
who presented for Aortic stenosis and underwent CABG
(SVG to right coronary artery). He was scheduled for
AVR and preoperative 3-D MDCTA was performed to
assess the mediastinum. MDCTA showed previous SVG
traversing the midline and was densely attached to the
sternum. Preoperative strategy was changed to axillary
artery cannulation and femoral vein cannulation. We had
prepared to go on CPB prior to sternal reentry. During
sternal reentry, SVG injury was encountered and CPB
was established immediately. There were no hemody-
namic instability or ST changes. Mediastinal dissection
Figure 2. 70 years old, M status post CABG with prior SVG
graft to RCA. Graft was traversing the midline and was at-
tached to the sternum (white arrow). Knowing this, we per-
formed axillary artery cannulation and femoral vein cannula-
tion. There was a SVG graft injury during entry and cardio-
pulmonary bypass was established immediately. Mediastinal
dissection followed by SVG graft repair was performed primar-
ily. No hemodynamic instability was encountered as well as ST
changes.
was performed and SVG injury was repaired primarily.
Rest of the procedure was conducted without any prob-
lem.
3. DISCUSSION
3.1. Comparison to Chest X-Ray (CXR)
and Angiography
Gasparovic, et al. compared 3-D MDCTA and CXR/ co-
ronary angiography in their series of 33 patients who
underwent reoperative cardiac surgery after previous
CABG [13]. The correlation for distance of the LIMA
graft from midline and posterior sternum obtained by
CTA and CXR was poor (correlation coefficient: R =
0.56 and 0.49. The R for CTA and angiography was 0.54
for LIMA graft. One third of the CXR could not identify
LIMA graft (from the clips) and 85% of patients could
not obtain pertinent CTA finding from standard CXR and
Angiography.
No injury to the LIMA was reported in this case series.
This is significant improvement from 5.3% injury rate
from Cleveland clinic with perioperative infarction rate
of 50% which CXR and angiography was used for pre-
operative imaging [20].
3.2. Outcomes Using 3-D MDCTA
Early report from Cremer, et al. assessed their 99 adult
patient experiences that had reoperative coronary artery

T. Kaneko et al. / World Journal of Cardiovascular Diseases 3 (2013) 389-393 392
surgery with preoperative spiral CT without contrast an-
giography [21]. No MDCTA was used as well as 3-D
reconstructions, but safe sternal reentry was obtained in
98% and only 2% had injury to vital structures. This
opened the door for preoperative assessment using CT.
In their 33 patient series, Gasparovic, et al. reported
21% alteration in surgical strategy [13]. Two injuries to
vital structures were encountered, right ventricle in 1
case, vein graft in 1 case, but no IMA and aortic injury.
Operative mortality was 17%.
Kamder, et al. reported the largest series to date in
their 167 adult patients with reoperative cardiac surgery
who had prior CABG [14]. They used contrast enhanced
MDCTA for preoperative evaluation. The high risk find-
ings were <1 cm distance between chest wall and right
ventricle or aorta (24%) and graft crossing midline <1
cm anteroposteriorly (38%). 49% had 1 or more high risk
findings. 86% had alteration in surgical strategy: 8%
nonmidline incision, 5% deep circulatory hypothermic
arrest, 11% initiation of peripheral CPB, 53% extratho-
racic vascular exposure before incision and 4% cancella-
tion. Frequency of severe bleeding, injury to vital struc-
tures and 1-month mortality were 4.4%, 5% and 2.5%.
There was no difference in frequency of vital structure
injury and severe bleeding between high-risk and low-
risk MDCTA groups, likely from adoption of preventive
surgical strategies.
Maluenda, et al. reported 137 patients who underwent
MDCTA prior to reoperative cardiac surgery and was
compared to 227 patients who did not have MDCTA
prior to reoperative cardiac surgery [12]. MDCTA group
had lower incidence of perioperative myocardial infarc-
tion, shorter perfusion time and cross clamp time, total
time in intensive care unit and lower volume of postop-
erative transfusion.
3.3. Additional Information
Additional information can be found during CTA as-
sessment. In Gasparovic’s report, 42% had atheroscle-
rotic ascending aortic and aortic arch disease and 3% had
ascending aortic aneurysm. Persistent left superior vena
cava, innominate and superior vena cava stenosis was
found in 9.1%. Of these findings, atherosclerosis of the
aorta carries significance. In case of unclampable aorta,
cannulation stratery (peripheral cannulation) will be al-
tered and may require deep hypothermic circulatory ar-
rest to replace and clamp the aorta.
3.4. Cost
Goldstein, et al. reported their data on cost data in pa-
tients who underwent MDCTA prior to reoperative car-
diac surgery and compared to group who did not undergo
MDCTA [22]. $1150 was charged to MDCTA patients
for preoperative testing. MDCTA group had lower me-
dian recovery room and nursing charges, but higher op-
erative room charges. Despite the additional charge for
MDCTA, MDCTA group had lower mean total hospital
charges ($163,000 vs $192,000; p ≤ 0.06).
3.5. Upcoming Technology
Dynamic MDCTA has recently been reported <ref Mal-
guria, et al. “Static and Cine CT Imaging to Identify and
Characterize Mediastinal Adhesions as a Potential Com-
plication Undergoing Redo Sternotomy.” AJR ACCEP-
TED FOR PUBLICATION NOT YET IN PRESS. In
addition to the distance used in most MDCTA studies,
this technique will take structure’s movement into ac-
count, therefore called 4-D-MDCTA. For instance, if the
RV is in close proximity to sternum yet has free move-
ment, this means that it may not be adherent or that the
adhesions are not severe. If the RV does not move at the
attachment to the sternum, this likely indicates severe
adhesion.
4. CONCLUSION
Although no randomized prospective study has been
performed to prove the role of preoperative MDCTA, it
is obvious that new piece of information provided by
MDCTA will give important clues to the surgeons per-
forming reoperative sternotomy. High risk findings will
alter surgeon’s operative strategy and may even lead to
cancellation or seek other procedures such as transcathe-
ter procedure. In conclusion, 3-D MDCTA provides im-
portant information in a high risk reoperative cardiac sur-
gery and lowers the frequency of injury to vital struc-
tures.
REFERENCES
[1] Weintraub, W.S., Jones, E.L., Craver, J.M., Grosswald, R.
and Guyton, R.A. (1995) In-hospital and long-term out-
come after reoperativecoronary artery bypass graft sur-
gery. Circulation, 9, II50-II57.
doi:10.1161/01.CIR.92.9.50
[2] Shapira, I., Isakov, A., Heller, I., Topilsky, M. and Pines,
A. (1999) Long-termfollow-up after coronary artery by-
pass grafting reoperation. Chest, 115, 1593-1597.
doi:10.1378/chest.115.6.1593
[3] Yau, T.M., Borger, M.A., Weisel, R.D. and Ivanov, J.
(2000) The changingpattern of reoperative coronary sur-
gery: Trends in 1230consecutive reoperations. The Jour-
nal of Thoracic and Cardiovascular Surgery, 120, 156-
163. doi:10.1067/mtc.2000.106983
[4] Stephan, W.J., O’Keefe, Jr. J.H., Piehler, J.M., et al.
(1996) Coronaryangioplasty versus repeat coronary artery
bypass graftingfor patients with previous bypass surgery.
Journal of the American College of Cardiology, 28,
1140-1146. doi:10.1016/S0735-1097(96)00286-0
[5] Akl, E.S., Ozdogan, E., Ohri, S.K., et al. (1992) Early

T. Kaneko et al. / World Journal of Cardiovascular Diseases 3 (2013) 389-393 393
and long termresults of re-operation for coronary artery
disease. British Heart Journal, 68, 176-180.
doi:10.1136/hrt.68.8.176
[6] Fitzgibbon, G.M., Kafka, H.P., Leach, A.J., Keon, W.J.,
Hooper, G.D. and Burton, J.R. (1996) Coronary bypass
graft fate and patient outcome: Angiographic follow-up
of 5065 grafts related to survival andreoperation in 1388
patients during 25 years. Journal of the American College
of Cardiology, 28, 616-626.
doi:10.1016/0735-1097(96)00206-9
[7] Shimada, Y., Dixit, A., Fermanis, G. and Horton, D.
(1998) Reoperation forrecurrent coronary artery disease:
Results of 200 consecutivecases. Australian and New
Zealand Journal of Surgery, 68, 729-734.
doi:10.1111/j.1445-2197.1998.tb04661.x
[8] Fullerton, D.A., St Cyr, J.A., Fall, S.M. and Whitman,
G.J. (1994) Protection of the patent internalmammary ar-
tery bypass graft from subsequent sternotomy. The Jour-
nal of Cardiovascular Surgery (Torino), 35, 499-501.
[9] Kaneko, T., Nauta, F., Borstlap, W., McGurk, S., Couper,
G. and Cohn, L. (2012) The “no-dissection” technique is
safe for reoperative aortic valvereplacement with a patent
left internal thoracic artery graft. The Journal of Thoracic
and Cardiovascular Surgery, 144, 1036-1041.
doi:10.1016/j.jtcvs.2012.07.057
[10] Song, M.H., Ito, T., Watanabe, T., Nakamura, H. (2005)
Multidetectorcomputed tomography versus coronary an-
giogram in evaluationof coronary artery bypass grafts.
The Annals of Thoracic Surgery, 79, 585-588.
doi:10.1016/j.athoracsur.2004.08.014
[11] Anders, K., Baum, U., Schmid, M., et al. (2006) Coro-
nary artery bypassgraft (CABG) patency: Assessment
with high-resolution submillimeter16-slice multidetector-
row computed tomography (MDCT) versus coronary an-
giography. European Journal of Radiology, 57, 336-344.
doi:10.1016/j.ejrad.2005.12.018
[12] Maluenda, G., Goldstein, M.A., Lemesle, G., Weissman,
G., Weigold, G., Landsman, M.J., Hill, P.C., Pita, F.,
Corso, P.J., Boyce, S.W., Pichard, A.D., Waksman, R.
and Taylor, A.J. (2010) Perioperative outcomes in reop-
erative cardiac surgery guided by cardiac multidetector
computed tomographic angiography. American Heart
Journal, 159, 301-306. doi:10.1016/j.ahj.2009.11.005
[13] Gasparovic, H., Rybicki, F.J., Millstine, J., Unic, D.,
Byrne, J.G., Yucel, K. and Mihaljevic, T. (2005) Three
dimensional computedtomographic imaging in planning
the surgical approach for redo cardiac surgery after coro-
nary revascularization. European Journal Cardio-Tho-
racic Surgery, 28, 244-249.
doi:10.1016/j.ejcts.2005.03.024
[14] Kamdar, A.R., Meadows, T.A., Roselli, E.E., Gorodeski,
E.Z., Curtin, R.J., Sabik, J.F., Schoenhagen, P., White,
R.D., Lytle, B.W., Flamm, S.D. and Desai, M.Y. (2008)
Multidetector computed tomographic angiography in plan-
ning of reoperative cardiothoracic surgery. The Annals of
Thoracic Surgery, 85, 1239-1245.
doi:10.1016/j.athoracsur.2007.11.075
[15] Follis, F.M., Pett, Jr. S.B., Miller, K.B., Wong, R.S.,
Temes, R.T. and Wernly, J.A. (1999) Catastrophic hem-
orrhage on sternal reentry: Still a dreaded complication?
The Annals of Thoracic Surgery, 68, 2215-2219.
doi:10.1016/S0003-4975(99)01173-X
[16] Luciani, N., Anselmi, A., De Geest, R., Martinelli, L.,
Perisano, M. and Possati, G. (2008) Extracorporeal circu-
lation by peripheral cannulation before redo sternotomy:
Indications and results. The Journal of Thoracic and Car-
diovascular Surgery, 136, 572-577.
doi:10.1016/j.jtcvs.2008.02.071
[17] Byrne, J.G., Aranki, S.F., Adams, D.H., Rizzo, R.J., Cou-
per, G.S. and Cohn, L.H. (1999) Mitral valve surgery af-
ter previous CABG with functioning IMA grafts. The
Annals of Thoracic Surgery, 68, 2243-2247.
doi:10.1016/S0003-4975(99)01120-0
[18] Byrne, J.G., Aklog, L., Adams, D.H., Cohn, L.H. and
Aranki, S.F. (2001) Reoperative CABG using left thora-
cotomy: A tailored strategy. The Annals of Thoracic Sur-
gery, 71, 196-200. doi:10.1016/S0003-4975(00)02182-2
[19] Aviram, G., Sharony, R., Kramer, A., Nesher, N., Lober-
man, D., Ben-Gal, Y., Graif, M., Uretzky, G. and Mohr,
R. (2005) Modification of surgical planning based on car-
diac multidetector computed tomography in reoperative
heart surgery. The Annals of Thoracic Surgery, 79, 589-
595. doi:10.1016/j.athoracsur.2004.07.012
[20] Gillinov, A.M., Casselman, F.P., Lytle, B.W., et al. (1999)
Injury to a patent left internal thoracic artery graft at co-
ronary reoperation. The Annals of Thoracic Surgery, 67,
382-386. doi:10.1016/S0003-4975(99)00009-0
[21] Cremer, J., Teebken, O.E., Simon, A., Hutzelmann, A.,
Heller, M. and Haverich, A. (1998) Thoracic computed
tomography prior to redo coronary surgery. European
Journal Cardio-Thoracic Surgery, 13, 650-654.
doi:10.1016/S1010-7940(98)00087-6
[22] Goldstein, M.A., Roy, S.K., Hebsur, S., Maluenda, G.,
Weissman, G., Weigold, G., Landsman, M.J., Hill, P.C.,
Pita, F., Corso, P.J., Boyce, S.W., Pichard, A.D., Waks-
man, R. and Taylor, A.J. (2013) Relationship between
routine multi-de- tector cardiac computed tomographic
angiogra- phy prior to reoperative cardiac surgery, length
of stay, and hospital charges. The International Journal
of Cardiovascular Imaging, 29, 709-717.
doi:10.1007/s10554-012-0133-2