A RANDOMIZED STUDY OF The impact of perfusion technique and mode of pH management during

THE INFLUENCE OF cardiopulmonary bypass has not been well characterized with respect to
PERFUSION TECHNIQUE postoperative cardiovascular outcome. Methods: This double-blind, random-
AND PH MANAGEMENT ized study comparing outcomes after alpha-stat or pH-stat management and
STRATEGY IN 316 PATIENTS pulsatile or nonpulsatile perfusion during moderate hypothermic cardiopul-
UNDERGOING CORONARY monary bypass was undertaken in 316 patients undergoing coronary artery
ARTERY BYPASS SURGERY bypass operations. Results: Cardiovascular morbidity and mortality were not
affected by pH management, and the incidence of stroke (2.5%) did not differ
I. Mortality and between groups. Overall in-hospital mortality was 2.8%, eight of the nine
cardiovascular morbidity deaths occurring in the nonpulsatile group (5.1% versus 0.6%;p = 0.018). The
incidence of myocardial infarction was 5.7% in the nonpulsatile group and
0.6% in the pulsatile group (p = 0.010), and use of intraaortic balloon
pulsation was significantly more common in the nonpulsatile group (7.0%
versus 1.9%; p = 0.029). The overall percentage of patients having major
complications was also significantly higher in the nonpulsatile group (15.2%
versus 5.7%; p = 0.006). Duration of cardiopulmonary bypass, age, and use of
nonpulsatile perfusion all correlated significantly with adverse outcome.
Conclusions: Use of pulsatile perfusion during cardiopulmonary bypass was
associated with decreased incidences of myocardial infarction, death, and
major complications. (J TnORAC CARDIOVASCSURG 1995;110:340-8)

J. M. Murkin, MD, FRCPC,a J. S. Martzke, PhD, RPsych,b

A. M. Buchan, MD, FRCPC,c C. Bentley, RN,a and C. J. Wong, MSc,d
London and Ottawa, Ontario, and Vancouver, British Columbia, Canada

Ptimal perfusion characteristics during cardio- amine levels, 9' 10 renin activity, 11 angiotensin, aldo-
O pulmonary bypass (CPB) remain controver-
sial. 1'2 The use of pulsatile perfusion has been
sterone, 12 and lactate 13 levels, and to provide better
preservation of pituitary responsiveness. 14' 15
variously demonstrated to improve myocardial per- Many clinical studies have also demonstrated
fusion,3, 4 oxygenation,4-6 compliance, and indices of improved myocardial function and lower morbidity
contractility,7, 8 as well as to lower plasma catechol- and mortality after pulsatile compared with nonpul-
satile perfusion during CPB. 9' 16-19 In contradistinc-
From the Department of Anaesthesia, University Hospital, Uni- tion, however, a seemingly equal number of clinical
versity of Western Ontario, a and Clinical Trials Resources studies have been unable to detect benefit associ-
Group, Robarts Research Institute, d London, Ontario; the
Department of Psychology, Vancouver Hospital and Health
ated with pulsatile perfusion.2°-26
Sciences Centre, b Vancouver, British Columbia; and the Ideal pH management during moderate hypo-
Department of Neurology, Civic Hospital, University of Ot- thermia for CPB has also been the subject of
tawa, c Ottawa, Ontario, Canada. contention. Several studies have assessed the influ-
Supported by grant A1498 from the Heart and Stroke Founda- ence of pH management on brain and heart func-
tion of Ontario. tion, and most have demonstrated relative improve-
Presented in part at the Sixty-seventh Congress of the Interna- ments with alpha-stat in comparison with pH-stat
tional Anesthesia Research Society, San Diego, Calif., March
1993.
management.27-29 Some animal studies have also
Received for publication Aug. 2, 1994. shown improved myocardial performance and an
Accepted for publication Dec. 22, 1994.
elevated fibrillation threshold with alpha-stat man-
agement,27, 2s, 30, 31 whereas others have been un-
Address for reprints: John M. Murkin, MD, FRCPC, Depart-
ment of Anaesthesia, University Hospital, 339 Windermere able to detect differences in the functional capacity
Rd., London, Ontario, Canada N6A 5A5. of the myocardium whether the blood peffusate was
Copyright © 1995 by Mosby-Year Book, Inc. controlled by means of either alpha-stat or pH-stat
0022-5223/95 $3.00 + 0 12/1/63340 management strategies. 32

340
The Journal of ]thoracic and
Cardiovascular :Surgery Murkin et al. 34 1
Volume 110, Number 2

Prospective, randomized outcome studies assess- supplemented as necessary with volatile anesthetics, were
ing the differential impact of perfusion technique or used similarly in all groups. After sternotomy and hepa-
rinization, moderate systemic hypothermia (nasopharyn-
p H m a n a g e m e n t strategy on mortality and cardio-
geal temperature 26° to 28° C) with a hollow-fiber mem-
vascular morbidity have been conducted. However, brane oxygenator (Cobe CML, Cobe Stockert Inc.,
few have used sufficiently large sample sizes to Lakewood, Colo.; Terumo Capiox E, Terumo Corp.,
demonstrate possible outcome benefit for such low Tokyo, Japan) having a 40 /~m arterial line filter and
base rate events as death and myocardial infarction nonocclusive roller pump, was used for CPB in all pa-
tients. Flows of 2.0 to 2.5 L. rain - 1 , m -2 were used.
(MI). In most cardiac centers, use of nonpulsatile
Crystalloid cardioplegia was used by two surgeons, the
perfusion during CPB continues to be routine, al- other two used a 4:1 blood/crystalloid cardioplegic mix-
though use of alpha-stat p H m a n a g e m e n t has been ture, and all surgeons used topical cooling with saline
increasing over the past decade. We therefore un- slush. Ventricular venting was either through the ascend-
dertook a prospective, randomized, double-blind ing aorta or the pulmonary vein. After completion of all
study in 316 patients undergoing coronary artery distal coronary anastomoses, the crossclamp was released,
rewarming commenced, and proximal graft anastomoses
bypass (CAB) operations to assess differences in performed by means of side-clamping of the aorta.
clinical outcome after use of pulsatile or nonpulsa- In the pulsatile group a pump flow interrupter (Pul-
tile CPB and alpha-stat or pH-stat p H m a n a g e m e n t satile Flow Controller II, Cobe Stockert Inc.), initially
strategies. set at rate of 65 cps with an ejection time of 65% of
cycle time, was used from placement of the aortic
crossclamp until commencement of ventricular ejection
Methods after rewarming. Analog displays of pulse pressure and
This prospective, double-blind, clinical trial, random- mean arterial pressure were recorded continuously
ized with respect to pH management and perfusion tech- and analyzed specifically after 10, 30, and 60 minutes of
nique, was conducted as a part of a study assessing hypothermic CPB, at commencement of rewarming,
neurologic and cognitive outcomes in patients having and after rewarming was completed (i.e., rectal temper-
cardiac operations) 3 ature at or above 34 ° C). Hemodynamic data during
Study population. After institutional review board ap- hypothermia are reported as the means of all specific
proval and written informed patient consent were ob- time point values during stable hypothermia. Values for
tained, 316 :patients undergoing hypothermic CPB for normothermia were based on those obtained after
CAB (excluding concomitant open chamber procedures) completion of rewarming.
were enrolled in this study. To account for differences in Arterial blood gas analysis was performed with a Radi-
surgical technique, randomization was stratified by sur- ometer ABL2 analyzer (Radiometer Als, Copenhagen,
geon (four surgeons) and patients were randomly assigned Denmark). For alpha-stat management during CPB, arte-
to undergo either pulsatile or nonpulsatile perfusion and rial blood gas was measured at 37 ° C and carbon dioxide
alpha-stat or pH-stat pH management. was adjusted to produce an arterial carbon dioxide tension
Outcome measures. Each patient was assessed before of 40 mm Hg and an arterial pH of 7.4. For pH-stat
and after the operation by a blinded research nurse, while management, arterial blood gas was measured at 37°C
an independent study technician obtained intraoperative and corrected to the patient's nasopharyngeal tempera-
hemodynamic data. Mortality and major complications, ture with exogenous carbon dioxide adjusted to maintain
including MI, cerebrovascular accident (CVA), arrhyth- a temperature-corrected arterial carbon dioxide tension of
mia, use of an intraaortic balloon pump (IABP), and renal 40 mm Hg and a temperature-corrected arterial pH of 7.4.
impairment, were recorded before the operation for each During CPB, intermittent blood gas analyses, based on
patient. Arrhythmia was defined as onset of atrial or samples drawn at intervals corresponding to the specific
ventricular arrhythmias necessitating cardioversion or in- time points for hemodynamic measurements, were used to
stitution of specific antiarrhythmic therapy in the postop- maintain either temperature corrected (pH-stat manage-
erative period. CVA was diagnosed on the basis of clinical ment) or non-temperature corrected (alpha-stat manage-
presentation and either confirmatory brain imaging or ment) arterial pH 7.4 and arterial carbon dioxide tension
postmortem evidence of cerebral infarction. MI was diag- 40 mm Hg. Simultaneously with arterial blood gas sam-
nosed on the basis of development of new Q waves on the pling, blood was obtained for glucose and hemoglobin
electrocardiogram and an elevation of creatinine kinase measurements.
myocardial band levels greater than 10% of total creatine Statistical analysis. Log-linear model analysis was first
kinase. MI data do not include patients dying during the performed to ensure that no interaction existed between
operation, in whom such data could not be obtained, pH management, perfusion technique, and outcome
although these latter patients were included in mortality events. Outcome events were compared between treat-
statistics. IABP use was defined as successful insertion of ment groups by the X2 test. Fisher's two-tailed exact test
the IABP. Renal impairment was defined as elevations was used if the expected cell sizes were small. Demo-
of creatinine concentration 50% above the upper limit of graphic characteristics were assessed similarly for cate-
laboratory reference normal values. goric variables. Two-way factorial analysis of variance was
CPB. Routine monitoring, including radial and pulmo- used to examine continuous variables and to confirm that
nary artery cannulation, and high-dose opioid anesthesia, pH management did not interact with perfusion tech-
 The Journal of Thoracic and
342 Murkin et al. Cardiovascular Surgery
August 1995

Table I. CAB group demographics (numbers of patients, or mean +_ standard deviation)

Alpha-star pH-stat

Pulsatile Nonpulsatile Pulsatile Nonpulsatile CAB total

(n = 79) (n = 79) (n = 79) (n = 79) (n = 316)
Age (yr) 60.9 + 8.7 61.2 - 7.8 60.2 -+ 8.5 61.4 + 8.4 60.9 + 8.3
Gender (M/F) 66/13 65/14 70/9 63/16 264/52
CVA/TIA 8 11 9 4 32 (10%)
DM 20 17 8 13 58 (18%)
MI <1/12 9 9 9 6 33 (10%)
Prey. C A B 4 9 8 7 28 (9%)
LVEF <50% 23 23 29 22 97 (31%)
LVEF <35% 5 6 4 2 17 (5%)

CVA/TIA, History of cerebrovascular accident or transient ischemic attack; DM, diabetes mellitus; MI <1//2, myocardial infarction in previous month; Prec.
CAB, previous coronary bypass surgery; LVEF <50%, LVEF <35% patients with left ventricular ejection fraction less than 50% or 35%.

Table II. Distribution and intraoperative characteristics by CAB group (numbers of patients, or mean +_
standard deviation)
Alpha-stat pH-stat

PulsatiIe Nonpulsatile Pulsatile Nonpulsatile CAB total

(n = 79) (n = 79) (n = 79) (n = 79) (n = 316)
Surgeon 1 22 24 24 24 94
Surgeon 2 26 25 22 26 99
Surgeon 3 5 6 6 4 21
Surgeon 4 26 24 27 25 102
NPT (° C) 28.5 + 2.1 28.5 + 1.2 28.5 + 2.0 28.2 _+ 1.2 28.4 _+ 1.7
Hgb (gm/L) 78.4 _+ 18.4 78.9 _+ 11.5 81.5 _+ 14.5 78.4 _+ 12.3 79.3 -+ 14
Glu (mmol/dl) 10.6 -+ 3.5 11.3 -+ 3.0 10.4 _+ 3.1 10.4 -+ 3.0 10.7 + 3.1
CAB 2.9 _+ 0.6 2.9 _+ 0.8 3.0 +_ 0.8 3.0 +- 0.8 3.0 -+ 0.7
CAB > 4 10 16 22 17 65 (21%)
A o X C (min) 43.9 _+ 17.4 44.1 +_ 18.1 47.8 _+ 17.2 46.9 _+ 15.5 45.7 +_ 17
CPB (rain) 92.1 +_ 29.1 95.1 _+ 29.1 98.5 + 36.7 98.6 _+ 32.6 96.1 _+ 32
O R (min) 271 -4- 48 280 _+ 76 279 _+ 59 282 _+ 57 278 _+ 61
Hosp (days) 13.4 _+ 10.8 12.2 _+ 5.1 14.8 _+ 24.8 12.9 _+ 11.1 13.3 + 15

Surgeon N,, Number of patients in each group operated on by individual surgeon; NPT, mean nasopharyngeal temperature during hypothermic CPB; Hgb,
lowest mean hemoglobin concentration during hypothermic CPB; Glu, highest mean glucose values during hypothermic CPB; CAB, mean number of coronary
bypass grafts; CAB >4, patients having four or more coronary bypass grafts; AoXC, aortic crossclamp time; CPB, duration of cardiopulmonary bypass; OR,
duration of operation; Hosp, operation to discharge time with in-hospital deaths excluded.

nique. No adjustments were made for multiple compari- Alpha-stat versus pH-stat groups. For pH man-
sons. Logistic regression analysis was applied to examine agement groups, intended treatments during CPB
potential risk factors for adverse outcomes. were achieved such that there was a significant
difference in arterial pH of 0.11 __ 0.18 pH units and
Results of arterial carbon dioxide tension of 20.6 ± 7.8 mm
Characteristics of patients. Demographic and Hg between alpha-stat and pH-stat groups during
clinical characteristics of the 316 patients in the hypothermic CPB (Table III).
study are shown in Table I. Table II shows Pulsatile versus nonpulsatile groups. During hy-
distribution of patients into study group according pothermia, mean arterial pressure was similar be-
to surgeon and intraoperative and postoperative tween the two groups. Pulsatile perfusion was asso-
values. No significant differences were identified ciated with a significantly greater pulse pressure,
between groups in any parameters including de- averaging 14.8 ± 6.7 mm Hg higher during hypo-
mographics, CPB duration, ischemic time, ventric- thermic CPB and 6.8 ± 14 mm Hg higher during
ular function, or numbers of coronary vessels normothermic CPB than in the nonpulsatile group
bypassed. (see Table III). During CPB there were no differ-
The Journal of Thoracic and
Cardiovascular Surgery Murkin et al. 343
Volume 110, Number 2

Table III. Perfusion characteristics by CAB group (mean +_ standard deviation)

Alpha-stat pH-stat

CAB total PuIsatile Nonpulsatile Pulsatile Nonpulsatile

(n = 316) (n : 79) (n = 79) (n = 79) (n = 79)
PP-C (ram Hg) 16.9 _+ 6.1" 1.8 _+ 2.4 15.5 _+ 6.8* 1.6 ± 2.4
PP-NT (mm Hg) 11.1 ± 9.9* 5.5 ± 12.6 12.0 ± 9.3* 4.1 ± 7.3
MABP (mm Hg) 57.9 ± 12.3 61.0 ± 11.3 57.2 ± 10.1 57.6 ± 11.9
pHa'~ 7.42 ± 0.18:) 7.43 _+ 0.12:) 7.54 ± 0.13 7.54 ± 0.06
Paco2 (ram Hg)'~ 38.i ± 4.1:) 37.5 ± 4.1:) 57.5 ± 6.9 59.3 ± 6.4

PP-C, Mean pulse pressure during hypothermic CPB; PP-NT, pulse pressure during normothermic CPB; MABP, lowest mean arterial blood pressure during
hypothermic CPB; pHa, arterial pH; Paco2, arterial carbon dioxide tension.
*p < 0.001 versus nonpulsatile.
?Mean of all values as measured at 37° C during hypothermic CPB.
:~p = 0.00i versus pH-stat.

ences in urine output between pulsatile or nonpul- CVA with irreversible brain injury, and one patient
satile groups (367 __ 346 ml versus 378 _+ 358 ml, died 4 months after the operation of multisystem
respectively), consistent with what we had observed failure.
in a subset of 100 of these patients in whom frac- Cardiovascular risk factors. Univariate logistic
tional excretion of sodium and potassium, as well as regression was used to determine the relationship
renal failure index, did not differ between perfusion between the presence of adverse outcomes, defined
groups. 34 as death, MI, arrhythmia, or IABP insertion, and
Surgical morbidity and mortality. Data on car- potential risk factors, identified as insulin-depen-
diovascular and other major complications were dent diabetes mellitus, MI within the previous
assessed and patients with major complications are month, gender, left ventricular ejection fraction less
listed individually in Table IV. There was no differ- than 0.5, left ventricular ejection fraction less than
ence in cardiovascular outcomes related to pH 0.35, number of coronary arteries grafted, perfusion
management (e.g., mortality 3/158 versus 6/158; MI technique, pH management strategy, cardioplegia
3/158 versus 7/158; alpha-stat versus pH-stat, respec- type (blood or crystalloid), cardioplegia volume,
tively). A significant difference in cardiovascular age, duration of crossclamping, and duration of
outcomes was observed that strongly favored pulsa- CPB. Significant correlations were found for use of
tile perfusion, however. In the nonpulsatile group nonpulsatile perfusion (p = 0.0082), age (p =
there were significantly higher rates of mortality 0.0211), and CPB duration (p = 0.0006). When
(8/158 versus 1/158, p = 0.018), MI (9/158 versus examined multivariately, these factors remained sig-
1/158, p = 0.010), and IABP use (11/158 versus nificant risk factors (p = 0.0106,p = 0.0223, andp =
3/158, p = 0.029) than in the pulsatile group. These 0.0003, respectively) for adverse outcome. The other
factors contributed toward a decreased overall inci- factors listed did not correlate with adverse out-
dence of major complications in the pulsatile group come. Although cardioplegia type did not correlate
(24/158 versus 9/158, p = 0.006). The percentage of with adverse outcome, surgeons using crystalloid
patients who died or had major complications, bro- eardioplegia completed more CAB grafts (3.1 _+ 0.8
ken down by perfusion technique, is presented in versus 2.8 _ 0.6, respectively; p = 0.001) with
Fig. 1. TO assess the effect of patients with compli- shorter crossclamp times (41.5 _ 14.1 minutes ver-
cations disproportionately influencing morbidity sus 59.3 ± 13.3 minutes, respectively; p < 0.001)
data, we undertook a repeat analysis of complica- than those using blood cardioplegia.
tions involving only surviving patients. A trend (/9 =
0.062) for fewer MIs in the pulsatile group persisted Discussion
(1/157 versus 6/150, respectively). Of the patients The current study demonstrated significantly
who died, three died during the operation of cardiac lower incidences of mortality and cardiovascular
failure, three died within 14 days after the operation morbidity in patients undergoing pulsatile CPB but
of MI with irreversible low output state, one of no detectable influence of pH management strategy
whom also had a CVA, two died within 5 days of a on these outcome variables. The overall incidences
 The Journal of Thoracic and
344 Murkin et aL Cardiovascular Surgery
August 1995

Table IV. Individual mortality and major morbidity for pulsatile versus nonpulsatile CPB
Patient Death MI Arrhythmia CVA IABP Renal
Nonpulsatile CPB (n = 158)
1 X X X
2 X
3 X
4 X X
5 X X X
6 X X X
7 X X X
8 X X X X
9 X X X X X
10 X X
11 X X
12 X X
13 X X
14 X X
15 X
16
17 X
18 X
19 X
20 X
21 X
22 X
23 X X
24 X
Pulsatile CPB (n = 158)
1 X X
2 X
3 X
4 X X
5 X X
6 X X X
7
8 X
9 X X
p = 0.006* p = 0.018 p = 0.010 p = 0.157 p = 1.0 p = 0.029 p = 0.371

The incidence of death, myocardial infarction (MI), postoperative arrhythmia necessitating treatment, cerebrovascular accident (CVA), use of intraaortic
balloon pump (IABP), and renal insufficiency for those patients undergoing pulsatile versus nonpulsatile perfusion during cardiopulmonary bypass (CPB). The
p values comparing incidences of complications between the two groups are along the bottom of the table.
*This value refers to the relative incidences of any complications in the two groups.

of MI (3.2%), IABP use (4.4%), CVA (2.5%), and during which total carbon dioxide is kept constant
mortality (2.8%) reported in the current study are and pH and arterial carbon dioxide tension vary with
consistent with results reported in the literature. In body temperature. 29
a recent survey of 1513 patients undergoing CAB, In the current study, we did not find any influence
MI occurred in 5.5%, 4.7% required IABP, CVA of mode of pH management on cardiovascular
occurred in 2.8%, and overall mortality rate was morbidity or mortality. This observation likely re-
3.1%. 35 flects both the relatively insensitive clinical end
pit management. Two divergent strategies for points used (e.g., MI and IABP use) and, more
pH management have generally been used clinically: fundamentally, the fact that the heart is not perfused
(1) pH-stat--addition of exogenous carbon dioxide and is effectively excluded from the systemic circu-
to the oxygenator to achieve a temperature-cor- lation (and thus largely uninfluenced by mode of pH
rected arterial pH of 7.4 and an arterial carbon management) during hypothermia by virtue of the
dioxide tension of 40 mm Hg; (2) alpha-stat-- aortic crossclamp. Rewarming, and thus conver-
The Journal of Thoracic and
Cardiovascular Surgery Murkin et al. 345
Volume 110, Number 2

NON PULSATILE (n=158)

PULSATILE (n= 158)

p=0.029
7
p=0.157
6 p=0.01 N
0
Z
5
Cm
5 4
Z
N
3 ~=I.0

0
DEATH MI
/
ARRHYTHMIA OvA IABP

Fig. 1. The incidence of death, MI, postoperative arrhythmia necessitating treatment, CVA, and use of
IABP for those patients undergoing pulsatile versus nonpulsatile perfusion during CPB.

gence of both alpha-stat and pH-stat strategies, reveal any significant differences in preoperative risk
generally commences during performance of the last factors.
of the distal coronary anastomoses, while the heart To date only two other large-scale studies assess-
is still isolated from systemic perfusion. ing cardiovascular outcomes after pulsatile or non-
Pulsatile perfusion. The current study docu- pulsatile CPB have been presented, and they have
mented markedly lower mortality in patients under- similarly reported salutary effects of pulsatile perfu-
going pulsatile CPB. Although nonfatal adverse sion. Significantly decreased mortality rates with
outcomes (e.g., MI, IABP) and mortality over- pulsatile pcrfusion were reported by Taylor and
lapped, analysis of data from surviving patients still associates is in a series of 350 patients with cardiac
demonstrated a trend for a higher incidence of MI disease randomized to pulsatile or nonpulsatile per-
in the nonpulsatile group (6/150 versus 1/157). Rea- fusion; Taylor's group used a roller-pump system
sons for these differences in outcome remain spec- similar to the one we employed. In that series
ulative. They are unlikely to reflect differences in significantly lower requirements for postoperative
surgical technique, because stratification by surgeon circulatory support modalities were demonstrated,
ensured that the number of patients in both pulsatile and mortality from low-output cardiogenic shock
and nonpulsatile groups was balanced between sur- was significantly decreased from 6.3% to 1.1%.
geons. Assignment into pulsatile or nonpulsatile More recently, Minami and coworkers~9 reported
groups was also undertaken in a blinded fashion, that in a retrospective comparison of 175 patients
and there was no subsequent crossover during the undergoing CPB for greater than 120 minutes, pul-
operations. Disproportionate weighting of patients satile perfusion decreased the incidence of cardio-
at higher risk into one group or the other is thus vascular morbidity, although overall mortality was
unlikely, although no attempt was made preopera- not affected.
tively to stratify patients into treatment groups The mechanism responsible for the decreased
according to cardiovascular or other risk factors. incidence of cardiovascular complications remains
Review of patient characteristics (see Tables I and speculative. However, it is possible that during the
II) confirms the randomization process and does not early reperfusion phase reductions in neurohumoral
 The Journal of Thoracic and
346 Murkin et aL Cardiovascular Surgery
August 1995

stress responses and improvements in myocardial proved clinical cardiovascular outcomes demon-
contractility and subendocardial blood flow are con- strated with pulsatile flow.
tributory.3, 4, 7-10
In fibrillating hearts, increased myocardial oxygen We are indebted for the support and cooperation of
and lactate extraction, improved subendocardial cardiac surgeons G. M. Guiraudon, F. N. McKenzie,
blood flow,3' 4, 36-38 and increased diastolic compli- A. M. Menkis, and R. L. Novick, perfusionists A. Cleland,
ance 4 have been observed with pulsatile perfusion. M. Henderson, R. Mayer and J. MacDonald, and anes-
thesia research technician P. Lok. The assistance of D. A.
Clinical studies also demonstrated increased ejec- Sim, D. Sharma, P. Campbell, and D. Giles is also
tion fractions, lower incidences of MI, and increased acknowledged.
coronary graft blood flow. 16' 17, 39 In the presence of
acute coronary stenoses, pulsatile perfusion is asso- REFERENCES
ciated with higher oxygen tension, lower tissue car- 1. Hickey PR, Buckley MT, Philbin DM. Pulsatile and
bon dioxide tension, and greater subendocardial nonpulsatile cardiopulmonary bypass: review of a
blood flow.s' 38 A recent study of dogs with coronary counterproductive controversy. Ann Thorac Surg
artery ligation and profound heart failure has con- 1983;36:720-7.
firmed increased coronary blood flow and an im- 2. Bartlett RH. Nonpulsatile flow--a noncontroversy. J
proved ratio of diastolic time index/tension time TI~ORACCARDIOVASCSURG [Letter] 1994;107:644-5.
3. Steed DL, Follette DM, Foglia R, Maloney JV,
index with pulsatile as compared with nonpulsa-
Buckberg GD. Effects of pulsatile assistance and non-
tile perfusion for left ventricular bypass. 4° Lower pulsatile flow on subendocardial perfusion during car-
levels of thromboxane and increased concentra- diopulmonary bypass. Ann Thorac Surg 1978;26:133-41.
tions of prostacyclin, changes favoring coronary 4. Habal SM, Weiss MB, Spotnitz HM, et al. Effects of
vasodilatation and augmented myocardial blood pulsatile and nonpulsatile coronary perfusion on per-
flow, have also been demonstrated with pulsatile formance of the canine left ventricle. J Tnop,ac
perfusion. 41 CARDIOVASC SURG 1976;72:742-5.
However, other studies have been unable to dem- 5. Schaff HV, Ciardullo RC, Flaherty JT, Brawley RK,
onstrate any significant influence of perfusion tech- Gott VL. Regional ischemia distal to a critical coro-
nique. 21-26 Overall, benefit does appear to be more nary stenosis during prolonged fibrillation: improve-
ment with pulsatile perfusion. Circulation 1977;
likely at lower rather than higher perfusion flow 56(Suppl):I25-32.
rates, 2 in fibrillating rather than beating hearts, 3' 4, 6 6. Bixler TJ, Magee PG, Flaherty JT, Gardner TJ, Gott
and in the presence of global ischemia, a ventricular VL. Beneficial effects of pulsatile perfusion in the
hypertrophy,7 and acute s rather than chronic 42 isch- hypertrophied ventricle during ventricular fibrillation.
emia. Circulation 1979;60(Suppl):I141-6.
Whether any of these experimental conditions are 7. Dunn J, Peterson A, Kitsch MM. Effects of pulsatile
directly comparable to the clinical situation encoun- perfusion upon left ventricular function. J Surg Res
tered during CAB operations, given that the heart is 1978;25:211-6.
8. Mori F, Ivey TD, Itoh T, Thomas R, Breazeale DG,
isolated and nonperfused for over half the duration
Misbach G. Effects of pulsatile reperfusion on post-
of CPB, is unclear. It is apparent that pulsatile ischemic recovery of myocardial function after global
perfusion can significantly increase subendocar- hypothermic cardiac arrest. J THOe,AC CARDIOVASC
dial myocardial perfusion during ventricular fi- StJRa 1987;93:719-27.
brillation, particularly in the presence of coronary 9. Minami K, Korner MM, Vyska K, Kleesiek K,
stenoses. It can also enhance subsequent load- Knobl H, Korfer R. Effects of pulsatile perfusion on
dependent performance. In clinical practice the plasma catecholamine levels and hemodynamics dur-
heart is often nonbeating or fibrillating for a ing and after cardiac operations with cardiopulmonary
bypass. J Taop,AC CAP,DIOVASCSUR6 1990;99:82-91.
variable period after removal of the aortic cross-
10. Philbin DM, Levine FH, Kono K, et al. Attenuation of
clamp, during which time the proximal graft anas-
the stress response to cardiopulmonary bypass by the
tomoses are being made. During this critical addition of pulsatile flow. Circulation 1981;64:808-12.
postischemic reperfusion phase, it may be that 11. Canivet JL, Larbuisson R, Damas P, et al. Plasma
improved subendocardial perfusion, in conjunc- renin activity and urine/32-microglobulin during and
tion with increased prostacyclin and decreased after eardiopulmonary bypass: pulsatile vs non-pulsa-
catecholamine levels, may be factors in the ira- tile perfusion. Eur Heart J 1990;11:1079-82.
The Journal of Thoracic and
Cardiovascular Surgery Murkin et al. 347
Volume 110, Number 2

12. Nagaoka H, Innami R, Arai H. Effects of pulsatile 24. Kaul TK, Swaminathan R, Chatrath RR, Watson DA.
cardiopulrnonary bypass on the renin-angiotensin- Vasoactive pressure hormones during and after
aldosterone system following open heart surgery. Jpn cardiopulmonary bypass. Int J Artif Organs 1990;13:
J Surg 1988;18:390-6. 293-9.
13. Dapper F, Neppl H, Wozniak G, et al. Effects of 25. Goto M, Kudoh K, Minami S, et al. The renin-
pulsatile ,and nonpulsatile perfusion mode during angiotensin-aldosterone system and hematologic
extracorporeal circulation--a comparative clinical changes during pulsatile and nonpulsatile cardiopul-
study. Thorac Cardiovasc Surg 1992;40:345-51. monary bypass. Artif Organs 1993;17:318-22.
14. Taylor KM, Wright GS, Bain WH, Caves PK, Beastall 26. Levine FH, Phillips HR, Carter JE, et al. The effect of
GS. Comparative studies of pulsatile and nonpulsatile pulsatile perfusion on preservation of left ventricular
flow during cardiopulmonary bypass. III. Response of function after aortocoronary bypass grafting. Circula-
the anterior pituitary gland to thyrotropin-releasing tion 1981;64(Suppl):II40-4.
hormone. J THORACCARDIOVASCSURG 1978;75:579- 27. Becker H, Vinten-Johansen J, Buckberg GD, et al.
84. Myocardial damage caused by keeping pH 7.40 during
15. Komoda T, Maeta H, Imawaki S, Shiraishi Y, systemic deep hypothermia. J THORAC CARDrOVASC
Tanaka S. Hematologic and endocrinologic effects of SURG 1981;82:810-20.
pulsatile cardiopulmonary bypass using centrifugal 28. Swain JA, White FN, Peters RM. The effect of pH on
pump. J Jpn Orthodont Soc 1992;40:901-11. the hypothermic ventricular fibrillation threshold. J
16. Maddoux G, Pappas G, Jenkins M, et al. Effect of THORACCARDIOVASCSURG 1984;87:445-51.
pulsatile and nonpulsatile flow during cardiopulmo- 29. Swan H. The importance of acid-base management for
nary bypass on left ventricular ejection fraction early cardiac and cerebral preservation during open-heart
after aortocoronary bypass surgery. Am J Cardiol operations. Surg Gynecol Obstet 1984;158:391-414.
1976;37:1000-6. 30. McConnell DH, White F, Nelson RL, et al. Impor-
17. Bregman D, Bowman FO, Parodi E, et al. An tance of alkalosis in maintenance of "ideal" blood pH
improved method of myocardial protection with during hypotherima. Surg Forum 1975;26:263-5.
pulsation during cardiopulmonary bypass. Circulation 31. Hering JP, Schroder T, Singer D, Hellige G. Influence
1977;56:157-60. of pH management on hemodynamics and metabo-
18. Taylor KM, Bain WH, Davidson KG, Turner MA. lism in moderate hypothermia. J THORACCARDIOVASC
Comparative clinical study of pulsatile and non-pul- SURG 1992;104:1388-95.
satile perfusion in 350 consecutive patients. Thorax 32. Sinet M, Muffat-Joly M, Bendaace T, Pocidalo J-J.
1982;37:324-30. Maintaining blood pH at 7.4 during hypothermia has
19. Minami K, Dramburg W, Notohamiprodjo G, no significant effect on work of the isolated rat heart.
Korfer R. Effects of pulsatile perfusion on periopera- Anesthesiology 1985;62:582-7.
tive morbidity and mortality in high-risk patients. In: 33. Murkin JM, Martzke JS, Buchan AM, Bentley C,
Minami K, Korfer R, Wada J, eds. Cardio-thoracic Wong CJ. A randomized study of the influence of
surgery. What is new in current practice. Amsterdam: perfusion technique and pH management strategy in
Elsevier, 1992:67-75. 316 patients undergoing coronary artery bypass sur-
20. Alston RP, Murray L, McLaren AD. Changes in gery. II. Neurologic and cognitive outcomes. J THO-
hemodynamic variables during hypothermic cardio- RAC CARDIOVaSCSURG 1995;110:349-62.
pulmonary bypass: effects of flow rate, flow character, 34. Badner NA, Murkin JM, Lok P. Differences in pH
and arterial pH. J THORAC CARDIOVASCSURG 1990; management and pulsatile/nonpulsatile perfusion dur-
100:134-44. ing CPB do not influence renal function. Anesth Analg
21. Singh RKK, Barratt-Boyes BGF, Harris EA. Does 1992;75:696-701.
pulsatile flow improve perfusion during hypothermic 35. Jones EL, Weintraub WS, Craver JM, Guyton RA,
cardiopulmonary bypass? J THORAC CARDIOVASC Cohen CL. Coronary bypass surgery: Is the operation
StJR6 1980;79:827-32. different today? J THORAC CARDIOVASCSURG 1991;
22. Louagie YA, Gonzalez M, Collard E, et al. Does flow 101:108-15.
character of cardiopulmonary bypass make a differ- 36. Moores WY, Hannon JP, Crum JD, Willford DS.
ence? J THORACCARDIOVASCSURG 1992;104:1628- Continuous and pulsatile extracorporeal perfusion
38. in the beating and fibrillating swine myocardium:
23. Alston RP, Singh M, McLaren AD. Systemic oxygen effects on left ventricular function. Surg Forum
uptake during hypothermic cardiopulmonary by- 1977;28:262-4.
pass: effects of flow rate, flow character, and arterial 37. Birder TJ, Magee PG, Flaherty JT, Gardner TJ, Gott
pH. J THORAC CARDIOVASCSURG 1989;98:757-68. VL. Beneficial effects of pulsatile perfusion in the
 The Journal of Thoracic and
348 Murkin et aL Cardiovascular Surgery
August 1995

hypertrophied ventricle during ventricular fibrillation. the pulsatility of a new pulsatile left ventricular
Circulation 1979;60:141-6. assist device--the integrated cardioassist cathe-
38. Ciardullo RC, Schaff HV, Flaherty JT, Donahoo JS, t e r - i n dogs. J THORAC CARDIOVASC SURG 1994;
Gott VL. Comparison of regional myocardial blood 107:569-75.
flow and metabolism distal to a critical coronary 41. Watkins WD, Peterson MB, Kong DL, et al. Throm-
stenosis in the fibrillating heart during alternate peri- boxane and prostacyclin changes during cardiopnlmo-
ods of pulsatile and nonpulsatile perfusion. J THORAC nary bypass with and without pulsatile flow. J THORAC
CARDIOVASCSURG 1978;75:193-205. CARDIOVASCSURG 1982;84:250-6.
39. Ishikura Y, Tokunaga H, Shimazu A, et al. Clinical 42. Chitwood WR Jr, Sink JD, Hill RC, Wechsler AS.
evaluation of pulsatile perfusion for long cardiopul- The comparative effects of pulsatile and non-pulsatile
monary bypass. Kyobu Gek 1987;40:311-5. myocardial perfusion during cardiopulmonary bypass.
40. Ide H, Yamaguchi A, Ino T, et al. Evaluation of J Surg Res 1981;30:553-62.

O N THE MOVE?
D o n ' t miss a single issue of the journal! To e n s u r e p r o m p t service w h e n y o u c h a n g e y o u r address,
please p h o t o c o p y a n d c o m p l e t e the f o r m below.

Please send your change of address notification at least six weeks before your move to ensure continued service.
We regret we cannot guarantee replacement of issues missed due to late notification.
JOURNAL TITLE:
Fill in the title of the journal here.
OLD A D D R E S S : NEW A D D R E S S :
Affix the address label from a recent issue of the journal here. Clearly print your new address here.

Name
Address

City/State/ZIP

COPY A N D M A I L T H I S F O R M TO: OR FAX TO: OR PHONE:

Journal Subscription Services 314-432-1158 1-800-453-4351
Mosby-Year Book, Inc. OutsidetheU.S.,call
11830 Westline Industrial Dr. 314-453-4351
St. Louis, MO 63146-3318 Mosby