Northfield Laboratories Releases Summary Observations from Its Elective Surgery TrialANH Procedure Decision Tree (Graphic: Busi

Northfield Laboratories Inc. (NASDAQ: NFLD) announced
today that in light of heightened interest in its Phase III elective
surgery trial with PolyHeme conducted in the late 1990s, the Company
would provide an extended summary of the major observations
immediately. This release is in lieu of the previously announced
presentation of an abstract at the Network for Advancement of
Transfusion Alternatives (NATA) meeting in April. The purpose of this
summary is to describe the objectives of the study, the study design,
the procedure of acute normovolemic hemodilution (ANH), the population
studied, and the major safety and efficacy observations.

   Key points about the study and summary observations are:

   --  The study was designed to assess whether the use of PolyHeme
        would allow an increase in the volume of autologous blood
        collected during ANH and therefore avoid transfusion of
        donated blood.

   --  The study demonstrated that a greater volume of the patient's
        blood could be collected during ANH with the use of PolyHeme.

   --  The study did not achieve its objective of avoiding
        transfusion of donated blood.

   --  The algorithm in the protocol resulted in multiple treatment
        differences between the two groups in the study.

   --  Ten patients in the treatment group experienced myocardial
        infarction versus none in the control group. Two of these
        patients died, at 7 and 32 days.

   --  The cardiovascular events noted are in sharp contrast to our
        trauma experience in which PolyHeme has been given in doses up
        to 20 units and has been well-tolerated.

   --  It cannot be determined whether the cardiovascular events were
        due to the more extensive ANH in the treatment group, to the
        reinfusion of more blood following surgery in the treatment
        group or to PolyHeme(R) itself.

   --  Northfield does not believe the cardiovascular events were due
        to a direct pharmacologic effect of PolyHeme, but to complex
        fluid management issues in these patients.

   Acute Normovolemic Hemodilution (ANH) Study
   Summary of Major Observations

   Background

   The Phase III elective surgery study discussed below was conducted
between 1998 and 2000. Northfield was concurrently conducting a Phase
II study with PolyHeme in trauma patients in the hospital setting,
with doses up to 20 units.
   Enrollment in the elective surgery study progressed slowly due to
a number of factors, as Northfield reported at the time. The protocol
was complex and required additional time in the operating room.
Endovascular stent grafting was being developed and was changing the
surgical approach to repair of abdominal aortic aneurysm. The study
was originally slated to enroll 240 patients. However, in August 1998,
FDA requested that the number of patients be increased to 600, based
on another sponsor's safety experience with a different
hemoglobin-based oxygen carrier in the trauma setting. A single
interim analysis was scheduled to be conducted by an Independent Data
Monitoring Committee (IDMC) after 120 patients were enrolled. At the
initial review, the IDMC noted certain differences between the two
study groups. The IDMC asked that the data be unblinded and additional
analysis be undertaken to assess whether the observed differences were
due to randomization failure, a treatment confounder, or a possible
study drug effect. At the time of this request, 138 patients had been
enrolled. The analysis revealed the myocardial infarctions (MIs) in
the treatment group, and also revealed that the incidence was not
evenly distributed among participating sites, which would have been
expected if the MIs were due to a direct effect of PolyHeme.
   The above findings were reported to FDA. Northfield also informed
FDA that in view of the complexity of the protocol and the slow rate
of accrual, it had begun to close slow enrolling sites and would
submit a study report to FDA.

   Study Objectives

   This was a randomized, controlled, open-label, multi-center,
parallel group, Phase III protocol designed to evaluate the safety and
efficacy of PolyHeme, an investigational human hemoglobin-based oxygen
carrier, used as a red blood cell substitute during acute normovolemic
hemodilution (ANH) and the peri-operative period in adult patients
experiencing planned acute blood loss during elective surgery for
abdominal aortic aneurysm (AAA). ANH is a procedure of withdrawing a
unit of blood from a patient and replacing the volume with colloid
solutions that replace the withdrawn volume, but do not carry oxygen.
The goal of ANH is to minimize the loss of the patient's own
(autologous) red cells during surgery and potentially reduce or avoid
the use of allogeneic (donated) transfusion. ANH is generally limited
to withdrawing only two units of the patient's own blood, because if
more is withdrawn, the patient's hemoglobin falls to an unacceptably
low level.
   The primary objective of this study was to evaluate whether using
PolyHeme during ANH would allow for the collection of larger volumes
of the patient's own blood to be stored for later return to the
patient, thus potentially reducing or avoiding transfusion of donated
blood throughout the first seven days postoperatively. The potential
to avoid transfusion exists because PolyHeme is a colloid that also
provides supplemental hemoglobin to replace that withdrawn during ANH.
Other protocol objectives were to evaluate the safety profile of
PolyHeme through assessment of patients' vital signs, laboratory
measurements, and adverse events. Safety was assessed through 12 weeks
(84 days) after surgery.
   Two hundred and forty patients were to be randomized to one of two
groups: either the treatment group (replacement with standard
solutions and PolyHeme) or control group (replacement with standard
solutions only).

   Study Design

   The study was organized into four distinct phases: the period
before surgery, the period during surgery, the period following the
surgery, and the follow-up period. Up to six units of PolyHeme could
be infused into the treatment group during the pre-operative and
intra-operative periods only.

   Before the Surgery

   In order to be eligible for this study, patients were required to
have a total hemoglobin concentration of at least 11 g/dL (normal
adult circulating hemoglobin concentration is approximately 12-14
g/dL). There were no specific exclusion criteria related to
cardiovascular disease or risk factors. All patients provided informed
consent. Patients underwent ANH in the operating room. ANH was
performed by removing a 500 mL-unit of blood (ANH unit) and replacing
it with a 500 mL colloid infusion. A maximum of approximately 60% of
the blood volume of each patient, usually about 6 units, could
potentially be collected based on body weight. The blood collected
during ANH was replaced with the colloid Hespan for the first two
units, and then with the colloid 5% albumin, as long as the total (Hb)
was greater than or equal to 9 g/dL. The use of Hespan was limited to
two units since more could produce coagulation disturbances. Total
hemoglobin concentrations were measured after each unit of blood had
been removed and the volume had been replaced.
   If the patient's hemoglobin concentration fell to less than 9 g/dL
before all planned units had been obtained, the treatment and control
groups were treated differently. In the control group, ANH was stopped
before the maximum numbers of ANH units were collected. In the
treatment group, ANH continued using PolyHeme because PolyHeme, in
contrast to Hespan or 5% albumin, provides supplemental hemoglobin to
replace the hemoglobin lost as blood was withdrawn during ANH.
   In both the control and the treatment groups, ANH was to be
stopped at any time if there was any evidence of unstable vital signs,
cardiac ischemia, or other clinical evidence of the patient's
inability to tolerate further anemia. Even if ANH were stopped before
the planned maximum number of units was collected, the patient would
remain in the study for evaluation.

   The following schematic illustrates the complexity of the ANH
portion of the study. (See graph.)

   During Surgery

   The transfusion decision during surgery was made in the
conventional manner in both groups, based on clinical signs, vital
signs, and hemoglobin levels. Total hemoglobin was measured hourly.
During the surgery, any blood the patient lost was also collected and
reinfused prior to each hourly hemoglobin determination. If the total
hemoglobin was greater than or equal to 8 g/dL, and the patient was
clinically stable, no transfusion was given. If the patient's total
hemoglobin was less than 8 g/dL, the control group received a
transfusion of either autologous blood (ANH units) or allogeneic blood
and the treatment group received either PolyHeme, ANH units or
allogeneic blood, one unit at a time, until total hemoglobin was
greater than or equal to 8 g/dL, even if the patient was clinically
stable.
   In the control group, the first transfusions given during surgery
were the ANH units. The units were transfused on a "last unit out,
first unit in" basis. In the treatment group, PolyHeme was transfused
first. If a total of six units of PolyHeme had been infused during the
pre-operative and operative period and further transfusion was
required, ANH units were given, as needed, on the "last unit out,
first unit in" basis.
   In both groups, if all collected ANH units had been transfused and
further transfusions were required, allogeneic blood was to be
administered unit-by-unit, as needed.

   Following Surgery

   When the operation had been completed, the collected ANH units or
allogeneic blood were transfused as needed for patients to achieve
hemoglobin concentrations greater than or equal to 9 g/dL. Any
autologous blood not used within twenty-four hours was discarded, in
accordance with current blood banking standards.

   Follow-up Period

   At six and twelve weeks, patients returned to the hospital for
assessment of vital signs, laboratory measurements, and adverse
events.

   Study Population

   One hundred and fifty two (152) patients were enrolled at 18
different sites. The number of patients enrolled at participating
sites ranged from one patient at some centers to 43 patients at the
highest enrolling site. There were 81 patients randomized to receive
PolyHeme (treatment group) and 71 randomized to the control group.
Nearly half of the 152 patients (71 patients or 47%) were enrolled at
two sites (43 patients and 28 patients). Males comprised 93% of the
patients in the treatment group and 83% of the control group. The mean
age in the treatment group was 70.1 years and 71.4 years in the
control group.

   Study Design Confounders

   The design of the protocol resulted in two different treatment
regimens in the treatment and control arms of the study. Different
volumes of blood were removed and replaced as a part of ANH (2.8
liters in the treatment group and 1.5 liters in the control group) and
different colloids were used as replacement fluids (Hespan, albumin,
and PolyHeme versus only Hespan and albumin). The protocol design
therefore resulted in the administration of two different total
volumes and different replacement fluids, representing two
substantially different colloid loads. In addition, the total volume
of oxygen-carrying blood products returned to the treatment and
control groups during the first 24 hours was markedly different (4.4
liters in the treatment group versus 2.2 liters in the control group).
As a result, the treatment difference between the groups was not
simply the infusion of PolyHeme in the treatment arm. This represents
a departure from the usual scientific design where only one variable
at a time is changed between two study groups.

   In summary, the treatment group experienced:

   (1) Removal of larger volumes of autologous blood during ANH;

   (2) Infusion of larger volumes of colloid replacement during ANH
        (Hespan, albumin, PolyHeme);

   (3) Reinfusion of larger volumes of autologous blood, PolyHeme,
        and allogeneic blood throughout the perioperative period; and
        finally

   (4) Infusion of up to six units of PolyHeme.

   Evaluation of the Primary Study Objective: Avoidance of Allogeneic
Blood

   The primary study objective was to increase the number of patients
who avoided allogeneic blood (donated) transfusion in the treatment
group compared to the control group. Overall 38% of the patients
assigned to receive PolyHeme avoided allogeneic blood compared to 40%
of the control group. This study did not meet its primary study
objective.

   Evaluation of Safety: Adverse Events

   Adverse events (AEs) are defined as any untoward medical
occurrences that occur in subjects administered any dose of a study
drug (in either group), regardless of whether there is a relationship
to the drug being studied. Serious adverse events (SAEs) are defined
as any adverse drug experience, occurring at any dose that results in
any of the following outcomes: (1) death (2) a life-threatening
adverse drug experience (3) inpatient hospitalization or prolongation
of existing hospitalization (4) a persistent or significant
disability/incapacity, or a congenital anomaly/birth defect or (5)
other important medical events as judged by the investigator.
   In any study, AEs and SAEs may be due to either the underlying
condition of the patient, the treatment setting, the specific
treatment, or the investigational product. Investigators report the
AEs and SAEs, characterize them as an AE or SAE based on prescribed
regulatory definitions, and ascribe a relationship of the
investigational product to the events ranging from "unrelated,"
"remotely related," "possibly related," to "probably related." In this
study, the investigators were not blinded to the treatment assignment.
The attribution of an event in a patient who received several types of
solutions, (Hespan, albumin, PolyHeme and autologous and donated
blood) would therefore be less straightforward than a patient
receiving only a single test solution.
   At least one AE was reported for 99% of patients who received
PolyHeme and 97% of patients who received control. In general,
specific adverse events occurred with similar incidence in the two
groups. Serious adverse events were reported in 54% of patients who
received PolyHeme and in 28% of patients who received control. The
principal difference was the occurrence of myocardial infarction (MI)
in the treatment group. None of the SAEs in either group was judged by
the investigator as "probably related" to study treatment. Only one
SAE (rash) in a patient receiving PolyHeme was reported as "possibly
related."
   Myocardial infarction was reported for 10 of the 81 patients
(12.3%) in the treatment group and none of the patients in the control
group. The patients who experienced MI ranged in age from 57 to 82
years. All of these patients had at least one risk factor for heart
disease, such as hypertension or hypercholesterolemia, and eight of
the ten patients were smokers. The MIs occurred within 0 to 7 days
after AAA surgery. Nine of the 10 patients in the treatment group who
had MIs had adverse events consistent with fluid overload, including
atrial fibrillation, congestive heart failure, respiratory
insufficiency, pulmonary edema and peripheral edema. Only one of the
35 patients in the treatment group who were enrolled at the two
highest enrolling sites experienced MI.
   Eight of the 10 patients with MI survived and two died (one at 7
days and one at 32 days). Four of the MIs were reported as "remotely
related" to study treatment and six as "not related."
   All patients had ECG tracings done in the operating room at
baseline and following completion of ANH. All of the ECG tracings were
read by an independent cardiologist who was blinded to the group
assignment. There were no differences in ECG tracings between the
treatment group and the control group. There were no differences in
ECG tracings before and after ANH following infusion of PolyHeme in
the treatment group. This is important because the historical concern
has been that some hemoglobin-based oxygen carriers have caused
coronary vasoconstriction that would reduce coronary perfusion and
result in myocardial ischemia.

   Assessment of Safety: Mortality

   A total of eight deaths (9.9%) in the treatment group and four
(5.6%) in the control group were reported. The time of death is
relevant. Surgical mortality rates are traditionally defined as deaths
that occur within 30 days following surgery. All four deaths in the
control group occurred during this period for a mortality rate of 5.6%
(at 3, 14, 17, and 22 days). In contrast, only three of the eight
deaths in the treatment group occurred during this period, for a
mortality rate of 3.7% (at 7, 15, and 18 days), while five were late
deaths (at 32, 42, 42, 75, and 96 days). Furthermore, five of these
eight patients died following discharge from the hospital. Although
the protocol defined the follow up to 84 days after surgery, any
deaths that were known by the investigators to have occurred after
that follow-up period are also included in this analysis.

   Key Observations

   There are a number of important conclusions that can be drawn
based on the experience described here. ANH itself is a non-routine
procedure, traditionally limited to removal of approximately 20% to
30% of a patient's blood volume. The goal in this study was to use
PolyHeme to extend the ANH to 60% of the patient's blood volume
because of the ability of PolyHeme to provide supplemental hemoglobin
to replace that lost with removal of larger quantities of the
patient's own blood. The study was, therefore, designed to measure the
avoidance of allogeneic blood and look for an improvement with the use
of PolyHeme.
   The results indicate that extended ANH, with removal of up to 60%,
or 6 units, of the patient's blood volume can be accomplished with
PolyHeme. However, the study did not achieve its endpoint of avoidance
of allogeneic blood.
   This was a complex protocol with multiple treatment confounders.
The protocol design resulted in the treatment group being different
from the control group in four important ways:

   (1) Removal of twice as much blood during the ANH procedure,
        namely six units of blood versus three units of blood and
        thus,

   (2) Infusion of twice as much fluid to replace the removed blood;

   (3) Reinfusion or return of more blood and oxygen-carrying
        products during and after surgery; and lastly

   (4) Infusion of up to six units of PolyHeme.

   Ten patients in the treatment group experienced myocardial
infarction versus none in the control group. Two of these patients
died, at 7 and 32 days. Because of the study design, it is not
possible to separate the influence of PolyHeme from the other
differences in the treatment group. We believe the accompanying larger
total volume infusions led to complex fluid management issues in these
patients. We do not believe that these events are due to a direct
pharmacologic effect of PolyHeme.

   About Northfield Laboratories

   Northfield Laboratories Inc. is a leader in developing an
oxygen-carrying resuscitative fluid, PolyHeme(R), for the treatment of
urgent, large volume blood loss in trauma and resultant surgical
settings. PolyHeme(R) is a solution of chemically modified human
hemoglobin that requires no cross matching and is therefore compatible
with all blood types. It has a shelf life in excess of 12 months.
Enrollment is currently underway in a pivotal Phase III trial of
PolyHeme(R) beginning in the pre-hospital setting. For further
information, visit www.northfieldlabs.com.

   This press release may contain forward-looking statements
concerning, among other things, Northfield's future business plans and
strategies and clinical and regulatory developments affecting our
PolyHeme(R) red blood cell substitute product. These forward-looking
statements are identified by the use of such terms as "intends,"
"expects," "plans," "estimates," "anticipates," "should," "believes"
and similar terms. These forward-looking statements involve inherent
risks and uncertainties. Our actual results may therefore differ
materially from those predicted by the forward-looking statements
because of various factors and possible events, including our ability
to obtain FDA approval to market PolyHeme commercially, the
availability of capital to finance our clinical trials and ongoing
business operations, our ability to obtain adequate supplies of raw
materials and to manufacture PolyHeme in commercial quantities, our
ability to market PolyHeme successfully, the possibility that
competitors will develop products that will render PolyHeme obsolete
or non-competitive, our ability to protect our intellectual property
rights, the possibility that we may be subject to product liability
claims and other legal actions, our dependency on a limited number of
key personnel, the uncertainty of third party reimbursement for our
product and other risks and uncertainties described from time to time
in our periodic reports filed with the Securities and Exchange
Commission, including our most recently filed quarterly report on Form
10-Q and annual report on Form 10-K. These forward-looking statements
speak only as of the date of this press release. We do not undertake
any obligation to update or publicly release any revisions to
forward-looking statements to reflect events, circumstances or changes
in expectations after the time such statement is made. All subsequent
written and oral forward-looking statements attributable to Northfield
or any person acting on our behalf are qualified by this cautionary
statement.