| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Clinical Trials |
Developmental Therapeutics Program, Clinical Trials Unit, Medicine Branch [R. A. M., A. M. S., E. A. S.], Biostatistics and Data Management Section [S. M. S.], Pharmacy Department [D. K.], Medicine Branch [D. H., W. D. F., H. C.], Biologic Resources Branch [S. C.], and Laboratory of Pathology [E. S. J., M. S-S.], National Cancer Institute, Bethesda, Maryland 20892-1906; Cancer Immunobiology Center, University of Texas Southwestern Medical Center, Dallas, Texas 75235 [E. S. V., J. S., V. G.]; and Science Applications International Corporation-Frederick, National Cancer Institute-Frederick Cancer Research and Development Center, NIH, Frederick, Maryland 21702-1201 [D. F. M.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Adult patients received a continuous infusion of Combotox at 10, 20, or 30 mg/m2/192 h. No intrapatient dose escalation was permitted.
Patients with 
50 circulating tumor cells (CTCs)/mm3 in
peripheral blood tolerated all doses without major toxicity. The
maximum level of serum IT (Cmax) achieved in this group was
345 ng/ml of RFB4-dgA and 660 ng/ml of HD37-dgA (1005 ng/ml of
Combotox). In contrast, patients without CTCs (<50/mm3)
had unpredictable clinical courses that included two deaths probably
related to the IT. Additionally, patients exhibited a significant
potential for association between mortality and a history of either
autologous bone marrow or peripheral blood stem cell transplants
(P2 = 0.003) and between mortality and
a history of radiation therapy (P2 =
0.036). In patients with CTCs, prior therapies appeared to have little
impact on toxicity.
Subsequent evaluation of the ITs revealed biochemical heterogeneity between two lots of HD37-dgA. In addition, HD37-dgA thawed at the study site tended to contain significant particulates, which were not apparent in matched controls stored at the originating site. This suggests that a tendency to aggregate may have resulted from shipping, storage, and handling of the IT that occurred prior to preparation for administration. It is not clear to what extent, if any, the aggregation of HD37-dgA IT was related to the encountered clinical toxicities; however, the potential to aggregate does suggest one possible basis for problems in our clinical experience with HD37-dgA and leads us to the conclusion that non-aggregate-forming formulations for these ITs should be pursued prior to future clinical trials.
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
ITs employ plant or bacterial toxins linked to targeting moieties
derived from immunoglobulin or antibody fragments (4)
. ITs
were developed to minimize the nonspecific toxicity observed with more
conventional chemotherapeutic agents, by maximizing drug delivery to
tumors expressing epitopes specifically recognized by the ligands
(5
, 6) . The RFB4 anti-CD22 and HD37 anti-CD19 MAbs were
coupled to dgA via the heterobifunctional, thiol-containing
cross-linker,
N-succinimidyl-oxycarbonyl-
-methyl--(2-pyridyldithio)toluene.
CD22 molecules expressing the RFB4 epitope and CD19 molecules
expressing the HD37 epitope are present on 60–70% and >90% of NHL
cells, respectively. Studies in SCID mice with human Daudi cell
lymphomas have demonstrated that Combotox is more active than either IT
alone in eliminating minimal disease (7)
. The basis for
this effect could relate to ability of combined toxins to kill cells
negative for the antigenic determinate recognized by only one of
the two ITs, to altered efficacy of processing one IT by the presence
of the other, or to active propagation of death-promoting signals by
the antibody-portion of one or both ITs that favorably influences the
effect of the other IT.
A previous Phase I trial using CI of RFB4-dgA showed evidence of antitumor activity with 24% PRs and one long lasting CR (8) . The IT was administered over 192 h, and the MTD was 19.2 mg/m2/192 h, with a DLT at 28.8 mg/m2/192 h consisting of VLS, which manifested as a spectrum of clinical disorders ranging from mild edema to respiratory failure requiring ventilator support.
A previous Phase I trial using CI of HD37-dgA also showed evidence of antitumor activity (9) , albeit to a lesser degree than that observed with RFB4-dgA (8) , concordant with the less potent action of this IT in vitro (10) and in SCID mice with Daudi lymphoma (11) . Administered by either intermittent bolus infusion or CI, the MTD of HD37-dgA was 19.2 mg/m2/192 h, with DLTs consisting of VLS, expressive aphasia, and rhabdomyolysis (9) .
Combotox is a 1:1 mixture of these two ITs and has been proposed for the therapy of patients with minimal residual disease. The rationale for using Combotox in this setting is to avoid the emergence of antigen-negative variants. In addition, preclinical studies had demonstrated the supra-additive activity of a 1:1 mixture of RFB4-dgA and HD37-dgA in causing cell kill in the Daudi disseminated lymphoma in SCID mice (7) . This Phase I study used Combotox in patients with advanced or refractory B cell lymphomas expressing CD19 and CD22. The design of the study was to administer previously well tolerated doses of each IT by CI up to a total dose 30 mg/m2 over 192 h.
|
PATIENTS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Pathology and Eligibility.
Eligible patients had a histological diagnosis that included B cell NHL
of any type except lymphoblastic lymphoma, B cell chronic lymphocytic
leukemia, B cell or pre-B cell acute lymphocytic leukemia, or hairy
cell leukemia. Histological diagnosis was confirmed at the Laboratory
of Pathology, NCI. Biopsy specimens from open biopsy or fine needle
aspiration were evaluated by immunohistochemical analysis for the
expression of CD19 and CD22. When feasible, expression of CD19 and CD22
(both pre- and posttreatment) was evaluated using the HD37 and RFB4
MAbs on cell suspensions prepared for flow cytometry. Peripheral blood
mononuclear cells, including tumor cells, where present, were stained
within 24 h of collection with a panel of antibodies by a
technique described previously (12)
. Five-parameter,
three-color flow cytometry was performed with a Becton Dickinson
FACScan flow cytometer equipped with a 15-mW argon laser (excitation at
488 mm). At least 5000 lymphocytes were acquired per tube. For
analysis, lymphocytes were gated by forward and side scatter, and the
CD19-, CD20-, CD3-, and CD34-positive populations were back-gated to
determine the appro-priateness of analysis gates. Identification of
a monoclonal B cell population (13)
expressing surface
light chains identical to the patient’s original lymphoma was
considered diagnostic for CTCs.
ITs.
The study was carried out under Investigational New Drug application
3539. The procedures used for the production of dgA, its coupling to
RFB4 or HD37, and the production of IT have been described
(14, 15, 16)
. Patients received Combotox derived from UTSWMC
lots VG-3 (HD37-dgA) and PT-8 (RFB4-dgA). IT from UTSWMC lot VG-1
(HD37-dgA) underwent biochemical analysis but was not administered to
patients.
Dosage and Administration.
The ITs were transported overnight on dry ice from UTSWMC and stored
separately in endotoxin-free glass vials at –70°C in the NCI
pharmacy. After thawing stock IT solutions at 4°C, an amount of each
IT sufficient for 2 days of the CI was removed and filtered through a
0.22 µm filter (the remaining stock IT solution was refrozen for
research purposes but not for clinical use). The thawed IT stock
solutions were then mixed to give the Combotox stock solution. The
appropriate amount of Combotox stock was then diluted to a final volume
of 480 ml with normal saline. The IT infusions for days 2, 4, 6, and 8
were prepared 1 day prior to use and were maintained at 4°C in the
pharmacy until administered. Cytotoxic activity against Daudi cells was
maintained for the 24-h period prior to use. Combotox was administered
by CI over 192 h unless a cycle was curtailed due to toxicity. The
three dose levels were 10, 20, and 30 mg/m2. In an attempt
to minimize the risk of allergic reaction, each patient received a test
dose on day 1 of each cycle. This test dose involved i.v.
administration of 0.1 mg of IT Combotox over 1 min, with subsequent
monitoring for 30 min for any allergic reactions. The test dose was not
included as part of the patient’s total dose calculation, and none of
the 22 patients had an allergic reaction to the test dose.
Pharmacokinetic Analysis.
Plasma samples were obtained for pharmacokinetic assessments at various
times after the initiation of the 192 h CI and after its
completion. To detect the two different ITs in the serum sample, a RIA
as used. Ninety-six-well microtiter plates were coated with affinity
purified rabbit antiricin A chain, washed, and blocked with BSA.
Dilutions of the patient sera or a standard Combotox solution were
added after washing the wells. The HD37-dgA was detected using an
affinity purified rabbit anti-HD37 Id, and the RFB4-dgA was
detected using an affinity purified rabbit anti-RFB4 Id. Each anti-Id
had been prepared by absorbing the anti-RFB4 or anti-HD37 antibodies
with mouse serum IgG and either RFB4 (for anti-HD37 Id) or HD37 (for
anti-RFB4-Id). Cross-reactivities were <5%. The two anti-Ids were
labeled with Na125I using the iodogen methods
(17)
. Plates were washed, and the wells were cut out and
counted. IT concentrations were determined from the standard curve and
expressed as ng/ml.
Pharmacokinetic parameters for each IT were calculated by weighted nonlinear least squares analysis fitting a one-compartment and two-compartment open linear model computed by ADAPT II (Biomedical Simulations Resource, University of Southern California, Los Angeles, CA). In addition, noncompartmental parameters were calculated using the area under the concentration curve as calculated by the trapezoidal rule. Model selection was determined based on Akaike’s Information Criterion and visual examination of the difference between the measured and fitted concentration. Pharmacokinetic parameters determined from previous trials were used for comparisons. In the present pharmacokinetic analysis, outlier points (+2 SDs outside the fitted line) were not disregarded, and all data points were included.
Assessment of Response.
Patients were evaluated for response between day 21 and day 28 of each
cycle. Clinical responses were characterized using the following
definitions (18)
: a CR was defined as the disappearance of
all known disease, determined by two observations not less than 28 days
apart. A PR involved a 50% or more decrease in cross-sectional area of
indicator lesions that were objectively evaluated by two observations
not less than 28 days apart. A MR signified a clear biological effect
of less than 50% reduction of tumor mass maintained for at least one
month, or a greater than 50% reduction maintained for less than one
month. Stable disease was defined as not meeting the criteria for CR,
PR, or PD, whereas PD included the appearance of a new lesion, a 25%
or greater increase in the sum of the areas of the lesions or an
increase of at least 50% in the area of any existing lesion, (as long
as that increase was at least 2 cm2). Patients
were eligible for retreatment if they achieved a CR, PR, or MR and had
less than 1 µg/ml of HAMA or HARA at the time of retreatment.
Assessment of Toxicity and HAMA/HARA.
The NCI Common Toxicity Criteria were used to assess toxicity
(19)
. In addition, VLS was characterized as follows: grade
1, asymptomatic, not requiring therapy; grade 2, symptomatic but not
requiring fluid support; grade 3, respiratory compromise or requiring
fluids; grade 4, life threatening, requiring pressor support and/or
ventilatory support. MTD was defined as the dose at which 
3 of at
least 6 evaluable patients experienced reversible DLT. The MTD was
considered to have been exceeded if even one patient experienced a
grade 4 toxicity clearly related to drug. DLT was defined as any
reversible grade 3 toxicity or nonreversible grade 2 toxicity. Levels
of HAMA and HARA were determined as described previously
(20)
.
Statistics.
Fisher’s exact test was used to test for potential of association
between a variety of clinical parameters and encountered toxicities.
t1/2 ß values were compared between patients
with and without CTCs using the Wilcoxon rank sum test. All tests were
two-sided.
Stability Testing of the ITs.
After filtration through a 0.22-µm filter, SEC-HPLC was
performed on 7 IT samples derived from all lots obtained from the NIH
pharmacy, as well as "never-thawed" controls stored at UTSWMC and
shipped overnight on ice packs. Samples for SEC-HPLC analysis were
chromatographed on a Waters HPLC system with Millennium software using
a TosoHaas TSK guard column followed by a TSK
G4000SWXL and a TSK
G3000SWXL connected in series (both 7.8-mm
internal diameter x 30-cm length).
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The mean age was
50.8 years, and the study group included 13 males and 9 females.
Eleven patients were originally diagnosed with a low-grade lymphoma, 10
patients with intermediate-grade lymphoma, and 1 with mantle cell
lymphoma. Patients had received a mean of 3.7 prior chemotherapy
regimens (range, 1–7) and eight had at least one prior course of
radiation therapy. All patients expressed CD19 (HD37) and CD22 (RFB4)
on at least 30% of their tumor cells.
|
. The protocol permitted dose
escalation after a minimum of three patients at a given dose level had
been followed for >28 days. Therefore, patient 4, who entered the
study 14 days after patient 3, was also started at dose level 1.
Patient 4, who had evidence of CTCs, developed hypoalbuminemia to 3.0
g/dl within 24 h of starting Combotox infusion. On day 3,
the patient exhibited 1+ pretibial edema, pulmonary congestion, cough,
and mild shortness of breath. Chest radiograph confirmed the presence
of pulmonary edema that was characterized as a grade 3 VLS toxicity.
Combotox administration was terminated, and a single 40-mg i.v.
dose of furosemide was administered, resulting in diuresis with
complete resolution of symptoms. The occurrence of this toxicity
dictated that additional patients be accrued at dose level 1, and these
patients tolerated level 1 without difficulty. All but one patient in
this group of six had evidence of CTCs.
|
Subsequently, patients with CTCs (patients 7, 8, 9, 14, 15, and 21) tolerated dose levels 2 and 3 without major toxicity with the exception of patient 14, who experienced mild shortness of breath on the fourth day of IT therapy. Additional manifestations of this readily reversible grade 3 VLS toxicity included bilateral pulmonary edema observed on chest radiograph with a concomitant 2-kg weight gain and mild hypoalbuminemia to 3.1 g/dl After administration of furosemide, all symptoms resolved within 24 h. The Cmax of serum IT concentrations achieved in these patients ranged from 37 to 345 ng/ml RFB4-dgA and from 128 to 660 ng/ml HD37-dgA (165–1005 ng/ml Combotox), with patient 14 achieving the highest serum levels of Combotox (345 and 660 ng/ml of RFB4-dgA and HD37-dgA, respectively, or 1005 ng/ml Combotox).
In the group of patients lacking CTCs, patient 10 tolerated the IT
without difficulty. Patient 11 developed grade 3 VLS, without evidence
of thrombocytopenia, that failed to improve after cessation of IT
therapy even after achieving only a modest Cmax
of 293 ng/ml RFB4-dgA and 463 ng/ml HD37-dgA (756 ng/ml Combotox).
Three days after the cessation of therapy, this patient developed
increasing dyspnea and a pericardial rub, with subsequent development
of adult respiratory distress syndrome and death. Autopsy
revealed massive pericardial, pleural, and abdominal tumor, as well as
marked pulmonary congestion and a dilated right cardiac ventricle.
Additionally, microscopic sections of the lung showed the presence of
lymphoma and Candida. Overt evidence of progression of
disease and uncontrolled infection therefore suggested a grade 3 VLS
toxicity with progression of disease complicating assessment of
reversibility. Accordingly, this cohort of patients was expanded to
six. Patients 12 and 13 achieved a Cmax of 109
ng/ml RFB4-dgA and 268 ng/ml HD37-dgA (377 ng/ml Combotox) and 264
ng/ml RFB4-dgA and 337 ng/ml HD37-dgA (601 ng/ml Combotox),
respectively, without major toxicity. Patient 16 tolerated Combotox
infusion without difficulty through the sixth day, when she was noted
to have increased pulmonary vascular markings on chest radiograph in
conjunction with a decrease in systolic blood pressure (from
120 mm Hg baseline to 100 mm Hg) and a 3-kg weight gain.
Treatment consisted of blood pressure support with infusion of normal
saline and concomitant furosemide administration to facilitate
diuresis. This readily reversible grade 3 VLS toxicity at a
Cmax of 304 ng/ml of RFB4-dgA and 320 ng/ml of
HD37-dgA (624 ng/ml Combotox) defined the MTD for the patient cohort
lacking CTCs. Patient 17 tolerated dose level 2 without major toxicity
with a Cmax of 216 ng/ml RFB4-dgA and 336 ng/ml
HD37-dgA (552 ng/ml Combotox). In contrast, patient 18 developed
hypotension and dyspnea 3 days into Combotox administration.
Shortly after discontinuing the infusion and ICU transfer, the
patient’s clinical deterioration was characterized by development of
oliguric renal failure and a progressively worsening anemia with
presence of microangiopathic changes on peripheral blood smear.
Thrombocytopenia, normal prothrombin, and partial thromboplastin
times and an elevated lactate dehydrogenase level in conjunction
with a normal neurological status and evidence of hemolytic anemia
indicated a diagnosis of HUS. The patient’s clinical status continued
to deteriorate until death occurred on the 15th day of the first cycle
of infusional Combotox. Review of the patient’s pharmacology
suggested that toxicity was probably related to the high
Cmax of 389 ng/ml RFB4-dgA and 1317 ng/ml
HD37-dgA (1706 ng/ml Combotox) and that this dose level therefore
actually exceeded the MTD for patients lacking CTCs.
After the characterization of this DLT, additional patients lacking CTCs were accrued at the lowest dose level. Patient 19 tolerated the IT without difficulty, and patient 20 developed grade 2 VLS with a Cmax of 156 ng/ml RFB4-dgA and 387 ng/ml HD37-dgA (543 ng/ml Combotox). Patient 22 developed VLS 5 days after starting the IT the infusion and subsequently required transfer to the ICU for respiratory support and management of acute renal failure. While in the ICU, the patient also exhibited a HUS-like clinical picture that included microangiopathic hemolytic anemia, thrombocytopenia with relatively normal partial thromboplastin time and prothrombin level, an elevated lactate dehydrogenase level, and a maintained neurological status until progression to multiple organ failure, acute development of ventricular tachycardia, and death. This outcome occurred at serum concentrations of 206 ng/ml RFB4-dgA and 315 ng/ml HD37-dgA (521 ng/ml Combotox). This concentration had been well tolerated by other patients. The trial was terminated at this point without clear delineation of a safe dose level in patients without CTCs.
Pharmacokinetics.
Tables 2
and 3
summarize the
pharmacokinetic parameters of the two ITs for cycle 1, separated
according to presence or absence of CTCs. For patients with CTCs, the
Cmax of RFB4-dgA achieved at the first dose level
was consistent with that of patients lacking CTCs at 117
versus 126 ng/ml (Table 2)
. However, at the second dose
level (20 mg/m2), the Cmax was
considerably reduced in patients with CTCs (92 ng/ml) versus
those without (293 ng/ml). For patients with CTCs, the elimination rate
constant (Ke) was consistent between the two
dose levels (median, 0.022 and 0.018 liters/h, respectively), but the
volume of distribution increased from a median of 13.9 to 41.5
liters/m2 (Table 3)
. For the three patients with
CTCs who received dose level 3, the Cmax was
higher than the median Cmax for dose level 2 (195
versus 92 ng/ml). The t1/2ß for
RFB4-dgA tended to be slightly longer for those patients with CTCs when
compared to those without [median for dose level 1, 32.1
versus 24.4; dose level 2, 39.2 versus 21.6
(P2 = 0.25 and
P2 = 0.29, respectively)]. In
patients without CTCs, the Cmax for RFB4-dgA
increased between dose levels 1 and 2 from 126 to 293 ng/ml,
respectively. The volume of distribution and elimination rate constant
were consistent between the two dose levels (median 9.4
versus 7.5 liters/h/m2 and 0.029
versus 0.032 1iters/h, respectively).
|
The t1/2 ß of HD37-dgA was similar for patients with and without CTCs [median, 24.9 and 46.8 h versus 22.5 and 32.0 h, respectively, at dose levels 1 and 2 (P2 = 0.72 and P2 = 0.35)].
Pharmacodynamic Correlation with Cycle 1 Toxicity.
Patients with CTCs tolerated the complete cycle of IT infusion at all
doses and AUC levels with minimal toxicity except for two occurrences
of readily reversible grade 3 VLS (Fig. 1A).
Patients with grade 3
toxicity (patients 4 and 14) were excluded from Fig. 1A
attributable to an interruption of the first cycle of infusion
preventing an accurate assessment of the AUC. In contrast, patients
without CTCs experienced severe toxicity at variable AUCs and at both
dose levels tested (Fig. 1B).
Patients 11 and 18, with
toxicity grades of 3 and 5, respectively, were excluded from Fig. 1
due
to interrupted C1 infusions preventing accurate definition of
AUC.
|
summarizes the adverse events (>
grade 2) overall cycles and dose levels in patients with and without
CTCs. Any toxicity of clinical importance was ultimately related to the
development of VLS or HUS, consistent with prior studies with the
individual ITs (8)
. Other adverse reactions were generally
constitutional in nature, transient, and easily manageable with
supportive care.
|
-methyl--(2-pyridyldithio)toluene
cross-linker. Typically these preparations contain >85% IT with one
dgA molecule per MAb [IgG-(dgA)1] and 15%
IT with two [IgG-(dgA)2] or three
[IgG-(dgA)3] dgAs per MAb. These ITs can exist
as monomers (IgG-dgA), or they may associate to form dimers
[(IgG-dgA)2], trimers
[(IgG-dgA)3], or even higher order aggregates.
To test the pharmaceutical stability of the ITs, vials from lots used
in the trial and vials that had never previously been shipped were sent
overnight from UTSWMC and were subjected to FT cycling; a process that
may have inadvertently occurred as part of IT shipping and storage at
the study site.
After thawing, NIH stock HD37-dgA (UTSWMC lots VG-3 and VG-1),
underwent A280 determination and
visual inspection. Vials were found to contain variable amounts of
precipitate (Table 5)
. Samples with
increased levels of precipitate displayed the largest decrease in
A280 after further freezing and
thawing, and some of the IT solutions were viscous; losses were
observed after filtration by binding to vials and tubes. The FT cycle,
which was precluded for ITs used in patients, was introduced into the
stability testing to duplicate potential storage temperature
variability that may have occurred at NCI. None of this FT-cycled IT
was used in the clinical trial.
|
, is representative SEC chromatograms for samples taken
from NIH stock HD37-dgA (lots VG-1 and VG-3, respectively) that
illustrate interlot variability in the amount of monomeric, dimeric,
and trimeric molecular species present in samples stored at NCI.
|
One lot (8 vials total) of RFB4-dgA (UTSWMC lot PT-8) was available for
analysis, so interlot variation was not examined. No visible
precipitate was seen before or after filtration. These findings were
consistent with control samples from UTSWMC lot PT-8 and indicated no
visible aggregation of RFB4-dgA as compared to HD37-dgA. SEC-HPLC
analysis of RFB4-dgA showed that (IgG-dgA)1 was
present at 84%, whereas dimers and higher molecular weight forms made
up 15% of the material, as reported in previous trials using this
agent. A representative SEC chromatogram for samples taken from
RFB4-dgA UTSWMC Lot PT-8 is shown in Fig. 2C.
After
filtration, RFB4-dgA was found to contain
RFB4-(dgA)1, RFB4-(dgA)2,
and RFB4-(dgA)3 species in a ratio of
59/33/8% (data not shown).
In summary, the analyses suggested that HD37-dgA but not RFB4-dgA had a propensity to form aggregates during thawing and that there were interlot variations. Aggregation of this material is now avoided by the addition of 0.1 mg/ml Tween 80 (Sigma).4
Statistical Analysis.
To better understand the relationship of patient characteristics,
pharmacological values, and the toxicities encountered during this
trial, a variety of factors were analyzed for potential of association
in a univariate fashion using Fisher’s exact test. Table 6
lists the factors that were tested for
association between toxicity grade >3 and death. Of the 10 sets of
factors tested for independence, only 2 showed a potential association
by having P2 < 0.05 for patients
undergoing CI of Combotox. Specifically, a potential for association
exists between either prior bone marrow transplant or PBSC transplant
and death (P2 = 0.003, in a study in
which 3 of 4 posttransplant patients died, compared to 0 of 18
patients without prior transplant) and a prior history of radiation
therapy and death (P2 = 0.036, in a
study in which 3 of 8 patients with a history of radiation therapy died
compared to 0 of 14 without a history of radiation therapy). The
Ps are unadjusted for the multiple factors examined. The
remaining factors failed to exhibit significant evidence of
association.
|
, HAMA/HARA levels of
greater than 1 µg/ml ultimately occurred in 6 of 22 (27%) patients,
in accord with results from previous trials with the individual ITs.
Repeat treatments of patients with HAMA/HARA levels of >1 µg/ml were
delayed so that 7 of 8 responding patients could eventually be
retreated. In one patient (patient 12), the presence of 315 µg/ml
HAMA prior to the third infusion cycle precluded retreatment despite
evidence of a clinical response.
|
lists the clinical response of the 22 patients entered
into the study. Two patients (9%) had a PR (95% confidence interval,
1–29%) lasting 1 and 5 months, respectively, prior to progression of
disease, whereas five patients had minimal response to therapy. Disease
progression occurred in 12 patients (including one death attributed to
progression of disease). Only one patient was removed from the trial
after developing HAMA/HARA antibodies, and two deaths were directly
attributable to administration of the IT. These values compare with
22% of patients with PRs and 17% with MRs when RFB4-dgA alone was
administered by CI and 11% PRs when HD37-dgA was administered alone by
CI. |
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
50/mm3) tolerated all three dose levels of
Combotox (10, 20, and 30 mg/m2/192 h). In contrast, in
patients lacking CTCs (<50/mm3), there was a
disparity between toxicity grade and total IT concentration with
observed VLS and HUS at both administered dose levels. (b)
Prior irradiation or bone marrow transplantation showed a potential for
association with severe toxicities and may predispose patients to the
most severe toxicities. (c) There was evidence of HD37-dgA
multimerization, which may have contributed to patient toxicity. One of the purposes of this study was to establish safe dose levels of Combotox in a Phase I trial in heavily pretreated patients with bulky disease, prior to using Combotox in the setting of minimal residual disease. The safe dose of Combotox in patients with CTCs was possibly as high as 30 mg/m2/192 h, although additional patients would be needed at that dose level to establish this point. In contrast, we were unable to establish a safe dose level of Combotox in patients without CTCs. Three deaths occurred during the trial, and none of these patients had CTCs as determined by flow cytometric analysis. The first death (patient 11) appeared to have been related to progression of disease, as confirmed by the findings at postmortem examination. The second death occurred in patient 18, a 54-year-old male with a history of follicular large cell lymphoma who had previously undergone three separate regimens of chemotherapy and three courses of radiation therapy in addition to a splenectomy and an ABMT. During the trial, the patient achieved high plasma concentrations of Combotox (at 1707 ng/ml, the highest in the trial) shortly before the development of HUS and death. The third death occurred in patient 22, a 43-year-old male with a history of diffuse large cell lymphoma who had received six prior regimens of chemotherapy and one course of radiation therapy and had previously received a peripheral-blood stem cell transplant. This patient developed a HUS-like clinical picture shortly before multisystem organ failure and death. Interestingly, the patient had achieved only average serum concentrations of Combotox (521 ng/ml), which had been generally well tolerated by other patients irrespective of the presence of CTCs.
The lack of association between toxicity grade and serum IT concentration led us to evaluate the pharmaceutical properties of the NIH stock ITs used in the trial and to review patient demographic and clinical features to identify any factors that may have been associated with poor clinical outcome after administration of Combotox.
Variables encountered in this analysis included the fact that both lots of HD37-dgA tested, of which only one was used in the trial, had a tendency to aggregate after thawing. Hence, pharmaceutical stability was affected, as the amount of drug lost to precipitation after thawing and filtering the IT could have led to differences in the dosage received from day to day. Despite the lack of clear evidence that an association exists between the severity of toxicity and the level of precipitation prior to filtration, variation in HD37-dgA could have led to toxicity at the previously well tolerated level I (10 mg/m2/192 h). RFB4-dgA was stable after thawing and showed no signs of visible aggregation.
Toxicities in this trial, including VLS and HUS, were similar to those reported in previous trials using other types of ITs. VLS, characterized by increased vascular permeability, can result in life-threatening edema and organ failure (21 , 22) and has been a DLT in trials using IL-2-DAB as well as ITs prepared with blocked ricin and PE (8 , 9 , 19 , 23, 24, 25, 26, 27, 28, 29, 30, 31) . The etiology for VLS is unknown and could include direct, nontargeted ricin dgA chain-mediated toxicity to endothelial cells, targeted (by low-level antigen expression) toxicity to endothelial cells, and cytokine-mediated mechanisms related to the release of cytokines from macrophages and/or endothelial cells. In the case of PE-ITs, VLS may have been due to cross-reaction of the MAb with vascular endothelial cells. Although the causes of VLS are poorly understood, a recent study has shown that an amino acid structural motif that is common to IL-2, dgA, and Pseudomonas exotoxin and diphtheria toxins may be responsible for the binding of ITs to endothelial cells followed by endothelial cell damage (32) . Preliminary studies in animals using ITs with a mutated version of this motif indicate a lack of VLS.5
Previous studies with dgA ITs have failed to document increase in several cytokines in patients developing VLS after treatment with RFB4-dgA (8) . However, improperly timed sampling errors that might miss a "burst" of circulating cytokines or that might not sample the correct compartment, e.g., in the damaged endothelium, do not exclude a potential contribution of cytokines to the development of VLS.
In prior studies with dgA-based ITs, we have attempted to define the factors that predict the development of severe VLS. A previous Phase I study using RFB4-dgA identified serum IT concentrations greater than 1 µg/ml as predictive of the severity of VLS symptoms (8) . In this trial, three patients (patients 10, 14, and 18) achieved this level of serum IT, resulting in toxicity grade levels of 1, 3, and 5, respectively. In addition, two of these patients (patients 10 and 18) lacked CTCs. A previous Phase I trial using RFB4-dgA showed that the presence of lymphoma cells in the blood at levels >1010 cells/liter was associated with reduced toxicity (23) , and another trial (8) suggested that the absence of CTCs was predictive of severe VLS toxicity. In this trial, several of the patients without CTCs experienced toxicities that were unrelated to the total concentrations of IT in sera. Additional reservoirs of tumor could include marrow and spleen. It is noteworthy that two of the deaths occurred in marrow negative or unknown patients, and grade 3 VLS (patient 16) also occurred in a marrow-negative patient without CTCs.
Although the etiology of HUS is unclear, it has been demonstrated that ricin holotoxin (containing both A and B chains) induces HUS in rats (33) . Previous human trials with other ITs have also resulted in HUS as a clinical toxicity. One such trial used the fusion toxin DAB486IL-2, which consists of two Diphtheria toxin domains spliced to the IL-2 gene (34) . The initial trials of this fusion protein toxin in patients with lymphoma demonstrated a DLT of elevated hepatic transaminase levels and a HUS-like constellation of symptoms, including renal insufficiency, anemia, and thrombocytopenia (35 , 36) . A subsequent trial used the modified toxin DAB386IL-2 (which differed in IL-2 receptor affinity and had greater in vivo stability) did not generate HUS (29) . It is tempting to speculate that the HUS observed during trials involving these ITs may reflect, in part, endothelial damage followed by shear stress fragmentation of RBCs and that this may be within the range of toxicities expected from endothelial cell binding of these ITs.
Importantly, the analysis of this trial showed evidence of a potential association between prior ABMT or PBSC transplant and death (P2 = 0.003), as well as between prior radiation therapy and death (P2 = 0.036). Although these findings should be considered tentative due to the exploratory nature of these analyses in a limited numbers of patients, one can speculate that prior radiation therapy and/or high-dose chemotherapy may damage a vascular bed and promote binding of IT aggregates, RBC damage, and hemolysis. Both radiation therapy and high-dose chemotherapy can damage endothelium (37 , 38) , predisposing patients to IT-mediated VLS (22 , 28 , 31) , which could result in consumption of platelets and the appearance of a thrombotic thrombocytopenic purpura-like state. In addition, the tendency of HD37-dgA to form higher molecular weight aggregates must be avoided, and a new formulation appears to circumvent this aggregation.4
Unlike HD37-dgA used in this trial, RFB4-dgA did not aggregate upon
thawing. Nevertheless, a previously unreported female patient who had
no CTCs and total body radiation therapy was treated with a RFB4-dgA
regimen and had an unexpected and fatal occurrence of HUS at levels of
IT well tolerated by other
patients.6
Additionally,
acrocyanosis and distal digital skin necrosis have been reported in
three patients undergoing infusion of the HD37-dgA (9)
,
and evidence from angiographic studies has suggested that unlike
RFB4-dgA, HD37-dgA caused occlusion of the distal limb vasculature
(9)
. It has also been noted that the unconjugated HD37 MAb
has a tendency to form dimers and trimers, whereas the RFB4 MAb does
not (39)
. These clinical phenomena parallel other
experiences with the pharmaceutical properties of the ITs, in which
aggregation detected by HPLC after FT cycles was readily apparent. Both
HD37-dgA and RFB4-dgA can react with
2-macroglobulin or albumin to form covalent
adducts of higher molecular mass (40
, 41)
. To what extent
this was related to toxicity in vivo is unclear. Detailed
toxicological analysis with IT preparations that differed in the amount
of aggregates will be necessary to understand the relationship between
aggregation or multitimer formation and toxicity. However, to the
extent that further preparation of ITs can control for this variable
prospectively, its emergence as a confounding factor in interpretation
of clinical data would be mitigated.
Combotox appears to be safe in patients with even minimal numbers of CTCs, and the MTD was potentially as high as 30 mg/m2/192 h, although further evaluation of more patients would be necessary to establish this point. Evidence of transient responses observed in this trial, albeit modest, coupled with more impressive responses reported in other trials with individual ITs (8 , 9 , 42) , argues for continued clinical evaluation of these ITs under more-defined pharmaceutical and clinical conditions. The potential for association between either a prior history of radiation therapy or ABMT/PBSC transplant and severe toxicity should be considered in future trial design and informed consent. Additional characterization of the factors that are predictive of the clinical toxicities observed in patients undergoing administration of ITs would further define safe dosing parameters for these biologically targeted agents.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported in part by Food and Drug
Administration Orphan Drug Agent FDR 001124–03 (to E. S. V.). 
2 To whom requests for reprints should be
addressed, at DTP Clinical Trials Unit, Medicine Branch, National
Cancer Institute, 10 Center Drive Building 10, Room 12N226, Bethesda,
MD 20892. Phone: (301) 496-1211; Fax: (301) 402-0831; E-mail: messmann{at}pop.nci.nih.gov
3 The abbreviations used are: NHL, non-Hodgkin’s
lymphoma; AUC, area under the curve; ABMT, autologous bone marrow
transplant; CI, continuous infusion; CTC, circulating tumor cell; dgA,
deglycosylated ricin A chain; DLT, dose-limiting toxicity; FT,
freeze-thaw; HUS, hemolytic uremic syndrome; HPLC, high-pressure liquid
chromatography; HAMA, human antimouse antibody; HARA, human antiricin
antibody; IT, immunotoxin; ICU, intensive care unit; Cmax,
maximum concentration; MTD, maximum tolerated dose; MAb, monoclonal
antibody; NCI, National Cancer Institute; UTSWMC, University of Texas
Southwestern Medical Center; PBSC, peripheral blood stem cell; PE,
Pseudomonas exotoxin; SCID, severe combined
immunodeficient; SEC-HPLC, size exclusion chromotography-HPLC; VLS,
vascular leak syndrome; P2, P
from two-sided test; PD, progressive disease; MR, minor response; Id,
idiotypic antibody; CR, complete response; PR, partial response.
4 V. Ghetie and E. S. Vitetta, unpublished
data.
5 E. S. Vitetta, unpublished data.
6 E. A. Sausville, personal communication.
Received 7/ 2/99; revised 12/20/00; accepted 2/ 4/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. E. Strome, E. A. Sausville, and D. Mann A Mechanistic Perspective of Monoclonal Antibodies in Cancer Therapy Beyond Target-Related Effects Oncologist, September 1, 2007; 12(9): 1084 - 1095. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Varki and T. Angata Siglecs--the major subfamily of I-type lectins Glycobiology, January 1, 2006; 16(1): 1R - 27R. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. A. Vallera, M. W. Brechbiel, L. J. Burns, A. Panoskaltsis-Mortari, K. E. Dusenbery, D. R. Clohisy, and E. S. Vitetta Radioimmunotherapy of CD22-Expressing Daudi Tumors in Nude Mice with a 90Y-Labeled Anti-CD22 Monoclonal Antibody Clin. Cancer Res., November 1, 2005; 11(21): 7920 - 7928. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. G. Maloney Immunotherapy for Non-Hodgkin's Lymphoma: Monoclonal Antibodies and Vaccines J. Clin. Oncol., September 10, 2005; 23(26): 6421 - 6428. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. A. Vallera, D. A. Todhunter, D. W. Kuroki, Y. Shu, A. Sicheneder, and H. Chen A Bispecific Recombinant Immunotoxin, DT2219, Targeting Human CD19 and CD22 Receptors in a Mouse Xenograft Model of B-Cell Leukemia/Lymphoma Clin. Cancer Res., May 15, 2005; 11(10): 3879 - 3888. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Zhang and A. Varki Cell surface sialic acids do not affect primary CD22 interactions with CD45 and surface IgM nor the rate of constitutive CD22 endocytosis Glycobiology, November 1, 2004; 14(11): 939 - 949. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Meng, J. E. Smallshaw, L. M. Pop, M. Yen, X. Liu, L. Le, M.-A. Ghetie, E. S. Vitetta, and V. Ghetie The Evaluation of Recombinant, Chimeric, Tetravalent Antihuman CD22 Antibodies Clin. Cancer Res., February 15, 2004; 10(4): 1274 - 1281. [Abstract] [Full Text] [PDF]
|
) and RFB4-dgA (
) for
individual patients with (A) and without
(B) evidence of CTCs at each dose level.
Numbers within the plotted symbols indicate the maximum
cycle 1 toxicity (grades 1–5) experienced by each patient undergoing
complete cycle 1 infusion at corresponding AUC and dose levels. The
horizontally staggered symbols are used to facilitate reading
the graph and do not indicate variations in the administered doses.