| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Experimental Therapeutics, Preclinical Pharmacology |
The CRC Centre for Cancer Therapeutics, Institute of Cancer Research, Sutton, Surrey SM2 5NG, United Kingdom
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
10–14%; Refs.4
and 5
). The latter,
however, may be life threatening if copresenting with neutropenia. In
one Center, this has been reported as 3% (4 of 127 patients, no
deaths; Ref. 7
). Although this toxicity is generally
manageable with immediate and appropriate supportive care such as
hydration and antibiotics, deaths have been reported in some studies.
For example, a recent United Kingdom Medical Research Council study
reported a 4% toxic death rate (6)
. Similarly, reports
have highlighted the relationship between renal impairment and
increased RTX plasma levels (8)
and the danger of failing
to reduce the dose in cases of renal impairment (9)
.
Increasing experience with RTX and further clinician and patient
education regarding the use of this drug is likely to lower this
incidence. Nevertheless, it is appropriate to examine the potential for
rescue agents that may be given on presentation of these symptoms that
may improve the management of these toxicities. In considering potential rescue agents, it is necessary to understand the cellular biochemistry of RTX and other appropriate antifolates where rescue agents have been successfully used. An example of this is the administration of LV after high-dose MTX, usually at 24 h, to prevent otherwise lethal effects of the drug (10 , 11) . These protocols have been developed to promote penetration of MTX into pharmacological sanctuary sites and drug-resistant tumors and to selectively rescue normal proliferating tissues from the toxic effects of the drug. LV is commonly used later if patients on standard dose therapy have an unexpected or protracted toxic response. A number of mechanisms, probably acting in concert, have been suggested to explain LV rescue of MTX cytotoxicity at the cellular level (12) . These include: (a) circumvention of DHFR inhibition by replenishment of the folate pool; (b) displacement, by FH2 polyglutamates, of MTX from DHFR; and (c) in the continued presence of extracellular MTX, prevention of further drug uptake and polyglutamation.
Previous studies demonstrated that coincubation of mouse L1210 leukemia
cells with RTX and LV reduces the growth-inhibitory activity of the
drug primarily by competing for drug uptake and polyglutamation
(1
, 13)
. Polyglutamation of RTX has greater
pharmacological consequences than that of MTX. RTX polyglutamates of
high chain length (tri-hexaglu) form the majority of the intracellular
drug pool and are approximately two orders of magnitude more potent
inhibitors of TS than the parent drug (Ki RTX, 60 nM;
tetraglu, 1 nM). Thus, cells expressing low levels
of, or mutant FPGS are resistant to RTX (reviewed in Ref.
14
). A property of RTX polyglutamates shared with MTX
polyglutamates is a reduced rate of efflux giving rise to drug
retention and cytotoxicity under short exposure conditions. Indeed,
polyglutamation of RTX is so rapid that if LV is added to cultured
L1210 cells just 4 h after the drug, it is considerably less
effective in reducing the cytotoxicity (13)
.
Studies in mice have also shown rapid RTX polyglutamation in normal tissues and tumors (2 , 15 , 16) , which is responsible for: (a) slow clearance of the drug from the tissues relative to plasma; and (b) activity by intermittent bolus dosing. Coadministration of LV (20 mg/kg twice daily) prevented both antitumour activity and toxicity (1) . Studies have not investigated the effect of delaying LV administration by, for example, 4 or 24 h, or reducing the LV dose and examining any potential selective rescue effect that this may have for normal tissues. Moreover, the studies described below also do not address this issue but rather the role that LV may have in rescuing patients presenting with severe gastrointestinal toxicity/myelosuppression, which may be 8–15 days after RTX administration. This will be at a time when a proportion of the proliferating cell fraction will have either died or been committed to cell death pathways, but in contrast with cells in culture, a proportion should also have escaped damage by, for example, not being in an active phase of the cell cycle (G0) during the time of maximal drug exposure. Such cells may still have accumulated low cytostatic or growth-inhibitory concentrations of RTX polyglutamates and may be sensitive to LV administration if it could reduce the concentration and stimulate cell division and tissue regeneration.
RTX polyglutamates are slowly lost from tissues (2 , 16) , believed to be partly through the action of FPGH. This enzyme hydrolyses high chain length folate or antifolate forms to shorter polyglutamates or parent drug that can be effluxed (17 , 18) . Recently, M. Rhee and J. Galivan have confirmed that raltitrexed tetraglutamate is a substrate for FPGH.6 It may be speculated that the hydrolysis products become substrates for repolyglutamation, thereby reducing the rate of drug loss. Thus, a dynamic situation may exist in tissues offering potential points of intervention for LV. For example, its metabolic products (reduced folate cofactors and polyglutamates) could compete or regulate the formation or hydrolysis of RTX polyglutamates already formed inside cells.
A further potential site for interaction of delayed LV is competition
with RTX for drug influx via the RFC. RTX undergoes a third phase of
elimination in mice and humans that is very slow (mice, 3 h; humans,
257 ± 63 h; Refs. 8
and 16
).
During this time viable cells, or quiescent cells stimulated to
re-enter the cell cycle, may be exposed for some days to
cytostatic/growth-inhibitory concentrations that may inhibit normal
tissue function. The concentration in human plasma 5 and 15 days after
a standard dose of 3 mg/m2
RTX is 2–10
nM and 2–4 nM, respectively
(19)
,7
a growth-inhibitory concentration for many cell lines in culture, and
could therefore inhibit drug uptake and promote recovery.
To investigate whether delayed administration of LV might be a clinical option for patients presenting with a severe pattern of toxicity, in vitro and in vivo models have been used to study the effects of LV on RTX polyglutamation and cytotoxicity. Additionally, the effects of LV have been investigated on the drug-induced histopathological changes to BALB/c mouse small intestinal epithelium and the rate of recovery of body weight in mice. This mouse strain has been shown previously to be highly sensitive to the gastrointestinal effects of RTX (20) and provides a useful model for studying the potential of rescue agents. Data are presented herein that supports the clinical use of LV as part of more generalized supportive care, immediately after the relatively rare severe pattern of life-threatening toxicity is recognized.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Measurement of RTX Polyglutamates in Human W1L2 Cells.
Duplicate 10-ml cultures of W1L2 cells in the logarithmic growth phase
(3–5 x 105
/ml) were treated with
5-[3
H]RTX (specific activity, 8.9–10.8
Ci/mmol) to give a final concentration of 100 nM. Full
details have been published for RTX purification, treatment of W1L2
cells, extraction, and estimation of [3
H]
polyglutamate forms by an ion-pairing HPLC method (23)
.
LV (10 µM) was included in some experiments. This was added 4 h after RTX, and the cells were incubated for an additional 20 h. In one experiment, the RTX-treated cells were centrifuged, washed two times with fully supplemented RPMI 1640 at 37°C, resuspended in fresh medium containing 10 µM LV, and incubated for an additional 20 h. The cellular concentration is given in µM and was calculated taking into account the volume of the cells (1 x 106 cells, 0.89 µl; Ref. 23 ).
RTX Administration to Mice.
Male BALB/c mice, 7–9 weeks of age, were used as a sensitive model for
RTX-induced gastrointestinal toxicity. Details of this experimental
model have been published recently (20)
. RTX requires
injection of relatively high doses over 4–5 days to induce a
significant level of toxicity to the proliferating tissues (weight
loss, diarrhea, and myelosuppression). This is attributable to the high
levels of plasma thymidine and plasma folates in mice that partially
mask the TS inhibitory effects (24
, 25)
.
RTX (provided by AstraZeneca plc as a powder) was dissolved in 0.05
M NaHCO3 at a concentration of 1 or
10 mg/ml and adjusted to pH 9.0 with 1 M NaOH and
injected i.p. into mice at a volume of 0.1 ml/10 g of body weight (10
and 100 mg/kg) at 9.00 on days 1–4 or 1–5. Control mice received
injections of 0.05 M NaOH NaHCO3 or
saline. LV (powder purchased from Lederle Labs Gosport) was dissolved
in sterile water at 10 or 20 mg/ml. Mice received injections i.p. (0.1
ml/10 g of body weight) with LV (9.00 and 17.00 h), starting 24 h
after the 4th or 5th injection of RTX (days 5 or 6) and continuing
until day 7. To eliminate potential effects of administered fluid
volume on recovery, saline was given in place of LV in some groups of
RTX-treated mice. Daily weights of individual mice were recorded, and
mice were inspected twice daily for signs of distress. Mice were
sacrificed if they could not freely access food and water and/or had
>30% body weight loss.
Histopathology of the Small and Large Intestine.
Mice (six per group) received injections of 100 mg/kg RTX daily x
4 days (days 1–4) and saline, or LV was administered at a dose of 100
mg/kg (days 5–7) as described above. Mice were killed by cervical
dislocation 24 h after the final RTX dose, which is also the
staring time for the rescue (day 5). Mice from each "rescue" group
were sacrificed 48 and 72 h after the start of rescue (days 7 and
8, respectively), and the whole length of the intestine was removed.
Tissue was fixed, embedded in paraffin, sectioned, and stained with H&E
as described previously (20)
. Damage to the small
intestine was graded in a blinded fashion as follows: 0, normal; +,
minimal (minimal blunting and fusion of villi and vacuolated epithelium
at the villi tips); ++, mild (some blunting and fusion of villi and
decreased mitotic activity); +++, moderate (blunting and fusion of
villi, decreased mitotic activity, and increased inflammatory cell
infiltration of the lamina propria of the mucosa). Damage to the large
intestine was graded as: 0, normal; +, minimal (occasional necrotic
mucosal epithelial cells); and ++, mild (some necrosis of mucosal
epithelium and increased inflammatory cells in mucosa). Higher grade
damage was not seen to this organ.
Hematology.
Mice (six per group) received injections of 100 mg/kg RTX and 200 mg/kg
LV in the protocol described above. Control mice received saline. Blood
was taken for full blood count and differential white cell count as
described recently (20)
. In one experiment, 200 mg/kg LV
were given instead of saline to control mice. There was no significant
difference in the concentration of peripheral blood elements between
these two control groups.
Measurement of Plasma and Tissue RTX (Polyglutamate) Levels by RIA.
Mice (five or six per group) were injected with RTX and either saline
or LV in the same schedule as described above. Blood was collected by
open cardiac puncture under oxygen/halothane anesthesia (without
recovery) into a 1-ml syringe and transferred to a 1.5-ml
microcentrifuge tube containing 10 µl of heparin (50 units/ml; Leo
Laboratories, Buckinghamshire, United Kingdom). This was immediately
centrifuged at 13,000 rpm in a microfuge for 2 min to separate the
plasma. Plasma was immediately frozen and stored at -70°C. Liver,
kidneys, and intestine were removed immediately after exsanguination.
The epithelium was scraped from the small intestine as described
previously (15)
. All tissues were stored at -70°C until
analysis.
The polyclonal RTX antibody (provided by AstraZeneca) was used to measure the concentration of plasma RTX or tissue RTX polyglutamates concentrations by RIA (16) . This antibody does not cross-react with LV or natural folates.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
10
nM. This growth-inhibitory effect was prevented if 25
µM LV was added to L1210 cells up to 18 h after the
addition of the drug (data not shown). However, growth inhibition
induced by a 10-fold higher concentration of RTX could only be
prevented if LV was added simultaneously. Further studies have
characterized the effect of delaying the LV addition on cell growth
using trypan blue exclusion as a viability measurement at 48 h.
Incubation with RTX (100 nM) alone reduced both the total
number of cells and the number of viable cells compared with controls
by 90 and 97%, respectively (Fig. 1A)
. Furthermore, the number
of viable cells had fallen to <50% of the initial cell count
(t = 0), consistent with a cytotoxic response. Studies
not reported here have shown that when cell cycle distribution is
measured by flow cytometry, an apoptotic sub-G1
population is present at this time. When 25 µM
LV was added simultaneously with RTX there was no drug-induced growth
inhibition, but when it was delayed for 4 or 8 h, the total cell
number was 50 and 20% of control, respectively. Nevertheless, the
cell population had expanded during the 48 h and remained largely
viable. Delaying the LV for 18 or 24 h markedly reduced both the
growth rate of the culture (10% of control by 48 h) and the
number of these cells remaining viable at 48 h (20–30% of
population). In contrast, if dThd was added at 4 or 8 h, the cell
numbers were 100 and 50% of control, respectively, at 48 h, and
all were scored as viable (data not shown).
|
4) was the same as for the L1210 cells]. Seventy-two h
of incubation with 100 nM RTX led to almost complete growth
inhibition and loss of cell viability (Fig. 1B)
.
Furthermore, this effect was characterized by the presence of apoptotic
cells and the loss of clonogenicity (data not shown). Coincubation with
LV gave a pattern of results similar to that seen for L1210 cells with
no reduction in culture growth rate or viability at 72 h. If LV
was added after 4 or 8 h, a reduction in the viable cell
population was observed (70 and 25% of control, respectively, at
72 h). Although the addition of LV at 18 h reduced the cell
number to <20% of control, there was nevertheless an expansion in the
population of viable cells over the 72-h period (approximately one
population doubling), and the number of viable cells was three times
greater than with RTX alone. Rescue effects were not seen when LV was
added after 24 h and particularly after 48 h. In contrast,
dThd could be added 18–24 h after the addition of 100
nM RTX without affecting viability (90%),
although the number of cells was 50% of controls, consistent with
growth inhibition before the addition of dThd, i.e., one
cell doubling (data not shown).
Polyglutamation of RTX in W1L2 Cells.
The rate of drug uptake and polyglutamation during continuous exposure
to 100 nM RTX is shown in Fig. 2
. RTX polyglutamates accumulated inside
cells rapidly so that a 40-fold higher intracellular concentration was
measured at 4 h compared with the extracellular environment (4
µM). Mono-diglutamates accounted for just 2% of the
total drug pool, and the predominant drug forms were tetraglutamates
and pentaglutamates. Accumulation continued more slowly after this
time, giving levels of RTX polyglutamates of 6.5 and 11
µM at 18 and 48 h, respectively. The rate of
conversion to polyglutamates appeared to slow with time as the mono and
diglutamate fractions increased to 9 and 15% of the total drug pool,
respectively. The ratio of the cellular RTX polyglutamates (>diglu) to
the extracellular parent RTX pool at 48 h was 110. These data
represent uptake in a dividing cell population because dThd was present
during the incubation to prevent cell death. Data not shown
demonstrated an intracellular:extracellular RTX ratio of 240 when the
extracellular concentration was 10 nM (99% >diglu). dThd
was not included in these experiments because growth was only inhibited
50% at this concentration of drug, and all cells were viable, as
measured by exclusion of trypan blue.
|
. Furthermore, a 30%
reduction in tetra and pentaglutamate levels was observed, suggesting
hydrolysis to lower chain length polyglutamates that had been effluxed.
Because the 4-h exposure to 100 nM was only 50%
growth inhibitory, there was expansion in cell numbers during this
period; therefore, the intracellular concentration was in fact lower
but has been appropriately adjusted for the purpose of reflecting the
amount of drug truly lost from the cells. A striking effect was seen
when cells that had been incubated with RTX for 4 h were placed in
drug-free medium for 20 h (Fig. 3B)
. In the absence of
LV, there was 25% loss of the intracellular total drug pool in this
drug-free period. These results again have been normalized for the
increase in cell numbers and hence "dilution" of the drug pool. The
distribution of polyglutamates also changed during the drug-free
period, leading to expansion of the pentglu, hexaglu, and >hexaglu
pool. Tetra-hexaglutamates accounted for 97% of the intracellular drug
compared with 90% at 4 h. LV prevented this chain elongation in
the drug-free period and reduced the concentration of the intracellular
polyglutamate pool by 26% compared with cells incubated with drug
alone. This led to a markedly different distribution of polyglutamates
with >tetraglutamate accounting for 86 and 28% of the polyglutamate
pool in the absence and presence of LV, respectively.
|
. Mice given 100 mg/kg/day
of RTX for 4 days (days 1–4) and saline "rescue" on days 5–7 lost
body weight from day 2 with a day 8–9 nadir (75% of starting
weight). Weight gain was rapid after this time, with the mice reaching
95% of their starting weight by day 14 (Fig. 4A)
. Mice that
had received 100 mg/kg of LV twice a day on days 5–7 lost the same
amount of weight as the saline "rescued" mice up to day 7 (78%
of starting weight). In contrast, however, further weight loss did not
occur in the LV-treated mice, and significant weight gain was seen on
day 9 (83% of starting weight) with recovery to at least 95% of
starting weight by day 12. LV-treated mice had higher body weights than
saline-treated mice on days 8–12 that were statistically significant.
However, excessive weight loss (>30%) was seen in some mice in both
groups; therefore, these mice were culled: 6 of 20 saline-rescued mice
and 3 of 21 LV-rescued mice.
|
. However, six of seven mice given LV (200 mg/kg twice
daily, days 6–8) responded well and gained weight from day 8 or 9
(74% of initial weight). This clearly demonstrates that LV can rescue
from potentially lethal doses of RTX in mice.
Effect of LV on RTX-induced Changes to Intestinal Histopathology.
RTX (100 mg/kg daily days 1–4) induced "mild" (++) to
"moderate" (+++) changes to the small intestinal epithelium on days
5 and 7, which included some blunting and fusion of villi, a marked
decrease in mitotic activity, and an inflammatory cell infiltrate (data
not shown). Saline or LV rescue (200 mg/kg twice daily) was given days
5–7. In the saline-treated group, changes were classified as
"moderate" (+++) on day 7 (Table 1)
.
Some recovery was seen by day 8 with one of six mice having
"minimal" changes (grade +), three of six grade ++, and two of six
having normal small intestinal histopathology. In comparison,
LV-treated mice had "mild" damage to the small intestine on day 7
and by day 8, three of six mice had grade +, and three of six had
normal histopathology. The colon was affected to a lesser extent by
RTX, and changes were characterized only as "mild" (day 7) necrosis
of the mucosal epithelium and some evidence of the presence of
inflammatory cells with normal by day 8. LV administration
reduced the effects on day 7 from grade ++ to +.
|
. LV (200 mg/kg days
1–4) did not significantly increase these values on day 8 but did on
day 9 (147 and 75% for neutrophils and platelets, respectively). In a
second experiment (day 9 only), the neutrophil and platelet count was
again higher in the LV group but only attained statistical significance
for the neutrophil count.
|
2 pmol/ml (21 and 1
ng/ml) by days 7 and 15, respectively (Fig. 5A)
. Liver RTX levels were
considerably higher (11 and 3 nmol/g on days 5 and 7, respectively),
and a concentration of 1 nmol/g was detected from days 8 to 15 (Fig. 5B)
. Thus, the liver:plasma ratio increased from 25 on
day 5 to 400 on day 15. Mice that received 200 mg/kg LV had
statistically significantly lower plasma (2–4-fold) and liver
(4–13-fold) drug levels on days 7–15 compared with mice that had
received saline (Fig. 5)
. The day 7 results are included in Table 3
, together with those obtained after
administration of 100 mg/kg LV. Compared with their respective
saline-rescued controls, the reduction in RTX in the plasma and RTX
(polyglutamates) liver was similar in both the 100 and 200 mg/kg
LV-rescued groups of mice. Measurements are also shown for small
intestinal epithelium and kidney at this time. There were 3–4-fold
lower levels in the intestinal epithelium and kidney at both 100 and
200 mg/kg LV, respectively.
|
|
30 pmol/ml. Plasma, liver, and kidney levels of RTX
on day 7 were 22 pmol/ml, 1200 pmol/g, and 290 pmol/g, respectively.
These were 3–4-fold lower than seen at the higher 100 mg/kg daily x 4 dose. LV significantly reduced (P < 0.05) the
levels in these tissues by 4-, 10-, and 3-fold for plasma, liver, and
kidney, respectively (data not shown). Intestinal epithelial levels
were not measured.
Additional experiments were performed with a single injection of 100
mg/kg RTX (day 1), with LV (100 mg/kg) being administered from days 4
to 7. This single administration is well tolerated by mice because of
the high concentration of plasma dThd; therefore, this study is not in
the context of that of a toxicity model. At the start of the LV rescue
(day 4), the RTX plasma and liver levels were 4 and 1000 pmol/g,
respectively. The plasma, liver, and kidney levels of RTX on day 7 were
4 pmol/ml, 630 pmol/g, and 80 pmol/g, respectively, which is 20-,
8-, and 15-fold lower, respectively, than that seen in the 100 mg/kg
daily x 5 schedule. After LV, no differences in the RTX levels
were observed in plasma and kidney on day 7 (two experiments) and days
9 and 11 (one experiment; data not shown). However a 2–3-fold lower
level was measured in the liver (Table 4)
. Similarly, a 4-fold lower liver
concentration was observed when mice were given 200 mg/kg LV. Small
intestinal epithelial levels were not measured because previous
experiments had shown that the concentration of RTX is rapidly
"diluted" in this tissue when noncytotoxic drug doses/schedules are
given to mice. For example, if day 5 drug levels are compared in the
two schedules, plasma, liver, and kidney were 30-fold lower in the
single administration schedule. However, the small intestinal
epithelial concentration was nearly 200-fold lower at this time (data
not shown).
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Potential rescue agents include dThd and LV. dThd, which circumvents TS inhibition induced by RTX through the dThd salvage pathway, is not available as a registered agent (1) . However, LV is available because it is used clinically as a rescue agent with high-dose MTX and as a promoter of the cytotoxicity of 5-FU (9 , 27) . The different function of TS compared with DHFR in folate metabolism and the importance of RTX polyglutamates as potent TS inhibitors suggest that delayed LV rescue may be less effective for RTX than MTX. Nevertheless, LV has been shown to reverse the growth-inhibitory effects of RTX in vitro when added to the culture medium at the same time as RTX because of the prevention of the formation of RTX polyglutamates (13) . This has now been explored further to gain insight into the relationship between the time interval between RTX and LV and the effects on cell viability. Even when 25 µM LV was given 8 h after RTX to L1210 or W1L2 cells, the cell population was able to expand (although not at control rate), with all cells remaining viable 48 or 72 h later. In the slower growing W1L2 cells, the LV could even be added as late as 18 h, and some population expansion was still observed.
W1L2 cells were also used to examine the effect of the delayed addition of LV on RTX polyglutamation and on the stability of preformed RTX polyglutamates. Polyglutamation of [3 H]RTX was rapid over the first 4 h, as described previously. However, because polyglutamation continued over the following 20 h, albeit at a slower rate, the addition of LV at 4 h prevented further accumulation of the drug as high chain length polyglutamates. In turn, this appeared to promote drug loss through efflux of low chain length polyglutamates or parent drug. In wash-out experiments, when the RTX-containing medium was exchanged at 4 h for LV-containing medium, loss of polyglutamates was greater than that seen without LV. Although the mechanism(s) involved are unknown, it can be hypothesized that expansion of the intracellular folate pool through the administration of LV inhibits FPGS (either by competition or feed-back inhibition) so that any hydrolysis of RTX occurring through the action of FPGH leads to the efflux of RTX from the cells. However, the in vitro loss of polyglutamates after the addition of LV was not large and may relate to low expression of FPGH in cultured cells compared with either tumor cells in vivo or some normal tissues (17) . Indeed, our own studies have failed to detect significant enzyme levels in tissue culture compared with tissues such as liver and small intestine (data not shown). This may lead to relatively high polyglutamate stability in vitro compared with in vivo. Overall conclusions from these in vitro studies are that, in cycling cells, the delayed addition of LV may: (a) reduce further uptake of RTX; and (b) reduce the level of preformed RTX polyglutamates. These results are useful in explaining results in context of an in vivo animal model.
A previously described BALB/c mouse model (20) was used to study the effect of delayed administration of LV on mouse body weight and tissue drug levels after RTX administration. Toxicity (weight loss and diarrhea) continued after completion of the course of daily 100 mg/kg injections of RTX (days 1–4). LV rescue, which commenced 24 h later (day 5), led to less weight loss and more rapid recovery of weight. Histological examination of the intestine confirmed that this effect was attributable, at least in part, to effects on the intestinal epithelium. Furthermore, the potentially lethal dose of 100 mg/kg RTX daily x 5 (days 1–5) were prevented by LV administration (commenced day 6). LV also effectively stimulated bone marrow recovery, as shown by recovery of peripheral neutrophils and platelets.
The measurement of RTX (polyglutamates) by RIA after 100 mg/kg RTX days 1–4 demonstrated lower levels in plasma, liver, kidney, and small intestinal epithelium in LV-treated mice. This difference was as much as an order of magnitude in the liver. These data are consistent with LV interrupting homeostatic regulation of RTX polyglutamates and promoting recovery of TS activity.
Although these preclinical data support the role of LV as a rescue
agent, it is worth considering the relevance of the experimental model.
The LV was given as a 3-day course starting 24 h (day 5) after the
final (day 4) RTX injection. This timing is largely determined by the
narrow "window" in which mice continue to lose weight before
recovering naturally. However, because mice require high doses of RTX
(100 mg/kg daily x 4) to induce the level of toxicity required,
plasma drug levels at the start of the "rescue" were still 400
pmol/ml (0.5 µM), despite the fact that the vast majority
of the drug has been eliminated from the plasma. This compares with
2–10 pmol/ml in humans 5 days after administration of a standard 3
mg/m2
dose (19)
. The question
therefore arises whether, in the mouse model, LV is largely preventing
further drug uptake and polyglutamation and consequently preventing
greater damage rather than reversing the effects of preformed RTX
polyglutamates. To address this question, two different experimental
designs were used. The RTX dose was reduced to 10 mg/kg daily x 4
(days 1–4). This resulted in plasma levels of 30 pmol/ml on day 5.
LV (30 mg/kg twice daily, days 5–6) significantly reduced the
concentration of RTX in the plasma, liver, and kidney on day 7. A
single injection of 100 mg/kg (not sufficient to induce toxicity) was
given to mice on day 1, and LV was given from days 4 to 6. The plasma
and liver concentrations of RTX at the start of the LV rescue were 3–4
and 1000 pmol/g, respectively. These are values similar to those
observed in human plasma and liver 5 days after treatment (Ref.
19
and data not shown). LV administration reduced the RTX
concentration in the liver by 3-fold (days 7, 9, and 11) but not in
the plasma and kidney. Unfortunately, intestinal epithelium
measurements could not be made because previous experiments had shown
that a single injection of RTX does not cause marked gastrointestinal
toxicity in mice, and hence the rapid proliferation of the intestinal
epithelium "dilutes" out the RTX in this tissue. These data suggest
that LV can reduce tissue levels of RTX in mice when the plasma
concentration at the start of the rescue are of the same magnitude as
those that may be seen in patients presenting with toxicity.
Although the exact mechanisms of the LV effect and relevance of the preclinical model have some uncertainties, the fact remains that LV can rescue mice from toxicity at a time when considerable damage has been done to the small intestinal epithelium and suppressive effects are being seen in the bone-marrow. This observation is therefore consistent with some cell populations in the intestinal epithelium or bone marrow, e.g., stem cells, being "resistant" to the drug by being quiescent for example. If quiescent Chinese hamster ovary cells in culture are exposed to RTX for 24 h before being stimulated to proliferate they are relatively resistant to the effects of the drug.8 This may be attributable to low activity of either the RFC or FPGS during quiescence, and previous reports concerning the effects of MTX in quiescent tumor cells demonstrated reduced MTX polyglutamate formation during this period (28 , 29) . Thus, quiescent cells may evade the cytotoxic effects of the transiently high drug concentrations seen in the plasma after bolus injection of RTX. However, these cells may still be exposed to, or have accumulated, growth-inhibitory concentrations of the drug that could still manifest as toxicity. Intervention with LV has the potential to overcome this effect and stimulate growth more quickly.
At 100 mg/kg RTX daily x 4 days, the toxicities observed in
BALB/c mice are gastrointestinal toxicity combined with
myelosuppression. Experiments described herein suggest that LV can
prevent the lethal combination of toxicities by preventing further RTX
uptake into cells and/or promoting RTX efflux. Although extrapolation
from the preclinical models is difficult, it is nevertheless
recommended that LV be given alongside vigilant supportive care in
patients presenting with the life-threatening combination of
gastrointestinal and bone marrow toxicities. LV rescue should be given
earlier if protocol violations, such as failure to reduced dose in
patients with poor renal function, are recognized soon after treatment.
It is difficult to recommended a clinical dose of LV to use because the
dose of RTX used in the mouse experiments (100 mg/kg x 4 days)
was necessarily higher than that in humans (3
mg/m2
x 1; equivalent to 1 mg/kg x 1).
For this reason, high concentrations of LV were also used (100 and 200
mg/kg twice daily x 3 days). Although this translates into a
human dose of 300–600 mg/m2
, it is probably not
necessary to give equivalent high doses to humans. It is therefore
recommended to give a similar LV dose to that used for high-dose MTX,
e.g., 25 mg/m2
every 6 h and
continue until symptoms resolve. Finally, it should be emphasized that
although LV has a place in the rescue of toxicity, its use should not
be advocated routinely because it is likely to induce similar rescue
effects in tumors. The possibility of selective effects on normal
proliferating tissues versus tumors, perhaps at lower LV
doses, should be the subject of further investigation.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 The work was supported by the Cancer Research
Campaign (United Kingdom) and AstraZeneca (United Kingdom). 
2 Present address: Medical Oncology Department,
St. Bartholomew’s Hospital, West Smithfield, London EC1A 7BE, United
Kingdom.
3 Present address: Sydney Cancer Centre,
Camperdown, New South Wales, Australia 2050.
4 To whom requests for reprints should be
addressed, at The CRC Centre for Cancer Therapeutics, Institute of
Cancer Research, 15 Cotswold Road, Sutton, Surrey SM2 5NG, United
Kingdom. Phone: 44-181-643-8901; Fax: 44-181-643-6940; E-mail: annj{at}icr.ac.uk
5 The abbreviations used are: RTX, raltitrexed;
TS, thymidylate synthase; 5-FU, 5-fluorouracil; LV, leucovorin,
5-formyl tetrahydrofolate; MTX, methotrexate; DHFR, dihydrofolate
reductase; FPGH, folylpolyglutamate hydrolase; dThd, thymidine.
6 M. Rhee and J. Galivan, personal
communication.
8 D. Blakely, personal communication
(AstraZeneca).
Received 2/11/00; revised 5/22/00; accepted 5/22/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-polyglutamates by folylpolyglutamyl hydrolase derived from various tumors and normal tissues of the mouse. Cancer Res., 46: 2230-2235, 1986.This article has been cited by other articles:
|
|
 |
|
  E. Van Cutsem, D. Cunningham, J. Maroun, A. Cervantes, and B. Glimelius Raltitrexed: current clinical status and future directions Ann. Onc., April 1, 2002; 13(4): 513 - 522. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. M. Pritchard, L. Bower, C. S. Potten, A. L. Jackman, and J. A. Hickman The Importance of p53-Independent Apoptosis in the Intestinal Toxicity Induced by Raltitrexed (ZD1694, Tomudex): Genetic Differences between BALB/c and DBA/2 Mice Clin. Cancer Res., November 1, 2000; 6(11): 4389 - 4395. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
) and viable cell number as determined by
trypan blue exclusion (
) were determined at 48 and 72 h for
L1210 and W1L2 cells, respectively. The horizontal broken
line is the initial cell density (5 x
104/ml). Cells were incubated continuously with 100
nM RTX; 10 µM LV was added either
simultaneously or at later times as given. Bars, range
of one to three separate experiments (each performed in duplicate).
); tri- to pentaglutamate (•).
) or 24 h (10
µM LV added at 4 h; 
), days 5–7
inclusive. Results are given as means; bars, SD. The
difference between saline- and LV-treated mice was statistically
significant (P < 0.05) at all time points.
A, plasma: n = 5 or 6 mice. Pooled
results from two experiments are shown for day 7, n = 10 or 11 mice. B, liver: n = 5 or
6 mice. Pooled results from two experiments are shown for days 7 and 9
(n = 11 or 12 mice).