|
Original Article
Cetuximab-induced hypomagnesaemia aggravates peripheral
sensory neurotoxicity caused by oxaliplatin
1Division of Gastroenterologic and General Surgery, Department of Surgery, Asahikawa Medical University, Asahikawa, Japan; 2Division of Chemotherapy, Higashi-Asahikawa Hospital, Asahikawa, Japan; 3Department of Hospital Pharmacy and Pharmacology, Asahikawa Medical University, Asahikawa, Japan
|
|
Abstract
Calcium and magnesium replacement is effective in reducing oxaliplatin-induced neurotoxicity. However, cetuximab treatment
has been associated with severe hypomagnesaemia. Therefore, we retrospectively investigated whether cetuximabinduced
hypomagnesaemia exacerbated oxaliplatin-induced neurotoxicity. Six patients with metastatic colorectal cancer who
were previously treated with oxaliplatin-fluorouracil combination therapy were administered cetuximab in combination with
irinotecan alone or irinotecan and fluorouracil as a second-line treatment. All patients had normal magnesium levels before
receiving cetuximab. The Common Terminology Criteria for Adverse Events version 3.0 was used to evaluate the grade of
neurotoxicity, hypomagnesaemia, hypocalcaemia, and hypokalemia every week. All six patients had grade 1 or higher hypomagnesaemia
after starting cetuximab therapy. The serum calcium and potassium levels were within the normal range at the
onset of hypomagnesaemia. Oxaliplatin-induced neurotoxicity occurred in all patients at the beginning of cetuximab therapy,
with grade 1 neurotoxicity in five patients and grade 2 in one patient. After cetuximab administration, the neurotoxicity worsened
in all six patients, and three progressed to grade 3. Among the three patients with grade 3 neurotoxicity, two required a
dose reduction and one had to discontinue cetuximab therapy. A discontinuation or dose reduction in cetuximab therapy was
associated with exacerbated oxaliplatin-induced neurotoxicity due to cetuximab-induced hypomagnesaemia in half of patients
who had previously received oxaliplatin. Therefore, when administering cetuximab after oxaliplatin therapy, we suggest serially
evaluating serum magnesium levels and neurotoxicity.
Key words hypomagnesaemia; cetuximab; oxaliplatin; neurotoxicity; colorectal cancer
J Gastrointest Oncol 2010; 1: 97-101. DOI: 10.3978/j.issn.2078-6891.2010.024
|
|
Introduction
Oxaliplatin (L-OHP)-f luoropyrimidine combinations
are widely used in the first-line treatment for metastatic
colorectal cancer (1-3). Due to recent advances in molecular
targeted therapies, cetuximab (Cmab), an anti-epidermal growth factor receptor (EGFR) antibody, is recommended
as the first-line therapy with L-OHP, leucovorin, and
f luorouracil (FOLFOX) or as second-line therapy after a
FOLFOX regimen for stage IV colorectal cancer patients
(4,5).
Peripheral sensory neurotoxicity (PSN) is a doselimiting
toxicity that is associated with L-OHP, which is
the key drug in the FOLFOX regimen. Therefore, a stopand-
go approach has been proposed to manage PSN (6).
PSN can either be transient and acute or chronic due to the
accumulation of L-OHP (2,7). The hallmarks of PSN are
dysesthesia and paresthesia in the limbs, which are triggered
by cold exposure and in some cases accompanied by cramps
(8). PSN occurs in 90% of patients who receive L-OHP
and persists in 30% of patients after one year of stopping
treatment (1). In addition, L-OHP must be discontinued when the cumulative dose reaches 800 mg/m² because
10-15% of cases develop grade 3 or higher functional
disorder (1,9).
Previous studies on the mechanism of PSN reported
that calcium and magnesium replacement effectively
reduced chronic PSN, suggesting that these supplements
are efficacious (10,11). Moreover, the prospec t ive
CONcePT study confirmed the effectiveness of calcium
and magnesium replacement (12). However, Cmab has
been reported to induce hypomagnesaemia (13-15). This
anti-EGFR antibody blocks EGFR in the nephron and
inhibits magnesium reabsorption from the convoluted
distal tubule, leading to magnesium loss from the kidneys
(13-15). Therefore, we retrospectively investigated whether
Cmab-induced hypomagnesaemia exacerbated the chronic
neurotoxicity associated with L-OHP therapy.
|
|
Methods
This study included si x patients with unresectable
metastatic colorectal cancer who had previously
received FOLFOX as a first-line treatment until disease
progression and were treated with Cmab in combination
with irinotecan alone or irinotecan-f luoropyrimidine
combination as a second-line treatment. None of the
patients had KRAS codon 12 and 13 mutations in the
tumor tissue or diabetes mellitus. The present study was
conducted in accordance with the Declaration of Helsinki
for the care for human study adopted by the ethics
committee of Asahikawa Medical University and Higashi-
Asahikawa Hospital. All patients prov ided written,
informed consent.
Patients received Cmab (initial dose of 400 mg/m²
infused over 2 hours, and 250 mg/m² weekly over 1 hour
thereafter) after receiving 1 hour of irinotecan（150 mg/
m2）alone or in combination with fluorouracil, leucovorin,
and irinotecan FOLFIRI (150 mg/m2 irinotecan infused
on day 1 over 2 hours; 200 mg/m2 leucovorin infused over
2 hours, followed by f luorouracil given as a 400 mg/m2
intravenous bolus and then 2400 mg/m2 continuously
infused over 44 hours on days 1 and 2) or Cmab alone
until the occurrence of progressive disease or unacceptable
toxicity. Adverse events were recorded during treatment.
Serum magnesium, calcium, and potassium levels were
assessed at baseline (i.e., within 1 week before starting
Cmab treatment) and then every week thereafter. The
Common Terminolog y Cr iter ia for Adverse Events
version 3.0 (CTCAE) was used to evaluate the grade of
neurotoxicity, hypomagnesaemia, hypocalcaemia, and
hypokalemia. Additionally, the following variables were
evaluated: 1. total L-OHP dose (mg/m²), 2. time (days) from the last L-OHP dose to first Cmab treatment, 3.
cumulative Cmab dose at the onset of hypomagnesaemia,
4. duration (day) and number of Cmab cycles until the
onset of hypomagnesaemia, 5. cumulative dose of Cmab
at the time of neuropathy aggravation, 6. number of
Cmab cycles until neuropathy aggravation, 7. severity of
hypomagnesaemia, hypocalcaemia, and hypokalemia at
the time of neuropathy aggravation, 8. grade of neuropathy
at the t ime of ag g rav at ion, 9. whether Cmab wa s
discontinued or reduced after the neurotoxicity worsened,
10. whether magnesium sulfate was administered, and 11.
whether any patients developed diabetes.
|
|
Results
Table 1 shows the characteristics of the six patients who
were primarily treated with L-OHP-f luoropyrimidine
combination therapy for metastatic colorectal cancer
and then secondarily treated with Cmab-irinotecan
combination therapy. The mFOLFOX6 regimen was
administered to all patients, and the median total dose
of L-OHP was 722.5 mg/m² (320-1105). The median age
at the time of the initial Cmab therapy was 67.5 years
(59-80), and the median time between the last L-OHP
administration and first Cmab administration was 232
days (202-1046). Grade 1 or higher hypomagnesaemia
was obser ved in all si x patients after starting Cmab
therapy, although all patients had normal magnesium
levels before starting cetuximab. The serum calcium and
potassium levels were within the normal range at the
onset of hypomagnesaemia. The cumulative dose, median
duration, and number of cycles of Cmab at the onset of
hypomagnesaemia were 1400 mg/m² (900-1650), 6 days
(29-42), and 5 cycles (4-6), respectively. Among the six
patients with hypomagnesaemia, five were treated with 1
mEq/mL of magnesium sulfate.
All patients experienced mFOLFOX6 regimen-induced
peripheral neuropathy at the beginning of Cmab therapy,
with grade 1 neuropathy in five patients and grade 2
neuropathy in one patient. After Cmab administration, PSN
worsened in all six patients, and three patients progressed
to grade 3. Among the three patients with grade 3 PSN,
two required a dose reduction and one had to discontinue
treatment. The cumulative dose, median duration, and the
number of cycles of Cmab at the time of PSN exacerbation
were 2150 mg/m² (1150-3150), 59.5 days (29-105), and 8
cycles (4-12), respectively. Five of the six patients (83%)
developed hypomagnesaemia prior to PSN progression.
Among these five patients, one had hypokalemia but none
had abnormal calcium and potassium levels. One patient
(17%) whose PSN was exacerbated before the onset of hypomagnesaemia had normal calcium and potassium
levels.
|
|
Discussion
In the MOSAIC study, 90% of the neurotoxic effects
occurred during active L-OHP therapy and the incidence
decreased over time after discontinuation. Up to 70% and
80% improvement was noted after one and two years of
discontinuation, respectively, which means that 20-30%
did not even have improvement (1). Although precise
mechanisms underlying the development of PSN have
not been fully clarified, PSN has been attributed to the
accumulation of platinum in the dorsal root ganglion based
on the results from animal experiments (8). Gamelin et
al. suggested that a possible mechanism may be the effect
of oxalate, a one of the breakdown products of L-OHP, on
neuronal sodium channels (11). Based on this hypothesis,
chelation of oxalate can be a possible candidate for
improvement of PSN. For this reason, L-OHP therapy is
proactively supplemented with calcium and magnesium
for chelat ion of ox a late. Therefore, administer ing
hypomagnesaemia-inducing Cmab to patients who have
been treated with L-OHP over a long period is thought to
aggravate PSN by depleting magnesium that is necessary
to chelate the breakdown the products of accumulated
L-OHP. Furthermore, long-term Cmab therapy has been
reported to influence not only magnesium levels but also
the levels of calcium, potassium, and other electrolytes
(9,14-16). These results suggest that hypomagnesaemia
may not be the only causative etiology. In our patients, low
serum magnesium exacerbated peripheral neuropathy, but
the neuropathic symptoms improved with IV magnesium
sulfate. Therefore, we postulate that hypomagnesaemia may
be pivotal in aggravating peripheral neuropathy. However,
patients whose neuropathy worsened before the onset
of hypomagnesaemia did not necessarily have abnormal
calcium and potassium levels. More studies are needed to
investigate the role of other causative factors besides an
electrolyte imbalance.
There are several reports on the timing of Cmab-induced
hypomagnesaemia. Despite the high degree of interpatient
variability, these reports show a correlation between the
severity and onset of hypomagnesaemia after a median of 3
months (1-6) for grade 2 and 5.5 months (1-14) for grade 3.
Furthermore, additional data clearly indicate a relationship
between the duration of Cmab exposure (6 months) and the incidence/grade of
hypomagnesaemia (9,15). In our study, hypomagnesaemia
appeared within 5 cycles and approximately one month
after initiating Cmab therapy; the neurotoxicity worsened after a median of 8 cycles and approximately 2 months of
therapy. Except for one patient, exacerbated neuropathy
occurred in all patients after a median of 3 cycles and within
one month after the onset of hypomagnesaemia. Aside from
one patient, all patients had grade 1 hypomagnesaemia
when the neurotoxicity began to worsen. Therefore, it is
important to monitor serum magnesium levels shortly after
initiating Cmab therapy. Based on our results, we were
unable to draw any conclusions regarding the relationship
between calcium/potassium levels and exacerbated
neurotoxicity.
Although most patients with grade 1 and grade 2
hypomagnesaemia after Cmab therapy are asymptomatic,
those with grade 3 or higher may present with fatigue
or hypocalcaemia (14). For the latter cases, the current
recommendation is to measure and correct the magnesium
levels if they are low (9); however, the decision to treat
low magnesium remains inconclusive (17,18). That is, if
the decreased QOL due to hypomagnesaemia outweighs
the clinical benefits, the magnesium imbalance should
be treated. If, on the other hand, the anticancer effects
override the hypomagnesaemia, then low magnesium
should be treated less v igorously. In our study, we
discontinued Cmab if hypomagnesaemia progressed, but
also noted that magnesium wasting could be resolved
within 2 weeks (unpublished data). The recovery rate of
less than 4 weeks is consistent with the half-life of Cmab
(9).
Based on extant reports, the incidence of Cmab-induced
hypomagnesaemia is approximately 50% after accounting
for all reported grades (19). However, contrary to our
study protocol, most studies did not measure magnesium
on a weekly basis, raising the possibility that the incidence
of hypomagnesaemia is underestimated, especially for
grade 1 (9). L-OHP-induced neurotoxicity was aggravated
in our study, albeit in a small sample size, with the onset of
grade 1 hypomagnesaemia. Therefore, the early detection
of hypomagnesaemia is essential and should be factored
into the design of large-scale, controlled studies in the
future.
|
|
Conclusion
Although our retrospective analysis was based on a small
sample size, we found that Cmab, as a second-line therapy in
patients with long-term L-OHP exposure, may exacerbate
residual L-OHP-induced neurotox icit y by inducing
hypomagnesaemia. Therefore, we recommend serially
evaluating serum magnesium levels and neurotoxicity when
initiating Cmab treatment after L-OHP therapy.
|
|
References
Cite this article as:
Kono T, Satomi M, Asama T, Ebisawa Y, Chisato N, Suno M, Karasaki H, Furukawa H, Matsubara K. Cetuximab-induced hypomagnesaemia aggravates peripheral
sensory neurotoxicity caused by oxaliplatin. J Gastrointest Oncol. 2010;1(2):97-101. DOI:10.3978/j.issn.2078-6891.2010.024
|
