|
Home:
Web Guide 2: David J Kerr, MD, FRCP
Edited comments by Professor Kerr
Rationale for combining capecitabine with irinotecan
When combining drugs, it is attractive to use those with different
molecular targets. For irinotecan, the target is topoisomerase-1,
and for capecitabine, a prodrug of 5-FU, it is thymidylate synthase.
Capecitabine, which starves the cell of thymidine, might synergize
with irinotecan, which inhibits DNA topoisomerase-1. If the cell
is depleted of thymidine, there is no doubt that irinotecan will
be more effective.
Both drugs are active as single agents. There are also data from
preclinical, in vitro models suggesting that capecitabine/irinotecan
may be synergistic. It makes sense to bring these two mechanistically
novel drugs together into a combination regimen.
Phase I/II trials with capecitabine/irinotecan
Working with colleagues in the Netherlands, we have entered 27
chemotherapy-naïve patients with advanced measurable colorectal
cancer in our phase I/II trial. In the dose-finding phase I study,
the recommended dose was 1,000 mg/m2 twice daily for 14 days for
capecitabine and 250 mg/m2 every three weeks for irinotecan.
At the recommended dose, there was about a 50% response rate with
tolerable side effects. The dose-limiting toxicities were myelosuppression
and diarrhea. At the recommended dose, a 10% grade III/IV toxicity
rate is projected. Capecitabine/irinotecan has manageable toxicity,
is active and is very convenient for patients.
There are two or three other European trials, which have come
to very similar conclusions in terms of dose finding. So, the doses
of irinotecan 250 mg/m2 every three weeks and capecitabine 1,000
mg/m2 twice daily for 14 days every three weeks seem to be the ones
that will be included in all the phase III randomized trials. The
fact that we have four studies showing very similar conclusions
adds weight and certainty to that regimen.
Capecitabine compared to 5-FU/leucovorin
As demonstrated in some very compelling randomized studies in
advanced disease, I think capecitabine is much more useful than
the bolus administration of 5-FU/leucovorin. There are toxicity
and convenience benefits associated with capecitabine.
Since we commonly use infusional 5-FU/leucovorin in Europe and
there are no direct comparisons with capecitabine, it would be interesting
to study. I would guess that capecitabine is probably as effective
with about the same level of toxicity as our rather complicated
48-hour infusional 5-FU/leucovorin regimens, which require pumps
and intravenous access.
If we can substitute capecitabine in the combinations with irinotecan
or oxaliplatin, then that may be an improvement. There are initiatives
in Europe to compare capecitabine/irinotecan to infusional 5-FU/leucovorin/irinotecan.
In pharmacologic terms, such as the degree of inhibition of thymidylate
synthase, there are some data suggesting that capecitabine and infusional
5-FU/leucovorin are approximately equivalent. From small patient
studies looking at pharmacodynamic end points, they do appear similar.
CASE 4:
66-year-old man treated with irinotecan/capecitabine followed
by hepatic resection |
History
This active and fit ex-Army brigadier was diagnosed about
two years ago, at a different cancer center, with Dukes’
C colorectal cancer. At that time, he did not receive adjuvant
chemotherapy.
He re-presented to our center with evidence of advanced
hepatic metastatic disease. He had bilobular disease with
25% of his liver replaced by tumor. His lungs and the rest
of his abdomen were clear. His CEA was ten times the upper
limit of normal. His performance status was 1, and he was
in good shape. After a multidisciplinary team meeting, it
was deemed that he had inoperable disease.
Follow-up
He enrolled on the capecitabine/irinotecan phase I/II trial.
He was entered on dose level 3, which consisted of irinotecan
250 mg/m2 and capecitabine 2,000 mg/m2/day for 14 days. After
four months of therapy, only two out of the original ten metastatic
lesions remained, and they were surgically removed.
Case discussion
After two months of therapy, a CT scan revealed an excellent
partial response. Within the first two months of treatment,
his CEA level came down very quickly to normal, consistent
with the excellent partial response. After another two months
of treatment, the CT scan demonstrated that his disease was
now unilobar. He only had two lesions remaining out of what
were initially ten.
The patient tolerated the treatment well. He did not require
any dose reductions throughout the four-month treatment period.
He did, however, have noticeable thinning of his hair.
His performance status was well maintained. He was able
to maintain his daily activities — chairman of a local
hospital trust and caring for his large house and garden.
Since he only came to the hospital once every three weeks,
rather than more frequently as is customary with the complicated
5-FU infusion regimens used in Europe, he was freed up. This
case illustrates the use of chemotherapy to downsize an inoperable
tumor.
The patient underwent resection of his residual disease
two months ago. The surgeons performed an intraoperative ultrasound
and found nothing to biopsy. His liver has started to regenerate,
his two-month postoperative CT scan was completely clear and
all tumor markers are normal.
We are not planning to restart chemotherapy at this point.
As far as we can tell, he has no residual tumor. Of course,
the likelihood is that there is residual disease, but we are
giving him a period of time off treatment.
He may be part of a group of patients who now have a five-year
survival rate of about 15% to 20%. I believe had we not removed
the residual disease, his median life expectancy would have
been 12 to 18 months.
Increasingly in patients with isolated hepatic metastases,
we may be able to downsize, rather than downstage, their disease
with these powerful chemotherapy regimens. Then, I think they
should have surgery. We are moving into a very different natural
history of this disease.
There are data from neoadjuvant trials with infusional 5-FU/leucovorin/oxaliplatin
that look very interesting. In one quite large surgical series
from the hospital Paul Brousse in France, patients with initially
inoperable disease that was made operable by chemotherapy
had as good a survival as patients with initially operable
disease. Although it is not a randomized trial, it is a sophisticated
anecdote from one of the largest surgical series in the world. |
Gene therapy for colorectal cancer
We have manipulated the common cold virus, the adenovirus, so
that it only divides in cancer cells — those with a mutant
P-53 or some other molecular switch. Then we put new genes into
the virus for enzymes that convert nontoxic, inactive prodrugs into
fairly cytotoxic species.
We clone the enzyme from a bacteria or a virus, so there is no
human equivalent. Then we put the gene for that enzyme into a viral
vector, in our case adenovirus. The enzyme we are interested in
is called nitroreductase (NTR). It converts the prodrug, CB1954,
into a very toxic bifunctional alkylating agent, which is active
in all the colorectal cancer cell lines tested. We have also worked
with cytosine deaminase, another bacterial enzyme, which converts
the antifungal drug 5-fluorocytosine (5-FC) into 5-FU. Another enzyme,
thymidine kinase, converts ganciclovir to a toxic antimetabolite.
We have also made a virus with two enzymes, NTR and cytosine deaminase.
We have shown synergistic cell kill through the generation of both
5-FU and the alkylating agent in the same cell. Therefore, we have
delivered combination chemotherapy with a single virus.
Phase I gene therapy trial
We have just completed our first clinical trial, which is not
published yet. In that trial, we injected the virus directly into
the cancers of patients with hepatic metastases. We saw very significant
expression of the virus and the enzyme, which was the proof of our
principle. Does the virus infect the cells? Are the enzymes made?
In the adjuvant setting, giving the virus via the hepatic artery/portal
vein or directly into the peritoneal cavity could be a clever idea.
I am a supporter of regional chemotherapy, and it seems logical.
No doubt, we are working on developing stealth viruses that can
be given intravenously.
In our phase I trial giving the virus directly into the liver
tumor nodules, there were no toxicities. Once the prodrug is activated
into the cytotoxic species, it has a half-life of only seconds.
The cytotoxic species is confined in the cell and does not diffuse
into the bloodstream. Therefore, it would be very unlikely to see
the systemic toxicity associated with conventional chemotherapy.
Since our virus-directed enzyme prodrug therapy (VDEPT) synergizes
with 5-FU/leucovorin and capecitabine, combination studies with
the virus, the prodrug and probably chemotherapy will be next. Perhaps
some day in the adjuvant setting, we will give a complicated postoperative
chemotherapy regimen and adjunctive gene therapy to the liver, administered
through the portal vein or into the peritoneal cavity. Gene therapy
holds some promise, but it is some ten years away.
Selected references
|
