History
This patient had about a six-month history of weight loss
and abdominal complaints — trouble with intermittent
diarrhea and constipation. He acknowledged being ill for a
few weeks before being admitted to the hospital with a high
fever.
His evaluation revealed large liver metastases, which appeared
abscessed, and a large colon cancer as the primary. He had
no personal or family history of colorectal cancer. The primary
tumor, in the right colon by the hepatic flexure, had seeded
bacteria into his metastases. He was treated with antibiotics,
and his liver abscess was drained.
When we saw him, he was on antibiotics with a resolved bacteremia
and not in very good shape. The primary tumor was minimally
symptomatic. It was not obstructed or bleeding actively. He
was minimally anemic with a hemoglobin of 12 g/dL. His liver
function tests and bilirubin were normal, but his LDH was
900 IU/L. The metastases occupied at least half of his liver.
Clinical course
This patient had a vision of being very aggressive with
his therapy. About three months ago, he enrolled on a phase
II clinical trial in which he received irinotecan/5-FU/leucovorin
and an angiogenesis inhibitor. He did not tolerate the chemotherapy,
he experienced tremendous fatigue such that he was bedridden
for two weeks after chemotherapy, and he was taken off of
the study. Subsequently, he received capecitabine and had
a major response.
Case discussion
Infected liver metastases
This patient either denied his symptoms or was in good enough
condition that the disease was quite advanced when it was
diagnosed. The patient’s primary tumor led to bacterial
seeding into the bloodstream, infection of the metastases
and an abscess. Since bacteria can seed from the bowel into
the bloodstream and then infect a heart valve, colon cancer
can cause Streptococcus bovis endocarditis. It is a wonder
that metastases do not get infected more often. Our group
probably sees two or three cases a year of infected liver
metastases.
Choice of systemic therapy
The only sites of tumor involvement were the liver and colon.
The first question faced was — should his primary tumor
be removed? Since his performance status was marginal, he
was not obstructed and he had recently been treated for bacteremia,
we decided to avoid surgery at that time.
The next question was — how do we optimally treat
this young, but relatively debilitated, colon cancer patient?
The options included irinotecan/5-FU/leucovorin, 5-FU/leucovorin
alone, capecitabine or a phase II study combining irinotecan/5-FU/leucovorin
with an angiogenesis inhibitor.
Originally, most of us made the assumption that irinotecan/5-FU/leucovorin
should be used in the sickest patients where you needed to
make a difference. Surprisingly, the sickest patients are
not those who derive the greatest benefit. In our practice,
probably a quarter of all patients are not candidates for
aggressive, upfront therapy. In those cases, I would use capecitabine
as an alternative.
Phase II clinical trial of irinotecan/5-FU/leucovorin
and Angiozyme®
This patient and his family wanted to be very aggressive
with the therapy. At the first visit, the treatment decision
was actually deferred because of his marginal performance
status. About two weeks later, the patient insisted he was
doing reasonably well, and he did not appear to have an active
infection. Therefore, he was offered aggressive treatment.
Barely making the cutoff for performance status, he enrolled
in the phase II trial with the angiogenesis inhibitor. Since
it was a phase II study, all of the patients were treated
with combination chemotherapy (irinotecan/5-FU/leucovorin)
and the angiogenesis inhibitor. The angiogenesis inhibitor
was Angiozyme® (antiangiogenic ribozyme), a ribozyme to
one of the VEGF receptors.
Consistent with what we have learned, he tolerated the treatment
extremely poorly. His disease burden made him unable to tolerate
the chemotherapy. He did receive an experimental agent, but
there was no evidence that Angiozyme® contributed to his
toxicity. There is concern that the angiogenesis inhibitors
may affect hemostasis and coagulation. This patient had none
of those problems. His main problems were horrendous fatigue
and some diarrhea. He was bedridden after two weeks of chemotherapy.
Predictors of irinotecan toxicity
It might have been a suboptimal decision to enroll this
patient on the clinical trial. In clinical research, the performance
status is a very important component of the patient’s
assessment, but it is a subjective measure. Either the physician
overestimates the performance status or the patient, wishing
to participate in clinical trials, boosts their performance
status and claims to be doing much better than they actually
are, in reality. This patient’s wife called a week later
and said, “He came to see you yesterday and said he
was doing fine, but the truth is he can’t get out of
bed.” This emphasizes how critical it is to make an
accurate assessment of the patient’s performance status.
When discussing the reasons not to use an aggressive three-drug
regimen (irinotecan/ 5-FU/leucovorin), many physicians use
age as an important criterion. Age is perhaps a minor criterion,
but the key issue is the patient’s performance status
and overall condition. As one might have expected from this
patient’s high LDH and marginal performance status,
he did very poorly with the irinotecan-based chemotherapy.
It is anticipated that patients treated with prior radiation
therapy will experience more irinotecan-related toxicities.
There is also concern, not clearly based on data, that patients
with a right colon resection may have more toxicity. The diarrhea
associated with irinotecan is a small-bowel diarrhea. If a
right hemicolectomy is performed, the ileocecal valve is removed;
hence, the flow of diarrheal stool into the colon is greater.
Therefore, there is more liquid stool without a valve at the
ileocecum. Some believe patients with a total colectomy may
also have more irinotecan-induced diarrhea.
Capecitabine monotherapy
He was taken off the study. After about three weeks, he
recovered to his baseline and was started on capecitabine
1,000 mg/m2 twice a day (two weeks on, one week off) to which
he had a major response. I consistently find the dose of 1,250
mg/m2 twice a day, to be more toxic than I am willing to accept;
therefore, I do not use it in my own practice. Occasionally,
I start with a dose of capecitabine as low as 800 mg/m2 twice
a day in patients whom I believe will not tolerate the higher
dose (i.e., patients who had pelvic radiation or elderly patients
who had prior chemotherapy).
This patient has had six cycles of capecitabine. He tolerated
the first four cycles well, the fifth cycle less well and
then developed relatively impressive hand-foot syndrome on
the sixth cycle. He is currently on a “vacation”
from his chemotherapy. His performance status is still not
normal, but he is fully functional. It turns out he had understated
his side effects all along. In retrospect, he had tingling
in his hands and feet.
The question becomes — should capecitabine be continued
at a lower dose? He is petrified about taking a “vacation”
from the treatment. In patients who achieve a response and
in his case a meaningful clinical response, it is often a
battle to discontinue the treatment, even for just a few weeks.
He had about an 80% reduction in his liver metastases. His
primary, however, is still in place. I ultimately convinced
him to stop the capecitabine in order to resect his primary.
Managing patients with a synchronous primary
tumor and metastases at presentation
In colon cancer, the introduction of new agents has somewhat
changed the management of patients with a synchronous primary
tumor and metastases. Anticipating problems from the primary,
for many years it was removed immediately and then the metastatic
disease was addressed.
While I have always been a proponent of that idea, I have
started edging in the other direction — particularly
for patients with bulky disease, because aggressive chemotherapy
may have a big impact. Losing a month of time while the patient
recovers from surgery may take away that window of opportunity
for the patient. The primary concern is the need to wait for
wound healing before initiating systemic therapy.
If the primary tumor is removed first, it can be a scheduled
operation. In a patient with a nearly obstructing sigmoid
colon cancer and metastatic disease, if the primary is not
removed, the patient may present late one night with an obstruction.
If they have received aggressive chemotherapy, they are neutropenic
and thrombocytopenic. Now, a diverting colostomy is required,
because the patient is bowel septic.
Obviously, this is a worse case scenario and an extreme
example of what happens if the primary is not removed. But,
chemotherapy may have enough of an impact that it is important
to give it. Previously, 5-FU/leucovorin provided a marginal
effect on the cancer. Now, the different chemotherapy options
have a major impact.
This patient’s tumor had already been a local problem;
he had already seeded his liver metastases from the primary.
In my mind, it was a problem waiting to happen. In general,
right colon primaries are much less problematic, because the
stool is liquid in the right side of the colon. So, very rarely
do they present as an obstruction. They may bleed, but this
patient really had not bled.
On repeat colonoscopy, he had an ulcerated tumor without
bleeding in his right colon by the hepatic flexure. It appeared
flattened and largely necrotic, compared to what was initially
a fungating mass.
After his operation, if all goes well, the decision will
be to either go back to irinotecanbased chemotherapy or continue
capecitabine. He had horrendous toxicity with irinotecan,
and there are no guarantees he would not have that toxicity
again. However, his performance status has improved to the
point where he may tolerate irinotecan. |
