Michael H. Levy, MD: This 38-year-old white male first came to
his physician in January of 1993 complaining of epigastric and low
back pain. In March of 1993, he was diagnosed with pancreatic cancer
that was metastatic to his liver and his retroperitoneal lymph nodes.
He came to Fox Chase for treatment in May and was placed on a phase I
protocol looking at topotecan (Hycamtin) as a chemotherapeutic agent.
We saw the patient in consultation at the time of his first
chemotherapy admission and he said that the pain was localized to his
abdomen, rated at a 5 out of 10 on MS Contin (morphine sulfate) 60 mg
every 12 hours, which he was supplementing with 15 mg of short-acting
morphine in the form of Roxanol (morphine sulfate) four times a day.
He described the pain as throbbing, cramping, stabbing, and radiating
into the back. He said there was a constant component to the pain,
but there were also unbearable spasms and exacerbations. The patient
reported urinary hesitancy and bladder spasms. In addition, perhaps
due to noncompliance or inappropriate laxative therapy, he was
alternating between constipation and diarrhea.
One of the things that we try to do in evaluating patients is to get
a sense of the pain etiology. We knew about his cancerwhere and
what it wasand how it was going to be treated. But, what did we
know about his pain? We thought it was predominantly visceral, and by
its central location and the radiation into the back location,
empirically we thought it probably involved the celiac plexus. We
also thought that perhaps the patients pain had either a
neurogenic or a neuropathic component. We thought that his urinary
hesitancy and the constipation/diarrhea cycles were due to opioid
toxicity, though it was possible that the diarrhea also might be
caused by pancreatic insufficiency.
As a first step, we decided to stick to simple things. We increased
his MS Contin to 90 mg q12h and his Roxanol to 30 mg q4h as needed.
He received his first dose of chemotherapy. Topotecan has very
minimal side effects, and does not cause a lot of nausea or sedation.
When the patient came in the next day for another reason, we saw him.
Unfortunately, he was sedated and nauseated. So, we said, OK, where
do we go from here?
Topotecan is a phase I drug, so it was not even clear that his cancer
was going to respond to it.
I would like to open this case to discussion. How would each of you
proceed at this point?
Russell K. Portenoy, MD: What did the CT scan show?
Dr. Levy: The CT scan showed a large mass in the head of the
pancreas. There was retroperitoneal lymphadenopathy. The scan did not
show any encasement of the celiac plexus. There was involvement of
the lymph nodes next to the pancreas, and there were a number of
Dr. Portenoy: The reason I asked that question is because the
obvious decision for most of us here, I think, would be whether or
not to proceed to celiac plexus neurolysis for this patient. However,
in patients who are totally socked in, whose tumor has spread
retroperitoneally, a celiac plexus block is not going to have much effect.
Dr. Levy: One of our concerns was not knowing how much of the
pain was due to the pancreas, how much was due to the retroperitoneal
lymph node involvement, and how much was due to the liver
involvement. We have had some difficulty obtaining an adequate
response from celiac plexus block in patients with pain due to liver
metastasis. On the other side, we have found celiac plexus block to
be the easiest, most effective neurolytic procedure, with the least
side effects, to relieve the pain from advanced pancreatic cancer.
The procedure might also help the patients constipation. We
have also found that, in a patient like the one in this case report,
the sooner we did the block, the betterthat is, before his
blood counts went down and he became infected from his chemotherapy.
Peter S. Staats, MD: Given the CT scan, proceeding with the
celiac plexus block has a lot of advantages. It would decrease the
opioid requirements. It also would relieve the constipation because
you would have a decrease in the sympathetic tone of the gut. About a
year ago, we published data showing that, across the board in
pancreatic cancer patients, celiac plexus block increases life
expectancy by about a month.
Dr. Portenoy: The issue in our center would be whether or not
to first try an alternative opioid. Two major points on our decision
tree would be 1) to change to a different opioid and administer it by
perhaps a nonoral routegiven the concerns that youd have
about absorption in someone with alternating diarrhea and
constipation, and possibly malabsorption due to pancreatic
insufficiencyor 2) go to a celiac plexus block. The decision,
to me, at that point would be dependent on what was going to happen
in terms of the antineoplastic therapy. If there was an expectation
that his blood counts might start to drop rapidly, that would be an
indication to choose celiac plexus block quickly. If there was
concern about chemotherapy toxicity, but not about opioid toxicity,
then you might want to hold off on the block to see whether or not
the pain cleared over the next few days, or perhaps wait until you
obtained a topotecan level to see whether he was metabolizing the
Elliot S. Krames, MD: In my opinion, celiac plexus block
absolutely would be indicated here, but on the other hand, because
the patient was sedated on 180 mg of MS Contin per day, which is not
an extremely high dose of MS Contin, switching to another agent might
control his pain and decrease the sedation.
Stuart Du Pen, MD: We have to remember that celiac plexus
block is not without complications. One of my patients developed
paralysis following the procedure due to arterial spasm and lack of
blood supply to the cord. I discuss this with each patient because I
think they must be aware of this possibility, even though it is very rare.
In addition, hypotension is common, due to the vasodilation. Diarrhea
is another possible side effect from celiac plexus block, although
patients with pancreatic cancer are usually accustomed to having
diarrhea because of the pancreatic involvement. There is also another
possible late side effect due to the vasodilationI have found
that, when patients who have had celiac plexus block become
dehydrated or cachectic, they develop postural hypotension earlier
than patients who have not undergone a block.
Dr. Portenoy: True, and there is also the possibility of other
rare but serious complications, like renal puncture with hemorrhage
and renal injury by phenol injection, or alcohol injection into the kidney.
Dr. Krames: It depends on the skill of the one who does the
procedure. When you talk about celiac plexus block, did you mean as a
local anesthetic or did you mean neurolysis?
Dr. Du Pen: I agree. In my 31 years of practice and 2,000
celiac plexus blocks with alcohol, I have never encountered a renal
injury. I attribute this to my training with Dr. D.C. Moore.
Dr. Levy: We made our assessment and we thought that he did
not have pancreatic insufficiency. He really was having more
constipation than diarrhea. His constipation and his urinary
hesitancy were due to opioid toxicity. In fact, it was quite clear
that this patient was having opioid toxicity and he was very much in
favor of having a block. So, on May 28 the celiac plexus block was
performed. Our anesthesiologist used a transaortic approach with 100%
alcohol. The procedure was fluoroscope-guided; there was good spread
of the dye; and 50 mL of 100% alcohol was injected on both sides.
Another question is what to do with the opioids around the time of
the block. We usually reduce them by 50% to 75%. We continue patients
on some opioid dosage to prevent opioid withdrawal and to control the
back pain that patients often have for a day or two after plexus
block. So the patient was continued on MS Contin 30 mg q12h, and the
next day his pain measured 1.5 on a scale of 10. His urinary flow was good.
His pain was still centered in the abdomen and back; it was not clear
whether it was from the procedure or if it was still the boring,
penetrating pain of pancreatic cancer.
Dr. Krames: This raises an important issue. The
anesthesiologist went ahead and did an alcohol block without a
diagnostic block, and that is controversial in the literature.
Personally, I support that approach. I think there is no good reason
for a diagnostic block in this setting. In pancreatic cancer, either
it works or it doesnt work, and submitting a patient to two
procedures is not cost-effective, in both money and quality of life.
It is a painful procedure and I see absolutely no reason to do two procedures.
Dr. Du Pen: The problem is we do not have an accurate
diagnostic tool. The local anesthetic agents will spread rapidly and
far. The effect of the high concentration of alcohol is limited to a
very small area. Therefore, a test dose with bupivacaine (Marcaine,
Sensorcaine) or lidocaine (Xylocaine) is not a good predictor of
response to full celiac plexus block; both drugs diffuse less
extensively than the alcohol block. The test will tend to provide
more relief than you will ever get with the alcohol.
Dr. Portenoy: Some people would say that the only advantage to
testing is for a negative result. If it does not work, then you do
not go ahead and do the neurolysis, but even if it does work, it is
unpredictable. It does not predict whether the full block will work.
Dr. Staats: In our institution, we used to always do a
diagnostic test first. But then we realized that, in pancreatic
cancer patients, we almost invariably proceeded to a neurolytic
block. The protocol that I use now is to place the two needles in the
right places. I inject some local anesthetic and sit there with the
patient until I see a change in the level of his or her pain. If
there is any kind of pain relief, I proceed immediately to the
neurolysis with 100% alcohol. If there is no pain relief from the
local anesthetic, I proceed to a full diagnostic block because a
negative result may mean that the needles are in the wrong place, not
where I think they are. If the patient cannot wiggle his toes after I
put in a little bit of local anesthetic, there is still opportunity
to abort the procedure if the local anesthetic has spread to the
somatic nerve roots. That is how I make use of the negative result,
and I watch for the possibility of side effects.
Dr. Portenoy: Are you saying that if you get a negative
response, you still go ahead with the alcohol?
Dr. Staats: No, if I get a negative response, I inject local
anesthetic and do a full diagnostic block, and pull the needles out.
But having said that, I need to point out that I have not had a
negative result with a pancreatic cancer patient in the last 2 years.
That is my protocol.
Dr. Krames: There are multiple ways to do celiac plexus blocks
and neurolysis. One way is the way you talked about, under
fluoroscopic guidance. This is the transaortic approach, going
through the aorta and injecting dye and seeing a crescent of dye
under fluoroscopy. But you also want to see spread of that dye. You
want to see spread in the right area on the fluoroscopy. Some people
do celiac plexus blocks under CT guidance. There are also
single-needle techniques. There are two-needle techniques. Some
people do splanchnic nerve blocks rather than celiac plexus block,
which accomplishes the same thing; basically this is a retrocrural block.
I used to do celiac plexus blocks alone, and now I do celiac plexus
blocks and retrocrural blocks (splanchnic) at the same time. On the
left side, my needle goes transaortic, and I inject the alcohol
transaortically. On the right side, I am at the anterolateral border
of the vertebral body, and I do a splanchnic nerve block. As I pull
my transaortic needle back into the retrocrural space, I inject 5 mL
of alcohol there. My success in combining these two techniques is
about 25% to 30% greater than when I perform a celiac plexus block alone.