Anda di halaman 1dari 4

In traarterial Infusion Chemotherapy for Hepa tic Carcinoma

Using a Totally Implantable Infusion Pump


HENRY BUCHWALD, MD, PHD, THEODOR B. GRAGE, MD, PHD, PERICLES P. VASSILOPOULOS, MD,
THOMAS D. ROHDE, MS, RICHARD L. VARCO, MD, PHD, AND PERRY J. BLACKSHEAR, MD, DPHIL*

Intraarterial infusion chemotherapy has several theoretical advantages over conventional therapy for
the treatment of unresectable malignancies. However, the catheter problems and patient restriction to
the hospital associated with its use have resulted in infrequent application and a notable lack of progress
in this field of oncology. This paper describes the use of a totally implantable, percutaneously refillable
infusion pump in 5 patients with primary or metastatic carcinoma of the liver. The infusion cannulae
were placed into the hepatic arteries under direct vision at laparotomy, and the pumps were placed in
subcutaneous pockets. Four patients received infusions of 5-fluorodeoxyuridine at rates of 0.2-0.5 mg/
kg/day for periods of three to 29 weeks; the pump in the fifth patient was defective and was removed. The
implanted pumps were well tolerated in these subjects, who received chemotherapy as outpatients; the
only adverse effects noted were related to FUDR toxicity. This implantable infusion pump appears to be a
practical means of delivering long-term intraarterial infusion chemotherapy to outpatients.
Cancer 45866-869, 1980.

I NTRAARTERIAL INFUSION of chemotherapeutic


agents has been used with varying success in the
treatment of many unresectable soiid tumors. Several
the delivery catheter, i.e., clotting, breaking, kinking,
leaking, dislodgement, arterial thrombosis, and air
embolism; fever and septicemia; hemorrhage; catheter
theoretical advantages have been cited in favor of this embolism; and o t h e r ~ . ~ * In
~ , addition,
’* the technique is
method of drug administration: 1) increased regional inconvenient for patients and hospital staff, neces-
drug concentration; 2) decreased systemic toxicity; 3) sitating the use of portable infusion pumps and per-
continuous exposure of tumor cells to drug; and 4) en- cutaneous infusion catheters, with their inherent
hanced antitumor activity of some drugs when ad- restrictions on patients’ activities.
ministered via the arterial route. Despite two decades We describe an attempt to minimize the problems
of use, however, the pharmacology of intraarterial associated with intraarterial infusion chemotherapy by
infusion chemotherapy has not been rigorously char- means of a totally implantable infusion pump. The de-
acterized.gFurthermore, this method of chemotherapy sign, animal trials, and our initial experience with this
has not been credited with “curing” of any form of device as a delivery vehicle for heparin for patients
cancer to date.8 Yet it is still used, though infrequently, with refractory thromboembolic disease have been
resulting in our current rudimentary understanding of described in detail.1+3-9+11 Laboratory trials using this
the pharmacokinetics involved. device as a means of infusing 5-fluorodeoxyuridine
One reason why this technique has not been ap- (FUDR) into the hepatic artery of normal dogs proved
plied more widely is the high rate of complications successful enough to warrant a small clinical trial.
accompanying its use, including: problems related to Five patients with inoperable primary or secondary
hepatic carcinoma were included in this initial study
From the Department of Surgery, University of Minnesota,
Minneapolis, Minnesota. and are the subjects of this report.
Supported by grants from the National Institutes of Health, CA-
20564, and the Metal Bellows Corporation, Sharon, Massachusetts. Materials and Methods
* Present address of Dr. Blackshear: Medical Services, Massachu-
setts General Hospital, Boston, Massachusetts 02 114. Infusion Pumps
The authors thank Philip D. Schneider, MD, Fay J. Goldenberg,
PA, Francine Leppanen, RN. and Patricia H. Bordewich, MS, for The infusion pumps used in this study were our
their technical assistance in the conduct of this project. standard clinical model,” described in detail in the
Address for reprints: Henry Buchwald, MD, Department of
Surgery, Box 290,420 Delaware Street S . E . ,University of Minnesota
references previously above.’-3 Briefly, the pump is
Hospitals, Minneapolis, Minnesota 55455.
Accepted for publication March 16, 1979. * Infusaid‘R’,Metal Bellows Corp.. Sharon, Massachusetts.

0008-543X/80/0301/0866 $0.70 0 American Cancer Society

866
No. 5 INFUSION
CHEMOTHERAPY Buchwald et al. 867

a titanium disc about the size and shape of a small ice


hockey puck (Figs. 1,2). It is divided into two chambers
separated by a welded titanium bellows, the inner con-
taining the infusate, the outer containing a volatile
fluorocarbon power source. Pump design utilizes the
physical principle that a vapor in equilibrium with its
liquid phase exerts a constant pressure at a given
temperature, regardless of volume. Vapor pressure
from the evaporating fluorocarbon pushes on the bel-
lows, which in turn propels the infusate through a capil-
lary pressure drop mechanism, and then into a silicone
rubber delivery catheter. Flow of a fluid through a
capillary flow restrictor is governed by the Poisseville
equation: Q = .rrD4AP/128pL;where Q = flow in ml/
sec; D = tubing inner diameter in cm; p = viscosity in
poise; AP = pressure in dyne/cm2; and L = length of
tubing in cm. The flow rate of each pump is set during FIG. 1. Implantable infusion pump. Dimensions: 2.82 cm in height
manufacturing by cutting the capillary to an appropriate X 8.43 cm in diameter Empty weight is 184 g.
length. Thereafter, the viscosity of the infusate can be
used to regulate flow. The pump is refilled by per- delivery cannulae was assessed by fluorescein dye in-
cutaneous injection through a self-sealing silicone rub- jection and direct inspection under a Woods lamp. The
ber septum, an operation which also recharges the cannulae were previously led through the peritoneum
power source through pressure-induced condensation and the rectus muscle, and the infusion pumps were
of the fluorocarbon vapor. implanted into a subcutaneous pocket in the abdominal
wall created by blunt dissection. The fascia and sub-
Operative Technique cutaneous tissue and skin were closed in layers with-
out drains.
The pumps’ delivery cannulae were implanted under The pumps were allowed to infuse sterile water
direct vision at laparotomy into the hepatic artery via with 3000-5000 U sodium hepariniml for the first
the gastroduodenal or gastroepiploic artery (Fig. 3), as several postoperative days; they were then filled with
described by Watkins et d.’’iCorrect positioning of the FUDR at 12-26 mgiml and heparin at 3000-5000 U/

7 9
\

PIG.2. Cross section of implantable infusion pump: 1 ) flow Restrictor: 2) silicone rubber coating: 3) bellows: 4) charging fluid chamber; 5 )
infusate chamber; 6 ) charging fluid fill tube (permanently sealed): 7) filter assembly; 8) inlet septum: 9) needle stop.
868 March I 1980
CANCER VOl. 45

returned for scheduled outpatient visits, where assess-


ments of tumor response, pump function, and possible
drug toxicity were made. During the outpatient visits,
the pumps were emptied and refilled with FUDR or
sterile water by percutaneous injection through the sub-
cutaneous septum (Fig. 4).

Patient Follow- U p

Tumor response was followed by assessment of liver


size on physical examination and liver scan, patient
weight, and serum bilirubin, alkaline phosphatase,
;i SCOT, SGPT, and prothrombin time. Drug toxicity
was assessed by historical and objective evidence of
stomatitis, anorexia, nausea and vomiting, and bone
Gastroduodenala. marrow suppression. Finally, pump function was meas-
Pump ured by calculating the amount of drug infused during
the preceding refill period from the residual pump
infusate volume.

I ('Cannula
I Results
The duration of intraarterial infusion chemotherapy
Ftc;. 3. Pump and cannulae placement for use in hepatic arterial for each patient is listed in Table 1. Total infusion
chemotherapy.
time ranged from five to 33 weeks with Patient 1 receiv-
ing 29 weeks of continuous FUDR infusion. All pa-
ml, calculated to deliver 0.20-0.50 mglkg body weight/ tients are dead at the time of this report. Five episodes
day into the hepatic artery at 1 mllday. After wound that were felt to represent FUDR toxicity were re-
healing was complete, the patients were allowed to corded, involving pancytopenia or nausea, vomiting,
return home, and continue their regular activities. They anorexia, and stomatitis. These symptoms cleared
rapidly when the FUDR in the pumps was replaced with
sterile water, and usually were preventcd when a lower
dose of drug was used subsequently. N o complications
relating specifically to the chronically implanted pumps
or delivery cannulae were noted, although the lack of
angiographic and autopsy data precludes conclusions
about hepatic artery patency or cannula thrombosis.
The continued flow through the delivery cannulae at all
times argues against hepatic artery occlusion. The fifth
pump, in Patient 5 , was defective and was removed in
less than 24 hours. Three of the four functional pumps
needed refills only about every 30 days; the fourth was
emptied after 16 days because FUDR toxicity de-
veloped in Patient 3 , and treatment was not resumed.

Discussion
The present results indicate that continuous, long-
term intraarterial infusion chemotherapy is practicable
in outpatients through the use of a totally implantable
- - infusion pump. The device was well tolerated by these
patients. They continued their normal daily activities
FIG.4. I n vivn refilling technique. A. Pump is being emptied by and returned for outpatient visits at approximately
puncturing the septum with a special needle and allowing pump's
pressure to force the remaining infusate into an empty syringe bar- monthly intervals, unless they sought attention for
rel. B. Fresh infusate is then injected into the pump's reservoir. another reason. We emphasize that their activities
No. 5 INFUSION CHEMOTHERAPY BuchwaEd et al. 869

TABLE1 . Characteristics of Intraarterial Infusions in Patients with Hepatic Carcinomas

Average time Episodes of


Weeks of Weeks of between refills FUDR dose Beneficial FUDR Other
Patient Indication infusion FUDR (days) (mg/kg/day) response toxicity complications

1 MHC 33 29 27 0.2-0.35 Yes 1 0


2 PHC 13 13 33 0.4 Yes 0 0
3 MHC 5 3 16 0.35 No 1 0
4 MHC 24 16 31 0.25-0.5 No 2 0

PHC = primary hepatic carcinoma: MHC = metastatic hepatic carcinoma.

were not limited in any way by the device, in con- A modification of the pump is being investigated to
trast to the percutaneous cannulae and carrying harnesses make pump operation even more versatile. A side port
necessary for currently available intra-arterial infu- with a second self-sealing septum added by the manu-
sion methods. facterer* to some experimental models was tested in
The purpose of this study was to determine whether our animal laboratory. These tests demonstrated that
the tested implantable infusion pump might be a useful bolus injections and angiography could be performed
means of providing intraarterial infusion chemotherapy through the delivery cannula while the pump was
in selected cancer patients. We did not intend to study implanted. This modification should facilitate intra-
the effect of FUDR infusion on hepatic malignancies. arterial infusion chemotherapy in future clinical studies.
The early deaths of 2 patients, and the progression of
metastatic disease in the other 2, reflect the current dis- REFERENCES
mal state of liver cancer chemotherapy and that of non- 1. Blackshear PJ, Dorman FD, Blackshear PL Jr, Varco RL,
Buchwald H. A permanently implantable self-recycling low flow
resectable solid tumors in general. There are scattered constant rate multipurpose infusion pump of simple design. Surg
reports in the medical literature, however, suggesting Forum 1970: 21:136.
that long-term remission or even cure might be achieved 2. Blackshear PJ, Dorman FD, Blackshear PL Jr, Varco RL,
Buchwald H. The design and initial testing of an implantable infu-
by appropriate drug combinations and schedules of ad- sion pump. Surg Gynecol Obstet 1972; 143:51-56.
ministration. For example, Oberfield and co-workerss 3. Blackshear PJ, Rohde TD, Varco RL, Buchwald H. One year of
reported remissions of seven and ten years in 2 patients continuous heparinization in the dog using a totally implantable infu-
sion pump. Surg Gynecol Obstet 1975; 141:176-186.
with metastatic colon cancer after long courses of 5- 4. Cady B. Hepatic arterial patency and complications after
fluorouracil infusion. Furthermore, Harada et al. re- catheterization for infusion chemotherapy. Ann Surg 1973; 178:
cently described a case in which combination chemo- 156- 161.
5. Eckman WW. Patlak CS, Fenstermacher JD. A critical evalua-
therapy by both intraarterial and intravenous routes tion of the principles governing the advantages of intra-arterial infu-
resulted in apparently complete necrotization of a hepa- sions. J Pharm Biopharm 1974: 2:257.
toma, although the patient died of disseminated intra- 6. Goldman ML, Bilbao MK. Rosch J , Dotter CT. Complications
of indwelling chemotherapy catheters. Cancer 1975; 36: 1983- 1990.
vascular coagulation three months after chemotherapy. 7. Harada T, Makisaka Y, Nishimura H , Okuda K . Complete
These case reports suggest that we eventually may ar- necrotization of hepatocellular carcinoma by chemotherapy and
rive at successful chemotherapeutic regimens for solid subsequent intravascular coagulation. A case report. Cancer 1978;
42:67-73.
tumors. To make this possible, properly controlled 8. Oberfield RA. Current status of regional arterial infusion
studies of drug combinations and administration therapy. Med Clin North Am 1975; 59:411-423.
schedules are needed. 9. Rohde TD, Blackshear PJ, Varco RL, Buchwald H. Chronic
heparin anticoagulation in dogs by continuous infusion with a totally
The main advantage of this device for proposed implantable pump. Trans Am SOC Artif Intern Organs 197.5: 21:
studies of this kind is that it is a versatile mode of ac- 510-514.
cess to veins, arteries, or body cavities. As exempli- 10. Rohde TD, Blackshear PJ, Varco RL, Buchwald H. One year
of heparin anticoagulation in an ambulatory subject using a totally
fied in this study, the dose of drug delivered can be al- implantable infusion pump. Minn Med 1977: 60:719-22.
tered easily by means of a percutaneous injection, and 11. Rohde TD, Blackshear PJ, Varco, RL, Buchwald H. Pro-
drug infusion can be stopped and replaced by placebo tracted parenteral drug infusion in ambulatory subjects using an im-
plantable infusion pump. Trans Am SOCArtif Intern Organs 1977:
or a second drug using the same procedure. Combina- 23:13-16.
tion chemotherapy is practicable through this pump be- 12. Stehlin JS Jr, Goepfert H , Jesse RH Jr, Smith JP.
cause most of the chemotherapeutic agents used for Regional chemotherapy by continuous intraarterial infusion. In:
Cole WH. ed: Chemotherapy of cancer. Philadelphia: Lea and
intraarterial infusion chemotherapy are available in Febiger, I970:248- 276.
high enough concentrations to make very low infusion 13. Watkins E Jr. Khazei A M , Nahra KS. Surgical basis for ar-
rates practical. Limiting factors requiring testing are terial infusion chemotherapy of disseminated carcinoma of the liver.
Surg Gynecol Obstet 1970: 130: 581-605.
the drugs’ reactivity with the titanium pump com-
ponents, their viscosity, and their heat lability. * Metal Bellows Corporation, Sharon, Mass. 02067.

Anda mungkin juga menyukai