Br. J. Surg. 1990, Vol. 77, November, 1238-1 240

J. A. D. J. J. H. C. S.

Goldberg, Kerr*, N. Wilmottj-, McKillopS and McArdle

University Departments of Surgery and $Medicine, The Royal Infirmary, Glasgow. *Cancer Research Campaign Department of Clinical Oncology, University of Glasgow. and +Department of Pharmacy, The University of Strathclyde, Glasgow, UK Correspondence to: Miss J. A. G o l d b e r g , University D e p a r t m e n t of S u r g e r y , T h e Royal Infirmary, Glasgow G31 2ER, U K

Regional chemotherapy for colorectal liver metastases: a phase II evaluation of targeted hepatic arterial 5-fluorouracil for colorectal liver metastases The results of systemic chemotherapy in patients with liver metastases f r o m colorectal cancer remain dismal. Regional chemotherapy has been advocated as a method of improving the delivery of cytotoxic drugs to tumour, while minimizing systemic toxicity. The use of vasoactive agents to redistribute arterial bloodflow towards tumour, and of biodegradable microspheres to slow tumour blood flow, have also been suggested as methods of further improving tumour exposure to drug. W e present 21 patients who received intrahepatic arterial chemotherapy f o r colorectal liver metastases. Combined treatment (angiotensin II, albumin microspheres and 5-fluorouracil) was administered 4-6 weekly, and bolus 5-fluorouracil was given in the intervening weeks. Toxicity was minimal. Responses were seen in seven patients. Fewer than harfof the deaths were f r o m liver metastases: a quarter of the patients died from non-cancer-related causes. Survival was prolonged in the treated group compared with historical controls. These results suggest that this regimen has activity in patients with colorectal liver metastases. Keywords: Liver metastases, 5-fluorouracil, microspheres

There are more than 6000 deaths from colorectal cancer per annum among patients aged less than 70 years in the UK. More than 60 per cent of patients with recurrent colorectal cancer die with liver metastases; in one in five patients, disease is restricted to the liver’. Survival after diagnosis of liver metastases is limited; in o u r institution the mean survival of patients with multiple bilateral metastases is 3 months2. In this situation, systemic chemotherapy using conventional cytotoxic agents has been shown to be relatively ineffective. In the absence of a more specific chemotherapeutic agent for colorectal cancer, attention has focused on attempts to improve tumour exposure to the anticancer drugs that are currently available. Three levels of tumour targeting have been described: ( 1 ) selective drug delivery to the tumour-bearing organ (i.e. regional chemotherapy); ( 2 ) drug delivery biased to tumour rather than to normal liver; and (3) enhancement of uptake of cytotoxic drug by the cell”. A combined approach to improve both first and second order targeting is now feasible. We describe our experience of regional chemotherapy in patients with colorectal liver metastases. All patients received angiotensin I1 and albumin microspheres in addition to intrahepatic arterial 5-fluorouracil. Both of these agents have been shown to improve second level targeting. Angiotensin I1 is a vasoactive agent known to redistribute arterial blood flow towards tumours4. Biodegradable microspheres have previously been shown to slow tumour blood flow and to enhance tumour drug uptake’. Therefore, by combining the two, it is possible that microsphere delivery to the tumour would be increased, thereby selectively reducing tumour blood flow and increasing the probability of drug uptake.

U K ) was dissolved in 800 p1 water containing 2 mg sodium dodecyl sulphate. This constituted the disperse phase of a water-in-oil emulsion. Water droplets were stabilized by the addition of 2 4 0 ~ 115 per cent glutaraldehyde solution, and the resulting spherical particles were separated and washed in petroleum ether and isopropanol. The microspheres were reconstituted in phosphate-buffered saline containing 0.5 per cent Tween 80 (Sigma). Following mesh filtration and differential centrifugation, microspheres of between 20 and 40 pm in diameter (50 per cent weight average) were obtained. Particles of this diameter were selected because they are known to embolize within the first capillary bed encountered and to biodegrade over 48 h (Reference

71.

Albumin microspheres were made ‘in house’ under sterile conditions by a technique involving stabilization by glutaraldehyde of a water-in-oil emulsion containing proteinb. Three hundred and eighty milligrams of human serum albumin (Sigma Chemical Co. Ltd., Poolc.

A pilot study looking at the toxicity of incremental doses of microspheres had shown that a hepatic arterial bolus injection of 300 mg albumin microspheres (containing approximately 6 x 10’ particles) produced mild right upper quadrant pain and nausea. Doses greater than this caused more severe hepatic pain; therefore 300 mg was used in subsequent treatments. Twenty-one patients with colorectal liver metastases (mean(s.d.) age 59 (8)years, range 41-71 years) have been included in the study. Two patients had between 10 and 25 per cent hepatic replacement; nine had between 25 and 50 per cent hepatic tumour burden; and in ten patients the volume of disease was greater than 50 per cent as measured by albumin colloid scan. Liver metastases were found in 11 patients at the time of initial presentation. In the remaining patients, liver metastases were discovered during follow-up; in four, these metastases were discovered while the patient was still asymptomatic. All patients had surgical placement of a hepatic arterial perfusion catheter which consisted of a silicone catheter and subcutaneous resealable injection port. In patients with a standard arterial anatomy, the catheter was implanted into the gastroduodenal artery so that the tip lay at its orifice from the hepatic artery. Thus, infusions through the catheter were delivered to the liver via the hepatic arterial stream. In patients with a dual arterial supply, two catheters were inserted, a saphenous vein graft being used t o hold the catheter where no suitable hepatic arterial side-branch existed*. Chemotherapy was commenced as an outpatient procedure. 5-Fluorouracil was given in conjunction with the vasoactive agent angiotensin I1 and albumin microspheres. An infusion of angiotensin I1 was commenced via the catheter at a rate of 10 pg per minute; 100 s later, a bolus injection of microspheres, followed by 1 g 5-fluorouracil, was given via the catheter. Microspheres were available for

1238

0007-1323/90/1 I 1 2 3 8 4 3

Patients and methods

~

c 1990 Butterworth -Heinemann Ltd

Regional chemotherapy f o r colorectal liver metastases: J. A. Goldberg et al. Table 1 Number of patients suffering .from nausea, vomiting or epigastric pain during treatment or from raised liver enzymes or myelosuppression at 1 week after combined therapy (angiotensin I I , albumin microspheres and S$horouraciI) and regional 5-fluorouracil

Nausea/vomiting Mild Moderate Acute epigastric pain Mild Increased liver enzymes Myelosuppression*

Combined therapy (n=21)

5-Fluorouracil (n=21)

8 (38) 4 (19)

0 0

9 (43) 0 2 (10)

0 0 2 (10)

Values in parentheses are percentages; * myelosuppression is defined here as white cell count < 3000/mm3and platelet count < 100000/mm3

administration at monthly intervals. Between combination treatment, 1 g 5-fluorouracil was given weekly as a bolus injection through the catheter. Side-effects were recorded; full blood-count and liver function tests were performed weekly. The response to therapy was assessed every 8-12 weeks by ultrasonography. Standard response criteria were used to measure response (complete response where no evidence of disease remained; partial response where a 50 per cent reduction in disease occurred by two diameters; minimal response when between 25 and 50 per cent reduction in disease was seen by two dimensions; static disease where there was no change; progression when an increase in tumour size was recorded).

Results The mean duration of therapy was 7.4 months. Toxic side-effects were few. Table 2 summarizes the toxicity associated with either combined regional therapy or regional 5-fluorouracil alone. Acute pain, nausea and vomiting occurred only during the combined treatment. One patient was of particular interest. Before operation angiography demonstrated normal arterial anatomy and a catheter was inserted in the standard fashion. Unfortunately, hepatic arterial perfusion scintigraphy after operation showed that only part of the liver was being perfused by this route; in effect, the right lobe of the liver received regional chemotherapy, whereas the left lobe was exposed only to drug that had passed through the right lobe and had entered the systemic circulation (i.e. systemic therapy). Ultrasound assessment demonstrated disease progression in the region of liver receiving systemic treatment, whereas the tumour in the area of liver receiving regional therapy responded. The patient remains alive after 35 months oftreatment. This case is described in detail elsewhereg. Of the remaining 20 patients, one had a partial response, five hadminimal response and three had stable disease. Disease progression occurred in 11 patients during therapy. Of the 18 deaths, eight patients died from progressive disease, four from extrahepatic tumour (periportal lymph nodes, small intestine, lung and bone) and one from local recurrence with ureteric involvement. One patient died from respiratory complications following a secondary laparotomy for intestinal obstruction due to adhesions, and four died from intercurrent disease, including myocardial infarction, gastrointestinal haemorrhage (peptic ulceration), pancreatitis and an acute febrile illness while abroad. The median survival from the time of diagnosis was 9 months; 20 per cent of patients survived for more than 18 months (Figure I). Three patients remain alive and well at 8,25 and 35 months, respectively.

Discussion In theory, regional chemotherapy should enhance tumour drug delivery while reducing systemic exposure and hence toxic

Br._J.Surg., Vol. 77, No. 11, November 1990

side-effects. However, bolus intra-arterial 5-fluorouracil alone has been shown to have no impact on systemic drug levels". Coadministration of biodegradable starch microspheres with a half-life of approximatly 20min has been shown to improve the regional advantage of arterially administered cytotoxic agents' The microspheres are thought to act by slowing arterial blood flow within the liver, thereby increasing the time available for drug uptake by tissue within the target organ. It has been suggested that improved drug uptake with arterial blockade might be improved by using embolic particles with a longer half-life. Thus the albumin particles used in this study might be expected to confer an additional advantage. Previous results in both patients and animal models have suggested that tumour blood flow may be temporarily increased by an infusion of vasoactive agents"-15. The mechanism is thought to involve arteriolar constriction in the normal tissue surrounding tumour, causing a passive increase in tumour perfusion'6. Sasaki and colleagues4 showed that blood flow to intrahepatic tumour could be increased by up to four times basal levels during an angiotensin I1 infusion. The peak effect was found to be after 100 s of a regional infusion of 10 pg/min. We have recently confirmed these findings"*'*. Although the median survival of patients with widespread liver metastases in the present study is approximately three times that of historical controls2, there are a number of factors that might bias the results. Clearly, there was an element of selection for regional therapy. Furthermore, the few asymptomatic patients included in the study might have caused some distortion of the results because of 'lead time' gained between the 'screening' investigation and the onset of symptoms. Nevertheless, there is probably a modest improvement in survival. In the patient described, in whom regional chemotherapy was compared with systemic treatment, with the patient acting as his own control, tumour regression occurred only in the segment of liver receiving regional chemotherapy. Furthermore, it would appear that the pattern of disease in this group ofpatients has been altered; fewer than half of the patients died from liver metastases. Further improvements in the therapeutic index may be possible. Cytotoxic drug-loaded particle formulations which act as slow-release preparations have been synthesized for a variety of chemotherapeutic agents, including mitomycin C, Adriamycin@(Farmitalia Carlo. Erba Ltd., St. Albans, UK), 5-fluorouracil and cisplatin. We have previously shown that Adriamycin-loaded microspheres are trapped in rabbit kidney and liver following regional admini~tration'~, leading to a marked reduction in systemic exposure and maintenance of high drug levels within the target organ'. The drug-release characteristics can be modified to optimize the pharmacological profile of cytotoxic particles. Escalation of drug dosage is clearly possible because of the reduced systemic toxicity associated

'.

I

6

12

18

24

30

36

Time ( m o n t h s )

Figure I Life-table derived ,from 21 patients with colorectal liver metastases who received regional 5-fluorouracil albumin microspheres and angiotensin I I . Median survival is estimated ut 9 months

1239

R e g i o n a l c h e m o t h e r a p y f o r c o l o r e c t a l liver m e t a s t a s e s : J. A. G o l d b e r g et al.

with microencapsulated chemotherapy. Eventually either rising systemic drug levels or hepatotoxicity are likely to limit this approach, although any improvement in tumour drug exposure may be clinically useful in view of the steep dose-response curve associated with cytotoxic drugs. This effect could be further amplified by tumour targeting with vasoactive agents. This study shows that targeted microsphere-based regional chemotherapy is associated with prolongation of survival compared with historical controls. The differential response to regional and systemic chemotherapy in the patient described and the alteration in the pattern of mortality suggest that this benefit, although modest, is probably real. Additional techniques to improve the therapeutic advantage of regional chemotherapy, including drug-loaded particles and vasoactive agents, require further evaluation to define their clinical role in the treatment of colorectal liver metastases.

Acknowledgements The authors are grateful to the Cancer Research Campaign for their financial support and to Ciba Laboratories for the provision of angiotensin 11.

References 1.

2. 3. 4. 5. 6.

Welch J P , Donaldson GA. The clinical correlation of an autopsy study of recurrent colorectal cancer. Ann Surg 1979; 189: 496-502. Wood CB, Gillis CR, Blumgart L H . A retrospective study of the natural history of patients with liver metastases from colorectal cancer. Clin Oticol 1976; 2: 285-8. Widder KJ. Senyei AE, Ranney D F . Magnetically responsive microspheres and other carriers for the biophysical targeting of antitumour agents. A t h Phtrrtnucol Clieniotker 1979: 16:213-71. Sasaki Y . Imaoka S, Hasegawa Y rr ul. Changes in distribution of hepatic blood flow induced by intra-arterial infusion of angiotensin I I in human hepaticcancer. Cm1t.e).1985:55: 31 1-16. Dakhil S. Ensminger WD, Cho K el a / . Improved regional selectivity of hepatic arterial BCNU with degradable microspheres. C ~ t 7 t ~1982: r 50: 631-5. Willmott N. Cummings J. Stuart JFB, Florence AT. Adriamycinloaded albumin microspheres: preparation. in 11iro distribution and release in the rat. Biophtrrtn Drug Dispos 1985; 6 : 91-104.

1240

7.

8.

9. 10.

1 I.

12. 13. 14. 15.

16.

17. 18.

19.

Kerr DJ. Willmott N, Lewi H. McArdle CS. The pharmacokinetics and distribution of Adriamycin-loaded albumin microspheres following intra-arterial administration. Cancer 1988; 62: 878-82. Goldberg JA, Leiberman D P , McArdle CS. A useful manoeuvre for hepatic arterial catheterisation in patients with metastatic hepatic disease and abnormal vascular anatomy. Surg Gynecol O h t e t 1989; 169: 71-2. Goldberg JA. Kerr DJ, Stewart I, McArdle CS. A comparison of regional and systemic chemotherapy for hepatic metastases. Eur J Surg Oncol 1990; 16: 464-7. Goldberg JA, Kerr DJ, Willmott N, McKillop JH, McArdle CS. Pharmacokinetics and pharmacodynamics of locoregional 5-fluorouracil (5FU) in advanced colorectal liver metastases. Br J Cancer 1988; 57: 186-9. Gyves JW, Ensminger WD, VanHarken D, Neiderhuber JE, Stetson P. Walker S. Improved regional selectivity of hepatic arterial mitomycin by starch microspheres. Clin Pharmacol Ther 1983; 34: 259-265. Abrams HL. Altered drug response of tumour vessels in man. Nature 1964; 4915: 167-70. Burton MA, Gray BN. Self G W et a / . Manipulation of experimental rat and rabbit tumour blood flow with angiotensin I I . Cancer Res 1985; 45: 539@3. Ackerman NB, Hechmer PA. Effects of pharmacological agents on the microcirculation of tumours implanted in the liver. Bib1 Anut 1977: 15: 301-3. Suzuki M, Hori K, Abe I et d.A new approach to cancer chemotherapy: selective enhancement of tumour blood-flow with angiotensin 11. J Nuti Cancer Ins/ 1981; 67: 663-9. Hafstrom L. Nobin A, Persson B, Sundqvist K. Effects of catecholamines on cardiovascular response and blood flow distribution to normal tissue and liver tumours in rats. Cancer Res 1980; 40: 481-5. Goldberg JA. Fenner J , Bradnam MS et al. Improved tumour targeting of biodegradable microspheres with angiotensin 11. Br J Surq 1988; 75: 1262. Goldberg JA, Kerr DJ, Willmott N, McArdle CS, Murray T, Hilditch T. Increased uptake of radio-labelled microspheres with angiotensin 11 in colorectal hepatic metastases. Eur J Surg Oncol 1988; 14: 715. McArdle CS, Lewi H, Hansel1 DJ et a/. Cytotoxic-loaded albumin microspheres: a novel approach to regional chemotherapy. Br J Surg 1988; 7 5 : 1 3 2 4 .

Paper accepted 28 May 1990

Br. J. S u r g . , Vol. 77, No. 11, N o v e m b e r 1990

Regional chemotherapy for colorectal liver metastases: a phase II evaluation of targeted hepatic arterial 5-fluorouracil for colorectal liver metastases.

The results of systemic chemotherapy in patients with liver metastases from colorectal cancer remain dismal. Regional chemotherapy has been advocated ...
346KB Sizes 0 Downloads 0 Views