Attitudes Towards Detection and Management of Hepatic Metastases of Colorectal Origin: A Second Look

In the present study we undertook an international postal survey to assess the current attitudes towards the detection and management of hepatic metastases in colorectal cancer patients, who have been operated on with curative intent. Results of this survey were compared to results of an earlier survey, held in 1985. Both surveys indicate that there is no consensus on the follow-up of patients at risk of hepatic metastases. Especially the interpretation of unexplained rises in carcinoembryonic antigen (CEA) levels leads to much controversy. Only 37% of the hospitals performing liver surgery were willing to perform second-look laparotomies based on CEA only. Also there is no agreement on the maximum number of liver metastases that will justify partial liver resection for cure. Clearly, there is a need for prospective randomized trials on which a more rational policy regarding hepatic metastases in colorectal cancer patients can be based.


stases in col
rectal cancer patients can be based.

INTRODUCTION

Colorectal cancer is the second leading cause of death from cancer1.Although 75% of colorectal cancer pa- tients will have a primary surgical resection for cure, nearly half of all patients with colorectal cancer still die of metastatic tumor2.Resection of liver metastases and adjuvant chemotherapy are currently the most promis- ing develop ents in the treatment of colorectal cancer patients.

Because there exists worldwide controversy among clinicians regarding optimal diagnostic and thera- * Correspondence to: J. Kievit, M.D., Ph.D., Medical Decision Making Unit, Building 43, Room 1.002, University Hospital Leiden, P.O.Box.9600, 2300 RC Leiden, The Netherlands, Telephone: NL-71276780, Fax: NL-71225520.peutic policies, we held in 1985 a postal survey of 290 hospitals in 13 Western countries3.The response rate was 98%, which is exceptionally high for a postal survey.Nearly all hospitals in this survey used a sys- tematic follow-up aimed at detecting liver metastases in colorectal cancer patients operated for cure and performe liver surgery when necessary.Approximately 75% of the hospitals treated patients with unresect- able liver metastases with chemotherapy, and 33% of the hospitals used adjuvant chemotherapy following curative resection of liver metastases.

In 1992 we held, encouraged by our earlier results, a secon

postal survey.We devel
ped an extended version of the 1985 questionnaire, which we sent to 1955 clinicians in 77 countries.Results of this survey may give a more detailed and complete view on current attitudes toward hepatic metastases.


MATERIALS AND METHODS

In 1992 we sent a postal questionnaire to all 1955 members of the 'World Association of Hepato-Pancreato-Biliary Surgery'.Addresses were obtained from the 1990 HPB membership registry.

The questionnaire consists of four parts:

(1) Detection of liver metastases by systematic fol- low-up: Which diagnostic tests were used with which frequency?Were diagnostic second-look laparotomies performed?(2) Resection of liver metastases: What was the maximum number of metastases considered for radical resection, and were they confined to one or two lobes?How many patients were operated for potentially resectable liver metastases in 1991, and in how m ny patients were liver meta- stases actually resected?Which preoperative imaging techniques were used to assess the number ofliver metastases?Was the resectability of liver metastases intraoperatively assessed us- ing ultrasound imaging?

(3) Adjuvant chemotherapy following potentially curati e resection of colorectal cancer or partial liver resection

s were adminis
ered by which route? (4) Chemotherapy of unresectable liver metastases: How many patients were treated in 1991?By which route were drugs administered?

A sample of the questionnaire is included in the Appendix.


RESULTS


Response rate

The 1990 HPB membership registry listed 1955 names and addresses of clinicians involved in the field of hepatic, pancreatic and biliary surgery.Of these 1955 clinicians 903 were listed by their private address and 1052 by their hospital address.The 1052.clinicians listed by their hospital add ess were employed in 653 different hospitals.Thus, in total, there were 1556 questionnaires sent.Fifty-three of these questionnaires were returned to us unopened, indicating that at least 3% of the ad- dresses we used were incorrect.

A total of 171 questionnaires were answered and returned to us.The 1992 q

stionnaire, however, was sent
o 77 countries worldwide, in contrast to the 1985 questionnaire, which was sent to 13 Western coun- tries* only.The response rate of the recent survey calculated for these 13 Western countries is 21%.


Detection of liver metastases

In the first section of the 1992 questionnaire we asked clinician to describe the follow-up they used for colorectal cancer patients operated for cure.Systematic routine follow-up was used in 150 (88%) of the hospitals.Of these hospitals 134 (89%) used follow-up schemes of five years or longer.

Scheduled history taking and physical examinations were used in 144 (96%) hospitals, routine liver bio- chemistry in 119 (79%), seru

tumor markers in 141 (94%), li
er ultrasonography in 111 (74%), and com- puterized tomography (CT) in 65 (43%).The average number of individual diagnostic tests per year are presented in Table 1.

Serum carcinoembryonic fintigen (CEA) assays were used by all 141 hospitals which used tumor markers.

Other serum tumor markers also used by these hospi- tal were CA 19-9 (carbohydrate antigen associated with gastrointestinal cancer) in 32 hospitals, AFP (al- pha-fetoprotein) in 9 hospitals, TPA (tissue polypept- ide antigen)in 4 hospitals, CA 50 (carbodydrate antigen associated with carcinomas) in 3 hospitals, CA 125 (carbohydrate antigen associated with ovarian cancer) in 2 hospitals, CA 195 (carbohydrate antigen associated with gastrointestinal cancer) in 1 hospital, ferritin in 1 hospital, and NCC-ST 439 (monoclonal antibody against gastric cancer cells) also in 1 hospital.

Second-look laparotomies solely based on signifi- cant and consecutive rises in CEA-levels were performed in 52 (37%) of the 141 hospitals.

Compared to the 1958 survey, fewer hospitals per- formed clinical evaluations and liver biochemistry tests Liver CT * Austria, Belgium, Canada, France, Germany, Israel, Italy, Spain, Sweden, Switzerland, the Netherlands, the United Kingdom (UK), and the United States of America (USA).on a routine basis in 1991 (Table 2).Consequently, serial CEA testing was the most frequently used screen- ing method in colorectal cancer follow-up.


Resection of liver metastases

In the second section of the 1992 questionnaire we asked about the surgical aspects of liver metastases.In 144 hospital , in which numerical data on colorectal cancer patients were available, liver metastases were detected in 3250 patients in 1991.Potentially resectable liver metastases were never considered for resection in 5 (3%) of the hospitals and were resected only when solitary in28(16%) of the hospitals.In

1 (47%) of the hospitals multiple metastases w
re resected when they were confined to one lobe, with a maximum of 3 meta- stases, while 57 (33%) of the hospitals also resected multiple metastases not confined to one lobe, with a maximum of 4 metastases.In 36 (21%) hospitals resectability was assessed without a reference to the number of metastases.In 1991 a total of 1238 (38%)

patients had surgery for liver metastases, and in 911 (74%) of these patients liver metastases were actually resected.

When liver metastases were suspected, resectability was assessed using preoperative imaging techniques.Ultrasonography of the abdomen was used by 135 (82%) hospitals, CT (computerized tomography) by 112 (68%) hospitals, CT combined with angiography in 103 (62

hospitals,
MRI (magnetic resonance imaging) by 31 (19%) hospitals, and scintigraphy by 9 (5%) hospitals.The resectability of liver metastases was assessed by intraoperative ultrasonography of the liver in 85 (58%) of the hospitals which performed liver resections in 1991.No increase in intraoperatively di.agnosed unresectable recurrences was reported by the hospitals which used this technique.

In comparison to the 1985 survey more hospitals resected multiple bilobar metastases in 1991 (Table 3).

The number of liver resections per hospital has also  increased (Table 4).Hospitals in the UK and the Netherlands are less in favor of liver surgery, in con- trast to hospitals in Germany, France, the USA and Canada, which employ a more aggressive approach towards liver metastases.


Adjuvant chemotherapy

T e third part of the questionnaire surveys adjuvant chemotherapy after resection of a primary tumor or a recurrence.In 79 hospitals (46%) adjuvant chemo- therapy was administered after a primary tumor resec- tion for cure.Fluorouracil (5-FU) was administered in 66 (84%) of the hospitals using adjuvant chemother- apy.Therefore it was the most frequently employed chemotherapeutic drug given alone or combined with other drugs.Combinations of 5-FU and levamisole were used in 34 (43%) of the hospitals, and combina- tions of 5-FU and folinic acid (leucovorin, citrovorum factor) in 24 (30%).Other chemotherapeutic drugs used were doxorubicin (Adriamycin, ADR), mitomycin (MMC), cisplatin (CDDP), methotrexate (MTX), fluorodeoxyuridine (FUdR), methylcyclohexylchloro- ethylnitrosurea (methyl-CCNU), and interferon.In 62 (78%) of the hospitals systemic administration was used to deliver adjuvant drugs to the patient.Hepatic arterial infusion was employed by 4 (5%) and portal venous infusion by 5 (6%) of the hospitals.Adjuvant chemotherapy after resection for cure of liver metastases was used in 49 (34%) of the hospitals which performed partial liver resections in 1991.Ad- ministrations of 5-FU with or without other chemo- therapeuticdrugs were used in 38 (78%) of the hospi- tals.The most frequently used combinations of 5-FU were with levamisole in 11 (22%) and folinic acid in 14 (29%) of the hospitals.Other chemotherapeutic drugs used were doxorubicin, farmorubicin, m tomycin, cis- platin, methotrexate, fluorodeoxyuridine, interferon, and interleukin 2. Systematic administrations were used by 30 (61%), hepatic arterial infusion by 14 (29%) and portal venous infusion by 2 (4%) of the hospitals.

Compared to the 1985 survey the overall use of adjuvant chemotherapy has not changed in Western countries (Table 5).Adjuvant therapy was used less in Germany and the Netherlands, in contrast to France and the UK where its use has increased.


Chemotherapy of unresectable liver metastases

In the fourth and final part of the questionnaire we asked about the use of chemotherapy in patients with unresectable liver metastases.In the 144 hospitals men- tioned earlier 2318 (71%) patients had unresectable liver metastases, of which 1172 (51%) were treated with chemotherapy.

The number of hospitals using chemotherapy has increased slightly compared to the 1985 survey (Table 6).This is mainly caused by an increase in the use of chemotherapy in France and the UK.The average number of patients treated with chemotherapy per hospital, however, has dropped from 18 to 12 per year (Table 7).In addition, a shift from local adminisatra- tion towards systemic administration of chemotherapy was observed.


DISCUSSION

The present study was undertaken as a sequel to a postal survey on attitudes toward detection and management of hepatic metastases held in 1985a.In contrast to the 1985 survey, the recent survey had a low response rate of only 21%.Several factors may have contributed to this low response rate: (1) The extent of the questionnaire.The recent questionnaire not only contained more questions, but also asked more de- tailed questions a d assumed that data on the number of patients with partial liver resections and chemother- apy were available.(2) Satiety.Two other postal sur- veys closely related to ours were held in 1992' '5.(3) Incorrect addresses.In the recent survey, using the 1990 HPB membership registry, at least 3% of the addresses were incorrect, compared to zero percent in 1985 survey.

Because only 171 questionnaires were returned to us, we decided against the use of inferential statistical techniques.Therefore, only observed frequencies are reported in this paper.Also, we feel it is inappropriate to compare the 1985 and the recent surveys using inferential statistics, because of the possible differences between the two groups of respondents.The results of the recent survey, however, could be of value in the light of a worldwide discussion on colorectal cancer follow-up, with special reference to the detection and resection of hepatic metastases.

Our survey indicated that systematic routine follow- up of colorectal cancer patients operated on with curative intent was used.by88% of the hospitals.The majority of these hospitals used follow-up schemes of five years or more.Scheduled history taking and physi- cal examinations, routine liver biochemistry and car- cinoembryonic antigen (CEA) monitoring were the most popular and frequently used tests.Although there is a National Institutes of Health (NIH) consensus statement on the use of CEA [6], there still remains controversy about the diagnostic and therapeutic policy following an unexplained rise in CEA.Only 37% of the hospitals using CEA were willing to perform sec- ond-look laparotomies based on significant and con- secutive rises of CEA-levels only, as advocated by Minton et al. 7 Alternatively, many hospitals may have responded to such a rise by increasing the intensity of follow-up.

The percentage of hospitals resecting liver meta- stases in the 1992 survey has increased only slightly compared to the 1985 survey.The number ofresections per hospital, however, has doubled from 6 to 12 resec- tions per year.Also, the extent of liver operations has increased.In the 1985 survey mainly solitary and multiple unilobar liver metastases were resected, com- pared to multiple unilobar and bilobar metastases in the 1992 survey.The majority of the hospitals (79%) regarded patients with more than 3 or 4 hepatic meta- stases in the absence of extrahepatic disease as un- resectable.This observation may give cause for concern.Since the introduct on ofmore sensitive imaging techniques in recent years, smaller hepatic meta- stases can be identified preoperatively.Approximately 70% of the hOspitals used extremely sensitive imaging techniques like CT-angiography,