[Peritoneal carcinomatosis of colorectal origin: results of cytoreductive surgery with peritonectomy and hyperthermic intraoperative chemotherapy]

Until recently peritoneal carcinomatosis (PC) arising from colorectal cancer (CRC) was considered to be a terminal disease manifestation. Despite palliative systemic chemotherapy (CHT) the majority of patients died within a few months. Nowadays cytoreductive surgery (CRS) of the peritoneal cavity in combination with hyperthermic intraperitoneal CHT and perioperative systemic CHT may offer a chance for long-term survival in selected groups of patients. In this study we report the results obtained with this treatment strategy in 30 consecutive patients. Data were assessed prospectively. After a median follow-up of 16.9 months the median survival time in all 30 patients reached 24.3 months. Favorable prognostic factors are a low extent of intraperitoneal metastases as characterized by a low peritoneal cancer index (median survival PCI ≤ 10: 33.2 months vs. PCI 11-19: 12.1 months) and a complete or nearly complete CRS (median survival CCR 0/1: 33.1 months vs. CCR2/3: 12.1 months). The 2-year overall survival was 89% for patients with a PCI ≤ 10 and 65% for those with surgical CCR 0/1 cytoreduction. As not every patient with CRC and PC may profit from this relatively aggressive therapy an interdisciplinary patient selection (tumor board) and treatment in experienced surgical oncology centers is recommended.

[Multimodal therapy for colon cancer: state of the art]

In UICC stage I a selected group of patients with T1 tumours and a low risk profile regarding simultaneous lymph node metastases can be treated by endoscopic resection alone, if the tumour is thereby completely removed. In UICC stage II an adjuvant chemotherapy (CT) should not be routinely performed. However, in high risk UICC stage II patients (T4 tumour, less than 12 examined lymph nodes, emergency surgery, intraoperative tumour perforation), an adjuvant CT with infusional 5-FU/FA should be recommended. The state of the art in UICC stage III is an adjuvant CT with FOLFOX. In this tumour stage no beneficial effect of CT involving irinotecan or monoclonal antibodies has been documented. Due to CT-induced side effects an infusional 5-FU/FA protocol or oral capecitabine should be given in patients older than 70 years. In stage UICC IV with resectable liver metastases, surgical resection of the primary tumour and the metastases should be implemented. Since no conclusive data are currently available regarding the beneficial effect of neoadjuvant, perioperative or adjuvant CT in this setting, the therapeutic strategy should be individually discussed between surgeons and oncologists (tumour board). In cases of non-resectable liver metastases a neoadjuvant CT should be performed, preferentially with a FOLFOX protocol in combination with targeted therapies, i.e., the monoclonal antibody cetuximab, aimed at tumour regression with radical metastasectomy as the secondary intent (R0). Patients with UICC stage II colon cancer and microsatellite instability (MSI) apparently experience a better prognosis but do not profit from an adjuvant CT with 5-FU/FA alone. If a CT is under consideration for these patients, the MSI status should be determined on tumour tissue. In cases of a positive result a combination CT, i.e., with FOLFOX, should be given. The relevance of the MSI status in other tumour stages is as yet unknown. Before targeted therapies, i.e., cetuximab or panitumumab, are initiated, the KRAS status needs to be determined, since therapies with antibodies against the epithelial growth factor receptor (EGFR) are only effective in tumours bearing the KRAS wild-type.

Adjuvant chemoradiotherapy of advanced resectable rectal cancer: results of a randomised trial comparing modulation of 5-fluorouracil with folinic acid or with interferon-α

Background:

Standard adjuvant chemoradiotherapy of rectal cancer still consists of 5-fluorouracil (5-FU) only. Its cytotoxicity is enhanced by folinic acid (FA) and interferon-α (INFα). In this trial, the effects of FA and IFNα on adjuvant 5-FU chemoradiotherapy in locally advanced rectal cancer were investigated.

Methods:

Patients with R₀-resected rectal cancer (UICC stage II and III) were stratified and randomised to a 12-month adjuvant chemoradiotherapy with 5-FU, 5-FU+FA, or 5-FU+IFNα. All patients received levamisol and local irradiation with 50.4 Gy.

Results:

Median follow-up was 4.9 years (n=796). Toxicities (WHO III+IV) were observed in 32, 28, and 58% of patients receiving 5-FU, 5-FU+FA, and 5-FU+IFNα, respectively. No differences between the groups were observed for local or distant recurrence. Five-year overall survival (OS) rates were 60.3% (95% confidence interval (CI): 54.3–65.8), 60.4% (54.4–65.8), and 59.9% (53.0–66.1) for 5-FU, 5-FU+FA, and 5-FU+IFNα, respectively. A subgroup analysis in stage II (pT3/4pN0) disease (n=271) revealed that the addition of FA tended to reduce the 5-year local recurrence (LR) rate by 55% and increase recurrence-free survival and OS rates by 12 and 13%, respectively, relative to 5-FU alone.

Conclusions:

Interferon-α cannot be recommended for adjuvant chemoradiotherapy of rectal cancer. In UICC stage II disease, the addition of FA tended to lower LR and increased survival. The addition of FA to 5-FU may be an effective option for adjuvant chemoradiotherapy of UICC stage II rectal cancer.

[Diagnosis and treatment of colorectal liver metastases - workflow]

In this review, standards of diagnosis and treatment of colorectal liver metastases are described on the basis of a workshop discussion. Algorithms of care for patients with synchronous / metachronous colorectal liver metastases or locoregional recurrent tumour are presented. Surgical resection is the procedure of choice in the curative treatment of liver metastases. The decision about the resection of liver metastases should consider the following parameters: 1. General operability of the patient (comorbidity); 2. Achievability of an R 0 situation: i. if necessary, in combination with ablative methods, ii. if necessary, neoadjuvant chemotherapy, iii. the ability to eradicate extrahepatic tumour manifestations; 3. Sufficient volume of the liver remaining after resection ("future liver remnant = FLR): i. if necessary, in combination with portal vein embolisation or two-stage hepatectomy; 4. The feasibility to preserve two contiguous hepatic segments with adequate vascular inflow and outflow as well as biliary drainage; 5. Tumour biological aspects ("prognostic variables"); 6. Experience of the surgeon and centre! Extrahepatic disease does not contraindicate hepatectomy for colorectal liver metastases provided a complete resection of both intra- and extrahepatic disease is feasible. Even in bilobar colorectal metastases and 5 or more tumours in the liver, a complete tumour resection has been described. The type of resection (hepatic wedge resection or anatomic resection) does not influence the recurrence rate. Preoperative volumetry is indicated when major hepatic resection is planned. The FLR should be 25 % in patients with normal liver, 40 % in patients who have received intensive chemotherapy or in cases of fatty liver, liver fibrosis or diabetes, and 50-60 % in patients with cirrhosis. In patients with initially unresectable colorectal liver metastases, preoperative chemotherapy enables complete resection in 15-30 % of the cases, whereas the value of neoadjuvant chemotherapy in patients with resectable liver metastases has not been sufficiently supported. In situ ablative procedures (radiofrequency ablation = RFA and laser-induced interstitial thermotherapy = LITT) are local therapy options in selected patients who are not candidates for resection (central recurrent liver metastases, bilobar multiple metastases and high-risk resection or restricted patient operability). Patients with tumours larger than 3 cm have a high local recurrence rate after percutaneous RFA and are not optimal candidates for this procedure. The physician's experience influences the results significantly, both after hepatectomy and after in situ ablation. Therefore, patients with colorectal liver metastases should be treated in centres with experience in liver surgery.

Riboswitches in eubacteria sense the second messenger cyclic di-GMP

Cyclic di-guanosine monophosphate (di-GMP) is a circular RNA dinucleotide that functions as a second messenger in diverse species of bacteria to trigger wide-ranging physiological changes, including cell differentiation, conversion between motile and biofilm lifestyles, and virulence gene expression. However, the mechanisms by which cyclic di-GMP regulates gene expression have remained a mystery. We found that cyclic di-GMP in many bacterial species is sensed by a riboswitch class in messenger RNA that controls the expression of genes involved in numerous fundamental cellular processes. A variety of cyclic di-GMP regulons are revealed, including some riboswitches associated with virulence gene expression, pilus formation, and flagellum biosynthesis. In addition, sequences matching the consensus for cyclic di-GMP riboswitches are present in the genome of a bacteriophage.

Prognostic factors influencing the survival of patients with colon cancer receiving adjuvant 5-FU treatment

AIM

Adjuvant chemotherapy is recommended for stage III colon cancer. The aim of this study was to identify important prognostic factors among patients with colon cancer receiving adjuvant 5-FU-based treatment.

METHODS

Data sets of 855 colon cancer patients treated between 1992 and 1999 within a multicenter adjuvant trial comparing 5-FU modulation with folinic acid or interfereron-alpha were examined. Backward elimination in a proportional hazards model was used to identify prognostically relevant clinical and pathological factors.

RESULTS

Tumor recurrence (p<0.001), duration of adjuvant treatment (p<0.001), tumor substage (p=0.004), age (p=0.005), grading (p=0.016), treatment-related toxicity (p=0.021), and treatment (p=0.031) were identified in descending order of importance as prognostic factors for overall survival.

CONCLUSIONS

Adjuvant 5-FU-based treatment should be performed for at least 6months with a stepwise adjustment of 5-FU doses until toxicity >WHO II. Substages should be reported separately and used for stratification in future trials due to their broad variation in outcome. In the future, this may result in adjuvant treatment of stage III colon cancer adjusted for the risk of substages.

Colon cancer: survival after curative surgery

Several new aspects have evolved during the past years concerning factors that influence survival in surgically and medically treated colon cancer patients that are relevant to the treating team for the treatment strategy and patient's choice. The 5-year-survival rates dependent on UICC stages/substages (I: 68%-100%, II: 58%-90%, III: 33%-76%, IV: <5%-9%) show remarkable variations between published reports, surgical hospital units, individual surgeons, and continents (USA vs Europe). Those variations may be due to surgical techniques, training status, hospital and individual case volume, and, also, referral patterns and statistical evaluation methods. Survival times and cure rates are significantly improved by adjuvant chemotherapy in UICC III and in substages of UICC II (e.g. UICC II B) by 5%-12%, when compared with surgical controls. In three recently published trials standard adjuvant chemotherapy was further improved by increased survival rates, e.g. from 59% to 71% in stage III and IIB patients. Molecular and genetic factors, such as thymidylate synthase (TS), microsatellite instability (MSI) or loss of chromosome 18q/"DCC" might have an independent impact on prognosis in the spontaneous course, and TS could help to better select patients for adjuvant chemotherapy.

?Bodypacker? als chirurgischer Notfall

Body packing is a well recognized method of drug trafficking by smuggling drug containers in the gastrointestinal tract. Medical professionals might get involved with body packers after presentation by law enforcement or in case of medical emergencies such as drug overdose or mechanical intestinal obstruction due to the containers within the gastrointestinal tract. Besides the medical aspects in treating these patients, physicians must be aware of all the different legal specifics in dealing with body packers. In case of medical emergencies, drug traffickers have the legal status of regular patients with respect to professional medical discretion. The question remains of what physicians should do with the drugs after surgical removal? Even though the body packer remains the legal owner of the drugs, physicians may not return the drugs, since that constitutes the criminal offence of dealing in narcotics. Returning the drugs to law enforcement authorities is also prohibited because of professional medical discretion. The only way out of this predicament is for physicians to destroy the drugs under the observation of witnesses.

Intraperitoneal chemotherapy with mitoxantrone in malignant ascites

A retrospective analysis of intraperitoneal mitoxantrone instillation therapy for malignant ascites in advanced breast and gynecologic pelvic cancers was performed to confirm the efficacy and safety of this therapy. Several smaller phase II trials had suggested good palliative effects. In 143 patients (37 breast cancer and 106 gynecologic cancers), 257 instillations were registered. Response in breast cancer was induced in 49% and in 63% with gynecologic cancer. Severe or life-threatening clinical or laboratory side effects related to intraperitoneal mitoxantrone occurred in 2.7% (clinical) or 1.9% (laboratory) of the 257 instillations. Induction of adverse side effect was dose dependent. Intraperitoneal chemotherapy with mitoxantrone for treatment of malignant ascites in breast cancer and gynecologic malignancy is effective and well tolerated. For this treatment 30 mg mitoxantrone in > or = 1000 mL carrier solution (e.g., saline) is recommended. A minimal concentration of at least 10 micrograms/mL should be achieved.

[Interferon-alpha in adjuvant treatment of colorectal carcinoma]

Based on preclinical and clinical studies, in this German three-arm adjuvant multicenter trial the FOGT (Forschungsgruppe Onkologie Gastrointestinale Tumoren) studied whether one of the 5-FU modulations with either folinic acid(FA) or Interferon alpha-2a (IFNa) is superior to the recommended standard of adjuvant treatment in R0-resected colon cancer, 5-fluorouracil (5-FU) plus levamisole (LEV) for 12 months, in terms of overall survival rates.

PATIENTS/METHODS

From 7/92 to 10/99 813 patients with resected colon cancer stage II (only T4N0M0, 63 pts.) and stage III (750 pts.) were randomized into three treatment groups and stratified according to N-stage and participating centers (64 hospitals). The patients received a postoperative loading course with 5-FU [450 mg/m2 d1-5 (arms A and C)] or 5-FU [450 mg/m2 plus folinic acid (Rescuvolin, medac, Hamburg, Germany), 200 mg/m2 d1-5 (arm B)]. After completion of the first chemotherapy cycle LEV was administered orally at 150 mg/d d1-3, every 2 weeks. After a 4-week chemotherapy-free interval the treatment was continued weekly for up to 52 weeks. The standard group, arm A (279 pts.) was treated with 5-FU i.v. (450 mg/m2 at d 1, q 1 w) plus LEV. 5-FU plus LEV was modulated in arm B (283 pts.) with FA (200 mg/m2 d1, q 1 w), and in arm C (251 pts.) with IFNa at 6 million units 3x/week, q 1 w. Chemotherapy doses were adjusted to toxicity if toxic events > WHO 2 occurred. The patients were followed-up to determine relapse rates and--patterns and survival. Survival rates were calculated according to Kaplan-Meier, and treatment costs and immune effects were analysed.

RESULTS

Toxic event(s) > WHO2, mainly leukopenia, diarrhea and nausea, occurred in 113 pts. (14%), in arms A (8%), B (13%) and C (32%). Discontinuance rates were 28% (all), 29% (A), 21% (B), 34% (C), but 80% of patients received > or = 6 months treatment. Overall relapse rates were 27% (all), 30% (A), 24% (B) and 28% (C). Tumors relapsed either locally (2% each) or distant (A: 22%, B: 20%, C: 22%). 4-year overall survival rates in arms A, B and C were 66%, 77%, 66%, respectively. The 4-year survival rate in arm B was significantly superior to arms A and C (p < 0.02, log-rank). There were no signs of a superior immune function in either treatment arm (skin test, proliferation, cytotoxicity, flow cytometry). Treatment costs per patient were 2,500 [symbol: see text](arm A), 3,500 [symbol: see text](arm B) or 10,850 [symbol: see text](arm C), respectively.

CONCLUSION

Adjuvant therapy with 5-FU plus FA plus LEV for 12 months is superior to the recommended standard (5-FU + LEV, 12 m). IFNa-modulation of 5-FU (plus LEV) adds toxicity and high treatment costs without therapeutic benefit.

Regional celiac artery infusion in the adjuvant treatment of pancreatic cancer

We performed adjuvant celiac artery infusion in pancreatic cancer, to find out whether this treatment prolongs survival and changes the biology of the disease after resection, especially by reducing liver metastasis.

PATIENTS AND METHODS

Thirty-one patients received cyclic celiac artery infusions (CAI) after resection of their pancreatic cancer (27 ductal, 4 cystadenocarcinoma). The treatment consisted of 6 cycles (1 cycle = 5 days treatment) intra-arterial infusion using Seldingers technique with mitoxantrone A (Novantron) 10 mg/m2 d1, 5-fluorouracil + folinic acid 600 mg/m2 + 170 mg/m2 d2-d4 and cis-platinum 60 mg/m2 d5. Four to 5-week intervals between each cycle of chemotherapy were scheduled. The patients were monitored for toxicity, development of disease progression and survival.

RESULTS

The median survival time was 21 months. During an observation period of 19 months, 70% of the patients developed disease progression. In 50% of cases the progression was local, in 40% intraperitoneal while in 15% liver metastases developed. The median survival time of the CAI (celiac artery infusion)-treated patient group compared favorably to the median survival of 9.3 months in a matched historical control group, being significantly longer (p < 0.0003).

CONCLUSION

Adjuvant celiac artery infusion seemed to prolong median survival and the occurrence of liver metastases appeared to be delayed or reduced.

Conversion of locally inoperable primary rectal cancer with multiple liver metastases to an option for cure after local down-staging and hepatic arterial infusion chemotherapy

AIMS

This case report presents a patient with local unresectable primary rectal cancer and multiple synchronous liver metastases.

METHODS

The treatment consisted of primary tumor resection after down-staging by local irradiation followed by hepatic arterial infusion chemotherapy.

RESULTS

The patient is now without any signs of tumor growth 44 months after beginning of treatment.

Surgery, radio- and chemotherapy for multimodal treatment of rectal cancer

The possibilities and results of multimodal treatment in rectal cancer were reviewed with respect to the results of surgical treatment only. Based on the results of 4 studies, reducing local relapse rates and increasing long term survival rates significantly, postoperative radiochemotherapy (RCT) + chemotherapy (CT) should remain the recommended standard for R0 resected UICC II and III rectal cancers. The addition of RT to adjuvant CT reduces local relapses without significant impact on survival (NSABP R-02). Vice versa, the addition of CT to RT or an improved CT in the RCT-concept prolongs survival. Preoperative neoadjuvant radiotherapy (RT) reduced local relapse rates in 9 studies, and extended survival in one study that evaluated all eligible patients. Preoperative RT reduced local relapse rates in addition to total mesorectal excision (TME) but did not extend survival. The preoperative RCT + CT downstages resectable and nonresectable tumors and induces a higher sphincter preservation rate. Phase III data justifying its routine use in all UICC II + III stages are not yet available. This treatment may be routinely applied in nonresectable primary tumors or local relapses. Preoperative RCT (or RT) may evolve as standard, if the patient selection is improved and postoperative morbidity and long term toxicity reduced. Intraoperative RT could be added to this concept or be used together with preoperative/postoperative RT at the same indications. Postoperative adjuvant RT reduced local relapses significantly in a single trial, and no impact on survival time is reported. Since postoperative RT is inferior to preoperative RT, this treatment cannot be recommended, if RT is chosen as a single treatment modality in adjunction to surgery. The results of local tumor excisions may be improved with pre- or postoperative RCT + CT. In the future, multimodal treatment of rectal cancer might be more effective, if individualized according to prognostic factors.

Regional chemotherapy of nonresectable colorectal liver metastases with mitoxantrone, 5-fluorouracil, folinic acid, and mitomycin C may prolong survival

BACKGROUND

Regional chemotherapy of isolated, nonresectable colorectal liver metastases (CRLMs) by hepatic artery infusion (HAI) has the advantages of high response rates and the possibility of downstaging and resection of CRLMs. 5-Fluorodeoxyuridine (5-FUDR) has been the drug studied in most Phase II and III trials. The meta-analysis of the Phase III trials comparing HAI with systemic or supportive therapy confirmed an advantage for response and even survival for HAI. Hepatic artery infusion with 5-FUDR, however, is hepatotoxic, inducing sclerosing cholangitis (SC). The authors have introduced 5-fluorouracil (5-FU) with folinic acid for HAI and found equal effectivity but no SC when compared with HAI with 5-FUDR. Now, they report a new combination chemotherapy protocol based on HAI with 5-FU with FA and on in vitro Phase II studies suggesting mitoxantrone and mitomycin C as active drugs for HAI in CRLM. PATIENTS AND METHODS Between February 1993 and August 2000, 63 patients with CRLM were treated with HAI using mitoxantrone, 5-FU with FA, and mitomycin C (MFFM) via port catheters with a protocol planing up to 11 cycles of treatment. Toxicity and response were analyzed according to World Health Organization (WHO) criteria, and survival was analyzed according to Kaplan-Meier. All patients were treated with more than two HAI cycles.

RESULTS

The objective response rate (complete remission and partial remission) was 54% and primary intrahepatic progression (progressive disease) occurred in 4.8%, whereas in 41.3% of the patients the intrahepatic disease was evaluated as no change. Median survival times from the first diagnosis of CRLM or start of HAI were 25.7 months and 23.7 months, respectively, and 7 patients lived longer than 40 months. Grade 3 toxicity according to WHO occurred in 34.9%, and Grade 4 occurred in 3.2%. No toxic death or SC occurred.

CONCLUSIONS

Our new HAI protocol with MFFM seems to be superior to HAI with 5-FUDR, 5-FU with FA, and systemic chemotherapy with 5-FU and FA at acceptable toxicity. Currently, HAI with MFFM is compared with systemic chemotherapy using 5-FU and FA intravenously in a randomized Phase III trial.

Regional chemotherapy of nonresectable colorectal liver metastases with mitoxantrone, 5-fluorouracil, folinic acid, and mitomycin C may prolong survival

BACKGROUND

Regional chemotherapy of isolated, nonresectable colorectal liver metastases (CRLMs) by hepatic artery infusion (HAI) has the advantages of high response rates and the possibility of downstaging and resection of CRLMs. 5-Fluorodeoxyuridine (5-FUDR) has been the drug studied in most Phase II and III trials. The meta-analysis of the Phase III trials comparing HAI with systemic or supportive therapy confirmed an advantage for response and even survival for HAI. Hepatic artery infusion with 5-FUDR, however, is hepatotoxic, inducing sclerosing cholangitis (SC). The authors have introduced 5-fluorouracil (5-FU) with folinic acid for HAI and found equal effectivity but no SC when compared with HAI with 5-FUDR. Now, they report a new combination chemotherapy protocol based on HAI with 5-FU with FA and on in vitro Phase II studies suggesting mitoxantrone and mitomycin C as active drugs for HAI in CRLM. PATIENTS AND METHODS Between February 1993 and August 2000, 63 patients with CRLM were treated with HAI using mitoxantrone, 5-FU with FA, and mitomycin C (MFFM) via port catheters with a protocol planing up to 11 cycles of treatment. Toxicity and response were analyzed according to World Health Organization (WHO) criteria, and survival was analyzed according to Kaplan-Meier. All patients were treated with more than two HAI cycles.

RESULTS

The objective response rate (complete remission and partial remission) was 54% and primary intrahepatic progression (progressive disease) occurred in 4.8%, whereas in 41.3% of the patients the intrahepatic disease was evaluated as no change. Median survival times from the first diagnosis of CRLM or start of HAI were 25.7 months and 23.7 months, respectively, and 7 patients lived longer than 40 months. Grade 3 toxicity according to WHO occurred in 34.9%, and Grade 4 occurred in 3.2%. No toxic death or SC occurred.

CONCLUSIONS

Our new HAI protocol with MFFM seems to be superior to HAI with 5-FUDR, 5-FU with FA, and systemic chemotherapy with 5-FU and FA at acceptable toxicity. Currently, HAI with MFFM is compared with systemic chemotherapy using 5-FU and FA intravenously in a randomized Phase III trial.

Toxicity and effects of adjuvant therapy in colon cancer: results of the German prospective, controlled randomized multicenter trial FOGT-1

In this adjuvant three-arm multicenter trial, we studied whether modulating the standard 5-fluorouracil (5-FU) treatment with either folinic acid (FA) or interferon-alpha-2a (IFN-alpha) was superior to the recommended standard of adjuvant treatment in R0 resected colon cancer, 5-FU plus levamisole (LEV) for 12 months, in terms of toxicity and outcome. From July 1992 to October 1999, a total of 813 patients with resected colon cancer in stage II (T4N0M0; n = 63) or stage III (TxN1-3M0; n = 750) were randomized into three treatment groups and stratified according to N stage and participating centers (64 hospitals). The patients received a postoperative loading dose of 5-FU (450 mg/m2 on days 1 to 5 [arms A and C]) or 5-FU (450 mg/m2) plus FA (Rescuvolin, Medac, Hamburg, Germany, 200 mg/m2 on days 1 to 5 [arm B]). After completion of the first chemotherapy cycle, LEV was administered orally at a dosage of 150 mg per day on days 1 to 3, once every 2 weeks. After a 4-week chemotherapy-free interval, the treatment was continued weekly for 52 weeks. Treatment in one arm A ("standard") (n = 279) consisted of 5-FU intravenously (450 mg/m2 on day 1, once a week) plus LEV. 5-FU plus LEV was modulated in arm B (n = 283) with FA (200 mg/m2 on day 1, once a week) and in arm C (n = 251) with IFN-alpha at 6 million units three times a week repeated weekly. Treatment dosages were adjusted if toxic events above WHO grade 2 occurred. Patients were closely followed to determine recurrence and survival; the latter was calculated according to Kaplan-Meier analysis. Toxic events above WHO grade 2, mainly leukopenia, diarrhea, and nausea, occurred in 113 (14%) of 649 patients who had completed treatment in arms A (8.4%), B (13.5%), and C (31.7%). Discontinuance rates were as follows: 28% for all patients, 29% in arm A, 21% in arm B, and 34% in arm C. Overall relapse rates were 27% for all patients, 30% in arm A, 24% in arm B, and 28% in arm C. Relapses were local (8%), distant (78%), or combined (12%). Four-year overall survival rates in arms A, B, and C were 66.1%, 77.5%, and 66.2%, respectively. The 4-year survival rate in arm B was significantly higher compared to arm A (P <0.02, log-rank test) with arm A being equal to arm C. Adjuvant therapy with 5-FU plus FA plus LEV for 12 months is superior to the recommended standard (5-FU + LEV for 12 months). IFN-alpha modulation of 5-FU (plus LEV) adds to the toxicity with no therapeutic benefit.

Follow-up in colorectal cancer: cost-effectiveness analysis of established and novel concepts

BACKGROUND

Follow-up programs in colorectal cancer have been under question recently due to their low efficacy. Some meta-analyses found a survival benefit and four randomized studies suggested the opposite. Therefore, this paper analyzes existing and proposed follow-up programs for costs and efficacy.

METHODS

One thousand and fifty-four colorectal cancer patients intensively followed-up in Ulm (endoscopy, chest radiography, abdominal ultrasound, computed tomography (CT) pelvis) were prospectively evaluated for recurrence, salvage surgery rate, survival and costs. Costs were compared among existing and proposed international follow-up programs retrieved by a MedLine search.

RESULTS

A total of 350/1,054 colorectal cancer patients (33.2%) relapsed asymptomatically, as detected by follow-up. 56/350 (local 47%, distant 53%) recurrences were resectable (16%), 21/350 patients survived (6%). The efficacy of follow-up was 21/1,054 patients (2%). Costs were, per patient, 2,220 euros (colon) or 4,851 euros (rectum). Costs in five randomized studies varied between 616 euros for minimal and 5,049 euros for intensive follow-up. Four proposed follow-up concepts include risk-adaption, which claim to reduce costs by 50-60% and range between 889 (colorectal) and 5,910 (high-risk, rectum) euros. The recommended German follow-up guideline costs 610 euros (low-risk, colorectal), 1,120 euros (high-risk, colon) or 2,252 euros (high-risk, rectum).

CONCLUSION

Follow-up concepts for colorectal cancer should aim at the identification of curable patients with recurrence. Evidence-based concepts, including life quality tests, remain to be defined, but currently abdominal ultrasound, endoscopy and carcino-embryonic antigen (CEA) determination at 6-month intervals for 2 years and annual intervals for 3 years seem to identify this patient sub-group.

Oxaliplatin exerts potent in vitro cytotoxicity in colorectal and pancreatic cancer cell lines and liver metastases

BACKGROUND

Oxaliplatin displays potent activity in advanced colorectal cancer. The aim of this study was to estimate the potential efficacy of oxaliplatin for hepatic arterial infusion (HAI) chemotherapy.

MATERIALS AND METHODS

The anti-proliferative effects of oxaliplatin in human HT29 and NMG64/84 colon and COLO-357 MIA PaCa-2 and PMH2/89 pancreatic cancer cell lines and in fresh liver metastases from patients with colorectal and pancreatic cancer were investigated using the human tumor colony forming assay.

RESULTS

Oxaliplatin significantly inhibited the colony formation in all cell lines in a concentration- and time-dependent manner. All liver tumors displayed a significant concentration-dependent inhibition of colony formation after exposure to oxaliplatin for 2 hours. The IC50 of oxaliplatin of 9 of the 10 tumors was < 10 micrograms/ml.

CONCLUSION

Oxaliplatin is suitable for HAI therapy phase II studies. Due to the low IC50 values of most tumors we suggest that patients with colorectal or pancreatic liver metastases may benefit from HAI with oxaliplatin.

Pre-clinical evaluation of the activity of gemcitabine as a basis for regional chemotherapy of pancreatic and colorectal cancer

AIMS

To estimate the potential activity of gemcitabine for hepatic arterial infusion (HAI) chemotherapy in pancreatic and colorectal cancer.

METHODS

The anti-proliferative effects of gemcitabine were determined in MIA PaCa-2 and PMH2/89 pancreatic and HT29 and NMG64/84 colon cancer cell lines and in fresh tumours from patients with liver metastases of colon, rectal and pancreatic cancer in vitro using the human tumour colony forming assay.

RESULTS

Gemcitabine showed concentration and time-dependent cytotoxic effects in all tested cell lines. The IC(50)of gemcitabine in MIA PaCa-2, PMH2/89, HT29 and NMG64/84 cells at 2 h exposure time were >100, 18, 100 and 2.5 microg/ml, respectively, at 4 h 15, 1.2, 45 and 0.5 microg/ml, respectively, and at 24 h 0.2, 0.1, 1.8 and 0.1 microg/ml, respectively. All tumours displayed concentration dependent inhibition of colony formation after exposure to gemcitabine for 2 h. The IC(50)values of gemcitabine in six of the 10 metastases were </=100 microg/ml.

CONCLUSIONS

Based on our results and theoretical considerations regarding hepatic arterial infusion therapy gemcitabine seems to be suitable for HAI therapy phase II studies. Moreover, patients with colorectal or pancreatic tumours that demonstrated in vivo sensitivity may benefit from regional chemotherapy with gemcitabine.

Thymidylate synthase quantitation and in vitro chemosensitivity testing predicts responses and survival of patients with isolated nonresectable liver tumors receiving hepatic arterial infusion chemotherapy

BACKGROUND

Patients with isolated, nonresectable liver tumors may receive regional hepatic arterial infusion (HAI) chemotherapy with response rates of about 50%. The objective of this study was to investigate the value of thymidylate synthase (TS) determination in combination with in vitro chemosensitivity testing to predict the responses and survival of patients receiving HAI.

METHODS

TS mRNA expression was quantitated using a reverse transcription-polymerase chain reaction technique with beta-actin as the internal standard. In vitro chemosensitivity testing was performed with tumor cell suspensions using the human tumor colony-forming assay (HTCA).

RESULTS

An analysis of the test combination in 24 consecutive patients revealed that 77% (10 of 13) of the sensitive and 9% (1 of 11) of the resistant patients had complete or partial clinical responses. Sensitive patients were 8.5-fold more likely to respond (P = 0.0036) and displayed with 32 months (range, 5-75 months) a longer median survival than resistant patients with 17 months (range, 3-28 months, P = 0.003). Analysis of the Kaplan-Meier curves revealed that sensitive patients had a higher overall survival probability, as determined by the log rank test (P = 0.044).

CONCLUSIONS

These results suggest that the clinical outcomes of patients receiving HAI therapy may be predictable with TS quantitation and HTCA. It is possible, therefore, that this combination may be used in the future to select patients with liver tumors who will benefit from HAI before the start of regional chemotherapy.

Theoretical considerations and in vitro concentration response studies with two human colorectal carcinoma cell lines. The rational experimental base for clinical studies in regional chemotherapy

The theoretical advantage of regional vs. systemic chemotherapy was calculated, based on pharmacokinetic considerations. The relevance of exposure time and high local concentrations of chemotherapeutic drugs for regional chemotherapy was elucidated in time dependent concentration response curves with two human cell lines. The theoretical pharmacological advantage of regional vs. systemic chemotherapy was defined by the formula Rd = (AUCi. a. 1 + AUCi. a. 2)/(AUCi. v.) and in hepatic artery infusion is for adriamycin (ADM) 5.8-.6, cis-platinum (CDDP) 8, epirubicine (EPI) 6.3, 5-fluorouracil (5-FU) 22-58, mitomycin C (MMC) 4.6, mitoxantrone (NOV) 6.3. All drugs but 5-FUDR exerted concentration response behaviour in the cell line-experiments. In the cell lines cytotoxicity depended on exposure time so that concentration chi time products at (IC50), (c chi t (IC50)), were calculated to determine an optimal in vitro exposure time. Based on these results and clinical considerations, optimal clinical exposure times could be defined for regional chemotherapy. The results may be of high relevance for e.g. hepatic artery infusion at the lower and chemoembolization or intraperitoneal instillation at the higher test concentration, respectively.

Intraarterial adjuvant chemotherapy after pancreaticoduodenectomy for pancreatic cancer: significant reduction in occurrence of liver metastasis

The clinical benefit of adjuvant chemotherapy in pancreatic cancer patients is still questionable. Phase II studies using radiochemotherapy based on 5-fluorouracil (5-FU) provided evidence of an increase in median survival times. Because palliative chemotherapy by celiac artery infusion (CAI) led to an increase in survival in pancreatic cancer, we treated 24 patients with adjuvant CAI following resection of the head of the pancreas for pancreatic cancer (21 patients with Union Internationale contre le Cancer (UICC) stage III, 2 with UICC stage II, 1 with UICC stage I). Catheters were placed angiographically into the celiac artery and remained there for 5 consecutive days. One cycle of chemotherapy consisted of mitoxantrone, 5-FU, folinic acid, and cisplatinum. This treatment was repeated five times at monthly intervals. CAI was well tolerated, and World Health Organization (WHO) grade III toxicities were observed in 8%; WHO grade IV was seen in none of the treatment cycles. Furthermore, we observed pain reduction in nearly all patients under CAI. Median survival times in patients who received CAI were 23 months for all patients, whereas in patients who did not receive adjuvant treatment the median survival was 10.5 months. With Kaplan-Meier regression analysis of the patients who were curatively resected (R0 resection) and received CAI, the overall 4-year survival was 54%, whereas in patients without CAI the 4-year survival was 9.5%. The occurrence of liver metastases in the CAI group went down to 17%. These results demonstrate that CAI is well tolerated, reduces the risk of liver metastasis, and increases the survival time of pancreatic cancer patients.

Downstaging by regional chemotherapy of non-resectable isolated colorectal liver metastases

AIMS

To improve the course of isolated non-resectable colorectal liver metastases (CRLM) by hepatic arterial infusion treatment. Patients with CRLM have a worse prognosis than those whose liver metastases are resectable. Systemic (i.v.) chemotherapy for CRLM/colorectal metastases with 5-fluorouracil+folinic acid (5-FU+FA) i.v. may result in median survival times of 6.4-14.3 months. Hepatic artery infusion (HAI) with 5-fluorodeoxyuridine (5-FUDR) has been demonstrated in a meta-analysis of randomized trials to be superior to i.v. treatment/palliative care (median survival 15 vs. 10 months). The benefit of HAI with 5-FUDR, although recommended as treatment for CRLM, is severely compromised by the 5-FUDR induced hepatotoxicity, leading eventually to sclerosing cholangitis (SC)/liver cirrhosis. We have developed a stepwise protocol for HAI in CRLM, which is superior to HAI with 5-FUDR and to systemic chemotherapy.

METHODS

Between 1982 and 1997, 168 CRLM patients were treated within the following protocols. In protocol A, 48 CRLM patients received HAI with 5-FUDR. In protocol B, 46 patients received 5-FUDR i.a. (HAI)+i.v. In protocol C 5-FU+FA were delivered via HAI in 24 patients with CRLM. In protocol D, based on in vitro phase II studies and the results of protocol C, mitoxantrone and mitomycin C were added to 5-FU+FA (MFFM). Fifty (50) CRLM patients received HAI with HFFM.

RESULTS

The response rates, median survival time, systemic toxicity and SC rate were: 42%, 20.8 months, 0-19% and 38% for protocol A; 46%, 20.8 months, 0-20% and 41% for protocol B; 45%, 19.8 months, 4-25% and 0% for protocol C; and 66%, 27.4 months, 2-26% and 0% for protocol D. The surgically placed ports for HAI in protocols C and D functioned in 90%, 82% and 76% of patients, 6, 9, and 11 months after beginning HAI. Quality of life in protocol D was high. Nine patients from protocols C and D with either partial (PR, seven patients) or complete (CR, two patients) remissions received a secondary liver resection without hospital mortality, and seven of nine patients are alive 2-58 months after liver resection. The other two died 11 and 22 months after resection.

CONCLUSIONS

Optimal treatment of CRLM was found to be protocol D: HAI with MFFM. The results of this protocol, including high remission rate, long median survival time, good port function, good quality of life and, interestingly, the possibility of downstaging and resecting primarily non-resectable metastases, seem to be superior to HAI with 5-FUDR or 5-FU+FA and to systemic chemotherapy with 5-FU+FA. This hypothesis is currently being examined in a phase III study (HAI with MFFM vs. 5-FU+FA i.v.).

Immunotherapy in pancreatic cancer - current status and future

BACKGROUND

In pancreatic cancer, multimodal protocols, involving chemotherapy, radiation, or regional treatment, are initiated to improve oncological outcome. Since pancreatic adenocarcinoma has been shown to be susceptible to immune stimulation, several immunotherapy approaches have been investigated to define the role of immunotherapy in pancreatic cancer.

METHOD

A review of current and past data concerning experimental and clinical immunotherapy in pancreatic cancer is presented in the context of basic immunotherapeutic principles. Past pitfalls and future developments are analyzed and a synthesis of immune stimulation and immune suppression is deduced on the basis of published data.

RESULTS

Preclinical and initial clinical studies with monoclonal antibodies CO17-1A, BW494/32 and anti-epidermal growth factor receptor (EGFR) have been conducted, and various targets suitable for immunotherapy have been identified involving new molecular and gene technology. Targets on pancreatic cancer cells currently under investigation are mucins (MUC-1), glycoproteins (GA733), ras peptides and EGFRs. Side effects are minor and rarely auto-immune reactive. Another approach combines randomized regional with systemic chemoimmunotherapy (mitomycin C, 5-fluorouracil, folinic acid, carboplatin, epirubicin; interferon-gamma, interleukin-2) in nonresectable pancreatic cancer and obtains significant differences in median survival rates (14 months vs 4.5 months in controls) and quality of life.

CONCLUSION

Although single remarkable improvements in the immunological approach to treatment of pancreatic cancer have been made, immunotherapy in pancreatic cancer is still experimental. On the basis of reliable preclinical data, new immunotherapy protocols will have to be evaluated clinically. Careful monitoring of immune responses and side effects, and assessment of quality of life will ensure identification of effective immunotherapy protocols for human pancreatic cancer in the near future.

Regional chemotherapy of non-resectable liver metastases from colorectal cancer - literature and institutional review

BACKGROUND

Cure is possible by resecting colorectal isolated liver metastases. In non-resectable isolated colorectal liver metastases (CRLM), regional chemotherapy has been advocated to optimize the disease control in the liver in order to improve the results of the alternative, systemic chemotherapy. The drugs are delivered by means of hepatic artery infusion (HAI) via ports or pumps; pharmacological modifications of the hepatic arterial blood-flow-like HAI with starch microspheres or stop-flow and perfusion techniques were applied to improve HAI.

METHODS

We reviewed the literature and report our progress, up to May 1999, in analyzing the validity of HAI for CRLM therapy.

RESULTS

In the majority of phase-II and -III trials, the response rates to HAI were significantly higher than those from systemic chemotherapy, and local disease control could be achieved even when HAI was used second line to systemic chemotherapy. The meta-analysis of randomized trials comparing HAI with either systemic chemotherapy (five trials) or, optionally, either 5-fluorouracil (FU) or symptomatic treatment (two trials) showed a significant advantage of HAI in response (41% vs 14%, P<10(-10)) and median survival time (15 months vs 11 months, P<0.0009). The active anabolite of 5-FU, 5-fluordeoxyuridine (5-FUDR), the drug of choice for HAI in those trials, may cause severe hepatotoxicity. To avoid this toxicity, we developed a HAI protocol using mitoxantrone, 5-FU plus folinic acid (FA) and mitomycin C (MFFM). The response rates of HAI with 5-FU plus FA or MFFM were 45% and 66%, the interim median survival times 19.8 months and 27.4 months. 5-Year survivors were observed in all our protocols. Since no severe hepatotoxicity occurred, 9 of 74 patients were resected after response to HAI with 5-FU plus FA or MFFM, without surgical mortality and with survival times from 2+ months to 58+ months.

CONCLUSION

The high response rates, the long survival times, the possibility of achieving 5-year-survival either by HAI alone or by resection after down staging with HAI all sum up to the evidence that HAI could be the primary choice of treatment for CRLM. Phase-III trials are conducted to compare the protocols with optimal regional versus systemic chemotherapy.

Downstaging by regional chemotherapy of non-resectable isolated colorectal liver metastases

AIMS

To improve the course of isolated non-resectable colorectal liver metastases (CRLM) by hepatic arterial infusion treatment. Patients with CRLM have a worse prognosis than those whose liver metastases are resectable. Systemic (i.v.) chemotherapy for CRLM/colorectal metastases with 5-fluorouracil+folinic acid (5-FU+FA) i.v. may result in median survival times of 6.4-14.3 months. Hepatic artery infusion (HAI) with 5-fluorodeoxyuridine (5-FUDR) has been demonstrated in a meta-analysis of randomized trials to be superior to i.v. treatment/palliative care (median survival 15 vs. 10 months). The benefit of HAI with 5-FUDR, although recommended as treatment for CRLM, is severely compromised by the 5-FUDR induced hepatotoxicity, leading eventually to sclerosing cholangitis (SC)/liver cirrhosis. We have developed a stepwise protocol for HAI in CRLM, which is superior to HAI with 5-FUDR and to systemic chemotherapy.

METHODS

Between 1982 and 1997, 168 CRLM patients were treated within the following protocols. In protocol A, 48 CRLM patients received HAI with 5-FUDR. In protocol B, 46 patients received 5-FUDR i.a. (HAI)+i.v. In protocol C 5-FU+FA were delivered via HAI in 24 patients with CRLM. In protocol D, based on in vitro phase II studies and the results of protocol C, mitoxantrone and mitomycin C were added to 5-FU+FA (MFFM). Fifty (50) CRLM patients received HAI with HFFM.

RESULTS

The response rates, median survival time, systemic toxicity and SC rate were: 42%, 20.8 months, 0-19% and 38% for protocol A; 46%, 20.8 months, 0-20% and 41% for protocol B; 45%, 19.8 months, 4-25% and 0% for protocol C; and 66%, 27.4 months, 2-26% and 0% for protocol D. The surgically placed ports for HAI in protocols C and D functioned in 90%, 82% and 76% of patients, 6, 9, and 11 months after beginning HAI. Quality of life in protocol D was high. Nine patients from protocols C and D with either partial (PR, seven patients) or complete (CR, two patients) remissions received a secondary liver resection without hospital mortality, and seven of nine patients are alive 2-58 months after liver resection. The other two died 11 and 22 months after resection.

CONCLUSIONS

Optimal treatment of CRLM was found to be protocol D: HAI with MFFM. The results of this protocol, including high remission rate, long median survival time, good port function, good quality of life and, interestingly, the possibility of downstaging and resecting primarily non-resectable metastases, seem to be superior to HAI with 5-FUDR or 5-FU+FA and to systemic chemotherapy with 5-FU+FA. This hypothesis is currently being examined in a phase III study (HAI with MFFM vs. 5-FU+FA i.v.).

[Effect of therapy on survival of patients with pancreatic carcinoma]

PURPOSE

To identify the impact of treatment factors on overall survival in patients with pancreatic carcinoma.

PATIENTS AND METHODS

We performed a follow-up study on 38 patients with adenocarcinoma of the pancreas treated from 1984 to 1998. 18/38 patients were resected. Irradiated volume included the primary tumor (or tumor bed) and regional lymph nodes. Thirty-seven patients received in addition chemotherapy consisting of mitoxantrone, 5-fluorouracil and cis-platin, either i.v. (14/38) or i.a. (23/38). The influence of treatment related factors on the overall survival was tested. Biologically effective dose was calculated by the linear-quadratic model (alpha/beta = 25 Gy) and by losing 0.85 Gy per day starting accelerated repopulation at day 28.

RESULTS

Treatment factors influencing overall survival were resection (p = 0.02), overall treatment time (p = 0.03) and biologically effective dose (p < 0.002). Total dose and kind of chemotherapy had no significant influence. Treatment volume had a negative correlation (r = -0.5, p = 0.06) with overall survival, without any correlation between tumor size, tumor stage, and treatment volume. In multivariate analysis only biologically effective dose remained significant (p = 0.02).

CONCLUSIONS

Among with surgery, biologically effective dose strongly influences overall survival in patients treated for pancreatic carcinoma. Treatment volume should be kept as small as possible and all efforts should be made to avoid treatment splits in radiation therapy.

p53 in relation to therapeutic outcome of locoregional chemotherapy in pancreatic cancer

Since celiac artery infusion (CAI) led to an increase in survival in palliative chemotherapy in pancreatic cancer, we treated 26 patients with adjuvant CAI following resection for advanced pancreatic cancer. Catheters were placed angiographically into the celiac artery and remained there for five consecutive days. One cycle of chemotherapy consisted of mitoxantrone, 5-fluorouracil (5-FU), folinic acid, and cis-platinum. This treatment was repeated five times in monthly intervals. Median survival times in patients who received CAI are 21 months for all patients, whereas in patients who did not receive adjuvant treatment median survival is 10.5 months. In all patients p53 expression of the carcinomas was determined by immunohistochemistry. In 11/26 patients a p53 overexpression was observed. Although p53 overexpression turned out to be associated with poor prognosis in the patients who underwent adjuvant regional cancer treatment, p53 is not a sufficient prognostic parameter in pancreatic carcinoma, since p53 overexpression was more frequent in undifferentiated tumors and in palliative resected tumors.

Basic research supported developments of chemotherapy in nonresectable isolated colorectal liver metastases to a protocol of hepatic artery infusion using mitoxantrone, 5-FU + folinic acid and mitomycin C

OBJECTIVE

Since the developments in systemic chemotherapy of metastasized colorectal cancer have not resulted in substantial gains in survival times, we wished to improve the course of isolated nonresectable colorectal liver metastases (CPLM) by hepatic arterial infusion treatment.

BACKGROUND

Patients (pts) with CRLM have a worse fate than those pts whose liver metastases could be resected. Systemic (i.v.) chemotherapy for CRLM/colorectal metastases does not improve survival to a relevant level (median survival time (med. surv.) after 5-Fluorouracil + Folinic Acid (5-FU + FA) i.v.: 6.4-14.3 months (m)). Hepatic artery infusion (HAI) with 5-Fluorode-oxyuridine (5-FUDR) has been demonstrated in a metaanalysis of randomized trials to be superior to i.v. treatment/palliative care (med. surv.: 15 vs. 10 m). The benefit of HAI with 5-FUDR, although recommended as treatment for CRLM, is severely compromised by the 5-FUDR induced hepatotoxicity, leading eventually to sclerosing cholangitis (SC)/liver scirrhosis. We have stepwise developed a protocol for HAI of CRLM, which is superior to HAI with 5-FUDR, and, most evidently, to systemic chemotherapy.

PATIENTS/METHODS

Between 1982-1997, 222 CR (L) M patients were treated within subsequent protocols (Table). In protocol A, 68 CRLM pts received HAI with 5-FUDR (A1: nonrandomized pts; A2: randomized pts). In protocol B (randomized pts.), 46 pts received 5-FUDR i.a. (via HAI) + i.v. In protocol C, systemic chemotherapy with 5-FU + FA was conducted in 34 pts with metastasized colorectal cancers, including CRLM. In protocol D 5-FU + FA was delivered via HAI in 25 pts with CRLM. In protocol E, based on in vitro phase II studies and the results of protocol D, Mitoxantrone and Mitomycin C were added to 5-FU + FA (MFFM). Fifty (50) CRLM pts received HAI with MFFM.

RESULTS

The response rates, med. surv. times, systemic toxicity and SC rates are shown in the table. HAI with MFFM produced objective responses in 66%, the med. surv. was 27.4 m, and no SC occurred. The ports surgically placed for HAI, e.g., in protocols D and E, functioned in 90%, 82%, and 76% 6, 9, and 11 m after start of the HAI. Quality of life in protocol E was high. Nine pts from protocols D + E with either partial (PR, 7 pts) or complete (CR, 2 pts) remissions received a secondary liver resection without hospital mortality, and 7/9 pts are living 2-58 m after liver resection, 2/9 pts died 11 and 22 m after resection. [table: see text]

SUMMARY/CONCLUSIONS

Our learning curve to achieve optimal treatment of CRLM resulted in a protocol using HAI with MFFM. The results of this protocol (E) including the high remission rate, long median survival time, good port function, high quality of life, and, most interestingly, the possibility to downstage and resect primarily nonresectable metastases, seem to be superior to HAI with 5-FUDR of 5-FU + FA and to systemic chemotherapy with 5-FU + FA. This hypothesis is currently examined in a phase III study (HAI with MFFM vs. 5-FU + FA i.v.).

Surgery and multimodal treatments in pancreatic cancer--a review on the basis of future multimodal treatment concepts

BACKGROUND

To improve the surgical outcome after resection of pancreatic adenocarcinomas, multimodal treatment concepts need to be applied and improved. In spite of several positive studies, and the fact that multimodality treatment is the standard concept in major centers for pancreatic cancer surgery, a recent trial shed some doubt on the positive effect of adjuvant radiochemotherapy, so that the majority with reservations about multimodal treatment feel confirmed in their opinion that surgical treatment alone is sufficient therapy for resectable pancreatic cancer. The controversy among those for and against adjuvant treatment need an up-to-date review of the indications and results achievable with various treatment modalities.

PATIENTS/METHODS

The literature on the indications and results of adjuvant/neoadjuvant therapies in pancreatic cancer was reviewed to provide a solid base for current recommendations and future developments. A special view was concentrated on the biology of the disease in the spontaneous course, after surgery and during/after various palliative and adjuvant/neoadjuvant treatment modalities, to characterize the disease for an optimally targeted treatment in conjunction with surgical removal of the tumor. The results of systemic and regional chemotherapy and radiotherapy either alone or in combination, before, during, and after surgery, were critically analyzed with respect to the oncological possibilities and pitfalls of each treatment method.

RESULTS

In two randomized trials, one testing postoperative radiochemotherapy (GITSG), and one postoperative chemotherapy (Bakkevold), the adjuvant treatment achieved a significant prolongation of the median survival time. The 5- and 10-year survival rates were improved in the GITSG study. The EORTC-GITCCG trial could not confirm the benefit of adjuvant radiochemotherapy. This study had a different design than the GITSG trial. Several historical control studies supported the beneficial effect of postoperative radiochemotherapy. In three historical control trials using regional chemotherapy, one with intraoperative radiotherapy, the survival times were improved vs. surgery alone. Intraoperative or postoperative radiotherapy as single modalities might reduce local relapses, but a survival advantage is still debated. Preoperative neoadjuvant radiochemotherapy has several advantages (downstaging, devitalizing margins and lymph node metastases, compatibility of treatment vs. postoperative radiochemotherapy), and does not seem to increase the postoperative morbidity. Several trials have confirmed the feasibility of this concept, but no survival advantage has yet been proven. Systemic and regional chemotherapy is able to downstage primarily nonresectable pancreatic cancers.

CONCLUSIONS

Postoperative adjuvant radiochemotherapy with up-to-date protocols can be recommended for routine treatment, if the surgeon or the patient desires to improve the usually remote prognosis after surgery alone. For those being undecisive or against adjuvant therapy, the participation in trials, e.g., ESPAC 1 and 2 studies, is strongly recommended. Regarding our own positive experience with adjuvant regional chemotherapy and in view of the postresectional progression pattern, we currently favour adjuvant radiochemotherapy, with the chemotherapy delivered regionally via the celiac axis. This concept will be tested vs. surgery alone in the ESPAC 2 trial. Neoadjuvant therapies have a great potential, but should be conducted within studies, such as pre-, intra-, or postoperative radiotherapy.

[The coexistence of 2 different neuroendocrine tumors of the upper gastrointestinal tract and pancreas]

HISTORY AND CLINICAL FINDINGS

A 41-year-old obese patient presented with cramp-like abdominal pain, watery diarrhoea with partly digested food particles, projectile vomiting and newly diagnosed diabetes mellitus. For the preceding 6 years he had been treated for recurrent gastric and duodenal ulcers. Although the fasting gastrin level was raised and Zollinger-Ellison syndrome suspected, computed tomography (CT), magnetic resonance imaging (MRI) and coeliac angiography at another hospital had failed to discover a tumor.

INVESTIGATIONS

Biochemical tests were unremarkable except for an increased GPT concentration, slight fasting hyperglycemia and hypertriglyceridemia. The gastrin and chromogranin A levels were markedly elevated (15,590 pg/ml and 584.2 U/l, respectively). Gastroscopy revealed, in addition to multiple small duodenal ulcers, a round polypoid mass (diameter of 0.7 cm) lateral to the papilla of Vater, histologically an APUDoma. Endoscopic retrograde cholangiopancreatography (ERCP) revealed a 0.5 cm long compression of the duct of Wirsung in the region of the head of the pancreas. Liver metastases were excluded by magnetic resonance imaging and computed tomography. Endosonography showed a ca. 4 mm space-occupying lesion in the region of the body of the pancreas. Octreotide scintigraphy demarcated two foci at the level of the head of the pancreas (somatostatin-receptor positive).

TREATMENT AND COURSE

After a pylorus-preserving partial duodenopancreatectomy with lymph node dissection N1/N2, histology confirmed a gastrinoma of the duodenum and a glucagonoma of the pancreas (pT3pN1pMx). Postoperatively the patient became symptom-free and both the blood sugar level and the tumor marker were normal.

CONCLUSION

Combined ERCP, endosonography and scintigraphy are more sensitive than other radiological examinations (CT and MRI) in diagnosing and localizing neuroendocrine tumours of the gastrointestinal tract. Despite the low incidence of such tumours, the possible synchronous occurrence of several such tumour should not be ignored.

Adjuvant radiochemotherapy--what is the patients benefit?

BACKGROUND

Local relapse is a major problem after potentially curative rectal cancer surgery. Although the incidence of local recurrences may be reduced by specialized surgical techniques such as total mesorectal excision (TME), local relapse rates of 20% or higher are the surgical reality today. Studies using adjuvant postoperative radiotherapy, chemotherapy, radiochemotherapy or immunotherapy have tried to reduce local relapse rates and distant progression. Postoperative radiochemotherapy has been the recommended standard, after complete resection of Union Internationale Contra la Cancrum (UICC) stages II and III rectal cancers. In view of recent positive results with preoperative radiotherapy of TME without adjuvant therapy, we found it important to review the literature to update the recommendable adjuvant procedure in rectal cancer.

METHOD/PATIENTS

The literature from 1985 to May 1998 was reviewed for studies trying to either confirm or improve adjuvant therapy in rectal cancer. Only randomized controlled trials were analyzed with regard to their effectiveness in reducing the absolute rates of local recurrence and improving survival.

RESULTS

Two trials applying adjuvant radiotherapy were able to demonstrate the reduction of local relapse rates, one trial with marginal significance, both without impact on survival. Four trials involving 1104 patients with rectal cancer stages UICC II-III compared postoperative radiochemotherapy with either surgical controls, adjuvant radiotherapy or conventional radiochemotherapy. In these trials, local relapse rates were significantly reduced by 11-18%, and survival rates significantly improved by 10-14%. Severe acute toxicities occurred in 50-61% of the patients, compromising compatibility, and caused death in 0-1%. Small-bowel obstruction leading to surgery was noted in 2-6% and to death in up to 2% of the patients. Intraoperative radiotherapy (IORT) improved local control and survival after surgery of locally advanced disease/local relapse.

CONCLUSION

In view of four trials demonstrating a significant benefit of postoperative radiochemotherapy and with regard to recent still-debatable results of preoperative short-term radiotherapy optimal surgery with lowest local relapse rates plus postoperative radiochemotherapy remains the actual recommendable standard for rectal cancer surgery in R0 resected tumors stages UICC II+III.

Multimodal therapy in rectal cancer

The possibilities and results of multimodal treatment in rectal cancer were reviewed with respect to the results of surgical treatment only. Based on the results of 4 studies, reducing local relapse rates and increasing long-term survival rates significantly, postoperative radiochemotherapy (RCT) + chemotherapy (CT) should remain the recommended standard for R0 resected UICC II and III rectal cancers. Preoperative neoadjuvant radiotherapy (RT) reduced local relapse rates in 8 studies, and extended survival in one study that evaluated all eligible patients. Preoperative RT may evolve as standard, if the patient selection is improved and postoperative morbidity and long-term toxicity are reduced. Postoperative adjuvant RT reduced local relapses significantly in a single trial, and no impact on survival time is reported. Since postoperative RT is inferior to preoperative RT, this treatment cannot be recommended, if RT is chosen as a single treatment modality as an adjunct to surgery. Preoperative RCT + CT downstages resectable and nonresectable tumors and induces a higher sphincter preservation rate. This treatment may be routinely applied in nonresectable primary tumors or local relapses. Intraoperative RT could be added to this concept or be used together with preoperative/postoperative RT with the same indications. The results of local tumor excisions may be improved with pre- or postoperative RCT + CT.

Adjuvant regional chemotherapy in advanced pancreatic cancer: results of a prospective study

BACKGROUND/AIMS

In order to improve the dismal prognosis of patients suffering from advanced pancreatic cancer we treated 20 patients with adjuvant regional chemotherapy following resection of the tumor.

METHODOLOGY

All tumors were classified UICC stage III (TxN1M0). Regional chemotherapy consisted of Mitoxantrone 10 mg/m2 day 1, Folinic Acid 170 mg/m2 and 5-FU 600 mg/m2 days 2-4, and cis-Platinum 60 mg/m2 day 5 for up to 6 cycles. In a total of 101 cycles toxicities WHO III occurred in 6%, WHO IV in 0%.

RESULTS

The median survival times, compared to institutional historical controls (treated vs. controls) were 18.5 vs. 9.3 months (p < 0.0006). Hepatic disease progression seemed to be suppressed.

CONCLUSION

In conclusion regional chemotherapy seems to be effective for adjuvant treatment in resected pancreatic cancer.

In vitro concentration response studies and in vitro phase II tests as the experimental basis for regional chemotherapeutic protocols

The theoretical pharmacologic benefit of regional vs. systemic chemotherapy is defined and the concentration response behavior of cytostatic drugs and their optimal exposure times are described with human cancer cell lines (HT29, NMG64/84) and fresh human tumor cell suspensions in the human tumor colony assay (HTCA). The theoretical pharmacological advantages are 5.8 to 6 for adriamycin (ADM), 8 for cisplatinum (CDDP), 6.3 for epidoxorubicin (EPI), 22 to 58 for 5-fluorouracil (5FU), 4.6 for mitomycin C (MMC), and 6.3 for mitoxantrone (NOV). The drugs differed in their cytotoxic potency in vitro and thus also potential efficacy for regional chemotherapy; however, all but 5-fluorodeoxyuridine (5FUDR) exerted cytotoxicity dependent on exposure time and concentration. On average, elevation of the test concentrations by 1 lg doubled responses in fresh human tumor cell suspensions. From these results and clinical considerations, optimal times were defined for the regional chemotherapy strategies of hepatic artery infusion, intraperitoneal instillation, and chemoembolisation as performed at our institution.

Resection and radiochemotherapy of pancreatic cancer--the future?

BACKGROUND

To improve the surgical outcome after resection of pancreatic adenocarcinomas, multimodal treatment concepts need to be applied and improved. The controversies among those being pro and contra adjuvant treatment need an up-to-date review of the indications and results achievable with various treatment modalities.

PATIENTS/METHODS

The literature regarding the indications and results of adjuvant/neoadjuvant therapies in pancreatic cancer was reviewed to provide a solid base for current recommendations and future developments. The biology of the disease in the spontaneous course, after surgery and during/after various palliative and adjuvant/neoadjuvant treatment modalities was focussed on, to characterise the disease for an optimally targeted treatment in conjunction with surgical removal of the tumour. The results of systemic and regional chemotherapy and radiotherapy, either alone or in combination, before, during and after surgery were critically analysed with respect to the oncological possibilities and pitfalls of each treatment method.

RESULTS

In two randomised trials, one testing postoperative radiochemotherapy (GITSG), and one postoperative chemotherapy, the adjuvant treatment achieved a significant prolongation of the median survival time. The 5-year and 10-year survival rates were improved in the GITSG study. The EORTC-GITCCG trial could not confirm the benefit of adjuvant radiochemotherapy. This study had a different design than the GITSG trial. Several historical control studies supported the beneficial effect of postoperative radiochemotherapy. In three historical control trials using regional chemotherapy, one with intraoperative radiotherapy, the survival times were improved compared with surgery alone. Intraoperative or postoperative radiotherapy as single modalities might reduce local relapses, but a survival advantage is still debated. Preoperative neoadjuvant radiochemotherapy has several advantages (downstaging, devitalising margins and lymph node metastases, compatibility of treatment vs. postoperative radiochemotherapy), and does not seem to increase the postoperative morbidity. Several trials have confirmed the feasibility of this concept, but no survival advantage has yet been proven. Systemic and regional chemotherapy is able to downstage primarily nonresectable pancreatic cancers.

CONCLUSIONS

Postoperative adjuvant radiochemotherapy with up-to-date protocols can be recommended for routine treatment, if the surgeon or the patient desires to improve the usually remote prognosis after surgery alone. For those being indecisive or against adjuvant therapy, the participation in trials, e.g. the ESPAC 1 and 2 studies, is strongly recommended. Regarding our own positive experience with adjuvant regional chemotherapy and in view of the postresectional progression pattern, we currently favour adjuvant radiochemotherapy, with the chemotherapy delivered regionally via the celiac axis. This concept will be tested against surgery alone in the ESPAC 2 trial. Neoadjuvant therapies have a great potential, but should be conducted within studies, such as pre-, intra-, or postoperative radiotherapy.

In vitro concentration response studies and in vitro phase II tests as the experimental basis for regional chemotherapeutic protocols

The theoretical pharmacologic benefit of regional vs. systemic chemotherapy is defined and the concentration response behavior of cytostatic drugs and their optimal exposure times are described with human cancer cell lines (HT29, NMG64/84) and fresh human tumor cell suspensions in the human tumor colony assay (HTCA). The theoretical pharmacological advantages are 5.8 to 6 for adriamycin (ADM), 8 for cisplatinum (CDDP), 6.3 for epidoxorubicin (EPI), 22 to 58 for 5-fluorouracil (5FU), 4.6 for mitomycin C (MMC), and 6.3 for mitoxantrone (NOV). The drugs differed in their cytotoxic potency in vitro and thus also potential efficacy for regional chemotherapy; however, all but 5-fluorodeoxyuridine (5FUDR) exerted cytotoxicity dependent on exposure time and concentration. On average, elevation of the test concentrations by 1 lg doubled responses in fresh human tumor cell suspensions. From these results and clinical considerations, optimal times were defined for the regional chemotherapy strategies of hepatic artery infusion, intraperitoneal instillation, and chemoembolisation as performed at our institution.

Regional celiac artery infusion as adjuvant treatment after pancreatic cancer resection

BACKGROUND/AIMS

The dismal course of pancreatic adenocarcinoma patients after resection is determined by the biology of the disease preventing R-0 resections. In the spontaneous course after resection, patients frequently develop either local recurrences, liver metastases and/or peritoneal metastases. Postoperative radiochemotherapy may extend survival and reduce local recurrences without influence on hepatic progression. We performed adjuvant celiac artery infusion in pancreatic cancer, to find out whether this treatment prolongs survival and changes the biology of the disease after resection, especially by reducing liver metastasis.

METHODS

20 patients received cyclic celiac artery infusions (CAI) after resection of their pancreatic cancer (18 ductal, 2 cystadenocarcinoma). The treatment consisted of 6 cycles intraarterial infusion using Seldinger's technique with mitoxantrone (Novantron, Wyeth-Lederle, Germany) 10 mg/m2 dl, 5-fluorouracil + folinic acid (Fluroblastin, Farmitalia, Germany + Leucovorin, Wyeth-Lederle) 600 mg/m2 + 170 mg/m2 d2-4, and cis-platinum (Cisplatin, Bristol, München, Germany) 60 mg/m2 d5. The patients were monitored for toxicity, development of disease progression and survival.

RESULTS

The median survival time was 21 months, and only 15% of the patients developed liver metastases. The median survival time of the CAI-treated patient group compared favorably to the median survival of 9.3 months in a matched historical control group.

CONCLUSION

Adjuvant celiac artery infusion seemed to prolong median survival and the occurrence of liver metastases appeared to be delayed or reduced.

Palliative and adjuvant regional chemotherapy in pancreatic cancer

To improve the dismal prognosis of patients (pts) with pancreatic cancer we treated 32 patients with non-resectable (UICC III, 17 pts; UICC IV, 15 pts--group 1) and 20 patients with resected (UICC I, 1 pt; UICC II, 3 pts; UICC III, 16 pts--group 2) pancreatic cancer with palliative (group I) and adjuvant post-operative (group II) coeliac axis intra-arterial cyclic infusions (CAI). CAI consisted of mitoxantrone 10 mg/m2 on day 1, folinic acid 170 mg/m2 and 5-FU 600 mg/m2 during days 2-4, and cis-platinum 60 mg/m2 on day 5 for up to 11 (group I) or six (group II) cycles. In a total of 211 cycles toxicities at the level of WHO III occurred in 0-6% and of WHO IV in 0%. The median survival times, compared with institutional historical controls (treated vs controls), were 12 vs 4.8 months in UICC III (P < 0.006) and 4 vs 2.9 months in UICC IV (P < 0.05) group I pts, and 21 vs 9.3 months in group II (P < 0.0003). Hepatic disease progression appeared to be suppressed with CAI, which also appears to be effective for palliative and adjuvant treatment in non-resectable and resected pancreatic cancer.

Thymidylate synthase is a predictor for response and resistance in hepatic artery infusion chemotherapy

The value of intratumoral thymidylate synthase (TS) quantitation as a predictive parameter for hepatic artery infusion (HAI) chemotherapy in patients with colorectal liver metastases was investigated. Relative TS mRNA levels were determined in 29 tumor samples using a quantitative RT-PCR amplification method. The median level of expression was 3.0 x 10(-3) (no units) and varied considerably among the tumors over a range of 135-fold. Patients with low TS levels were 4.1-fold more likely to respond (P < 0.03) compared to patients with high TS levels. Our results indicate that TS quantitation is a valuable predictive marker for tumor response to HAI therapy.

[Duodenum-preserving pancreatic head resection--a standard method in chronic pancreatitis]

In patients with chronic pancreatitis the inflammatory process in the pancreatic head is frequently the pacemaker of the disease. In these cases an inflammatory tumor develops which leads to local complications in half of the patients. Duodenum-preserving pancreatic head resection, contrary to procedures used in the past, offers the possibility to preserve stomach, duodenum, biliary tree, and the insulin secretory capacity. Duodenum-preserving pancreatic head resection is a subtotal resection of the pancreatic head. In a series of 380 patients the hospital mortality rate was 0.8%, the frequency of reoperation 5.3%, and the median hospitalisation time 13.9 days. The early postoperative glucose metabolism was deteriorated in 2% and improved in 9% of cases. After a median follow-up time of 6 years, 88% of the patients were completely painfree or suffered pain rarely. Sixty-three percent were gainfully employed; the late mortality was 8.9%. Only 10% of the patients had further bouts of pancreatitis. The decisive advantage of duodenum-preserving pancreatic head resection over Kausch-Whipple resection is preservation of the endocrine pancreatic function and of neighbouring organs.

Celiac artery adjuvant chemotherapy. Results of a prospective trial

CONCLUSION

Celiac artery infusion (CAI) seems to be a qualified and successful method for adjuvant treatment of pancreatic cancer. To improve the dismal prognosis of resected pancreatic cancer patients, we performed postoperative regional chemotherapy via the celiac axis.

BACKGROUND

From 1994-1995, 20 patients with pancreatic cancer (18 ductal adenocarcinoma, 2 cystadenocarcinoma) received adjuvant celiac axis intra-arterial infusions (CAI) after resection of their tumors. Sixteen patients had macroscopically complete tumor removal (R0/R1 resection, 80% of the patients), whereas four patients had gross residual disease remaining after resection (R2 resection, 20% of the patients). Postoperative tumor stages were UICC I in 1 patient, UICC II in 3 patients, and UICC III in 16 patients.

METHODS

CAI was performed for six postoperative cycles via catheters placed into the celiac artery using Seldinger's technique. The chemotherapeutic protocol consisted of mitoxantrone (Novantron), Wyeth-Lederle (Münster, Germany) 10 mg/m2 (d 1), folinic acid (Leucovorin, Wyeth-Lederle, or Rescuvolin, Medac, Hamburg, Germany) 170 mg/m2 for 10 min, followed by 5-FU (Fluoroblastin, Farmitalia, Freiburg, Germany) 600 mg/m2 for 120 min (d 2-4), and Cisplatin (Cisplatin-medac, Medac) 60 mg/m2 (d 5). The cycles were repeated after a rest period of 4 wk. Cisplatin infusions were accompanied by supportive antiemetic (8 mg Tropisetrone i.v. [Navoban, Sandoz, Nürnberg, Germany] and 8 mg Dexametason i.v.) and diuretic measures.

RESULTS

Toxicity WHO III occurred in 8% of 100 cycles, and no toxic side effects WHO IV were encountered. The median survival of 21 mo in the treated group was nearly twice as long as the 9.3 mo of a historical matched control group (p < 0.0003). CAI seems to be a qualified and successful method for adjuvant treatment of pancreatic cancer.

Multimodal therapies in ductal pancreatic cancer. The future

CONCLUSIONS

Intra-arterial infusion chemotherapy via the celiac axis combined with external beam radiotherapy might be an effective method for palliative and perioperative multimodal treatment in pancreatic cancer. To improve the dismal prognosis in resectable and nonresectable pancreatic cancer, the results of multimodal palliative, adjuvant, and neoadjuvant therapies were reviewed and put into perspective with the results of two intra-arterial palliative and adjuvant treatment studies conducted at our department. The benefits and pitfalls of each method were outweighed, resulting in a concept for performing intra-arterial chemotherapy with radiotherapy in nonresectable stage UICC-III pancreatic cancer that eventually will be developed as a combined neoadjuvant/adjuvant treatment of all potentially resectable ductal pancreatic carcinomas.

CAM 17.1--a new diagnostic marker in pancreatic cancer

CAM 17.1-Ab is a recently described monoclonal antibody that detects a mucus glycoprotein with high specificity for intestinal mucus, particularly in the colon, small intestine, biliary tract and pancreas. We investigated the expression and release of CAM 17.1 in pancreatic carcinoma cell lines and tissue specimens of normal pancreas, chronic pancreatitis and pancreatic cancer. CAM 17.1 was weakly expressed on normal ductal cells and chronic pancreatitis, whereas it was overexpressed in pancreatic cancer. Serum analysis using a new enzyme-linked antibody sandwich assay (CAM 17.1/WGA) of patients with chronic pancreatitis, pancreatic cancer or other gastrointestinal cancer and of healthy blood donors revealed a high sensitivity (67%) and excellent specificity (90%) of CAM 17.1/WGA assay in pancreatic cancer. In comparison with the tumour marker CA19-9, the sensitivity of the CAM 17.1/WGA assay was similar to the sensitivity of CA 19-9 (67% and 76%, P = 0.22), whereas the specificity of CAM 17.1/WGA assay was higher than in CA 19-9 (90% compared with 78% in chronic pancreatitis, P > 0.05).

Regional chemotherapy directed by individual chemosensitivity testing in vitro: a prospective decision-aiding trial

A prospective decision-aiding trial was performed to select drugs for regional chemotherapy of various liver tumors (n = 36) by individual drug testing. The drugs were chosen for hepatic artery infusion according to the individual chemosensitivity of tumor biopsies in the human tumor colony-forming assay (HTCA). In vitro HTCA sensitivity correlated with complete response (CR) + partial response (PR) + no change (NC) 93% of the time and with CR + PR 55% of the time. The test sensitivity was 90%, and the specificity was 67% for CR + PR + NC versus progressive disease (PD), whereas the sensitivity and specificity were 89% and 28%, respectively, for CR + PR versus NC + PD. The overall predictive accuracy of the test was 86% for CR + PR + NC versus PD and 58% for CR + PR versus NC + PD. Overall, 83% of this heterogenous patient group with various tumors achieved CR + PR + NC and a 50% clinical response (CR + PR). In vitro-sensitive patients showed a significantly lower intrahepatic progression rate (7% PD) than in vitro-resistant patients (57%; P < 0.05). These results indicate that the HTCA could identify active drugs for individualized hepatic artery infusion, and patients may profit from the use of in vitro-sensitive drugs.

[Adjuvant regional chemotherapy in resected advanced pancreas carcinoma]

Systemic chemotherapy in patients who underwent resection of pancreatic carcinoma has not been significantly improved median survival times. We investigated whether regional chemotherapy administered via celiac trunc infusion increases survival rates in patients with locally advanced pancreatic carcinoma after resection of the primary. From 12/1992 to 2/1995 we treated 18 patients with regional chemotherapy consisting of mitoxantrone (d1), 5-FU + folinic acid (d2-4) and CDDP (d5) on five consecutive days. This cycle was repeated up to 6 times. Besides non-symptomatic GI-ulcerations (3/18) no severe side effects were observed. The median survival is actually 17.8 months (regression analysis). Compared to a historical control of our department survival times are significantly prolonged (17.8 vs. 9.3 months, p < 0,0003). In conclusion we state that adjuvant regional chemotherapy in patients with advanced pancreatic cancer is well tolerated and prolongs survival significantly.