A randomized study of KRAS-guided maintenance therapy with bevacizumab, erlotinib or metronomic capecitabine after first-line induction treatment of metastatic colorectal cancer: the Nordic ACT2 trial.

BACKGROUND: Maintenance treatment (mt) with bevacizumab (bev) ± erlotinib (erlo) has modest effect after induction chemotherapy in metastatic colorectal cancer (mCRC). We hypothesized the efficacy of erlo to be dependent on KRAS mutational status and investigated this by exploring mt strategies with bev ± erlo and low-dose capecitabine (cap). PATIENTS AND METHODS: Included patients had mCRC scheduled for first-line therapy, Eastern Cooperative Oncology Group (ECOG) 0-1 and no major comorbidities. Treatment with XELOX/FOLFOX or XELIRI/FOLFIRI + bev was given for 18 weeks. After induction, patients without progression were eligible for randomization to mt; KRAS wild-type (wt) patients were randomized to bev ± erlo (arms wt-BE, N = 36 versus wt-B, N = 35), KRAS mutated (mut) patients were randomized to bev or metronomic cap (arms mut-B, N = 34 versus mut-C, N = 33). Primary end point was progression-free survival (PFS) rate (PFSr) at 3 months after start of mt. A pooled analysis of KRAS wt patients from the previous ACT study was performed. RESULTS: We included 233 patients. Median age was 64 years, 62% male, 68% ECOG 0, 52% with primary tumor in situ. A total of 138 patients started mt after randomization. PFSr was 64.7% versus 63.6% in wt-B versus wt-BE, P = 1.000; and 75% versus 66.7% in mut-B versus mut-C, P = 0.579, with no significant difference in median PFS and overall survival (OS). In the pooled cohort, median PFS was 3.7 months in wt-B (N = 64) and 5.7 months in wt-BE (N = 62) (hazard ratios 1.03, 95% confidence interval 0.70-1.50, P = 0.867). The frequency of any grade 3/4 toxicities during mt was: 28%/58%/18%/15% (wt-B/wt-BE/mut-B/mut-C). CONCLUSIONS: Addition of erlo to bev as mt in KRAS wt mCRC did not significantly improve PFS or OS, but it did increase toxicity. KRAS status does not seem to influence the outcome of treatment with erlotinib. Metronomic cap warrants further investigation in mt strategies, given our explorative results. CLINICALTRIALSGOV: NCT01229813.

A randomized phase III trial on maintenance treatment with bevacizumab alone or in combination with erlotinib after chemotherapy and bevacizumab in metastatic colorectal cancer: the Nordic ACT Trial.

BACKGROUND: The main objective was to study the effect on progression-free survival (PFS) of adding erlotinib to bevacizumab as maintenance treatment following chemotherapy and bevacizumab as first-line treatment of metastatic colorectal cancer (mCRC). PATIENTS AND METHODS: Patients with untreated mCRC received doublet chemotherapy + bevacizumab during 18 weeks and those without tumor progression were eligible for randomization to bevacizumab + erlotinib (arm A) or bevacizumab alone (arm B), until progression or unacceptable toxic effect. RESULTS: Of the 249 patients enrolled, 80 started maintenance treatment in arm A and 79 in arm B. The rate of any grade 3/4 toxic effect was 53% in arm A and 13% in arm B. Median PFS was 5.7 months in arm A and 4.2 months in arm B (HR = 0.79; 95% confidence interval 0.55-1.12; P = 0.19). Overall survival (OS) from start of induction chemotherapy was 26.7 months in the randomized population, with no difference between the two arms. CONCLUSIONS: The addition of erlotinib to bevacizumab as maintenance treatment after first-line chemotherapy in mCRC did not improve PFS significantly. On-going clinical and translational studies focus on identifying subgroups of patients that may benefit from erlotinib in the maintenance setting. CLINICAL TRIALS NUMBER: NCT00598156.

A multinational phase II trial of bevacizumab with low-dose interferon-α2a as first-line treatment of metastatic renal cell carcinoma: BEVLiN.

BACKGROUND: Avastin and Roferon in Renal Cell Carcinoma (AVOREN) demonstrated efficacy for bevacizumab plus interferon-α2a (IFN; 9 MIU tiw) in first-line metastatic renal cell carcinoma (mRCC). We evaluated bevacizumab with low-dose IFN in mRCC to determine whether clinical benefit could be maintained with reduced toxicity. METHODS: BEVLiN was an open-label, single-arm, multinational, phase II trial. Nephrectomized patients with treatment-naive, clear cell mRCC and favourable/intermediate Memorial Sloan-Kettering Cancer Center scores received bevacizumab (10 mg/kg every 2 weeks) and IFN (3 MIU thrice weekly) until disease progression. Descriptive comparisons with AVOREN patients having favourable/intermediate MSKCC scores treated with bevacizumab plus IFN (9 MIU) were made. Primary end points were grade ≥3 IFN-associated adverse events (AEs) and progression-free survival (PFS). All grade ≥3 AEs and bevacizumab/IFN-related grade 1-2 AEs occurring from first administration until 28 days after last treatment were reported. RESULTS: A total of 146 patients were treated; the median follow-up was 29.4 months. Any-grade and grade ≥3 IFN-associated AEs occurred in 53.4% and 10.3% of patients, respectively. The median PFS and overall survival were 15.3 [95% confidence interval (CI): 11.7-18.0] and 30.7 months (95% CI: 25.7-not reached), respectively. The ORR was 28.8%. CONCLUSIONS: Compared with a historical control AVOREN subgroup, low-dose IFN with bevacizumab resulted in a reduction in incidence rates of IFN-related AEs, without compromising efficacy [NCT00796757].

Preoperative plasma soluble urokinase plasminogen activator receptor as a prognostic marker in rectal cancer patients. An EORTC-Receptor and Biomarker Group collaboration.

BACKGROUND AND AIMS: Since approximately 30% of patients with Dukes' stage B colorectal cancer will experience disease recurrence within five years of primary treatment, current staging of patients with early colorectal cancer apparently fails to adequately predict patient outcome. It has previously been shown that the preoperative plasma concentration of soluble urokinase plasminogen activator receptor (suPAR) is associated with the survival of patients with early colorectal cancer. In this study we sought to confirm the independent prognostic value of suPAR in rectal cancer. METHODS: suPAR was retrospectively determined by two different versions of a suPAR ELISA in preoperatively collected plasma samples from a Swedish (n = 354) and a Danish (n = 255) cohort of rectal cancer patients. RESULTS: In both cohorts the suPAR concentration was significantly higher in Dukes' stage D patients than in Dukes' stage A-C patients (p < 0.0001). Among Dukes' stage A-C patients, no differences in median suPAR values were seen. In univariate analysis, continuous suPAR was found to be associated with survival (p < 0.0001 in both cohorts). Of particular interest was that similar results were obtained for Dukes' stage A and B patients when analyzed separately. In multivariate analysis, continuous suPAR was found in both cohorts to be independent of Dukes' stage. CONCLUSIONS: This study confirms that the preoperative concentration of plasma suPAR contains independent prognostic information on patients with rectal cancer. This result was independent of the two different versions of an in-house suPAR ELISA used to perform the analyses. The next step in the evaluation of suPAR as a prognostic parameter in rectal cancer will be to launch an appropriately dimensioned prospective study where the benefit of applying preoperative plasma suPAR measurement to clinical decision-making regarding adjuvant therapy is assessed.

Preoperative plasma soluble urokinase plasminogen activator receptor as a prognostic marker in rectal cancer patients. An EORTC-Receptor and Biomarker Group collaboration.

BACKGROUND AND AIMS: Since approximately 30% of patients with Dukes stage B colorectal cancer will experience disease recurrence within five years of primary treatment, current staging of patients with early colorectal cancer apparently fails to adequately predict patient outcome. It has previously been shown that the preoperative plasma concentration of soluble urokinase plasminogen activator receptor (suPAR) is associated with the survival of patients with early colo-rectal cancer. In this study we sought to confirm the independent prognostic value of suPAR in rectal cancer. METHODS: suPAR was retrospectively determined by two different versions of a suPAR ELISA in preoperatively collected plasma samples from a Swedish (n=354) and a Danish (n=255) cohort of rectal cancer patients. RESULTS: In both cohorts the suPAR concentration was significantly higher in Dukes stage D patients than in Dukes stage A-C patients (p<0.0001). Among Dukes stage A-C patients, no differences in median suPAR values were seen. In univariate analysis, continuous suPAR was found to be associated with survival (p<0.0001 in both cohorts). Of particular interest was that similar results were obtained for Dukes stage A and B patients when analyzed separately. In multivariate analysis, continuous suPAR was found in both cohorts to be independent of Dukes stage. CONCLUSIONS: This study confirms that the preoperative concentration of plasma suPAR contains independent prognostic information on patients with rectal cancer. This result was independent of the two different versions of an in-house suPAR ELISA used to perform the analyses. The next step in the evaluation of suPAR as a prognostic parameter in rectal cancer will be to launch an appropriately dimensioned prospective study where the benefit of applying preoperative plasma suPAR measurement to clinical decision-making regarding adjuvant therapy is assessed. (Int J Biol Markers 2005; 20: 93-102).

Association between preoperative plasma levels of tissue inhibitor of metalloproteinases 1 and rectal cancer patient survival. a validation study.

The level of the tissue inhibitor of metalloproteinases 1 (TIMP-1) has previously been demonstrated to predict the survival of early stage colorectal cancer patients. The present study was undertaken to further validate plasma TIMP-1 as a prognostic marker in rectal cancer. Preoperative plasma from 352 rectal cancer patients were analysed using an immunoassay for TIMP-1. The TIMP-1 immunoassay demonstrated robustness and good reproducibility with low interassay coefficients of variation (CV). The rectal cancer patients had a mean plasma TIMP-1 level of 184 microg/l (standard deviation (SD): 70 microg/l). There were no significant differences in TIMP-1 levels between patients with Dukes' stage A, B or C disease, whereas Dukes' stage D patients had significantly increased TIMP-1 levels (P<0.0001); however, high levels of TIMP-1 were not restricted to those with advanced disease. Univariate analysis demonstrated an increasing risk of mortality with increasing TIMP-1 levels Hazard Ratio (HR)=2.9; 95% Confidence Interval (CI): 1.7-5.0; P<0.0001). Including additional covariates, multivariate analysis identified plasma TIMP-1 as an independent prognostic marker (HR=2.2; 95% CI: 1.2-4.1 (P=0.01). This study showed a highly significant and independent association between preoperative plasma TIMP-1 levels and survival in rectal cancer patients, thus confirming our previous findings. Furthermore, the TIMP-1 immunoassay proved to be stable and reproducible in this confirmatory study.

Prognostic importance of the soluble plasminogen activator receptor, suPAR, in plasma from rectal cancer patients.

Colorectal cancer is one of the most common tumour types with approximately one third of the tumours located within the rectum. Rectal cancer differs somewhat from colon cancer, e.g. regarding the method of operation and the use of preoperative radiotherapy due to a tendency for local tumour recurrence. Proteolytic enzymes have been identified as key molecules in tumour invasion and metastasis, and factors within the urokinase-plasminogen activation (uPA) system have been associated with prognosis in several tumour types, including colorectal cancer. Recently, methods have been developed to analyse the soluble fraction of the plasminogen activator receptor (suPAR) in blood samples. An association between elevated suPAR levels and poor prognosis has recently been demonstrated in colorectal cancer. We have measured suPAR levels in pretreatment plasma samples from 173 rectal cancer patients in order to confirm its prognostic strength in this clinical entity. suPAR levels were determined in ethylenediamine tetraacetic acid (EDTA) plasma by a kinetic enzyme-linked immunosorbent assay (ELISA) and analysed with respect to sex, age, Dukes' stage, tumour differentiation grade and survival. In a univariate analysis, continuous suPAR plasma levels were associated with survival (P<0.001) with shorter survival among patients with high suPAR values. Patients with suPAR values within the upper quartile had significantly shorter survival (hazard ratio (HR) 2.2, 95% confidence interval (CI) 1.3-43.7, P=0.002). In a multivariate Cox analysis, increasing suPAR values predicted shorter survival independent from Dukes' stage and tumour differentiation grade with an adjusted HR of 2.2 per ng/ml suPAR (95% CI 1.2-4.0, P=0.01). This study thus confirms that measurement of suPAR in preoperative plasma samples gives independent prognostic information in rectal cancer patients, higher values being associated with shorter survival.

Lack of activating c-SRC mutations at codon 531 in rectal cancer.

Cellular SRC (c-SRC), which is the human homolog of the Rous sarcoma viral oncogene, v-src, is highly activated in colorectal cancer. Recently, a subset of colon cancers have shown a nonsense mutation at codon 531, which truncates c-SRC directly C-terminal to the c-SRC kinase regulatory domain. This specific mutation has been demonstrated to be activating, transforming, and tumorigenic and to promote metastasis. We investigated 100 rectal cancers, half of which had tumor spread outside the rectum, for this mutation by using direct sequencing. None of these tumors displayed any genetic alteration at this locus, and we thus conclude that the codon 531 mutation is a rare cause of c-SRC activation in rectal cancer.

Microsatellite instability is rare in rectal carcinomas and signifies hereditary cancer.

We analysed microsatellite instability (MSI) in a consecutive series of 165 rectal carcinomas. Data on a personal and/or family history of cancer were collected from all patients and revealed metachronous cancer in 9 patients, 2 of whom had developed colorectal cancer, and a suspected familial aggregation of colorectal cancer in three families. Only three of the 165 (2%) rectal cancers showed MSI. The patients whose tumours displayed MSI had clinical histories suggesting hereditary cancer--a family history of colorectal cancer and/or synchronous colorectal cancers. Denaturing gradient gel (DGGE) analysis was used to screen the MSI+ patients for mutations in the hMLH1 and hMSH2 genes and revealed two new germline mutations; a 1 bp deletion in exon 10 of hMSH2 creating a premature stop-codon and a splice donor site mutation in intron 16 of hMLH1. Considering colorectal carcinomas as a group, MSI has been reported to occur in approximately 10-20% of the tumours and thus can not, per se be used for clinical detection of hereditary tumours. This study shows, however, that MSI is rare in rectal carcinomas and when present strongly suggests a hereditary predisposition for colorectal cancer development.

hMLH1, hMSH2 and hMSH6 mutations in hereditary non-polyposis colorectal cancer families from southern Sweden.

We have screened 17 Southern Sweden individuals/families with suspected hereditary non-polyposis colorectal cancer (HNPCC) for mutations in the DNA-mismatch repair genes hMLH1, hMSH2 and hMSH6 using denaturing gradient gel electrophoresis, protein truncation test and direct DNA sequencing. The families were selected on the basis of a family history of HNPCC-related tumors or the occurrence of metachronous colorectal cancer/endometrial cancer at young age in an individual with a weak family history of cancer. Furthermore, we required that tumor tissue from at least one individual in the family had to display microsatellite instability. We identified germ-line mutations in 9 individuals from 8 families. Five families had mutations in hMLH1, 4 of which were splice site mutations, 2 had frameshift mutations in hMSH2 and 1 patient with metachronous endometrial and rectal cancer but with a weak family history of cancer had a nonsense mutation in hMSH6. Our results present novel germ-line DNA-repair gene mutations, one of these in hMSH6, and demonstrate the diversified mutation spectrum in Sweden, where no founder mutation has so far been identified.