Annual meeting of the American Society of Clinical Oncology (ASCO), 2008.

Colorectal cancer: adjuvant treatment

Major progress has been made in the treatment of patients with high-risk colorectal cancer. For stage III disease, adjuvant treatment is usually offered and the FOLFOX regimen is the standard of care. Overall survival is improved by about 20% compared to surgery alone. When fluoropyrimidines are used alone, the cure rate is still increased by about 15%. High-risk stage II patients are usually offered adjuvant therapy as well, with an estimated benefit of about 10%. Recent adjuvant studies such as the MOSAIC trial indicate that survival curves continue to separate after 5 years of observation, suggesting that longer than 5-year observation times are necessary to appreciate the real survival benefit of modern adjuvant therapy. At this year's ASCO Meeting (Chicago, Illinois, US; May 30-June 3, 2008), de Gramont and co-workers reported on the time from relapse to death depending on the type of adjuvant treatment. Interestingly, the median time to relapse for patients who had received infusional 5-FU plus folinic acid alone was 24 months, compared to 21 months for patients who had received the FOLFOX-4 regimen. At the time of analysis 6.1% of patients who had received the FOLFOX regimen as adjuvant treatment were alive after relapse compared to 7.8% of patients who had received infusional 5-FU/folinic acid alone. Wolmark and co-workers observed very similar results from patients who had entered the CO-7 study. In this adjuvant trial, bolus 5-FU/folinic acid given as a weekly schedule according to the Roswell Park regime was compared to bolus 5-FU/leucovorin and oxaliplatin (FLOX regimen). At 5 years the percentage of patients who were alive differed by only 2%, favouring the FLOX regimen (78.3% versus 80.3%). However, at 6 years the survival curves had continued to separate and the difference was 4.2% (77.7% versus 73.5%). The survival difference just failed statistical significance [hazard ratio (HR)=0.85, confidence interval (CI) 0.72-1.10; P=0.06]. Interestingly, the survival of patients calculated after recurrence was also in favour of patients who had received 5-FU/folinic acid alone (median of 22.2 months), compared to patients who received the FLOX regimen (median of 17.6 months). This difference was statistically significant (P=0.02).

Both reports indicate that:

1. More effective adjuvant treatment with FOLFOX results in more patients who are alive by preventing relapse.

2. More effective palliative treatments were available for patients who had received 5-FU/folinic acid in the adjuvant setting and this probably led to longer survival times after relapse compared to patients who had received the FOLFOX regimen.

3. Monitoring for more than 5 years is necessary to see the true effect of a given adjuvant treatment on overall survival.

Ychou and co-workers reported on their adjuvant study of infusional 5-FU/folinic acid compared to the FOLFIRI regimen in patients who had received an R0 resection of colorectal liver metastases. A total of 153 patients were randomised into each arm. There was no difference in the disease-free survival for patients receiving the FOLFIRI regimen (HR=0.89, CI 0.66-1.19; P=0.43). Unfortunately, this study adds to the negative experience of irinotecan given in the adjuvant setting as reported for the PETACC-3 study at ASCO 2005, for the CALCB 89803 study [1], and by Mark Ychou at ASCO 2005, for high-risk stage III patients (i.e. patients with more than four positive lymph nodes). Thus, irinotecan-containing regimens do not have any role for the adjuvant treatment of patients with colorectal cancer.

Reducing oxaliplatin-induced neurotoxicity

Nikcevich and co-workers reported on their randomised study of patients receiving FOLFOX in the adjuvant setting. In order to reduce the oxaliplatin-induced neurotoxicity, patients were randomised to receive 1 g calcium gluconate and 1 g magnesium sulfate in 100 ml glucose 5% over 30 minutes immediately before and after oxaliplatin administration or placebo. The neurotoxicity was recorded in three different ways: according to the NCI/CTC version 3.0 grading; an oxaliplatin-specific scale; and a patient questionnaire. A total of 102 patients were treated in the study. According to the NCI/CTC scale 41% of the placebo group had at least grade 2 or higher neurotoxicity while only 22% of patients experienced neurotoxicity when treated with calcium and magnesium (P=0.0038). The difference was also significant when the oxaliplatin-specific scale was used with a total of 50.1% of patients in the placebo group and only 28% of patients in the treatment group experiencing greater than grade 2 neurotoxicity (P=0.018). This study indicates that calcium and magnesium infusions may be helpful in preventing the oxaliplatin-induced neurotoxicity.

Metastatic disease

For advanced disease, patients are usually treated with a sequence of regimens including FOLFIRI or FOLFOX, often in combination with an antibody such as bevacizumab or cetuximab. Using this approach, about 40-50% of patients respond to first-line treatment and experience a median progression-free survival of 10 months or more and a median overall survival of about 20 months. Punt et al. reported on the work of the Dutch Colorectal Cancer Study Group (DCCG) who randomised patients to receive either capecitabine and oxaliplatin (CapOx) plus bevacizumab or CapOx/bevacizumab plus cetuximab. The median progression-free survival was shorter for patients who also received cetuximab (9.6 months) compared to patients who did not (10.7 months) and was statistically significant (HR=1.21; P=0.018). This difference in progression-free survival did not translate into a difference in median overall survival which was 20.4 or 20.3 months, respectively (HR=1.15; P= 0.21). Patients receiving cetuximab in addition to CapOx/bevacizumab had a slightly higher rate of gastrointestinal toxicity. This experience is in accordance with the earlier reported PACCE study, where patients were randomised to receive FOLFOX plus bevacizumab alone versus FOLFOX plus bevacizumab plus panitumumab. The progression-free survival favoured patients who did not receive panitumumab with a median of 11.1 months compared to 9.6 months (HR=1.44; P=0.004). The results of both studies were unexpected, as added benefit of cetuximab and bevacizumab was observed in pretreated patients. Nevertheless, caution has to be exercised if both of the antibodies, bevacizumab directed against VEGF protein and cetuximab or panitumumab directed against the EGFR receptor, are combined. Further studies are needed to better understand this unexpected negative influence.

Grothey and co-workers investigated whether the modified FOLFOX-7 regimen, in combination with bevacizumab, given continuously until treatment failure, is advantageous compared to modified FOLFOX-7 plus bevacizumab but given on a 'stop-and-go' strategy for 4 months in combination with oxaliplatin, followed by 4 months without oxaliplatin. A total of 68 and 61 patients were randomised, respectively. The study was projected for a larger number of patients with a 2x2 factorial design to also study the influence of calcium and magnesium on oxaliplatin neurotoxicity. Because an interim analysis gave the impression that infusion of calcium and magnesium could reduce the response rate, this study was stopped early. According to the final analysis there is no indication that intravenous calcium or magnesium reduces the efficacy of the FOLFOX regimen plus bevacizumab, but the intermittent approach increased the time to treatment failure, compared to the continuous application (P=0.002). A possible explanation for this observation is that patients who have received FOLFOX continuously may have needed treatment interruptions due to oxaliplatin-induced neurotoxicity. Although this study is rather small, it is in line with other research which has not shown any benefit of treatment duration until progression for 5-FU, irinotecan-containing regimens or for the oxaliplatin-containing schedule.

Probably the most exciting data has come from the CRYSTAL study that was first presented last year. Patients receiving FOLFIRI plus cetuximab had a statistically significant prolongation of their progression-free survival with a median of 8.9 months compared to patients receiving FOLFIRI alone (8.0 months). This difference is statistically significant (HR=0.85; P=0.048). After 1 year, 31% of patients were without progression on the FOLFIRI plus cetuximab regimen compared to 23% of patients on FOLFIRI alone. It was clear from the 2007 presentation that the benefit was limited to a subgroup of patients that needed to be defined. The effects of antiproliferation of cetuximab may be circumvented by downstream proteins, such as K-ras, that in the case of a particular mutation will transform the cell into constant activation. Several retrospective analyses have demonstrated that tumours with wildtype K-ras have a better chance of responding to EGFR receptor antibodies compared to patients with mutant K-ras. Specimens from over 500 patients in this study (the largest cohort analysis so far) were analysed for K-ras status. Patients who had tumours with K-ras wildtype had a median progression-free survival of 9.9 months when cetuximab was added to FOLFIRI compared to 8.7 months in those patients who received FOLFIRI alone. This difference was statistically significant (P=0.017). After 1 year, 33% of patients in the combination arm were without progression compared to 25% of patients in the FOLFIRI-alone arm. In patients whose tumours were K-ras mutant (n=192) there was no difference in the median overall survival, which was 7.6 months for cetuximab plus FOLFIRI and 8.1 months for FOLFIRI alone (HR=1.07; P=0.47). When the response rate was analysed, 59% of patients with K-ras wildtype tumours responded to cetuximab and FOLFIRI compared to 43% of patients with FOLFIRI alone. In patients with K-ras mutant tumours, FOLFIRI led to a 40% response rate and FOLFIRI plus cetuximab resulted in a 36% response rate (P=0.46).

These data are in accordance with those reported by Bokemeyer and co-workers for FOLFOX [+ or -] cetuximab (OPUS study). In this study the objective response rate was the major study endpoint. A total of 96% of patients with K-ras wildtype tumours responded to FOLFOX plus cetuximab compared to 37% of patients receiving FOLFOX alone. This resulted in a longer progression-free survival for FOLFOX plus cetuximab compared to FOLFOX alone (HR=0.57). In patients with K-ras mutant tumours, the response rate appeared to be lower for patients with cetuximab in addition to FOLFOX (33%) compared to those patients who received FOLFOX alone (49%). The progressionfree survival was shorter for patients who received FOLFOX plus cetuximab compared to those patients who received FOLFOX alone (HR=1.83). The DCCG group also analysed their data on CapOx and bevacizumab [+ or -] cetuximab according to the K-ras genotyping. A total of 501 patients had tumours that could be analysed in this respect. Interestingly, the shortest median progression-free survival was observed for those patients who had a K-ras mutation and in addition to capecitabine, oxaliplatin and bevacizumab also received cetuximab (P=0.043). Although this did not result in a statistically significant difference in the overall survival, the median overall survival was lowest for patients with a K-ras mutation who received cetuximab in addition to chemotherapy and bevacizumab.

This data indicates that before administering an EGFR-receptor antibody the K-ras status of patients should be analysed. Those patients who have a wildtype K-ras are candidates for this treatment, while for those patients who have K-ras mutations, cetuximab or treatment with panitumumab may be detrimental, especially when a platinum compound is included in the chemotherapy regimen.

Loupakis and co-workers suggested that in addition to K-ras status, PTEN status may further add to the prognostic value. Patients who had PTEN-positive tumours and wildtype K-ras were more likely to respond to an EGFR-receptor antibody compared to those patients who did not have a PTEN-positive or K-ras wildtype tumour. As the estimation of PTEN expression is very subjective and not standard at the moment, these data need further studies.

In summary, these ASCO data indicate that more than 5 years of follow-up may be necessary in order to estimate the real benefit of a new adjuvant regimen. Also calcium and magnesium infusions may be effective in decreasing oxaliplatin-induced toxicity, at least in the adjuvant setting.

In metastatic disease, two strategies are available to reduce the oxaliplatin-induced neurotoxicity: magnesium and calcium infusions and a 'stop-and-go' strategy. Treatment until progression may not be necessary in most patients but should be carefully considered in those patients with aggressive tumours. Chemotherapy plus two antibodies directed against VEGF protein or the EGFR receptor may be harmful and needs further investigation before being used routinely in the clinic. The most important information is that K-ras-mutated tumours do not benefit from EGFR receptor-directed treatment. The K-ras mutation status should be determined before using these agents.

Upper-GI tumours Pancreatic cancer

Neuhaus and co-workers reported updated results of their adjuvant studies for patients with R0 or R1 resection of pancreatic cancer. A total of 175 patients were allocated to the observational group and 179 patients received gemcitabine, 1000 mg/[m.sup.2], on days 1, 8 and 15 every 4 weeks. The primary endpoint was disease-free survival and was reported at last year's ASCO meeting to be statistically significantly in favour of the gemcitabine treatment arm. This year updated results were reported. The median disease-free survival was 13.4 months for patients receiving gemcitabine compared to 6.9 months for the observational arm (P<0.001). The median overall survival was also statistically significantly prolonged for patients in the gemcitabine arm (22.8 months versus 20.2 months in the observation arm; P=0.005). More impressively, at 3 years 36.5% of patients were alive if treated with gemcitabine compared to 19.5% in the observation arm, and at 5 years 21% in the gemcitabine arm compared to 9% in the observation arm. These data confirm that gemcitabine is the standard of care for patients with R0 or R1 resection, independent of nodal disease or tumour stage.

Loehrer and co-workers reported on their study of gemcitabine plus radiotherapy in patients with localised, unresectable pancreatic cancer (study E4201). Patients were randomised to receive gemcitabine, 1000 mg/[m.sup.2], once weekly, six times, followed by 1000 mg/[m.sup.2], once weekly, three times, followed by 1 week of rest for five cycles. Patients in the experimental arm, in addition to 600 mg/[m.sup.2] gemcitabine, received radiotherapy of 180 cGy per day for 5 days per week over 6 weeks, for a total dose of 54.0 Gy. After radiotherapy, consolidation chemotherapy with gemcitabine followed. The primary endpoint was overall survival. However, the study had to be closed early due to a low accrual after inclusion of only 71 patients. Nevertheless, this study indicated that median overall survival could be improved for patients receiving radiotherapy in addition to gemcitabine compared to patients receiving gemcitabine alone (11.0 months versus 9.2 months; P=0.034). At 2 years, only 4% of patients were alive in the gemcitabine arm compared to 12% in the combination arm. Although the combination treatment of gemcitabine and radiotherapy was associated with a higher rate of toxicity, this study suggests that we should consider this kind of treatment in unresectable locally advanced pancreatic cancer.

Advanced pancreatic cancer

Several attempts have been made to improve the efficacy of gemcitabine in metastatic pancreatic cancer. Only erlotinib and capecitabine are able to modestly increase the overall survival, with erlotinib causing a statistically significant improvement. At last year's meeting it was reported that bevacizumab failed to improve the efficacy of gemcitabine (Kindler et al., ASCO 2007). This year, a randomised Phase II trial by Kindler and co-workers reported the use of gemcitabine, bevacizumab and cetuximab compared with gemcitabine, bevacizumab and erlotinib, with 61 and 71 patients randomised into each arm, respectively. The primary endpoint of this study was the overall response rate. The overall response rate, progression-free survival and overall survival were the same in both arms. While patients with an early rash had a significantly prolonged progression-free survival and overall survival, there was also a correlation between early hypertension and response that was not observed for overall survival. The authors concluded that gemcitabine and bevacizumab or cetuximab, and gemcitabine/ bevacizumab plus erlotinib, had similar activity and toxicity profiles, but did not recommend further evaluation of this kind of treatment.

Vervenne and co-workers reported on a large randomised trial including 607 patients randomised to gemcitabine plus erlotinib and bevacizumab compared to gemcitabine, erlotinib and placebo. There was no significant difference in the objective response rate, which slightly favoured the combination of gemcitabine, erlotinib and bevacizumab (8.6% versus 13.5%). The median progression-free survival was 1 month longer for the bevacizumab treatment arm (4.6 months versus 3.6 months; P=0.0002) but this did not translate into a difference of overall survival, which was 6.0 months for the placebo arm and 7.1 months for the bevacizumab arm (P=0.2087). In this study bevacizumab did not improve the efficacy of gemcitabine and erlotinib alone.

Metastatic pancreatic cancer: second line

Currently, while the first-line treatment with gemcitabine in metastatic pancreatic cancer is a well established treatment option, the value of second-line treatment and its effect on overall survival has never been randomly studied. Pelzer and co-workers reported on their randomised trials including 168 patients who had received gemcitabine as first-line treatment and received either infusional 5-FU/folinic acid (folinic acid, 200 mg/[m.sup.2], and 5-FU 2 g/[m.sup.2]) as a weekly 24-hour infusion or the same regimen plus oxaliplatin, 85 mg/[m.sup.2], on days 1 and 15. The median progression-free survival of 13 weeks favoured the oxaliplatin-containing regimen compared to 5FU/folinic acid alone (9 weeks; P=0.012). There was also a difference in the overall survival of 20 weeks versus 13 weeks (P=0.014) favouring the oxaliplatin arm as the regimen was feasible and tolerable with a greater clinical benefit. The authors suggested considering this regimen as a standard second-line treatment after progression on gemcitabine in metastatic pancreatic cancer.

Hepatocellular carcinoma

Sorafenib is approved as systemic therapy by the FDA and EMEA for the treatment of patients with hepatocellular carcinoma according to the SHARP study presented at last year's ASCO meeting. Cheng and co-workers reported on their experience on an Asian-Pacific population receiving placebo (n=76) versus sorafenib, 500 mg bid (n=150). Patients with advanced hepatocellular carcinoma, an ECOG performance status of at least 2 and a Child-Pugh A status without prior systemic treatment were randomised. The primary endpoint was overall survival, which favoured patients treated with sorafenib (6.5 months versus 4.2 months; HR=0.68; P=0.014). This Asian-Pacific population was compared to Western patients from the SHARP study. Asian-Pacific patients had a higher likelihood of having HBV or HCV infection as the underlying cause of the hepatocellular carcinoma (73% versus 80%, 8% versus 28%, respectively). Nevertheless, both populations appeared to benefit similarly from the sorafenib treatment. Sorafenib treatment was safe and well tolerated across patient populations in both these studies. This study confirmed the benefit of sorafenib for the treatment of HCC patients in varied geographic regions.

Gastro-oesophageal cancer

Stahl and co-workers updated their experience of chemoradiotherapy [+ or -] surgery in locally advanced oesophageal cancer. The overall survival favoured patients receiving surgery in addition to chemoradiotherapy in their 2005 publication [2] (P=0.004) because at 2 years 38.8% versus 35.4% of patients were alive. By December 2007, when data were re-analysed, the 5-year survival was 27.9% versus 17.0%, which no longer favoured the additional surgery to chemoradiotherapy alone (HR=1.15; P=0.36). Thus, also considering other studies of squamous cell carcinoma of the oesophagus, both chemoradiotherapy and surgery alone are valid treatment options.

Gastric cancer

Ridwelski and his group randomised patients with metastatic gastric cancer to receive docetaxel/cisplatin versus 5-FU/leucovorin and cisplatin as first-line treatment. A total of 273 patients were randomised to receive docetaxel, 75 mg/[m.sup.2] IV, and cisplatin, 75 mg/[m.sup.2] IV, on day 1, repeated every 3 weeks for six cycles. Patients in the other treatment arm received 5-FU 2000 mg/[m.sup.2] IV, given as a continuous infusion over 24 hours plus leucovorin, 500 mg/[m.sup.2] IV, as a 2-hour infusion on a weekly basis, cisplatin, 50 mg/[m.sup.2] IV, was added on days 1, 15 and 29. Cycles were repeated every 7 weeks for four cycles. The major study endpoint was time to progression. There was no difference in the median time to progression of patients receiving DC (5.8 months) versus patients receiving FLC (6.6 months, P=0.468). Also, the median survival with 8.2 months versus 9.6 months was not statistically significant (P=0.616). The overall response rate was 24% in both arms. Interestingly, the toxicity of the DC regimen as well as the FLC regimen was rather low, with a high rate of leukopenia and neutropenia in the DC arm, and a rate of febrile neutropenia of 5.5%. There was slightly more gastrointestinal toxicity for patients receiving FLC (29% versus 19%). The authors concluded that DC is not superior to FLC in contrast to the V325 study where D added to CF was superior to CF. FLC and DC were both well tolerated in the dose and schedule, and may be regarded as an additional treatment option for patients with metastatic gastric cancer.

References

(1.) Saltz LB, Niedzwiecki D, Hollis D et al. Irinotecan fluorouracil plus leucovorin is not superior to fluorouracil plus leucovorin alone as adjuvant treatment for stage III colon cancer: results of CALGB 89803. J Clin Oncol, 2007, 25, 3456-3461.

(2.) Stahl M, Stuschke M, Lehmann N et al. Chemoradiation with and without surgery in patients with locally advanced squamous cell carcinoma of the esophagus. J Clin Oncol, 2005, 23, 2310-2317.

Claus-Henning Kohne

Klinik fur Onkologie/Hamatologie, Klinikum Oldenburg, Oldenburg, Germany