LONDONEarly trial results from the Royal Marsden GI
Collaborators, in the United Kingdom, show that 12 weeks of
protracted venous infusion 5-fluorouracil (PVI 5-FU) results in the
same efficacy as 6 months of bolus 5-FU for colorectal cancer. This
results in less toxicity and impairment of quality of life (QOL).
Looking at these still preliminary results retrospectively, it also
appears that PVI 5-FU may be superior for patients with Dukes B
tumors and rectal primary tumors.
The goal of this multicenter study, Asha Saini, PhD, told those
attending the recent ASCO meeting, was to see if PVI 5-FU produced a
higher response rate and less toxicity than bolus 5-FU. It was
thought that perhaps this improved therapeutic index could enhance
the eradication of micrometastatic disease in the adjuvant setting.
The eligibility criteria included adenocarcinoma of the colon
or rectum; Dukes stage B or C; no residual or metastatic
disease as assessed by radiology or CEA measurement; normal
hematologic, renal, and hepatic function; and no serious concurrent illness.
The patients were randomized within 12-weeks of curative surgery. The
bolus group received 5-FU 425 mg/m² plus folinic acid 20
mg/m², on days 1 to 5, every 4 weeks for 24 weeks (Mayo
regimen). The PVI 5-FU group received 300 mg/m²/d via continuous
infusion over 12 weeks.
Survival was the primary study endpoint, with secondary endpoints of
relapse-free survival, toxicity, and quality of life. The accrual
goal was 716 patients. The trial is calculated to detect 10% overall
survival at 5 years (power 80%).
The treatment groups were well balanced in terms of age (range, 26 to
83). About a third of the tumors were located in the rectum, with 67%
being in the colon. Cancer stage was evenly split between Dukes
B and C, and most of the tumors were well to moderately
differentiated. The median follow-up at the time Dr. Saini reported
the findings at ASCO was 22 months.
While 95% of the patients experienced toxicity, there were no
treatment-related deaths. Treatment-related toxicities consisted
mostly of diarrhea and stomatitis. However, there was a significant
amount of grade 2 alopecia.
That is severe alopecia requiring a wig, which is not always
recognized as a toxicity associated with 5-FU therapy, Dr.
Saini said. In fact, when we analyzed female patients alone, we
found that 22% treated with the Mayo clinic schedule required a wig,
compared with under 1% receiving the PVI 5-FU regimen.
More of the patients on the Mayo regimen had grade 3-4 neutropenia
(55% vs 0.6% on PVI 5-FU). These toxicities, Dr. Saini said, also
reflect the achievable dose intensities of 73.7% for bolus 5-FU vs
89.7% for PVI 5-FU.
Hickman Line Complications
In the continuous-infusion arm, 395 patients had complications of
Hickman linesmost (23%) were associated with superficial skin
infections at the line insertion site11% had pain, 7% had
thrombotic events, and 8% had line removal or replacement. Most
patients received Warfarin at a prophylactic dose of 1 mg/d,
Dr. Saini said.
Quality of Life
In assessing quality of life, there was a maximal drop at the 2-week
time point, and it was significantly worse for those patients
receiving bolus 5-FU. However, at the end of the first month, the
scores were improved.
This is an extremely important point, because in many studies,
quality of life is measured on a monthly basis when this very early
important data point could be missed, she said.
Despite the early disadvantage for 5-FU, Dr. Saini continued,
the overall quality-of-life scores remained identical until the 12-week
time point, when they all return to baseline. At this point, of
course, the PVI patients had finished their therapy.
Although the median duration of follow-up is only 20 months, so
far, we already see that there are significantly fewer relapses among
patients who are treated with PVI 5-FU (P = .01), Dr.
Saini said. This difference has not yet developed into an
overall 5-year survival differential, she said, Survival
is about 74% regardless of treatment arm.
When relapse-free survival was looked at by Dukes stage, there was an
advantage to PVI 5-FU for Dukes stage B (P = .24). This
difference was not seen in Dukes stage C. Overall survival is
no different for either treatment arm or either stage.
If we separate patients according to site, the findings become
even more intriguing, because in those patients with colon tumors,
both relapse-free survival and overall survival are similar in both
treatment groups, Dr. Saini said. However, for the
subgroup of patients with rectal cancer, giving 5-FU by infusion
results in an enormous reduction in relapse rate, with a P
value of .003.
Time to Distant Relapse
While time to local relapse was identical between the two treatment
arms, the time to distant relapse showed a significant difference in
favor of PVI 5-FU (P = .0007).
Dr. Saini said this observation is suggesting that infusional
treatment predominantly acts on distant metastatic disease.
However, one oncologist attending the meeting countered that while
these data are interesting, it may be premature to so explicitly
extrapolate from the not-yet-completed trial. One drawback of the
study, he pointed out, is that the cost and inconvenience of the
continuous-infusion treatment were not part of the study design.
I absolutely agree with you that these are very early
results, Dr. Saini said. But we thought they were
interesting enough to present. She added that the cost data are
being included in the analysis and will be part of the final
published study. The subset analysis has been used to extend accrual
for rectal cancer patients to increase the statistical power of that
effect, she said.