Study ends adjuvant controversy?

February 1, 2008

Overall survival after pancreatic resection can be markedly improved with adjuvant chemoradiation therapy, according to investigative groups who presented encouraging data on two different regimens at the 2008 Gastrointestinal Cancers Symposium.

ORLANDO—Overall survival after pancreatic resection can be markedly improved with adjuvant chemoradiation therapy, according to investigative groups who presented encouraging data on two different regimens at the 2008 Gastrointestinal Cancers Symposium.

Survival after pancreaticoduodenectomy for pancreatic adenocarcinoma is only about 10% to 25% at 5 years, and the value of adjuvant chemoradiation therapy has been inconsistent in clinical trials. Results from a large cooperative study from Johns Hopkins and the Mayo Clinic (abstract 124), and from a smaller study of a novel regimen (abstract 125), suggest that adjuvant chemoradiation is beneficial rather than harmful, as a prior study suggested, possibly signaling an end to the controversy.

Charles C. Hsu, PhD, of Johns Hopkins Hospital, said, "At two high-volume centers for the treatment of pancreatic cancer, adjuvant chemoradiation therapy, compared to surgery alone, improved survival by 35% overall and consistently decreased risk by 17% to 43% among all subgroups. We also found that the status of resection should not significantly alter adjuvant chemoradiation therapy recommendations."

Dr. Hsu presented the results of a study of 1,092 patients treated with an adjuvant fluorouracil (5-FU)-based regimen plus radiotherapy at a mean dose of 50.4 Gy (n = 583) or by resection with curative intent only (n = 509).

Slightly more patients in the observation-only arm had negative surgical margins (R0) as well as low-grade histology. A propensity score analysis was conducted to account for biases associated with these differences. [Editor's note: See page 31 for a study of adjuvant chemoradiation in only R0 patients at Mayo.]

Dr. Hsu noted that the strengths of this study were its attempt to account for the nonrandom allocation and the fact that the data came from high-volume pancreatic cancer centers.

Odds of dying reduced 26%

Adjuvant chemoradiation significantly reduced the odds of dying by 26%, compared with surgery only, Dr. Hsu reported. Median overall survival in the observation arm was 15.5 months, increasing to 21.1 months with adjuvant chemoradiation therapy. Two-year survival was 34.6% vs 44.7%, respectively, and 5-year survival was 16.1% vs 22.3% (P < .001).

The propensity score analysis showed that the risk of dying was significantly improved by adjuvant chemoradiation therapy in all risk categories.

In a matched-pair analysis, the differences were striking. Adjuvant chemoradiation therapy reduced mortality by 41% (P < .001), with median survival being 14.3 months for the observation arm vs 21.9 months with adjuvant treatment. Two-year survival in this analysis was 31.4% vs 45.5%, while 5-year survival was 12.2% vs 25.4%.

In a matched-pair analysis by risk group, margin-negative and -positive patients received similar benefit (though margin-positive patients had worse outcomes in general). In margin-negative patients, adjuvant therapy reduced mortality by 39% (P < .001), with a median overall survival of 18.2 months for observation-only and 23.9 months with treatment. In the margin-positive group, mortality was reduced by 48% (P < .001), with a median survival of 8.4 months vs 18.2 months.

For poorly differentiated tumors, adjuvant treatment reduced risk by 52% (P < .001). In this group, 5-year survival was just 7.2% with observation but 25.1% with treatment.

In moderately- to well-differentiated tumors, relative reduction in risk was 26% (P = .056) with adjuvant therapy, and 5-year survival was 18.6% with observation vs 26.1% with adjuvant treatment.

For T1-2 tumors, adjuvant treatment reduced risk by 36% (P = .03) and for T3 tumors by 47% (P < .001).

Future trials should focus on sequencing of treatment modalities and even more aggressive systemic therapy, Dr. Hsu maintained.

The initial report from another phase II multicenter study, ACOSOG Z05031, found overall survival 20 months after surgery and adjuvant chemoradiation therapy of 67% and 2-year survival exceeding 40%. The problem was toxicity, which led to study termination prior to the target accrual of 93 patients.

Principal investigator Victor J. Picozzi, MD, of Virginia Mason Medical Center, Seattle, said, "The study achieved its initial primary endpoint (> 65% overall survival) in a prognostically unfavorable patient population, though toxicity was significant. Methods to reduce the impact of delivering this regimen are important."

Dr. Picozzi emphasized that the 67% survival at 20 months is "actual survival, not actuarial survival." He said that median survival is still being computed, and the results should be presented at ASCO 2008.

The study evaluated a unique chemoradiation protocol, consisting of:

• Radiation therapy on days 1-38 (XRT) (5,040 cGy total dose).

• Cisplatin 30 mg/m2 on days 1, 8, 15, 22, 29, and 36.

• 5-FU 175 mg/m2/d on days 1-38.

• Interferon-alfa 3 million units subcutaneously 3 days a week during XRT.

• Post-chemoradiation, 5-FU 200 mg/m2/d, was given during weeks 11 to 16 and 19 to 24.

"What is unique about this regimen is the use of chemotherapy with radiation specifically designed to be synergistic, and also the combination of agents. Interferon-alfa has no precedence in the pancreas. It was intended to enhance the chemoradiation but also to possibly be antiangiogenic or maybe serve as immunotherapy. How it works, we are not completely sure," Dr. Picozzi told ONI.

Investigators at 15 sites enrolled 89 poor-prognosis patients. At a median follow-up of 28 months, 64 of 80 evaluable patients (80%) had received more than 4,500 cGy XRT, while 55 (56%) had completed all three treatment cycles.

Increased toxicity

While survival was exceptional, toxicity was significant. Grade 3-4 toxicity was experienced by 96% of the group, including neutropenia (39%), nausea (36%), dehydration (27%), hypokalemia (26%), diarrhea (25%), and stomatitis (24%). Neutropenic fever, however, was seen in only 8%, and no irreversible toxicities or toxicity-related deaths were reported.

Dr. Picozzi added that the regimen was designed in 1995 before gemcitabine (Gemzar) was approved, and he believes that substituting gemcitabine for 5-FU might further enhance efficacy.

At this point, the major issue to overcome is toxicity, he said, noting that "because the drugs and radiation may be synergistic, unfortunately the side effects may be compounded."

Median survival for resected pancreatic cancer patients is only 16 to 22 months, he said, "and two-thirds of our patients have lived at least 20 months." These findings are "a third positive experience" to add to previous single-center experiences of Virginia Mason and Washington University, in which 5-year survival exceeded 40%.

"Based on the composite experience in approximately 200 patients, I am optimistic that this regimen, or some version of it, can be very effective," he said.

'Best data ever seen'

Jordon Berlin, MD, of the Vanderbilt Ingram Cancer Center and program chair for the meeting, commented on Dr. Picozzi's study for ONI.

"These are the best data we've ever seen for pancreatic cancer." The data confirm the "intriguing results of a previous study" but in a multicenter setting, he noted. The caveat is that patients were highly selected (ie, perhaps more fit than the average pancreatic cancer patient), which may have influenced the results, he added.

"At Vanderbilt, we enrolled patients on this study and selected them carefully because we knew this was a toxic regimen," he said. He believes the novel regimen needs refinement for tolerability, as well as further testing, but these issues notwithstanding, the survival results are very impressive.