PHOENIX – Seeing is not always believing when it comes to determining whether pancreatic adenocarcinomas are resectable.

In patients with locally advanced or borderline resectable pancreatic adenocarcinoma who underwent neoadjuvant therapy with the FOLFIRINOX regimen (leucovorin, fluorouracil, irinotecan, and oxaliplatin), presurgical imaging was not a reliable determinant of resectability, said Dr. Cristina R. Ferrone from the department of general and gastrointestinal surgery at Massachusetts General Hospital in Boston.

"It really is a paradigm shift that we can’t rely on CT scans and MRIs the way we did before," she said at the annual Society of Surgical Oncology Cancer Symposium.

An important case

For her team, the paradigm shift was sparked by the case of a 41-year-old woman who presented with a 3.6-cm pancreatic adenocarcinoma involving the superior mesenteric artery (SMA), and a cancer antigen 19-9 (CA 19-9) level of 985 U/mL (normal values for the nonspecific biomarker range from 0 to 37 U/mL, according to National Comprehensive Cancer Network guidelines).

Following 4 months of FOLFIRINOX and 50.4 Gy of chemoradiation, her CA 19-9 level normalized to 37 U/mL. But her postchemotherapy imaging studies appeared identical to her prechemotherapy studies, with no apparent evidence of tumor regression and apparent continued involvement of the SMA following FOLFIRINOX.

"We had a big debate about this woman, because she had a normalized CA 19-9. Do we attempt to resect a patient like this?" Dr. Ferrone said.

After considerable discussion, the decision was made to take the patient into the operating room, approach the pancreas from the left side, and perform intraoperative biopsies on the SMA. If the biopsies were positive, the surgical team would either perform intraoperative radiation therapy (IORT) or use a soft-tissue ablation system (Nanoknife). If the biopsies were negative, the team would proceed to resection.

The patient went on to resection, and was found to have only minimal residual tumor on postoperative pathology samples, Dr. Ferrone reported.

Case comparisons

To see whether other patients could benefit from a similar approach, the investigators compared outcomes for 87 patients who underwent upfront resection for presumed organ-confined disease and 38 patients who received neoadjuvant FOLFIRINOX before surgery. Of the latter group, 5 received FOLFIRINOX chemotherapy only; 16 received FOLFIRINOX and 50.4 Gy of radiation; 10 received FOLFIRINOX, radiation, and IORT; 4 received FOLFIRINOX and proton-beam therapy; and 3 had FOLFIRINOX with other drug therapies.

The median CA 19-9 level fell from 169 U/mL before FOLFIRINOX to 16 U/mL after (P = .003), and more than twice as many patients had levels below 40 U/mL after therapy (31% vs. 72%; P = .03)

Median tumor diameter on CT also decreased after chemotherapy, from 3.6 cm to 2.2 cm (P = .001).

Pathology review of images from 20 cases with the observer blinded to treatment status showed 14 cases judged to be locally advanced disease and 6 borderline pre-FOLFIRINOX, and 12 locally advanced cases, 3 borderline, and 5 cases deemed resectable post-FOLFIRINOX.

When the investigators compared clinical and operative results in the surgery-only patients vs. results in those who had neoadjuvant FOLFIRINOX, they found that the latter group had higher mean operative times (300 minutes for surgery-only patients vs. 393 minutes for FOLFIRINOX-treated patients; P less than .001) and higher mean blood loss (400 mL vs. 600 mL, respectively; P = .01).

However, patients who received neoadjuvant therapy had significantly fewer postoperative complications (63% vs. 35% of patients; P = .004) and fewer pancreatic fistulas (29% vs. 0%; P less than .001).

Patients who received FOLFIRINOX also had a lower proportion of readmissions within 90 days (30% vs. 21%); shorter lengths of stay (7 days vs. 6 days); and no deaths, compared with 1 in the up-front surgery group, but these differences were not significant.

Pathology results showed that 79% of the surgery-only patients had positive lymph nodes, compared with only 37% of those in the neoadjuvant group (P less than .001). The neoadjuvant group also had lower proportions of both lymphatic invasion (70% vs. 37%, P less than .001) and perineural invasion (P less than .001).

There were proportionally more R0 (clean margin) resections in the FOLFIRINOX group, but the difference was not significant.

After a median follow-up of 10 months for the up-front surgery patients and 12 months for the FOLFIRINOX-treated patients, there have been no differences in local or distant progression or deaths from pancreatic cancer. Dr. Ferrone acknowledged that a larger patient sample and longer follow-up may be needed to detect significant differences.

However, Kaplan-Meier overall survival estimates from the time of diagnosis out to 34 months significantly favor the FOLFIRINOX-treated patients (P = .02).

"With this therapeutic modality and intensive surgical interventions, we hope that we are making an impact in these patients’ lives and improving their survival," she said.

The study was internally supported. Dr. Ferrone reported having no financial disclosures.

Imaging studies don’t predict pancreatic resectability post-FOLFIRINOX