Do you recommend neoadjuvant chemotherapy or induction chemo + chemoRT for borderline resectable pancreatic cancer?

The definition of BRPC has expanded to include everything from minimal venous involvement to 180 degree abutment of the SMA.  It is important to determine if there is arterial involvement.  If there is, the best data show that XRT helps with R0 resections. In our experience at MDACC, 95% of patients with arterial invovement treated with CXRT have negative margins. These data do not exist with chemotherapy alone. That fact does not stop people from recommending neoadjuvant chemotherapy alone in thiese cases, but until there are similar data, that practice should be discouraged.  Radiation can be given with 50.4 Gy in 28 fractions as we do at MDACC or with  low dose SBRT (32Gy in 5 fractions). Either way the goal of treatment is R0 resection and both work to achieve that goal. There is no magical improvment in outcome with low dose SBRT, but it is more convenient and well tolerated. The notion that neaoadjuvant therapy improves resectablilty or "downstages" patients is a common misconception. In our experience, radiographic downstaging from T4 to T3 occurs 1% of the time. Response to therapy can influence a decision to operate, but that is not the same thing as improving resectability. The surgeon typically makes the decision to operate based on the biology of the disease and the PS of the patient after therapy. We favor the use of neoadjuvant chemotherapy before CXRT in these cases.

On the other end of the spectrum, patients with only minimal vein involvement probably can be treated with neoadjuvant chemotherapy alone with a high R0 resection rate. There are some emerging data to support that approach in these T3 patients with minimal vein involvment. As long as there is no tumor within 1cm of the SMA or celiac axis, omitting CXRT in this group is reasonable.

It is true that 54Gy of XRT did not improve OS in locally advanced patients in the LAP-07 trial, but that study is not relevant to the BRPC population with arterial involvement (also T4 but resectable) where the issue is improving margin negative resectability.  Margins are +80% of the time without XRT and <10% with XRT. Margin+ = death.

I would add the findings of Hammel et al. published in JAMA recently: patients with stage III locally advanced disease (so T4, unresectable at diagnosis) were randomized to chemo (gemcitibine, since the trial was started before FOLFIRINOX and Abraxane) vs. chemoRT (after an earlier randomization to gem vs. gem/erlotinib 4 months earlier). With adequate power, there was no difference in overall survival or progression-free survival from the time of first randomization, and there was an increase in grade 3-4 nausea in the chemoRT group. Locoregional progression was less frequent with chemoRT but was counterbalanced by more metastatic progression. Very few patients in both groups underwent surgical resection, and the trial was stopped early for futility.