Neoadjuvant treatment for resectable pancreatic adenocarcinoma: What is the best protocol?

Abstract Although upfront surgery has been the gold standard for pancreatic adenocarcinoma that is planned for resection, it should be compared with the alternative strategy of neoadjuvant therapy. Despite the many reports of the efficacy of neoadjuvant therapy, most of them were not comparative. Recently Prep‐02/JSAP05 study clearly demonstrated the significant survival benefit of neoadjuvant chemotherapy over upfront surgery for pancreatic adenocarcinoma that is planned for resection. These findings opened a new chapter of neoadjuvant therapy. Ongoing trials are expected to confirm the evidence. This review summarizes the past, present, and future perspectives of neoadjuvant therapy and its optimization.


| RE S EC TAB ILIT Y OF PAN CRE ATI C ADENOC ARCINOMA
Several definitions of resectability of pancreatic adenocarcinoma (PDAC) have been approved for determining the possibility for complete clearance (R0 resection) by surgery, taking into account oncological and general aspects. [1][2][3][4] Surgical resectability of PDAC is assessed by the evaluation of local tumor extension to vessels and distant metastases. Excluding tumor with distant metastases, which is defined as unresectable with metastases (UR-M), local resectability is classified in three categories: resectable (R), borderline resectable (BR), and unresectable (UR-LA). R PDAC shows no vascular infiltration to major vessels. Complete clearance of R tumor is required in standard pancreatectomy without combined vascular resection. BR PDAC is sub-classified into two categories: BR-PV showing PV distortion or narrowing, and BR-A showing semi-circumferential abutment with a major artery. There is a theoretical "borderline" between BR-PV and BR-A. Whereas PV resection is currently recommended for achieving R0 resection, 5,6 arterial resection remains controversial due to significantly increased rates of morbidity. 6 From the surgical perspective, BR-PV PDAC is borderline resectable, whereas BR-A PDAC is borderline unresectable. Considering surgical feasibility, R and BR-PV PDAC should be considered as candidates for "PDAC that is planned for resection (potentially resectable PDAC)." Potentially resectable PDAC has been treated by upfront surgery, 1,2 although neoadjuvant for BR PDAC might be considered given the poor oncological outcomes. 7

| P OTENTIALLY RE S EC TAB LE PDAC
Upfront surgery has been the gold standard for potentially resectable PDAC, as well as for most other solid cancers. Adjuvant therapy (adjuvant) is administered for macroscopically curatively resected PDAC with full recovery in the planned postoperative period and without immediate early recurrence ( Figure 1). This cohort benefits from recent advances of adjuvant chemotherapy. Randomized controlled trials (RCTs) of adjuvant chemotherapy reported that the median overall survival (OS) these selected patients reached was 46.5 months with S1 adjuvant 8 and 54.4 months with modified FOLFIRINOX adjuvant. 9 Adjuvant for patients with resected PDAC, who are eligible after selection for surgery (Figure 1), is fully accepted as the standard based on solid evidence. In contrast to eligible patients, patients with aggressive tumor (incompletely resectable, 10 immediately recurred, 11 or vulnerable for treatment (insufficiently recovered 12 ) who show a poor prognosis are excluded from analysis.
Unfortunately, it is hard to discriminate, before surgery, between eligible patients and ineligible patients for adjuvant. Since potentially resectable PDAC is not equal to resected PDAC eligible for adjuvant ( Figure 1), it is not convincing that upfront surgery is the optimal strategy for potentially resectable PDAC. The optimal strategy should be explored by a comparison between upfront surgery and the alternative strategy of neoadjuvant therapy (neoadjuvant) followed by surgery.

| PROS PEC TIVE S TUD IE S AND ME TA-ANALYS E S
As well as upfront surgery, neoadjuvant followed by surgery has patient selection during the neoadjuvant period, in addition to surgical selection ( Figure 2). In retrospective or case-series studies of neoadjuvant for PDAC, the survival outcome of only resected PDAC after neoadjuvant appeared to show a theoretically better trend because of exclusion of patients with a poor prognosis. This selection usually causes significant bias 13 even in large-scale studies. In contrast to retrospective analyses, a prospectively designed interventional study can provide low-biased survival data by intention-to-treat (ITT) analysis. Several prospective studies of neoadjuvant reported survival outcomes including data from ITT analyses.

| OP TIMAL PROTO COL FOR NEOADJ U VANT THER APY
Two cycles of the GS regimen, which was used in the Prep-02/ JSAP05 study, have been a standard regimen for NAC, at least in Japan, for potentially resectable PDAC. [38][39][40] Although several prospective trials using other regimens, which include radiotherapy, are ongoing, their results have yet been clearly reported. 36,37,[41][42][43][44] Considering recent progress in chemotherapy for UR PDAC, 45,46 a clinical question has been raised about the optimal protocol in the neoadjuvant setting.  24 Okano K 57 R, BR S1 + RT 91% N. R. N. R. 25 Motoi F 101 R, BR GEM/S1 73% 30.8 N. R. 26 Tsai S 130 R, BR FOLFIRINOX (n = 52) FOLFIRI (n = 26) GEM/Nab-P (n = 16) CAP/Nab-P (n = 15) *+RT (n = 83) 82% 38 45 27 Eguchi H 63 R GEM/S1 + RT that gemcitabine and oxaliplatin, which were used in a neoadjuvant study, showed a significantly higher response rate and longer PFS than gemcitabine alone. 49 Cunningham et al 50 also demonstrated a significantly higher response rate and longer PFS with combination gemcitabine plus capecitabine than with gemcitabine alone. These combination regimens might be candidates for the optimal NAC regimen, and they are summarized in Table 3.

| THE OP TIMAL REG IMEN FOR NEOADJ U VANT CHEMOTHER APY
In the adjuvant setting, the modified FOLFIRINOX was more active than gemcitabine. 9 In a similar setting, however, gemcitabine plus nab-paclitaxel showed marginal results that were not significant Abbreviations: CAP, Capecitabine; GEM, Gemcitabine; N, Number in the cohort; Nab-P, Nab-Paclitaxel; OS, median overall survival in months of each arm; OX, Oxaliplatin; PFS, median progression-free survival in months of each arm.
with respect to recurrence-free survival compared to gemcitabine. 51 The trial comparing gemcitabine plus capecitabine to gemcitabine in the adjuvant setting (ESPAC-4) showed positive results. 52 Murakami et al 53 reported, in a retrospective analysis, that the GS regimen was active in the adjuvant setting. Given the efficacy of FOLFIRINOX in various settings, 9,45 this regimen would be one of the most attractive candidates to be evaluated as NAC, compared with NAC-GS as a control. In any case, a well-designed RCT is necessary to explore the optimal regimen.

| CHEMOTHER APY AND/OR R ADIOTHER APY ?
Chemoradiotherapy is an attractive modality in the treatment of PDAC for local disease control. In the neoadjuvant setting, an increase of R0 resection, which would decrease local relapse after resection, is expected for NACRT. Many prospective non-randomized trials of NACRT have been reported. 14,15,[18][19][20][21][22]25,27,28 RCTs comparing NACRT with upfront surgery, however, have not yet been fully reported, [33][34][35][36][37] in contrast to NAC. 38 55 Their concept is "total neoadjuvant therapy (TNT)," an emerging approach with excellent outcomes for other cancers. 56 The group also reported the effect of TNT for UR PDAC, with excellent survival. 57 These strategies might also improve the survival outcome of R PDAC. The use of radiation should be examined, rather than assuming an either/or scenario in a prospective trial. Radiological response would be a candidate surrogate endpoint.

| OP TIMAL DUR ATI ON FOR NEOADJ U VANT
Radiological assessment, which could be performed before surgery, is superior to pathological assessment in clinical decision-making.
Radiological CR could reflect pathological CR, which could be a surrogate endpoint of survival in the other types of cancer described above. As well as pathological CR, radiological CR of PDAC is rarely obtained even after multi-modal treatment. 28,55 Serum tumor markers and their kinetics are other promising candidates as surrogate endpoints. CA19-9, which is increased in most PDACs at baseline, is widely used as a tumor marker. A decrease of tumor markers after therapy reflects a good response and longer survival for responders. 64 For resected PDAC, a decrease of CA19-9 to the normal range after surgery is associated with longer survival and a low hepatic relapse rate. 65,66 The CA19-9 level, which can be measured less-invasively and quantitatively, has several advantages as a surrogate endpoint. In a proportion of the cases with normal CA19-9 levels after surgery following neoadjuvant, CA19-9 levels would be a surrogate endpoint of survival to select an optimal regimen or duration of the treatment. However, further efforts are still needed to determine the optimal cut-off point of tumor marker as a surrogate endpoint.

| CON CLUS ION
Recently evidence opened a new chapter of the neoadjuvant era for PDAC. However, it was only a beginning, and further efforts are needed to optimize it with adequate surrogate markers.