Initial systemic chemotherapy for metastatic exocrine pancreatic cancer

INTRODUCTION — A majority of patients with exocrine pancreatic cancer are diagnosed with adenocarcinomas arising from the ductal epithelium. (See "Pathology of exocrine pancreatic neoplasms".)

Most patients with exocrine pancreatic cancer present with advanced disease, due to its aggressive natural history. Surgical resection offers the only chance of cure. However, only 15 to 20 percent of patients have resectable disease at initial diagnosis; the majority have either locally advanced or metastatic cancer. Neoadjuvant therapy may represent an appropriate therapeutic option for individuals with locally advanced, unresectable, or borderline resectable non-metastatic pancreatic cancer. (See "Initial chemotherapy and radiation for nonmetastatic, locally advanced, unresectable and borderline resectable, exocrine pancreatic cancer".)

For patients with metastatic cancer, palliative systemic chemotherapy can improve disease-related symptoms and prolong survival. Initial systemic chemotherapy in the treatment of metastatic pancreatic adenocarcinoma will be reviewed here. Second-line systemic therapy for subsequent treatment, multimodality management for locally advanced pancreatic cancer, specific methods for symptom palliation, and management of patients with metastatic islet cell (endocrine) tumors are discussed separately. (See "Second-line systemic therapy for advanced exocrine pancreatic cancer" and "Initial chemotherapy and radiation for nonmetastatic, locally advanced, unresectable and borderline resectable, exocrine pancreatic cancer" and "Supportive care for locally advanced or metastatic exocrine pancreatic cancer" and "Endoscopic ultrasound-guided celiac plexus interventions for pain related to pancreatic disease" and "Metastatic well-differentiated gastroenteropancreatic neuroendocrine tumors: Presentation, prognosis, imaging, and biochemical monitoring".)

PRETREATMENT EVALUATION AND COUNSELING

General principles — A multiphasic computed tomography scan of the chest, abdomen, and pelvis should be performed prior to initiating treatment to assess disease extent [1]. Other staging studies should be performed only as dictated by symptoms. Serum levels of the tumor marker carbohydrate antigen 19-9 (CA 19-9; also called cancer antigen 19-9) should be measured at the start of treatment, and, if elevated, may be used to follow disease status during chemotherapy. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Abdominal CT' and 'Response assessment' below.)

Multidisciplinary collaboration to formulate treatment and care plans is preferred. Goals of care, patient preferences, treatment response, symptom burden, and psychosocial issues (including the patient's support system) should guide decisions for treatment [1]. (See "Discussing goals of care".)

Patients with metastatic pancreatic cancer should be offered aggressive treatment for pain and other symptoms related to the cancer. Referral for a palliative care consultation should be considered early in the treatment course, especially for patients with a high symptom burden. Early initiation of palliative care services improves clinical and quality of care outcomes and may prolong survival. (See "Supportive care for locally advanced or metastatic exocrine pancreatic cancer" and "Benefits, services, and models of subspecialty palliative care", section on 'Rationale for palliative care'.)

All patients should be offered information about clinical trials. Enrollment in clinical trials is preferred, if available. If a trial is unavailable or participation is not feasible, conventional chemotherapy is an option for patients with an adequate performance status (PS), controlled comorbidity, and a sufficient support system.

Systemic chemotherapy in this setting is palliative and not curative. However, the available evidence from randomized trials and meta-analyses [2,3] suggests that systemic chemotherapy can improve symptoms and prolong survival over best supportive care alone. However, these trials have generally compared older fluorouracil (FU)-based combination chemotherapy regimens with supportive care alone for first-line treatment of advanced pancreatic cancer. In the Cochrane meta-analysis [3], which included eight prospective randomized trials conducted between 1974 and 2001 of chemotherapy versus supportive care alone for advanced pancreatic cancer [4-11], chemotherapy significantly improved one-year mortality by 73 percent (odds ratio 0.37, 95% CI 0.25-0.57).

There are no trials addressing the survival benefit of modern combination regimens versus best supportive care, but at least two trials have shown a clear survival benefit for combination therapy (leucovorin [LV] plus short-term infusional FU plus oxaliplatin and irinotecan [FOLFIRINOX], and gemcitabine plus nanoparticle albumin-bound paclitaxel [nabpaclitaxel], respectively) versus single-agent gemcitabine alone. (See 'FOLFOX and FOLFIRINOX' below and 'Gemcitabine plus nabpaclitaxel' below.)

Patients with an Eastern Cooperative Oncology Group (ECOG) PS ≥3 or poorly controlled comorbid conditions should be offered systemic chemotherapy only on an individualized, case-by-case basis; supportive care should be emphasized.

Germline and targeted tumor genomic testing — For patients with metastatic pancreatic cancer, both germline genomic testing (if not already known) and somatic genomic testing (gene profiling of tumor tissue using multigene panel-based assays) should be undertaken as quickly as possible after diagnosis because of the significant treatment implications both for first-line therapy and beyond [1,12]. A provisional clinical opinion from ASCO supports somatic tumor genomic testing in metastatic or advanced cancer if the finding of one or more specific alterations has regulatory approval as a biomarker to guide the use of, or exclusion from, certain treatments for their disease [12]. Given the tissue-agnostic approvals for any advanced cancer with a high tumor mutational burden (TMB) or DNA mismatch repair deficiency (dMMR; immune checkpoint inhibitor immunotherapy), or neurotrophic tyrosine receptor kinase (NTRK) fusion (TRK inhibitors), this provides a rationale for somatic genomic testing for all solid tumors, including advanced pancreatic cancer, if the individual would be a candidate for these treatments. For patients without tissue-based genomic test results, treatment may be based on actionable alterations identified in circulating free DNA (cfDNA).

Testing for potentially actionable germline genomic alterations should be preceded by a discussion between the clinician and patient as to the frequency of actionable findings, treatment implications of testing results, and the need for genetic counseling related to germline testing results.

The finding of pathogenic or likely pathogenic variants in homologous recombination repair (HRR) genes such as the breast cancer susceptibility (BRCA) or partner and localizer of BRCA2 (PALB2) genes (often referred to as "BRCA-like") is of particular relevance for first-line therapy of metastatic pancreatic cancer. (See 'Implications for first-line treatment' below.)

Both germline and somatic genomic testing for BRCA or PALB2 variants are necessary. Such alterations are present in approximately 10 percent of patients with pancreatic cancer. Approximately one-half will have somatic alterations only and one-half will be germline (ie, roughly 5 percent of patients with pancreatic cancer have a germline BRCA or BRCA-like pathogenic variant). However, the ranges reported for pathogenic germline or somatic variants in these and other well-characterized genes associated with homologous recombination deficiency such as ATM, ATR, CHEK2, RAD51, and the FANC genes, vary widely from 13 to 44 percent, depending on the sequencing method and average coverage depth [13]. (See "Familial risk factors for pancreatic cancer and screening of high-risk patients", section on 'Inherited cancer susceptibility syndromes'.)

In principle, all patients with a germline BRCA or PALB2 pathogenic variant should have that same pathogenic variant detected (if looked for) in the tumor. For patients whose germline BRCA or PALB2 pathogenic variant carrier status is not known, targeted tumor testing that identifies the presence of a BRCA or BRCA-like pathogenic variant might be interpreted as likely somatic or germline, but most tests cannot reliably distinguish between the two, and this may not be reported uniformly by all the different companies that provide this service. Furthermore, pathogenic germline variants (PGVs) may be missed by tumor testing alone for a variety of reasons, including technical limitations of tumor sequencing (especially small copy number deletions, and large or complex insertions or deletions), variant interpretation differences between tumor and germline tests, or differences in the genes tested in the tumor and the germline. In one report, in which 2023 patients with cancer unselected for family history received germline testing and previously had tumor DNA sequencing, PGVs were found in 617 (30.5 percent), and in 41.7 percent of the 187 pancreatic cancers [14]. Potentially more importantly, 8.1 percent of the PGVs were missed by tumor sequencing. Thus, tumor testing cannot substitute for germline testing in patients for whom a PGV, such as a BRCA or PALB2 pathogenic variant might influence treatment decisions, as described below. (See 'Implications for first-line treatment' below.)

Implications for first-line treatment

●Homologous recombination repair deficiency and sensitivity to first-line platinum agents and PARP inhibitors – Cells that lack BRCA1 or BRCA2 activity as well as those that have other defects in genes associated with HRR (eg, ATM, BAP1, BARD1, BLM, BRIP1, CHEK2, FAM175A, FANCA, FANCC, NBN, RAD50, RAD51, RAD51C, and RTEL1) have a deficiency in the repair of DNA double-strand breaks, which affects chemotherapy sensitivity and influences the choice of initial treatment (see 'Patients with homologous recombination repair deficiency' below):

•Regimens that include a platinum agent (eg, FOLFIRINOX, FOLFOX, gemcitabine plus cisplatin) should be considered standard regimens in all BRCA or BRCA-like pathogenic variant carriers with a good PS, and who can tolerate aggressive therapy. The deepest and most prolonged responses to platinum agents are seen in germline BRCA-mutant patients. (See 'Platinum-based chemotherapy' below.)

Furthermore, the finding of a germline BRCA or PALB2 pathogenic variant also selects those patients who might benefit from maintenance treatment with a poly(ADP-ribose) polymerase (PARP) inhibitor after an initial period of platinum-containing chemotherapy. (See 'PARP inhibitor maintenance therapy' below.)

Somatic BRCA pathogenic variants may also portend a deep and prolonged response to a platinum-containing regimen, and therefore, a platinum-containing regimen such as FOLFIRINOX is also preferred in patients with an unknown germline pathogenic variant status but with a tumor (somatic) BRCA or BRCA-like pathogenic variant, or a pathogenic variant in another gene associated with HRR (ie, ATM, BAP1, BARD1, BLM, BRIP1, CHEK2, FAM175A, FANCA, FANCC, NBN, RAD50, RAD51, RAD51C, and RTEL1). (See 'Basic biology' below.)

•On the other hand, first-line gemcitabine plus nabpaclitaxel could be a reasonable choice for initial therapy in those patients who are known to lack both germline and somatic mutations in HRR genes, and who have a good PS, a relatively favorable comorbidity profile, a serum bilirubin level that is <1.5 times the upper limit of normal (ULN), and a preference and a support system for aggressive medical therapy [1]. (See 'Gemcitabine plus nabpaclitaxel' below.)

Gemcitabine alone is an option for patients with an ECOG PS of 2, or a comorbidity profile that precludes more aggressive regimens. The addition of nabpaclitaxel or capecitabine to gemcitabine could be considered in this setting with proactive dose and schedule adjustments to minimize toxicities. (See 'Choice of regimen' below.)

For most patients with pending or unknown pathogenic variant status, we do not recommend waiting to initiate treatment until the germline or somatic pathogenic variant status is known, given the rapidity of progression in most patients with newly diagnosed metastatic pancreatic cancer. Therefore, FOLFIRINOX has become our standard regimen by default, as long as patients have an adequate PS and comorbidity, a desire for aggressive treatment, and access to port and infusion pump services. Often, we start LV plus short-term infusional FU and oxaliplatin (FOLFOX) in those patients who have a borderline PS, then add irinotecan when we know that the patient can tolerate it. (See 'Pathogenic variant status unknown' below and 'FOLFOX and FOLFIRINOX' below.)

●dMMR/MSI-H – Patients with deficient mismatch repair (dMMR)/microsatellite instability-high (MSI-H) tumors as well as those with a high TMB respond to checkpoint inhibitor immunotherapy. Although there are no phase II studies for first-line treatment, we believe that it is reasonable to consider first-line checkpoint inhibitor immunotherapy in this rare group of individuals. If first-line chemotherapy is chosen, is imperative that these patients be monitored closely and experience the potential benefits of checkpoint inhibitor immunotherapy in the event that chemotherapy is not effective, especially given the tendency of pancreatic cancer patients to decompensate quickly (eg, from biliary or duodenal obstruction).

Genetic considerations of identifying those with an inherited susceptibility — In addition to these therapeutic issues, genes associated with hereditary breast and ovarian cancer (HBOC; ie, BRCA mutations) and colorectal cancer predisposition (eg, deficient mismatch repair [dMMR]) have been shown to play a role in pancreatic cancer susceptibility. Growing evidence suggests that pancreatic cancer may be useful as a sentinel cancer to identify families that could benefit from HBOC or colorectal cancer surveillance even in the absence of a strong family history [15].

In 2018, ASCO published a provisional clinical opinion on the identification and management of patients and family members with a possible inherited predisposition to pancreatic cancer [16]. This included the recommendation that "all patients diagnosed with pancreatic adenocarcinoma should undergo assessment of risk for hereditary syndromes known to be associated with an increased risk for pancreatic adenocarcinoma." High-risk individuals should be referred for genetic counseling and germline genetic testing as appropriate. Access to clinical genetic testing has improved, and the cost is significantly lower and more affordable for multigene panel testing. Consistent with ASCO guidelines and updated guidelines for pancreatic cancer from the NCCN, germline genetic testing may also be offered to patients with pancreatic cancer and an unremarkable family history if an informative result could directly benefit the patient or his or her family members [16,17]. (See "Familial risk factors for pancreatic cancer and screening of high-risk patients" and "Genetic testing and management of individuals at risk of hereditary breast and ovarian cancer syndromes" and "Lynch syndrome (hereditary nonpolyposis colorectal cancer): Cancer screening and management".)

Because of these important treatment issues, and the emerging role of molecularly targeted therapy in those who have potentially targetable germline or somatic molecular abnormalities, our approach includes early germline testing and gene profiling of tumor tissue for all pancreatic cancers (as per ASCO [1] and National Comprehensive Cancer Network [NCCN] guidelines [17]). (See "Second-line systemic therapy for advanced exocrine pancreatic cancer", section on 'Molecularly-targeted therapy'.)

INITIAL TREATMENT — Our approach to first-line systemic therapy, which is consistent with clinical practice guidelines from the American Society of Clinical Oncology (ASCO) [1], the National Comprehensive Cancer Network (NCCN) [17], and the European Society for Medical Oncology (ESMO) [18], is outlined in the algorithm (algorithm 1).

Patients with homologous recombination repair deficiency — For patients with metastatic pancreatic cancer, genomic testing and targeted tumoral testing of the pancreatic cancer (ie, with next-generation sequencing) should be undertaken as quickly as possible since the finding of a pathogenic variant in a gene involved in homologous recombination repair (HRR) may influence the choice of initial treatment. These pathogenic variants are associated with a defective DNA damage response, which increases sensitivity to DNA damaging agents, such as platinum agents, and to drugs targeting the DNA damage response pathway, including poly(ADP-ribose) polymerase (PARP) inhibitors. (See 'Germline and targeted tumor genomic testing' above and 'General principles' above and "Familial risk factors for pancreatic cancer and screening of high-risk patients".)

Overview of our approach

●For advanced-stage pancreatic cancer that arises in the setting of a known germline or somatic pathogenic variant involving an HRR gene (BRCA1/2, PALB2, ATM, BAP1, BARD1, BLM, BRIP1, CHEK2, FAM175A, FANCA, FANCC, NBN, RAD50, RAD51, RAD51C, and RTEL1), we suggest a platinum-based rather than non-platinum-based chemotherapy regimen. It is expected that germline mutations will confer the greatest benefit.

●For most patients with an excellent Eastern Cooperative Oncology Group (ECOG) performance status (PS; 0 or 1 (table 1)) who are able to tolerate an intensive approach, who have a serum bilirubin level that is <1.5 times the upper limit of normal (ULN), a preference and a support system for aggressive medical therapy, and access to chemotherapy port and infusion pump management services, we prefer leucovorin (LV) plus short-term infusional fluorouracil (FU) plus oxaliplatin and irinotecan (FOLFIRINOX (table 2)) or modified FOLFIRINOX (table 3). Alternatively, gemcitabine plus cisplatin (table 4) can be used and probably has similar benefit. Another alternative regimen in this setting is LV plus short-term infusional FU plus oxaliplatin (FOLFOX (table 5)), with reassessment for adding irinotecan at a later time. For patients with an ECOG PS of 2, a comorbidity profile that precludes intensive therapy, or a serum bilirubin that remains >1.5 times ULN despite stenting, we would choose FOLFOX (table 5).

●In addition, for patients who have a germline BRCA or PALB2 pathogenic variant, after at least 16 weeks of initial platinum-based chemotherapy, for those without disease progression, we suggest discontinuation of chemotherapy and initiation of maintenance therapy using the PARP inhibitor olaparib, based on findings from the Pancreas cancer OLaparib Ongoing (POLO) trial. The optimal timing of initiation of the PARP inhibitor is not established. The POLO trial started olaparib in patients who lacked progression after 16 weeks of chemotherapy. Another option is to treat with chemotherapy for at least 16 weeks to maximal response and then switch to olaparib. Another option is maintenance therapy with the PARP inhibitor rucaparib.

The decision to continue treatment with chemotherapy or to switch over to a maintenance PARP inhibitor should be based on a discussion with the clinician and patient, emphasizing the following issues [1]: whether a maximum response and plateau in response to chemotherapy has been achieved; the level of cumulative toxicities associated with chemotherapy; patient preference and goals of care; convenience and cost; and clinical evidence, including the lack of an overall survival benefit demonstrated in the POLO trial.

For patients with somatic (ie, non-germline) mutations in BRCA or PALB2, the benefit of PARP inhibitors is still not known and is under investigation. (See 'Pathogenic variant status unknown' below.)

Basic biology — Pancreatic cancer occurs at increased frequency among carriers of certain pathogenic germline genetic alterations including those in the BRCA1, BRCA2, and PALB2 genes. The NCCN [17] and ASCO [16] recommend genetic counseling and consideration of germline genetic testing for all patients with a diagnosis of pancreatic ductal adenocarcinoma. (See "Familial risk factors for pancreatic cancer and screening of high-risk patients", section on 'Genetic predisposition syndromes' and "Familial risk factors for pancreatic cancer and screening of high-risk patients", section on 'Referral for genetic evaluation'.)

Mutations in these genes result in defective DNA damage repair. Each cell is equipped with DNA damage response mechanisms that guard the genome against mutational insults. Double-strand breaks are a particularly hazardous form of DNA damage, and they are repaired by two major repair pathways: error-free (high-fidelity) homologous recombination and nonhomologous (low fidelity) end-joining (NHEJ) [19,20]. HRR deficiency was initially described in cancers that arose in the setting of germline mutations in the tumor suppressors BRCA1 and BRCA2. It is now understood that genetic and epigenetic inactivation of other genes can lead to HRR deficiency in sporadic cancers as well.

Platinum-based chemotherapy — HRR is required for the repair of double-strand breaks that are generated during DNA interstrand crosslinking, which occurs during treatment with platinum-type chemotherapy drugs. For this reason, cells that have HRR deficiency may be particularly sensitive to platinum-containing chemotherapy. Pathogenic somatic or germline HRR mutations that predict better outcomes with platinum rather than non-platinum-containing chemotherapy include those arising in BRCA1/2, PALB2, ATM, BAP1, BARD1, BLM, BRIP1, CHEK2, FAM175A, FANCA, FANCC, NBN, RAD50, RAD51, RAD51C, and RTEL1 [21].

There are emerging data from small case series, retrospective analyses, and at least two randomized trials that support benefit for initial platinum-containing regimens in individuals with HRR deficiency, particularly biallelic and germline BRCA pathogenic variant carriers [22-29]. The benefits of this strategy can be illustrated by the following:

●In the POLO trial, 154 patients with germline BRCA-mutated metastatic pancreatic cancer who had not progressed during at least 16 weeks of first-line platinum-based therapy (a variant of FOLFIRINOX was used in 129) were randomly assigned to placebo or maintenance olaparib alone [30]. The objective response rate in the maintenance arm was 23 versus 10 percent with placebo; the median duration of response was 24.9 months with maintenance olaparib versus 3.7 months with placebo. Notably, 10 percent of patients who received placebo after an initial period of chemotherapy were still progression free after two years. (See 'PARP inhibitor maintenance therapy' below.)

●Unprecedented response rates to cisplatin plus gemcitabine were noted in a phase II trial in which 50 patients with germline BRCA or PALB2-mutated locally advanced or metastatic pancreatic cancer were randomly assigned to 21-day cycles of cisplatin (25 mg/m² on days 3 and 10) plus gemcitabine (600 mg/m² also on days 3 and 10) alone or with veliparib (80 mg orally on days 1 through 12 every 21 days) [31]. The addition of veliparib to cisplatin plus gemcitabine did not significantly improve objective response rate (74 versus 65 percent), disease control rate (100 versus 78 percent), median PFS (10.1 versus 9.7 months), or overall survival (15.5 versus 16.4 months). The two-and three-year survival rates for the entire cohort were 31 and 18 percent, respectively. Grade 3 or 4 hematologic toxicities for arm A versus arm B were 48 versus 30 percent for neutropenia, 55 versus 9 percent for thrombocytopenia, and 52 versus 35 percent for anemia. There was a similar rate of grade 3 or 4 nonhematologic toxicities in both arms.

Neither of the randomized trials were randomized trials of initial platinum- versus non-platinum-containing chemotherapy. Nonetheless, the high response rates observed with platinum-containing regimens (especially cisplatin plus gemcitabine) in individuals with germline BRCA mutations, and the long-term survival seen in both of the trials are unprecedented. Despite the lack of prospective comparative studies, updated consensus-based guidelines from the NCCN suggest consideration of a first-line platinum-based regimen (gemcitabine plus cisplatin or FOLFIRINOX/modified FOLFIRINOX) for advanced-stage pancreatic cancer that arises in the setting of a known BRCA or PALB2 pathogenic variant [17]. They endorse this approach for patients with either germline or somatic pathogenic variants.

PARP inhibitor maintenance therapy — After at least 16 weeks of initial platinum-based chemotherapy, for those without disease progression, we suggest discontinuation of chemotherapy and initiation of maintenance therapy using the poly(ADP-ribose) polymerase (PARP) inhibitor olaparib for patients with a germline BRCA pathogenic variant, based on findings from the POLO trial. We also suggest this approach for patients with a germline PALB2 pathogenic variant. The optimal timing of olaparib in this setting is not established. The POLO trial started the drug in patients without disease progression after 16 weeks of initial chemotherapy. Another option is to treat with chemotherapy for at least 16 weeks to maximal response and then switch to olaparib.

PARP activity is essential for the repair of single-strand DNA breaks via the base excision repair pathway, which is a primary repair pathway for DNA single-strand breaks in cells with HRR deficiency [32,33]. Cancer cells with deleterious mutations in BRCA1 or BRCA2 have defective HRR function, and the unrepaired DNA breaks that result after treatment with PARP inhibitors eventually lead to cancer cell death [34,35]. This process is referred to as "synthetic lethality," in which two conditions that would independently not cause cell death, when present in combination, cause lethal injury to the cancer cell [36,37].

Benefit — There is accumulating evidence for the benefits of PARP inhibitors in BRCA-associated breast and ovarian cancers. (See "Medical treatment for relapsed epithelial ovarian, fallopian tube, or peritoneal cancer: Platinum-sensitive disease", section on 'PARP inhibition in BRCA carriers' and "ER/PR negative, HER2-negative (triple-negative) breast cancer", section on 'Patients with previous exposure to chemotherapy'.)

In advanced pancreatic cancer, there appears to be no benefit to administering a PARP inhibitor concurrent with systemic chemotherapy [31]. However, accumulating data suggest benefit for maintenance therapy with a PARP inhibitor in patients with advanced pancreatic cancer and a germline BRCA and PALB2 pathogenic variant after an initial period of platinum-containing chemotherapy [30,38-43]. (See 'Platinum-based chemotherapy' above.)

The following data are available:

●A potential role for maintenance therapy using a PARP inhibitor after an initial phase of first-line platinum-containing chemotherapy in BRCA pathogenic variant carriers was suggested in the POLO trial, in which 154 patients with germline BRCA-mutated metastatic pancreatic cancer that had not progressed during at least 16 weeks of first-line platinum-based chemotherapy were randomly assigned to olaparib (300 mg twice daily) or placebo [30]. Maintenance olaparib was associated with a significant improvement in median PFS, the primary endpoint (7.4 versus 3.8 months, hazard ratio [HR] for disease progression or death 0.53, 95% CI 0.35-0.82), and twice as many patients were still progression free at two years (22 versus 9.6 percent); overall survival was similarly high in both groups (median 18.9 versus 18.1 months). Notably, crossover to olaparib was not allowed in the placebo group. The incidence of grade 3 or higher adverse events was higher with olaparib (40 versus 23 percent with placebo), with the most common grade 3 or worse toxic effects being anemia, fatigue, and anorexia. Health-related quality of life was preserved with maintenance olaparib, and there were no clinically meaningful differences compared with placebo [42]. In a later analysis, there was no significant improvement in median overall survival with PARP inhibitor maintenance therapy (19 versus 19.2 months, HR 0.83, 95% CI 0.56-1.22), but the survival curves separated after 24 months, and almost twice as many patients in this group remained alive at three years (34 versus 18 percent) [44].

Based on the initial report, olaparib was approved by the US Food and Drug Administration in December 2019 for first-line maintenance therapy in patients with germline BRCA gene pathogenic variants whose cancer has spread beyond the pancreas and whose tumors did not worsen after chemotherapy of at least 16 weeks [45].

●The safety and efficacy of maintenance therapy with a different PARP inhibitor (rucaparib) in a broader group of patients with HRR deficiency was shown in a phase II trial in which 46 patients with advanced pancreatic cancer and a germline or somatic pathogenic variant in BRCA or a germline pathogenic variant in PALB2 who were progression-free after at least 16 weeks of platinum-based chemotherapy without disease progression all received rucaparib 600 mg orally twice daily until disease progression [43]. The PFS rate at six months was 60 percent, median PFS was 13.1 months, and median overall survival was 23.5 months; these values all compare favorably with those seen in the maintenance olaparib arm of the POLO trial. Objective responses were noted in 41 percent of those with germline BRCA2 (11 of 27 patients), and in 50 percent (3 of 6 patients) of those with germline PALB2 pathogenic variants, and in 1 of 2 patients with a somatic BRCA pathogenic variant, but the small number of patients precludes assessment of this strategy in those with germline PALB2 or somatic BRCA pathogenic variants.

●Additional information is available from the phase II trial of gemcitabine plus cisplatin with or without veliparib, described above [31] (see 'Platinum-based chemotherapy' above). In an exploratory analysis of a subset of 10 patients who received four or more months of platinum-based therapy and then continued or received a PARP inhibitor as an immediate next line of therapy in the absence of disease progression (eight with stage IV disease), the median overall survival was 23.4 months.

Although responses to single-agent PARP inhibitors for patients with somatic BRCA pathogenic variants have been reported [41,43], the benefit of PARP inhibitors as maintenance therapy for those with somatic (ie, non-germline) pathogenic variants in BRCA or PALB2 is still not known and we consider this to be an investigational approach.

Ongoing trials are testing the benefit of adding immune checkpoint inhibitor immunotherapy to the maintenance PARP inhibitor [46], but whether this approach is superior to a PARP inhibitor alone will require randomized trials. Such a trial (NCT 04548752) is underway.

Pathogenic variant status unknown — It can take four to eight weeks for molecular analysis of either the tumor tissue or the germline to be resulted. For most patients with an unknown HRR deficiency pathogenic variant status, we do not recommend waiting until the germline or somatic pathogenic variant status is known, given the rapidity of progression in most patients with newly diagnosed metastatic pancreatic cancer. Therefore, we treat these patients like HRR deficiency pathogenic variant carriers until results of genetic testing are available (algorithm 1).

For patients with an excellent PS (ECOG 0 or 1 (table 1)), FOLFIRINOX has become our standard regimen by default. Often, we start FOLFOX in those patients who have a borderline performance status, then add irinotecan when we know that the patient can tolerate it. Gemcitabine plus cisplatin is another option. (See 'Overview of our approach' above.)

For patients with a borderline PS, the decision as to whether to start gemcitabine plus nabpaclitaxel or a platinum-containing regimen (gemcitabine/cisplatin or FOLFOX) is more difficult. There are phase III data confirming a survival benefit for gemcitabine plus nabpaclitaxel compared with gemcitabine alone in a BRCA/PALB2-unknown setting. Yet, the benefit of platinum-containing regimens in HRR pathogenic variant carriers (particularly those with a germline pathogenic variant) is clear. If HRR deficiency status is unknown, it is not wrong to start with gemcitabine plus nabpaclitaxel, and clinicians should feel comfortable with either choice. However, once pathogenic variant status is known, it is important that patients with HRR deficiency be offered a platinum-containing regimen and a PARP inhibitor at some point in their treatment.

Non-pathogenic variant carriers with ductal adenocarcinoma

Choice of regimen

Patients with deficient mismatch repair — Patients with dMMR/MSI-H tumors respond to checkpoint inhibitor immunotherapy; all of the available data in advanced pancreatic cancer are in the setting of second-line therapy or beyond. (See "Second-line systemic therapy for advanced exocrine pancreatic cancer", section on 'Deficient mismatch repair or high tumor mutational burden'.)

However, despite the lack of phase II studies for first-line treatment, we believe that it is reasonable to consider first-line checkpoint inhibitor immunotherapy in this rare group of individuals. If first-line chemotherapy is chosen, is imperative that these patients be monitored closely and experience the potential benefits of checkpoint inhibitor immunotherapy in the event that chemotherapy is not effective, especially given the tendency of pancreatic cancer patients to decompensate quickly (eg, from biliary or duodenal obstruction).

Other patients with a good ECOG PS who are able to tolerate an intensive approach — Outside of a clinical research protocol, for most patients with metastatic pancreatic ductal adenocarcinoma who have a good ECOG performance status (PS; 0 or 1 (table 1)), a favorable comorbidity profile (table 6)), and a serum total bilirubin level that is <1.5 times ULN and who are otherwise able to tolerate an intensive approach, we suggest FOLFIRINOX (table 2) or dose-modified FOLFIRINOX (table 3) rather than gemcitabine or a gemcitabine-based doublet. Patients receiving FOLFIRINOX should also have a patient preference and a support system for aggressive medical therapy, and access to chemotherapy port and infusion pump management services [1]. (See 'FOLFOX and FOLFIRINOX' below and "Treatment protocols for pancreatic cancer".)

Gemcitabine plus nabpaclitaxel (table 6) represents an acceptable and potentially less toxic alternative to FOLFIRINOX as long as patients have a relatively favorable but still adequate comorbidity profile (table 7), a serum bilirubin level that is <1.5 times ULN, and patient preference and a support system for aggressive medical therapy [1]. (See 'Gemcitabine plus nabpaclitaxel' below.)

For patients with serum bilirubin >1.5 times ULN despite placement of a stent, we prefer FOLFOX (table 5) over a gemcitabine-containing regimen as gemcitabine is associated with greater toxicity in those with hepatic impairment. (See 'FOLFOX and FOLFIRINOX' below.)

Other patients with ECOG PS 2 or a comorbidity profile that precludes intensive therapy — Outside of a clinical research protocol, for patients with an Eastern Cooperative Oncology Group (ECOG) PS other than 0 to 1 who have a serum bilirubin that is <1.5 times ULN and less than favorable but still adequate comorbidity (table 7) and who wish to pursue disease-directed therapy we suggest monotherapy with gemcitabine alone (table 8), gemcitabine plus capecitabine (table 9), or gemcitabine plus S-1, where available.

Where available, S-1 monotherapy represents a reasonable alternative to gemcitabine monotherapy for patients who prefer the convenience of an oral regimen. (See 'Fluoropyrimidines' below.)

Highly selected patients with an ECOG PS of 2 that is due to a heavy tumor burden could be considered for gemcitabine plus nabpaclitaxel because of its higher objective response rate [47].

For patients with serum bilirubin >1.5 times ULN despite placement of a stent, we prefer FOLFOX over a gemcitabine-containing regimen as gemcitabine is hepatically metabolized.

Acinar cell carcinomas — Acinar cell carcinomas (ACCs) are rare malignant neoplasms that constitute approximately 1 percent of all malignant nonendocrine pancreatic tumors. Histologically, there is clear evidence of acinar cell differentiation, which can be identified immunohistochemically by staining for trypsin, chymotrypsin, elastase, or lipase or can be ultrastructurally identified by the presence of zymogen granules. (See "Pathology of exocrine pancreatic neoplasms", section on 'Acinar cell carcinoma'.)

The overall prognosis for patients with ACC is better than that for patients with ductal adenocarcinoma [48-51]. In one large series, five-year survival for those with stage IV disease for ACC relative to pancreatic ductal adenocarcinoma was 17 versus 3 percent [50].

Very limited data exist to guide treatment for patients with advanced ACC. Published single-case reports and small retrospective series suggest higher response rates to fluoropyrimidines, particularly for combination regimens that include oxaliplatin, as compared with gemcitabine-based regimens [52-61]. For most patients who are able to tolerate it, we would use an oxaliplatin-based regimen, such as FOLFOX or FOLFIRINOX.

Patients with ECOG PS 3 or poorly controlled comorbidity — Regardless of histology or BRCA/PALB2 pathogenic variant status, patients with a European Cooperative Oncology Group (ECOG) PS ≥3 or poorly controlled comorbid conditions should only be offered systemic chemotherapy on an individualized, case-by-case basis; supportive care should be emphasized [1].

EFFICACY OF INDIVIDUAL REGIMENS

FOLFOX and FOLFIRINOX — Modern FU-based combination regimens that combine LV-modulated FU with oxaliplatin with or without irinotecan (agents that are highly active for metastatic colorectal cancer) have been associated with higher response rates than single-agent regimens [62-67], and one of these regimens, LV plus short-term infusional FU plus oxaliplatin and irinotecan (FOLFIRINOX), has been shown to improve survival over gemcitabine alone.

The superiority of FOLFIRINOX over gemcitabine monotherapy was established in the phase III ACCORD 11 trial, in which a total of 342 patients with chemotherapy-naïve metastatic pancreatic cancer, a performance status (PS) of 0 or 1 (table 1), and a serum bilirubin <1.5 times ULN were randomly assigned to gemcitabine alone (1000 mg/m² weekly for seven weeks, one week of rest, then weekly for three of every four weeks) versus FOLFIRINOX [65]. The classic FOLFIRINOX regimen is outlined in the table (table 2). (See "Treatment protocols for pancreatic cancer".)

The trial was stopped after enrolling only 250 patients when a preplanned interim analysis demonstrated that the primary trial endpoint (improved overall survival) had been met. The objective response rate was significantly higher with FOLFIRINOX (32 versus 9 percent), as was median PFS (6.4 versus 3.3 months) and overall survival (11.1 versus 6.8 months).

Treatment-related toxicity was also significantly worse with FOLFIRINOX, including grade 3 or 4 neutropenia (46 versus 21 percent), febrile neutropenia (5.4 versus 1.2 percent), thrombocytopenia (9.1 versus 3.6 percent), sensory neuropathy (9 versus 0 percent), vomiting (15 versus 8 percent), fatigue (23 versus 18 percent), and diarrhea (13 versus 2 percent). However, despite greater toxicity as compared with gemcitabine alone, FOLFIRINOX significantly improved global health status compared with gemcitabine alone [68]. In addition, the time to deterioration in quality of life by ≥20 points was significantly longer for FOLFIRINOX compared with gemcitabine for global health status; physical, role, social, and cognitive functioning; and six-symptom domains (fatigue, nausea/vomiting, pain, dyspnea, anorexia, and constipation).

There was no information on rates of infection, including cholangitis-complicating treatment. However, only approximately one-third of enrolled patients had tumors in the head of the pancreas, and thus, the number of patients who had biliary stents in place was limited compared with other populations in which pancreatic head lesions predominate.

The contribution of irinotecan to the efficacy of the FOLFIRINOX regimen is unclear. Similar response rates have been reported in a phase II study of 30 patients with locally advanced or metastatic pancreatic cancer who were treated with LV plus short-term infusional FU and oxaliplatin (FOLFOX) [64]. There were eight partial responses (27 percent), but the median PFS and overall survival durations were only 4 and 7.5 months, respectively.

Nevertheless, these data established FOLFIRINOX as a preferred regimen for initial systemic treatment of patients with metastatic pancreatic cancer who have a good PS and a normal serum bilirubin level.

A modified FOLFIRINOX regimen has been described (table 3) that was associated with similar outcomes to classic FOLFIRINOX (response rate 38 percent, median overall survival 11.2 months) in a single-arm phase II study, but with a modestly more favorable toxicity profile (grade 3 or 4 neutropenia in 48 percent, febrile neutropenia in 9 percent, diarrhea in 10 percent, fatigue in 6 percent, sensory neuropathy in 6 percent) [66].

Oxaliplatin "stop and go" strategy — The main toxicities with FOLFIRINOX are hematologic and a cumulative sensory peripheral neuropathy, which can be dose limiting. At least in the setting of metastatic colorectal cancer, attempts to mitigate or delay the onset of oxaliplatin-induced neuropathy have employed periods of "maintenance" oxaliplatin-free chemotherapy ("stop and go" strategy). (See "Systemic therapy for metastatic colorectal cancer: General principles", section on 'Continuous versus intermittent therapy'.)

The randomized phase II PANOPTIMOX-PRODIGE 35 trial directly compared six months of FOLFIRINOX versus FOLFIRINOX for four months followed by LV-modulated FU maintenance treatment until progression in 275 patients with previously untreated metastatic pancreatic cancer [62]. In the maintenance arm, transition to FU/LV was allowed for patients with an objective response or stable disease, and FOLFIRINOX was reintroduced at the time of disease progression. The primary endpoint was six-month PFS rate. Six-month PFS rates were similar with maintenance versus ongoing treatment (43 versus 47 percent), as was median overall survival (11.2 versus 10.1 months). The median duration of FU/LV maintenance therapy was 4.5 months. Unexpectedly, rates of severe (grade 3 or 4) neurotoxicity were actually higher in the maintenance arm (19.8 versus 10.2 percent), although it occurred later. This finding was attributed to a higher cumulative oxaliplatin dose in this group overall.

While a phase III study is needed to confirm these results, they support the discontinuation of oxaliplatin in patients who develop neuropathy after four months of treatment with standard FOLFIRINOX. The exact type and role of maintenance chemotherapy for patients who have stable disease after four months of FOLFIRINOX is unclear.

For most patients, we would still continue with irinotecan plus short-term infusional FU and LV (FOLFIRI) in this situation. Switching to a different regimen at the time of disease progression rather than reintroducing oxaliplatin is one option, particularly if neuropathy persists, although in the absence of persistent neuropathy, reintroducing a platinum agent is also a reasonable option.

Gemcitabine combinations — Gemcitabine has been combined with many other active cytotoxic agents, including nabpaclitaxel, FU, cisplatin, docetaxel, oxaliplatin, and irinotecan.

Despite the large number of randomized trials that have been conducted, very few have demonstrated a benefit for a gemcitabine combination compared with gemcitabine alone; superiority for combined therapy is best documented for nabpaclitaxel plus gemcitabine. However, there are no trials directly comparing first-line gemcitabine plus nabpaclitaxel versus FOLFIRINOX. (See 'FOLFOX and FOLFIRINOX' above.)

In general, compared with single-agent gemcitabine, gemcitabine-based combination regimens are associated with higher response rates, but a significant survival benefit has only been conclusively shown for gemcitabine plus nabpaclitaxel [69,70]. Combination regimens are also more toxic, with higher rates of nausea, diarrhea, neutropenia, and thrombocytopenia.

Gemcitabine plus nabpaclitaxel — Activity for gemcitabine in combination with nanoparticle albumin-bound paclitaxel (nabpaclitaxel) was initially suggested in a phase I/II study conducted in 67 patients with previously untreated metastatic pancreatic cancer [71]. Among the 44 patients treated at the maximum tolerated dose (nabpaclitaxel 125 mg/m² followed by gemcitabine 1000 mg/m² on days 1, 8, and 15 of every 28-day cycle), the response rate was 48 percent, and median survival was 12.2 months. At this dose, grade 3 or 4 toxicities included fatigue in 27 percent, neuropathy in 20 percent, and neutropenia in 49 percent.

Superiority for the combination of nabpaclitaxel (125 mg/m²) followed by gemcitabine (1000 mg/m²), each given on days 1, 8, and 15 every 28 days, over gemcitabine alone (1000 mg/m² weekly for seven weeks, then on days 1, 8, and 15 every four weeks) was shown in the multinational MPACT trial of 861 patients with previously untreated metastatic pancreatic adenocarcinoma [69]. Combined therapy was associated with a significantly higher objective response rate (23 versus 7 percent) and significantly longer median overall survival (8.5 versus 6.7 months) and PFS (5.5 versus 3.7 months). Grade 3 or 4 adverse events that were seen more often with combined therapy included neutropenia (38 versus 27 percent), febrile neutropenia (3 versus 1 percent), fatigue (17 versus 7 percent), diarrhea (6 versus 1 percent), and neuropathy (17 versus 1 percent). Longer follow-up identified some long-term (greater than three years) survivors in the nabpaclitaxel/gemcitabine arm [72].

In September 2013, nabpaclitaxel in combination with gemcitabine was approved in the United States for the first-line treatment of patients with metastatic adenocarcinoma of the pancreas.

A modification in the regimen has been proposed (gemcitabine 1000 mg/m² and nabpaclitaxel 125 mg/m² every two weeks) that was associated with a median overall survival of 10 months and less severe (grade 3 or 4) neutropenia (19 percent) and peripheral neuropathy (2 percent) than were seen with the original regimen [73]. However, a less favorable toxicity profile was noted for this biweekly regimen as compared with standard nabpaclitaxel plus gemcitabine given weekly for three of four weeks in a combined phase I/II trial conducted in patients with advanced pancreatic cancer and an ECOG PS of 2 [47].

Alternating gemcitabine/nabpaclitaxel with FOLFOX — The SEQUENCE trial evaluated whether survival could be improved by alternating courses of gemcitabine plus nabpaclitaxel with an oxaliplatin-containing regimen [74]. A total of 157 patients with newly diagnosed metastatic pancreatic cancer were randomly assigned to standard gemcitabine plus nabpaclitaxel on days 1, 8, and 15 of each four week cycle (table 6), or to the same regimen and doses of gemcitabine and nabpaclitaxel followed by a course of modified FOLFOX6 (table 5) on day 29, with cycles repeated every six weeks. In a preliminary report presented at the 2022 annual ASCO meeting, sequential treatment was associated with a significantly higher median (13.2 versus 9.7 months, HR 0.676, 95% CI 0.438-0.937) as well as 12-month (55 versus 35 percent) and 24-month (22 versus 8 percent) overall survival, but also higher rates of treatment-related toxicity, including grade ≥3 neutropenia (46 versus 24 percent) and thrombocytopenia (24 versus 8 percent), and there were two treatment-related deaths in this group as well (versus none in the control group). Whether these results are better than could be achieved with sequential administration of gemcitabine/nabpaclitaxel following by an oxaliplatin containing regimen at the time of progression (or vice versa) is not established.

Other combinations

●Gemcitabine plus FU – Gemcitabine has been combined with both bolus as well as infusional FU. Although promising results have been suggested in phase II trials, at least three phase III trials comparing gemcitabine plus FU, with or without LV modulation, have failed to demonstrate any advantage over gemcitabine alone [75-77].

●Gemcitabine plus capecitabine – Early uncontrolled trials suggested a greater degree of efficacy when gemcitabine was combined with capecitabine than with bolus FU [78,79].

Benefit for combined therapy over gemcitabine alone could not be confirmed in two subsequent multinational phase III trials:

•In a trial in which 319 patients with locally advanced or metastatic pancreatic cancer were randomly assigned to gemcitabine with or without oral capecitabine, median survival was not significantly improved by the addition of capecitabine (8.4 versus 7.3 months) [80].

•A second European trial comparing gemcitabine alone versus gemcitabine (1000 mg/m² weekly for three of every four weeks) plus capecitabine (830 mg/m² twice daily for 21 days of every 28-day cycle) in 533 patients with locally advanced or metastatic pancreatic cancer showed a significantly higher response rate (19 versus 12 percent) but no significant overall survival benefit from the addition of capecitabine (median overall survival 7.1 versus 6.2 months, p = 0.08) [81]. Combination therapy resulted in more grade 3 or 4 neutropenia (35 versus 22 percent) but not febrile neutropenia, and there was more hand-foot syndrome as well, but only 4 percent had grade 3 or 4.

•These same investigators performed a meta-analysis of the two phase III trials [80,81] and a third randomized phase II trial of gemcitabine with or without capecitabine [79]. When all three trials were analyzed together, there was a statistically significant survival benefit favoring combined therapy with capecitabine plus gemcitabine (HR 0.86, 95% CI 0.75-0.98). Despite the fact that the schedules and dose intensities of gemcitabine and capecitabine in the three trials were completely different, there appeared to be no intertrial heterogeneity. The authors concluded that the combination of gemcitabine plus capecitabine should be considered a standard first-line option for locally advanced and metastatic pancreatic cancer.

●Gemcitabine plus S-1 — Where available, gemcitabine plus S-1 is a reasonable treatment alternative to gemcitabine alone, but the main benefits are higher objective response rates and longer time to progression; combination treatment is also associated with greater treatment-related toxicity, and the survival advantage, if there is one, appears to be limited to those with locally advanced, and not metastatic, disease.

At least four randomized trials have directly compared gemcitabine plus S-1 with gemcitabine alone; all have shown that combination therapy significantly improves PFS and objective response rate; three of the four do not show a survival advantage to combined therapy [82-86].

Two meta-analyses have been conducted and have come to disparate conclusions:

•A meta-analysis of these four randomized trials (plus a retrospective report [87]) concluded that there was benefit to combined therapy in terms of overall response rate (risk ratio [RR] 2.52, 95% CI 1.85-3.42) and one-year survival rate (43 versus 31 percent, RR 1.62, 95% CI 1.12-2.33) [88].

•On the other hand, a meta-analysis of three of the randomized trials [82,85,86] concluded that there was no significant survival advantage for combined therapy over gemcitabine alone for metastatic pancreatic cancer (577 patients, median survival 9.43 versus 8.02 months, HR for death 0.872, 95% CI 0.738-1.032) and that combined therapy was associated with significantly higher rates of grade 3 or 4 rash, vomiting, and neutropenia [89]. By contrast, there did appear to be a survival benefit for patients with locally advanced disease (193 patents, median overall survival 16.41 versus 11.83 months, HR 0.708, 95% CI 0.527-0.951), with significantly higher rates of rash and thrombocytopenia with combined therapy.

Single-agent chemotherapy — Single-agent therapy has been the mainstay of treatment for metastatic pancreatic cancer. However, among active single agents, none has been consistently associated with objective response rates above 10 percent or median survival durations above six to seven months.

Gemcitabine — For patients with metastatic pancreatic cancer who are not candidates for a more intensive first-line chemotherapy regimen (eg, those with a PS other than 0 or 1 (table 10)), we suggest monotherapy with gemcitabine alone.

Gemcitabine is a nucleoside analog with structural similarity to cytarabine. Initial studies suggested a low objective response rate (6 to 11 percent) in chemotherapy-naïve patients administered single-agent gemcitabine (800 mg/m² intravenous [IV] weekly for three of every four weeks) [90,91].

However, clinical benefit has been observed in patients who lack an objective response. In a pivotal phase II study of gemcitabine (table 8), "clinical benefit" was defined as an improvement in pain, PS, or weight without a deterioration in any other factor [92]. Although the objective response rate for patients with measurable disease was only 11 percent, a clinical benefit was observed in 27 percent.

Based on these data, the clinical benefit response was used as the primary endpoint in a follow-up trial that included 126 previously untreated patients with locally advanced or metastatic pancreatic cancer who were randomly assigned to FU (600 mg/m² weekly) or gemcitabine (using the same regimen as was tested in the phase II trial) [93].

Although there were no confirmed objective responses in either group, gemcitabine was associated with a significantly better clinical benefit response (24 versus 5 percent), median overall survival (5.6 versus 4.4 months), and one-year survival (18 versus 2 percent), although grade 3 or 4 neutropenia was more frequent (23 versus 5 percent). This study was criticized for its failure to use a prospectively validated quality of life instrument and the failure to blind treatment assignment to those assessing clinical benefit. In fact, a later analysis of patients enrolled in a randomized trial of gemcitabine with or without bevacizumab concluded that response to gemcitabine treatment in advanced pancreatic cancer was not associated with an appreciable improvement in health-related quality of life [94].

Nevertheless, on the basis of significant improvements in clinical benefit and survival, gemcitabine was approved for first-line therapy of metastatic pancreatic cancer.

Fluoropyrimidines

●Fluorouracil – FU has been extensively studied since the 1950s, and objective response rates vary widely from 0 to 67 percent [95]. However, contemporary studies of LV-modulated FU in patients with advanced disease suggest a very low response rate (0 to 9 percent) using infusional as well as bolus administration schedules and a median survival that ranges from 2.5 to 6 months [96-98].

●Capecitabine — Capecitabine is a rationally designed oral fluoropyrimidine that is reliably absorbed, intact, through the intestinal wall and then converted into FU in three sequential enzymatic reactions. The final requisite enzyme, thymidine phosphorylase, is present at consistently higher levels in tumor (rather than normal) tissue, thereby providing the basis for enhanced selectivity and better tolerability [99].

The efficacy of capecitabine monotherapy (1250 mg/m² orally, twice daily for 14 of every 21 days) was shown in a study of 42 patients with chemotherapy-naïve advanced pancreatic cancer [100]. Although only three patients (7 percent) had an objective response, the clinical benefit response rate was 24 percent. Treatment was well tolerated, with grade 3 hand-foot syndrome (also called acral erythema or palmar-plantar erythrodysesthesia), nausea, and diarrhea in 17, 10, and 12 percent, respectively. (See "Infusion reactions to systemic chemotherapy".)

●S-1 – Where available, S-1 monotherapy represents a reasonable alternative to gemcitabine monotherapy for patients who are not candidates for a more intensive first-line regimen and who prefer the convenience of an oral regimen.

S-1 is an oral fluoropyrimidine that includes three different agents: ftorafur (tegafur), gimeracil (5-chloro-2,4 dihydropyridine; a potent inhibitor of dihydropyrimidine dehydrogenase), and oteracil (potassium oxonate; which inhibits phosphorylation of intestinal FU, thought to be responsible for treatment-related diarrhea). It is approved in Japan for adjuvant treatment of gastric cancer and in Europe for treatment of advanced gastric cancer; it is not available in the United States.

In the phase III GEST trial, S-1 (80 to 120 mg daily on days 1 to 28 with cycles repeated every 42 days) was directly compared with gemcitabine monotherapy (1000 mg/m² on days 1, 8, and 15 and repeated every 28 days) or S-1 plus gemcitabine in 834 patients with locally advanced (n = 202) or metastatic (n = 630) pancreatic cancer and an ECOG PS of 0 or 1 (table 1) [82]. For the primary endpoint of overall survival, results with S-1 were noninferior to those with gemcitabine monotherapy (median 9.7 versus 8.8 months), and the objective response rate was significantly higher with S-1 (21 versus 13 percent). Rates of grade 3 or 4 hematologic toxicity were higher with gemcitabine, but rates of severe nonhematologic toxicities were similarly low in both groups.

RESPONSE ASSESSMENT — Response assessment during treatment typically relies on periodic cross-sectional radiographic imaging with computed tomography or magnetic resonance imaging (typically every two to three months after treatment initiation). If they were initially elevated, serum levels of the tumor marker carbohydrate antigen 19-9 (CA 19-9) may be measured at the start of treatment and then every one to three months during therapy. Suspected disease progression based on rising CA 19-9 levels should be confirmed radiographically. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Role of tumor markers'.)

●Radiographic assessment — Traditional two-dimensional or unidimensional tumor measurements to determine objective response to therapy are often inadequate when evaluating the primary site in the pancreas due to the characteristic desmoplastic reaction and inflammatory response that develop there [101]. Despite this, single-agent chemotherapy trials in pancreatic cancer have commonly used objective response as a primary endpoint. This endpoint is likely more accurate for assessment of the liver and other metastatic disease sites than it is for assessing the primary tumor. The first response assessment using cross-sectional imaging should be offered at two to three months after initiation of therapy [1,102].

●Tumor marker response – In many studies of patients receiving chemotherapy for advanced pancreatic cancer, there has been a correlation between duration of patient survival and decline in levels of the tumor marker CA 19-9 [103-109]. The magnitude of marker reduction that best predicts improved outcomes (any reduction, >50 percent reduction, >75 percent reduction, >89 percent reduction) is not clear.

Furthermore, whether decreases in serum CA 19-9 can act as a reliable surrogate for survival in patients receiving chemotherapy for advanced pancreatic cancer has been called into question by others. As an example, in a series of 175 patients who were assessable for tumor marker response during treatment with gemcitabine with or without capecitabine, an early 50 percent or greater decrease in CA 19-9 levels after the first eight weeks of therapy was not associated with a significant prolongation of survival compared with lesser degrees of CA 19-9 response (median 10.1 versus 8.6 months) [110].

The short timeframe of CA 19-9 assessment (after eight weeks of therapy) may have contributed to these negative results. Among patients who enter complete remission from chemotherapy, tumor marker normalization is a gradual process, with an estimated half-life of 15 to 33 days [105].

CA 19-9 requires the presence of the Lewis blood group antigen (a glycosyl transferase) to be expressed. Among individuals with a Lewis-negative phenotype (an estimated 5 to 10 percent of the population), CA 19-9 levels are not a useful tumor marker.

Serum CA 19-9 levels are not considered an optimal substitute for imaging for assessing treatment response [1]. Suspected disease progression based on rising CA 19-9 levels should be confirmed radiographically.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Pancreatic cancer".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topic (see "Patient education: Pancreatic cancer (The Basics)")

●Beyond the Basics topic (see "Patient education: Pancreatic cancer (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●General principles

•All patients should undergo a multiphasic computed tomography (CT) scan of the chest, abdomen, and pelvis to assess disease extent. Serum levels of carbohydrate antigen 19-9 (CA 19-9) should be measured at baseline and if elevated, can be used to monitor therapy. (See 'General principles' above.)

•Multidisciplinary collaboration to formulate treatment and care plans is preferred, with decision making guided by goals of care, patient preferences, symptom burden, and psychosocial issues. (See "Discussing goals of care".)

•Especially for patients with a high symptom burden, early initiation of palliative care services improves clinical and quality-of-care outcomes and may prolong survival. (See "Benefits, services, and models of subspecialty palliative care".)

•Germline genomic testing and targeted tumoral (somatic) testing should be undertaken as quickly as possible, since findings may influence the choice of initial treatment, and prompt referral for genetic counseling. (See 'Germline and targeted tumor genomic testing' above.)

●Our approach to treatment

•Enrollment in clinical trials is preferred. Outside of a clinical research protocol, our approach to first-line systemic therapy is outlined in the algorithm (algorithm 1) and summarized below. Enrollment in clinical trials is preferred. (See 'Initial treatment' above.)

•For advanced-stage pancreatic cancer that arises in the setting of a known germline or somatic pathogenic variant involving an homologous recombination repair (HRR) gene we suggest a platinum-based rather than a non-platinum-based chemotherapy regimen (Grade 2C). (See 'Patients with homologous recombination repair deficiency' above.)

-For most patients with an excellent Eastern Cooperative Oncology Group (ECOG) performance status (PS; 0 or 1 (table 1)) who are able to tolerate an intensive approach, a serum bilirubin level <1.5 times the upper limit of normal (ULN), and a preference and support system for aggressive medical therapy we suggest leucovorin (LV) plus short-term infusional fluorouracil (FU) plus oxaliplatin and irinotecan (FOLFIRINOX (table 2)) or modified FOLFIRINOX (table 3) over other platinum-containing regimens (Grade 2C). Other alternatives are gemcitabine plus cisplatin (table 4) and FOLFOX (table 5). For patients with an ECOG PS of 2, a comorbidity profile that precludes intensive therapy, or a serum bilirubin >1.5 times ULN despite stenting, we prefer FOLFOX (table 5) over FOLFIRINOX. (See 'Platinum-based chemotherapy' above.)

-For patients with a germline BRCA or PALB2 pathogenic variant who have no disease progression after at least 16 weeks of initial platinum-based chemotherapy, we suggest maintenance therapy using a poly (ADP-ribose) polymerase (PARP) inhibitor rather than continued chemotherapy (Grade 2C). Either olaparib or rucaparib are acceptable alternatives. (See 'PARP inhibitor maintenance therapy' above.).

For those with somatic (ie, non-germline) pathogenic variants in BRCA or PALB2, the benefit of PARP inhibitor maintenance therapy is unknown, and we suggest not pursuing this approach (Grade 2C).

-Initial treatment for patients with an unknown (or pending) HRR deficiency status is the same as for HRR deficiency pathogenic variant carriers until results of genetic testing are available. (See 'Pathogenic variant status unknown' above.)

•For patients with advanced adenocarcinomas who lack a pathogenic variant in one of the HRR pathway genes, the following represents our general approach (see 'Non-pathogenic variant carriers with ductal adenocarcinoma' above):

-For the rare patient with deficient mismatch repair (MMR) we suggest checkpoint inhibitor immunotherapy rather than systemic chemotherapy (Grade 2C). (See 'Patients with deficient mismatch repair' above.)

-For most patients with MMR proficient tumors who have a good ECOG PS (0 or 1 (table 1)), a favorable comorbidity profile, a serum total bilirubin level <1.5 times ULN, and a preference and support system for aggressive medical therapy we suggest FOLFIRINOX (table 2) or modified FOLFIRINOX (table 3) rather than gemcitabine monotherapy (Grade 2B). We also suggest FOLFIRINOX or modified FOLFIRINOX rather than a gemcitabine-based doublet (Grade 2C). (See 'FOLFOX and FOLFIRINOX' above and "Treatment protocols for pancreatic cancer".)

Gemcitabine plus nabpaclitaxel (table 6) is an acceptable and potentially less toxic alternative to FOLFIRINOX for patients with an ECOG PS 0 or 1, an adequate comorbidity profile, a serum bilirubin level <1.5 times ULN, and a preference and support system for aggressive medical therapy. (See 'Gemcitabine plus nabpaclitaxel' above and 'Other combinations' above.)

-For patients with serum bilirubin ≥1.5 times ULN despite stenting, we prefer FOLFOX (table 5) over a gemcitabine-containing regimen, given concerns for gemcitabine toxicity.

-For patients with an ECOG PS of 2, favorable or adequate comorbidity, and a total serum bilirubin level <1.5 times ULN we suggest monotherapy with gemcitabine alone (table 8) (Grade 2C); other options are gemcitabine plus capecitabine (table 9) or S-1 plus gemcitabine, where available. (See 'Gemcitabine' above.)

Highly selected patients with an ECOG PS 2 because of heavy tumor burden could be considered for gemcitabine plus nabpaclitaxel because of its higher response rate. Proactive dose and schedule adjustments should be pursued to minimize toxicities. (See 'Gemcitabine plus nabpaclitaxel' above.)

•For patients with advanced acinar cell carcinoma we suggest an oxaliplatin-based regimen rather than a non-oxaliplatin regimen (Grade 2C). (See 'Acinar cell carcinomas' above.)

•Patients with an ECOG PS ≥3 or poorly controlled comorbidity should only be offered chemotherapy on an individualized, case-by-case basis; supportive care should be emphasized. (See 'Patients with ECOG PS 3 or poorly controlled comorbidity' above.)

●Response assessment – Response is assessed radiographically with periodic cross-sectional imaging, typically every two to three months. If initially elevated, serum levels of the tumor marker CA 19-9 may be repeated every one to three months during therapy, but suspected disease progression based on rising CA 19-9 levels should be confirmed radiographically. (See 'Response assessment' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges David P Ryan, MD, who contributed to earlier versions of this topic review.