|A systematic review of resectability and survival after concurrent chemoradiation in primarily unresectable pancreatic cancer
|Morganti AG, Massaccesi M, La Torre G, Caravatta L, Piscopo A, Tambaro R, Sofo L, Sallustio G, Ingrosso M, Macchia G, Deodato F, Picardi V, Ippolito E, Cellini N, Valentini V
The review concluded that, after neoadjuvant radiochemotherapy, radical surgery should be considered in patients with primarily unresectable locally advanced pancreatic cancer, since complete tumour removal was achieved in almost a quarter of patients, a third of whom survived for more than three years. Due to potential limitations, mainly arising from the analysis, the authors’ conclusions should be treated with caution.
To assess the impact of preoperative combined chemoradiotherapy on resection rate and survival in patients with locally advanced pancreatic carcinoma.
PubMed, Cochrane Central Register of Controlled Trials (CENTRAL), and National Cancer Institute and Clinical Trial Registry PDQ (Physician Data Query) databases were searched, from 2000 to March 2009, for publications in any language. Search terms were reported. Abstracts of conference proceedings of the American Society of Clinical Oncology and the American Society for Therapeutic Radiology and Oncology were also searched from 2000 to 2008. Only published clinical trials were eligible for inclusion.
Clinical trials (randomised or non randomised) that included an evaluation for surgery, systematically performed after combined chemoradiotherapy, on patients with a diagnosis of pancreatic adenoma were eligible for inclusion. Pancreatic adenoma diagnosis had to be based on histological or cytological findings (investigations on body tissue or cells). Trials that included patients with locally advanced/unresectable pancreatic adenocarcinoma were also eligible for inclusion. For radiotherapy, the eligible interventions included mega-voltage external-beam radiotherapy based on standard radiotherapy doses and fractionation (total dose 45Gy or more, dose/1.8 to 2Gy daily fraction) delivered concurrently with chemotherapy. For chemotherapy, eligible interventions included both single agent and combination therapy with all cytotoxic or anti-neoplastic drugs, regardless of dose or schedule, excluding hormonal or biological therapies (e.g. interferon, somatostatin). Trial were excluded if they used rapid-fractionated radiotherapy, had low total radiation doses, did not clearly specify the treatment regime, or did not systematically evaluate surgery after chemoradiation.
The primary outcomes were resection rate using an intention-to-treat analysis and survival (including median survival time and three-year survival rate). Secondary outcomes included: survival rates of unresected patients and in the overall population; clinical response; resectability rate; rates of radically (tumour free margins) resected tumours; and perioperative mortality.
Multi-drug chemotherapy was used in most of the included studies. Five studies used 5-fluorouracil-based regimes, mostly combined with cisplatin or oxaliplatin. Four studies used gemcitabine-based combinations. One study used 5-fluorouracil and gemcitabine. Sequential chemotherapy was administered before surgery in five series. Radiotherapy regimes were also reported.
Two independent reviewers performed the selection, with any disagreements resolved by a third reviewer.
Assessment of study quality
Evidential quality was assessed using the criteria of the Scottish Intercollegiate Guidelines Network (SIGN), where level of evidence 1 was provided by meta-analyses, level 2 by case-control or cohort studies, level 3 by non-analytical studies (such as case series and case reports), and level 4 by expert opinion. Levels of evidence 1 and 2 were subgraded according to the risk of bias or confounding in the study or studies.
The authors did not report how many reviewers performed the quality assessment.
Two reviewers performed the data extraction using standardised forms, with disagreements resolved by discussion with a third reviewer. The numbers of events for each outcome were extracted in order to calculate the percentage of patients responding, resection rate (%), median survival and three-year survival (%).
Methods of synthesis
Central tendency (median and weighted average of the medians) and dispersion (range) measures were calculated for all studies and for selected groups. Differences between groups were assessed with χ2 for qualitative variables and Mann-Whitney tests for quantitative variables, where statistical significance was set at p≤0.05.
Results of the review
Thirteen relevant studies were identified (n=510 patients, range 14 to 68). Ten studies were case series, with seven prospective case series (n=245 patients) and three retrospective case series (n=152 patients); their quality was evidence level 3 for all. Three studies were prospective cohort studies (n=113 patients); their quality was evidence level 2 for all (due to substantial heterogeneity among groups of enrolled patients). The highest overall level of evidence provided for the review was 2, according to the SIGN classification.
Resection rate (13 studies)
A median resection rate of 26.5% (range 8.3 to 64.2%; weighted average 27%) was reported, with a median of 23.1% (range 8.3 to 57.1%; weighted average 22.1%) enrolled patients having resection surgery with tumour-free margins (R0). Of the operated patients, a wide range had total resection with tumour-free margins (median 87.5%, range 57.1 to 100%). Five of ten studies reported that some patients had a complete pathological response (3.0 to 8.8%); the remaining five studies reported that no patients had a complete response.
There was no significant difference in resection rate between patients treated with higher radiation doses (24.5%; dose of 50Gy or more) and those treated with lower doses (26.3%; dose of less than 50Gy). The difference in resection rate between gemcitabine-based treatment (median resection rate 25.5%; weighted average 29.3%) and gemcitabine-free treatment (median resection rate 26.5%; weighted average 23.5%) showed that, after weighting, there was a higher resection rate with gemcitabine-free treatment (12 studies). The use of induction therapy in addition to chemoradiotherapy (13 studies) increased the resection rate (median resection rate 29%; weighted average 31.3%) compared with chemoradiotherapy alone (median resection rate 22.5%; weighted average 23.3%). The significance of the latter two analyses was not reported.
The median survival in all patient groups was 13.3 months (range 9.0 to 23.6 months; weighted average 14.9 months; nine studies). In resected patients, the median survival was 23.6 months (range 16.4 to 32.3 months; weighted average 25.4) compared with 10.0 months (range 7.4 to 21.2 months; weighted average 13 months) in unresected patients (eight studies). For the three prospective cohort studies with level 2 evidence, the median survival in resected patients was 29 months (range 21 to 32.3 months; weighted average 29 months) compared with a range of 10 to 21.2 months (weighted average 17.9 months) in unresected patients.
The three-year survival of resected patients was 42.5% (range 32 to 73%; weighted average 44%).
Gemcitabine-based treatment did not significantly increase survival compared with gemcitabine-free treatment in all patients, resected patients or non-resected patients.
Perioperative mortality after chemoradiotherapy ranged from 0 to 12.5%, with most series reporting no perioperative deaths (six case series). There was no significant difference for operative mortality and morbidity between the intervention and control groups in the three prospective cohort studies.
The finding of a high proportion of resections with tumour-free margins (R0) among all the resections performed confirmed the activity of neoadjuvant radiochemotherapy. After neoadjuvant chemoradiation, complete tumour removal was achieved in almost a quarter of patients, with a third of them surviving for more than three years. This was positive evidence for the consideration of radical surgery after chemoradiotherapy in patients with unresectable pancreatic cancer without disease progression.
The review addressed a well-defined question in terms of participants, interventions, and relevant outcomes, but the study design of eligible studies was less clearly defined. Relevant databases were searched without language restrictions. Conference abstracts were also searched, but other unpublished studies were not considered. Publication bias was not assessed. Efforts to reduce error and bias were reported for study selection and data extraction, but not reported for validity assessment.
Validity assessment was based on study type, giving levels of evidence rather than assessing the quality of individual studies. The study design of the included studies was not clear and level of evidence was not taken account of in the synthesis. Relevant study details were reported, but with no details of the age and gender of the patients. The reporting of results was not very clear; the analysis may have been inappropriate. Although statistical heterogeneity was mentioned and reported to be present, no relevant data were presented. The authors provided a discussion of confounding factors related to resection; they also acknowledged that there was extreme heterogeneity in selection for resection and a lack of stated criteria for resection in most studies. The data for median survival in all patient groups differed in the text.
In view of potential limitations, mainly arising from the analysis, the authors’ conclusions should be treated with caution.
Implications of the review for practice and research
Practice: The authors suggested that if no disease progression occurs after chemoradiotherapy, patients should be re-evaluated by a skilled surgical team.
Research: The authors identified a need for prospective high-quality trials for patients with locally advanced unresectable pancreatic cancer to determine whether neoadjuvant chemoradiotherapy provides sustained benefits and to identify the best approach to down-staging tumours for resection. Homogeneous and well-defined resectability criteria should be used in these trials (specific suggestions are made). Investigations should also be made of the use of preoperative chemoradiotherapy in patients with primarily resectable pancreatic cancer, particularly in patients with borderline respectable tumours at high risk of undergoing an incomplete (R1-R2) resection.
Morganti AG, Massaccesi M, La Torre G, Caravatta L, Piscopo A, Tambaro R, Sofo L, Sallustio G, Ingrosso M, Macchia G, Deodato F, Picardi V, Ippolito E, Cellini N, Valentini V. A systematic review of resectability and survival after concurrent chemoradiation in primarily unresectable pancreatic cancer. Annals of Surgical Oncology 2010; 17(1): 194-205
Subject indexing assigned by NLM
Chemotherapy, Adjuvant; Combined Modality Therapy; Humans; Neoadjuvant Therapy; Pancreatic Neoplasms /mortality /therapy; Radiotherapy, Adjuvant; Survival Rate
Database entry date
This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.