Beyond tumor biology: can host nutritional status refine prediction of early recurrence after pancreatic cancer surgery? Authors: Takehiro Okabayashi Published: 2026 This Editorial discusses the recent study by Oh et al., published in Gland Surgery (1), which examined whether preoperative serum prealbumin can predict early recurrence after pancreatic cancer surgery. By highlighting the potential role of host nutritional status in postoperative oncological outcomes, this study provides an opportunity to reconsider recurrence risk beyond tumor-centered characteristics and to explore a more integrated host-tumor perspective in pancreatic cancer. Pancreatic ductal adenocarcinoma (PDAC) remains one of the most lethal human malignancies despite continuous advances in surgical techniques, perioperative management, systemic chemotherapy, and molecular oncology (2,3). Surgical resection offers the only chance for long-term survival; however, recurrence develops in approximately 70–80% of patients even after apparently curative resection, and recurrence within the first postoperative year is particularly associated with extremely poor survival (3,4). Consequently, identifying patients at high risk for early recurrence has become one of the major challenges in contemporary pancreatic cancer management. Current risk stratification largely depends on tumor-related factors, including pathological stage, lymph node metastasis, resection margin status, tumor differentiation, vascular invasion, serum carbohydrate antigen 19-9 (CA19-9), and more recently, circulating tumor DNA (ctDNA) and molecular profiling (3-6). Nevertheless, these variables incompletely explain the remarkable heterogeneity in postoperative recurrence patterns. Patients with apparently favorable pathological features sometimes experience rapid systemic relapse, whereas others with advanced pathological disease remain recurrence-free for years. These observations suggest that recurrence cannot be explained solely by tumor biology. In this issue of Gland Surgery, Oh and colleagues address this clinically relevant question by investigating whether preoperative serum prealbumin, a routinely available nutritional biomarker, predicts early recurrence after curative-intent pancreatic cancer surgery (1). Their study introduces an important perspective by shifting attention from tumor characteristics toward host-related biological factors that may influence oncological outcomes. A simple biomarker with potential clinical utility The authors retrospectively analyzed 57 patients who underwent curative resection for PDAC and evaluated factors associated with recurrence within 1 year after surgery (1). Although lymph node metastasis and close resection margins were associated with early recurrence in univariable analyses, only preoperative prealbumin levels below 20 mg/dL remained independently associated with early recurrence in multivariable analysis (1). Although limited by sample size, this finding raises an intriguing possibility that host nutritional status may influence postoperative oncological behavior independently of conventional pathological variables. From a practical standpoint, prealbumin possesses several attractive characteristics. Unlike genomic sequencing, transcriptomic profiling, radiomics, or liquid biopsy technologies, serum prealbumin measurement is inexpensive, universally available, and routinely performed in clinical practice. Moreover, because of its relatively short biological half-life of approximately 2 days, prealbumin is considered more sensitive than albumin for detecting acute changes in nutritional status (7). These practical advantages make prealbumin an attractive candidate biomarker for perioperative risk stratification if its prognostic value can be validated in larger prospective studies. Nutritional assessment has become increasingly important in gastrointestinal oncology. Current European Society for Clinical Nutrition and Metabolism (ESPEN) guidelines recommend routine nutritional screening and optimization before major cancer surgery because malnutrition is associated with increased postoperative complications, impaired tolerance of adjuvant chemotherapy, and worse survival (8). Cancer cachexia, characterized by systemic inflammation, metabolic dysregulation, and progressive skeletal muscle loss, further complicates perioperative management and adversely affects long-term outcomes (9). Within this context, the study by Oh et al. extends the role of nutritional assessment beyond perioperative recovery by suggesting that nutritional biomarkers may also identify patients at increased risk for early oncological failure. Looking beyond tumor biology Perhaps the greatest contribution of this study is conceptual rather than statistical. During the past decade, pancreatic cancer research has largely focused on tumor biology. Advances in molecular classification, KRAS-driven signaling, stromal interactions, immune microenvironment, and ctDNA have substantially improved our understanding of tumor aggressiveness and treatment resistance (5,6,10). Precision oncology has therefore become largely synonymous with genomic characterization of the tumor. However, recurrence represents the consequence of continuous interactions between tumor biology and host biology rather than the intrinsic properties of cancer cells alone. Successful metastatic colonization requires not only aggressive tumor clones but also a permissive host environment. Nutritional status, systemic inflammation, immune competence, skeletal muscle mass, frailty, and metabolic resilience may collectively influence postoperative recovery, tolerance of systemic therapy, immune surveillance, and ultimately the progression of occult micrometastatic disease (8,9,11). From this perspective, prealbumin should not simply be regarded as an isolated nutritional marker. Rather, it may represent an integrated surrogate reflecting multiple biological processes, including nutritional reserve, inflammatory activity, hepatic protein synthesis, and overall physiological resilience. The observation reported by Oh and colleagues therefore supports a broader conceptual shift—from a tumor-centered model of recurrence toward an integrated host-tumor model of cancer progression. This concept is increasingly supported by evidence demonstrating that systemic inflammation profoundly influences cancer progression. Chronic inflammatory activation promotes immune suppression, angiogenesis, extracellular matrix remodeling, and metastatic dissemination, while simultaneously impairing antitumor immunity (11). Because prealbumin functions as a negative acute-phase reactant, reduced concentrations may reflect not only inadequate nutritional intake but also persistent systemic inflammation. Consequently, its prognostic significance may extend beyond nutrition alone. Interpreting the findings with appropriate caution Although the findings reported by Oh and colleagues are clinically intriguing, several limitations should be carefully considered before prealbumin can be incorporated into routine clinical decision-making (1). First, the study included only 57 patients, with merely 20 early recurrence events. Such a limited number of outcome events inevitably restricts the statistical robustness of multivariable analyses and increases the possibility of model overfitting. The wide confidence intervals observed in the multivariable model further illustrate the uncertainty surrounding the estimated effect size (1). Therefore, the present findings should be regarded as hypothesis-generating rather than definitive evidence. Second, low prealbumin should not necessarily be interpreted as a causal determinant of recurrence. Instead, it is more likely to represent a surrogate biomarker reflecting multiple overlapping biological conditions, including malnutrition, cancer cachexia, systemic inflammation, hepatic dysfunction, biliary obstruction, advanced tumor burden, and impaired physiological reserve (7-9,11). Distinguishing these biological mechanisms remains an important challenge because therapeutic interventions differ substantially depending on the underlying cause. Third, prealbumin is increasingly recognized as both a nutritional biomarker and a negative acute-phase reactant. Consequently, reduced serum concentrations may reflect persistent inflammatory activation rather than nutritional depletion alone (7,11). Future studies should therefore incorporate complementary inflammatory biomarkers, including C-reactive protein (CRP), neutrophil-to-lymphocyte ratio (NLR), Glasgow Prognostic Score (GPS), modified GPS, and the Prognostic Nutritional Index (PNI), to better characterize the biological significance of low prealbumin (12,13). Treatment-related variables deserve further investigation. Modern perioperative treatment strategies increasingly involve neoadjuvant chemotherapy, perioperative FOLFIRINOX-based regimens, and individualized adjuvant treatment (14). Nutritional status may influence chemotherapy tolerance, dose intensity, treatment completion, and postoperative recovery, thereby indirectly affecting recurrence risk. Whether prealbumin independently predicts recurrence after adjustment for treatment adherence remains uncertain and should be evaluated in future prospective studies. From tumor-centered to host-tumor integrated precision oncology Despite these limitations, the present study highlights an important conceptual evolution in pancreatic cancer research. Precision oncology has historically focused on tumor-derived information, including genomic alterations, transcriptomic signatures, molecular subtypes, ctDNA, and artificial intelligence-assisted imaging analyses (5,6,10). These approaches undoubtedly improve our understanding of tumor aggressiveness; however, they incompletely capture the biological complexity of cancer recurrence. Increasing evidence suggests that recurrence represents the dynamic interaction between tumor aggressiveness and host resilience. While tumor biology determines metastatic potential, host biological factors—including nutritional reserve, systemic inflammation, immune competence, skeletal muscle mass, frailty, and functional status—may determine whether disseminated tumor cells successfully establish clinically detectable metastases (8,9,11,15). This perspective supports the development of integrated predictive models combining tumor- and host-derived biomarkers. Future perioperative risk assessment may simultaneously incorporate pathological stage, ctDNA, molecular profiling, inflammatory biomarkers, nutritional indicators, sarcopenia assessed by computed tomography, frailty evaluation, and patient-reported outcomes. Such multidimensional models are likely to provide substantially greater predictive accuracy than reliance on pathological variables alone (15,16). Importantly, this integrated strategy also has therapeutic implications. Unlike genomic alterations, host biological factors are potentially modifiable. Nutritional intervention, structured exercise, multimodal prehabilitation, optimization of systemic inflammation, and individualized supportive care may improve physiological resilience before surgery (17,18). Although definitive evidence demonstrating reduced cancer recurrence is still lacking, these interventions represent promising strategies for improving both short-term recovery and long-term oncological outcomes. Future directions The study by Oh and colleagues raises several important questions that warrant future investigation (1). First, can preoperative nutritional optimization increase serum prealbumin concentrations and subsequently reduce early recurrence? Although causality has not yet been established, prospective interventional studies could determine whether nutritional improvement translates into better oncological outcomes. Second, should prealbumin be evaluated as part of a multimodal host assessment rather than as an isolated biomarker? Combining prealbumin with inflammatory indices, sarcopenia measurements, frailty assessment, and functional performance may more accurately characterize host resilience than any single parameter alone (12,15,17). Third, can host-derived biomarkers be integrated with emerging molecular technologies? Combining prealbumin with ctDNA, circulating tumor cells, molecular subtyping, radiomics, and artificial intelligence-based prediction models may enable personalized recurrence prediction with substantially improved discrimination (5,10,16,19). External validation using large prospective multicenter cohorts is essential before clinical implementation. Standardization of prealbumin cut-off values, nutritional assessment protocols, and perioperative treatment strategies will be necessary to determine the true clinical utility of this inexpensive biomarker. The study by Oh and colleagues provides more than evidence supporting the prognostic value of a readily available laboratory parameter (1). More importantly, it reminds us that successful pancreatic cancer treatment depends not only on understanding the biology of the tumor but also on understanding the biological condition of the host (20). Prealbumin is unlikely to become a standalone predictor of recurrence. Nevertheless, it symbolizes a broader conceptual transition in pancreatic oncology—from tumor-centered precision medicine toward integrated host-tumor precision oncology. Future advances in recurrence prediction will likely depend not on replacing conventional pathological factors, but on integrating molecular tumor characteristics with comprehensive assessments of host resilience, nutritional status, inflammation, and functional reserve. Such an approach may ultimately enable more individualized perioperative management and improve long-term outcomes for patients with pancreatic cancer. Supplementary The article’s supplementary files as