Clinical characteristics and prognosis of liver metastases with unknown primary site

Introduction

The liver is susceptible to metastasis from various malignant tumors and stands out as one of the primary sites for metastatic cancer (1). Malignant tumors originating in the gastrointestinal tract, pancreas, lung, breast, kidney, ovary, among others, have the propensity to metastasize to the liver, with the digestive system being the most prevalent source (2,3). Identifying the primary site holds significant clinical value, guiding subsequent treatment steps and enhancing the survival prognosis for patients with tumors. While most patients with liver metastases can pinpoint the primary site through various examinations, some cannot, leading to the categorization of these cases as cancers of unknown primary (CUP), constituting approximately 3–5% of all malignant tumors (4). Even in autopsy cases, primary foci remain elusive in 20–50% of subjects (5).

Given the low incidence and heterogeneity of CUP, coupled with the scarcity of evidence-based medicine and limitations in clinical practitioners’ understanding of the disease, diagnosing and treating CUP pose substantial challenges. CUP carries an exceedingly poor prognosis, with an average survival time of 6–16 months, primarily due to concurrent metastases from other sites upon diagnosis (6,7). Notably, there are no extensive studies analyzing the characteristics and prognosis of patients with liver metastases from cancer of unknown primary (CUPL). To gain comprehensive insights into unexplained liver metastases, we utilized the Surveillance, Epidemiology, and End Results (SEER) database to elucidate the clinicopathologic characteristics and prognostic factors for these patients. On the contrary, given the clinical rarity of patients under 50 years old with liver metastases of unknown primary origin (8), we also conducted a comparative analysis of clinical characteristics and survival disparities among various age groups in this study. This study aims to offer timely interventions for such cases, with the ultimate goal of improving patient survival. We present this article in accordance with the STROBE reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-24-136/rc).

Methods

Study population and variables

Patients with liver metastasis of unknown origin were identified from the SEER database for the period spanning 2010 to 2020 using SEER*Stat version 8.4.2.

Selection criteria involved isolating patients categorized as having an unknown primary site, denoted by the primary site code “C80.9-Unknown primary site”, and specifically setting the SEER Combined Mets at DX-liver (2010+) to “YES”. This limitation arises from the fact that the SEER database has been collecting data on liver metastases only since 2010. A total of 6,528 patients were retrieved for analysis. Exclusion criteria were applied as follows: (I) cases lacking pathological information (N=1,829); and (II) cases without recorded survival months (N=8). Ethical approval was not deemed necessary as the SEER database is openly accessible to researchers worldwide. This study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

Demographic data encompassed various factors, such as age, gender, race, marital status, year of diagnosis, histologic differentiation, presence of bone, brain, and lung metastases, information on radiotherapy and chemotherapy, as well as follow-up details. We classified patients into two age groups: <50 and ≥50 years, primarily considering the notable increase in diagnosed cases of liver metastases of unknown primary site after the age of 50 years (Figure 1) and the division between early-onset and late-onset cancers at the age of 50 years (9). The definition of overall survival (OS) encompassed the time elapsed from the date of diagnosis to the date of death from any cause or the date of the last follow-up. Cancer-specific survival (CSS) was defined as the time from diagnosis to cancer-specific death (CSD) or the date of the last follow-up.

Figure 1 Number of cases diagnosed across various age groups among patients with liver metastases of unknown primary site.

Statistical analysis

Statistical analyses were performed using R 4.3.2 software. Count data were assessed through the χ² test or Fisher’s exact probability method, while measures were analyzed using Student’s t-test. To address disparities in variables between the age <50 years group and the age ≥50 years group, we utilized propensity score matching (PSM). Prognostic factors influencing patients were assessed through both univariate and multivariate Cox regression analyses for OS, and Fine and Gray’s competing risk regression analysis for CSS. OS curves were generated using the Kaplan-Meier method, and statistical differences between the survival curves were assessed with the log-rank test. The cumulative incidence function (CIF) was employed to illustrate the probability for each event, with differences in CIFs between groups estimated using Gray’s test. A P value of less than 0.05 indicated statistical significance.

Results

Patients characteristics

The study encompassed a total of 4,691 patients, with 319 (6.8%) classified in the age <50 years group and 4,372 (93.2%) in the age ≥50 years group. Figure 1 demonstrates a notable rise in the diagnosis of patients with liver metastases of unknown primary site beyond the age of 50 years. The median age across all patients was 69 years (range, 1–90+ years), with a median survival time of 1 month (range, 0–121 months) and a mean survival time of 5.61±12.47 months. The majority of patients in both the age <50 and age ≥50 years groups identified as White race (72.1% vs. 80.2%, P=0.003), and the predominant histological type was adenocarcinoma not otherwise specified (NOS) (36.4% vs. 41.4%, P=0.004). In comparison, the age <50 years group exhibited a higher incidence of lung metastasis than the age ≥50 years group (41.7% vs. 29.9%, P<0.001). Regarding treatment modalities, a significant proportion of age <50 years group patients received chemotherapy compared to the age ≥50 years group (56.1% vs. 24.8%, P<0.001). Refer to Table 1 for a detailed breakdown of these findings.

PSM

To effectively mitigate confounding bias between the age <50 and age ≥50 years groups, we employed the PSM method. Consequently, a total of 610 patients were included in the evaluation, with 305 patients in each of the subgroups. Following 1:1 matching with a caliper value of 0.01, no significant differences were observed between the variables in the two groups (refer to Table 2). The matching process encompassed key factors such as sex, marital at diagnosis, race, grade, histology, year of diagnosis, bone metastases, brain metastases, lung metastases, as well as details regarding chemotherapy, and radiation.

Cox regression analysis for OS

Table 3 presents the comprehensive results of both univariable and multivariate Cox analyses for patients with liver metastasis of unknown origin before PSM. In the univariable analysis of OS, factors such as age, race, gender, histology, grade, presence of brain and lung metastases, as well as the administration of chemotherapy and radiotherapy, emerged as prognostic indicators for unexplained liver metastasis. In the multivariable analysis of OS, several variables were identified as significantly associated with improved survival, including age <50 years, neuroendocrine tumor, well-differentiated and moderately differentiated histology, absence of lung metastases, and the utilization of radiotherapy and chemotherapy. These findings highlight the multifaceted nature of factors influencing the prognosis of patients with liver metastasis of unknown origin. Even after PSM, the results of multifactorial Cox regression still indicated that age remained an independent prognostic factor influencing OS [hazard ratio (HR) =0.78, 95% confidence interval (CI): 0.65–0.93, P=0.006] (Table 4).

Competing-risk analyses for CSS

Table 5 provides a detailed presentation of both univariable and multivariate competing-risk analyses results for patients with liver metastasis of unknown origin before PSM. In the univariable analysis of CSS, prognostic factors encompassed age, race, histology, grade, presence of bone and brain metastases, lung metastases, as well as the administration of chemotherapy and radiotherapy. In the multivariable analysis of CSS, several variables were identified as significantly associated with improved survival, including age <50 years, neuroendocrine tumor, well-differentiated histology, and the use of radiotherapy and chemotherapy. These results underscore the diverse set of factors that play a role in determining the CSS outcomes for patients with liver metastasis of unknown origin. Following matching, no significant survival difference was observed between the age <50 and age ≥50 years groups (P=0.056) (Table 4).

Survival analysis after PSM

The OS of patients with liver metastasis of unknown origin was 14.7% at 1 year, while CSS stood at 23% at 1 year. Individuals in the age <50 years group exhibited a 1-year OS rate of 27.5% and a 1-year CSS rate of 35%. In contrast, those in the age ≥50 years group demonstrated a 1-year OS rate of 13.8% and a 1-year CSS rate of 22%. Upon conducting a comparative analysis following PSM, it was evident that the age <50 years group demonstrated superior OS and CSS outcomes in comparison to the age ≥50 years group, as illustrated in Figure 2. Further subgroup analyses before and after PSM demonstrated that the age <50 years group displayed improved OS and CSS compared to the age ≥50 years group among patients treated with chemotherapy (Figure 3). Additionally, within patients subjected to radiotherapy, the age <50 years group demonstrated superior OS compared to the age ≥50 years group, although no significant difference was observed in CSS (Figure 4). These findings highlight age-specific variations in both overall and CSS rates, especially within distinct treatment subgroups.

Figure 2 Comparison of overall survival and cancer-specific death analysis of patients in age <50 years and age ≥50 years groups before (A,B) and after (C,D) propensity score matching.

Figure 3 Comparison of overall survival and cancer-specific death between age <50 years and age ≥50 years groups in chemotherapy patients before (A,B) and after (C,D) propensity score matching.

Figure 4 Comparison of overall survival and cancer-specific death between age <50 years and age ≥50 years groups in radiation patients before (A,B) and after (C,D) propensity score matching.

Discussion

CUP refers to metastatic foci diagnosed as malignant tumors via pathological examination, yet the primary site eludes identification despite thorough medical history, physical examination, and pre-treatment tests (10). Factors contributing to the unknown primary site include limitations in detection methods, inadequate pathological tissue sampling, primary site removal, extensive metastasis hindering primary site identification, unique modes of tumor dissemination, diminutive primary site size, spontaneous regression of the primary site, among others (8,11,12). Recent years have witnessed a decline in CUP incidence, potentially attributable to enhanced success rates in identifying primary site tumors (13,14). While liver metastases with known primary sites have been extensively studied (15,16), liver metastases with unknown primary sites remain scarcely explored. Presently, this phenomenon is primarily documented in a limited number of case reports (17). This study, to the best of our knowledge, represents the largest sample size investigation to date, probing into the clinical characteristics and prognosis of patients with CUPL. The survival prognosis for CUP patients is already dismal, and this is further exacerbated when liver metastases are present (18,19). In our study, the 1-year OS rate for patients with liver metastases of unknown primary origin was a mere 14.7%, with a corresponding 1-year CSS rate of 23%. Hence, understanding the clinical characteristics of these patients and identifying viable therapeutic approaches are crucial for improving survival outcomes.

Our findings revealed that the majority of liver metastases of unknown primary origin occurred in age ≥50 years’ patients (93.2%), with only a minor proportion (6.8%) aged <50 years. Comparable observations have been reported by others, such as Australian authors noting the rarity of CUP in individuals under 40, with a mean diagnosis age of 75 in 2011, identifying age as a robust risk factor for CUP (10). Our study further underscores age as a common independent prognostic factor influencing both OS and CSS in patients with liver metastases of CUP. Both univariate Cox regression and competing analyses indicated a reduced risk of death in the age <50 years group [HR =0.59, 95% CI: 0.52–0.67; subdistribution HR (SHR) =0.70, 95% CI: 0.62–0.78], and multivariate analyses consistently supported this conclusion (HR =0.72, 95% CI: 0.63–0.82; SHR =0.78, 95% CI: 0.70–0.88). Even after the matching process, age continued to be identified as an independent prognostic factor for OS. This may be attributed to elderly patients' inherent susceptibility to various comorbidities, limited physical tolerance, and reduced accessibility to radiotherapy and chemotherapy, ultimately contributing to a poorer prognosis.

Beyond age, tumor histology and differentiation grade are pivotal factors influencing the prognosis of patients with unexplained liver metastases. Previous studies identified adenocarcinoma as the most prevalent CUP histologic subtype, constituting 50%, with poorly differentiated and undifferentiated carcinoma at 30%, and 15% classified as squamous-cell carcinomas, and 5% as undifferentiated neoplasms (3,6,8,13). Our study aligns with these findings, with adenocarcinoma NOS representing approximately 41% of the total population. Both Cox regression and competing analyses results indicated that any tumor, excluding neuroendocrine tumors, predicted a poorer prognosis for patients with liver metastases of CUP.

In the absence of a standard treatment regimen for CUP, general consensus recommends surgical resection if feasible, with postoperative chemotherapy, and for unresectable cases, systemic treatment options may be considered (20-22). Our findings from univariate and multivariate Cox regression, as well as competing risk analyses, suggested that patients with liver metastases receiving radiotherapy and chemotherapy had reduced death risks compared to those without such treatments. Notably, among patients undergoing chemotherapy, the age <50 years group exhibited superior overall and CSS compared to the age ≥50 years group. No significant difference in CSS was observed between younger and older groups among patients receiving radiotherapy. Thus, chemotherapy emerges as a crucial method for improving the prognosis of CUP patients with liver metastases. Other therapeutic avenues, including molecular targeting, immunotherapy, and organ- and target-specific therapies, warrant further investigation.

Acknowledging the limitations of our retrospective study, such as potential selection bias, incomplete treatment data, and lack of information on tumor recurrence in the SEER database, underscores the need for large-scale clinical randomized trials to validate our findings.

Conclusions

In summary, our study unveils that the majority of patients with CUPL are elderly, facing an exceedingly grim survival prognosis. Despite the age <50 years group comprising a small fraction of patients, their overall and CSS outcomes are relatively favorable compared to the age ≥50 years group. Furthermore, survival benefits are associated with chemotherapy, with superior overall and CSS in the age <50 years group receiving chemotherapy compared to their age ≥50 years counterparts.

Acknowledgments

The authors express their gratitude for the valuable efforts undertaken by the SEER Program in establishing and maintaining the SEER database.

Funding: This work was supported by Science and Technology Project of Changshu Health Committee (No. CSWS202014), and Suzhou 23rd Science and Technology Development Program Project (Clinical Trial Organization Capacity Enhancement) (No. SLT2023006).

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-24-136/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-24-136/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-24-136/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Wang P, Jie Y, Yao L, et al. Cells in the liver microenvironment regulate the process of liver metastasis. Cell Biochem Funct 2024;42:e3969. [Crossref] [PubMed]
2. Bosch FX, Ribes J, Díaz M, et al. Primary liver cancer: worldwide incidence and trends. Gastroenterology 2004;127:S5-S16. [Crossref] [PubMed]
3. Lee MS, Sanoff HK. Cancer of unknown primary. BMJ 2020;371:m4050. [Crossref] [PubMed]
4. Siegel RL, Miller KD, Fuchs HE, et al. Cancer statistics, 2022. CA Cancer J Clin 2022;72:7-33. [Crossref] [PubMed]
5. Pentheroudakis G, Golfinopoulos V, Pavlidis N. Switching benchmarks in cancer of unknown primary: from autopsy to microarray. Eur J Cancer 2007;43:2026-36. [Crossref] [PubMed]
6. Varadhachary GR, Raber MN. Cancer of unknown primary site. N Engl J Med 2014;371:757-65. [Crossref] [PubMed]
7. Qi P, Sun Y, Liu X, et al. Clinicopathological, molecular and prognostic characteristics of cancer of unknown primary in China: An analysis of 1420 cases. Cancer Med 2023;12:1177-88. [Crossref] [PubMed]
8. Vajdic CM, Goldstein D. Cancer of unknown primary site. Aust Fam Physician 2015;44:640-3. [PubMed]
9. Ben-Aharon I, van Laarhoven HWM, Fontana E, et al. Early-Onset Cancer in the Gastrointestinal Tract Is on the Rise-Evidence and Implications. Cancer Discov 2023;13:538-51. [Crossref] [PubMed]
10. Kato S, Alsafar A, Walavalkar V, et al. Cancer of Unknown Primary in the Molecular Era. Trends Cancer 2021;7:465-77. [Crossref] [PubMed]
11. Boys EL, Gao B, Grimison P, et al. Retrospective analysis of clinical characteristics and outcomes of patients with carcinoma of unknown primary from three tertiary centers in Australia. Cancer Med 2024;13:e7052. [Crossref] [PubMed]
12. Laprovitera N, Riefolo M, Ambrosini E, et al. Cancer of Unknown Primary: Challenges and Progress in Clinical Management. Cancers (Basel) 2021;13:451. [Crossref] [PubMed]
13. Binder C, Matthes KL, Korol D, et al. Cancer of unknown primary-Epidemiological trends and relevance of comprehensive genomic profiling. Cancer Med 2018;7:4814-24. [Crossref] [PubMed]
14. Rassy E, Pavlidis N. The currently declining incidence of cancer of unknown primary. Cancer Epidemiol 2019;61:139-41. [Crossref] [PubMed]
15. Chen C, Lu C, Viswanathan V, et al. Identifying primary tumor site of origin for liver metastases via a combination of handcrafted and deep learning features. J Pathol Clin Res 2024;10:e344. [Crossref] [PubMed]
16. Clark AM, Ma B, Taylor DL, et al. Liver metastases: Microenvironments and ex-vivo models. Exp Biol Med (Maywood) 2016;241:1639-52. [Crossref] [PubMed]
17. Lazaridis G, Pentheroudakis G, Fountzilas G, et al. Liver metastases from cancer of unknown primary (CUPL): a retrospective analysis of presentation, management and prognosis in 49 patients and systematic review of the literature. Cancer Treat Rev 2008;34:693-700. [Crossref] [PubMed]
18. Horn SR, Stoltzfus KC, Lehrer EJ, et al. Epidemiology of liver metastases. Cancer Epidemiol 2020;67:101760. [Crossref] [PubMed]
19. Pauli C, Bochtler T, Mileshkin L, et al. A Challenging Task: Identifying Patients with Cancer of Unknown Primary (CUP) According to ESMO Guidelines: The CUPISCO Trial Experience. Oncologist 2021;26:e769-79. [Crossref] [PubMed]
20. Krämer A, Bochtler T, Pauli C, et al. Cancer of unknown primary: ESMO Clinical Practice Guideline for diagnosis, treatment and follow-up. Ann Oncol 2023;34:228-46. [Crossref] [PubMed]
21. Olivier T, Fernandez E, Labidi-Galy I, et al. Redefining cancer of unknown primary: Is precision medicine really shifting the paradigm? Cancer Treat Rev 2021;97:102204. [Crossref] [PubMed]
22. Tomuleasa C, Zaharie F, Muresan MS, et al. How to Diagnose and Treat a Cancer of Unknown Primary Site. J Gastrointestin Liver Dis 2017;26:69-79. [Crossref] [PubMed]