Abstract
Background/Aim: The aim of this study was to investigate frailty as a prognostic factor in patients with colorectal liver metastasis undergoing hepatectomy. Patients and Methods: Eighty-seven patients who underwent hepatectomy at our institution were enrolled. Frailty was defined as a score of ≥4 on a clinical frailty scale. Patients were divided into frailty (n=29) and non-frailty (n=58) groups. Results: Overall and cancer-specific survival rates were significantly worse in the frailty group compared with the non-frailty group, and multivariate analysis revealed frailty as an independent prognostic factor. Disease-free survival tended to be worse in the frailty group. Fifty-eight patients relapsed after the first hepatectomy. Twenty-one of 58 recurrent patients were allocated to the frailty group. After recurrence, chemotherapy was significantly more frequently performed in the non-frailty group compared with the frailty group. Conclusion: Frailty can predict the prognosis of patients with colorectal liver metastasis undergoing hepatectomy.
Survival in patients with colorectal cancer (CRC) has improved over the last decades, primarily due to efforts in screening and early detection, and the recent advances in systemic and local therapies. However, the main cause of mortality in CRC patients is metastasis. The liver is the most common anatomic site for CRC metastasis (1). Approximately 35%-55% of CRC patients have liver metastases as the disease progresses (2). Surgical resection is currently the treatment of choice for CRC liver metastases, and has been proven as the only curative treatment (3, 4).
Due to aging, the number of elderly patients with colorectal liver metastasis (CRLM) will increase significantly in the near future. Age-related complications, such as cardiopulmonary complications, delirium, transfer to a rehabilitation facility, and dependency are major problems after hepatectomy in the elderly patients (5). Given the increasing number of elderly patients with comorbidities undergoing surgery, there is increased interest in preoperatively identifying patients who are at high risk of morbidity and mortality after hepatectomy (6). Frailty is a multidimensional and heterogeneous syndrome associated with instability that differs from disability or aging alone (7). Frailty is commonly measured using summative impairment lists and algorithms based on clinical assessment (8-10). Frailty is very important in predicting surgical outcomes in elderly patients (10). Recently, McIsaac et al. suggested that the clinical frailty scale (CFS) could be used to predict new disability or death after elective noncardiac surgery (11). The CFS is a nine-point global frailty scale based on a clinical evaluation in the domains of mobility, energy, physical activity, and function (12, 13). The CFS is a highly acceptable, feasible, and convenient instrument for clinical perioperative use.
There have been several reports on the relationship between hepatectomy and frailty (5, 6, 14-18). Louwers et al. (14) reported that frailty was a significant factor determining morbidity and mortality after hepatectomy. Tanaka et al. (5) reported that frailty could predict age-related complications after hepatectomy. Recently, we also reported a significant relationship between frailty and surgical outcomes after hepatectomy in older patients with hepatocellular carcinoma, and frailty was identified as an independent prognostic factor (19). Chen et al. (20) reported that frailty was significantly associated with 30-day morbidity in patients who underwent colorectal and liver resection, but this report only examined short-term outcomes. There are currently no reports on the relationship between frailty and long-term outcomes of CRLM patients undergoing hepatectomy. Therefore, we investigated the clinical significance of frailty as a prognostic factor in patients with CRLM undergoing hepatectomy.
Patients and Methods
Patients. Eighty-seven patients with CRLM who underwent hepatectomy at our institution from 2005 to 2018 were enrolled. All patients underwent curative liver resection. The mean follow-up period after hepatectomy was 46.2 months (range=4.4-158.5 months).
Patients’ backgrounds and disease baseline characteristics, such as sex, age, comorbidities, and liver function were collected from medical records. Tumor factors, including tumor markers, number of lesions, maximum diameter, and the timing of liver metastasis were also assessed. Regarding metastatic tumor status, the maximum diameter × number (MDN) was defined as the product of maximum tumor diameter and number of liver metastases according to our previous report (21).
Patients’ immuno-nutritional statuses were also assessed using the neutrophil-lymphocyte ratio (NLR), prognostic nutritional index (PNI), C-reactive protein (CRP)-albumin ratio (CAR), and modified Glasgow Prognostic Score (mGPS) (22-25). Data were collected from preoperative blood tests. Of all 87 patients with CRLM, NLR and PNI were analyzed in 83 cases, mGPS and CAR in 85 cases, and albumin and CRP in 86 cases, respectively. Postoperative complications were defined according to the Clavien–Dindo classification (26). This study was approved by the Institutional Review Board of Tokushima University (No. 3215-1). All patients involved in the study provided written informed consent to participate.
Assessment of CFS. The CFS, a nine-point global frailty scale, is based on a clinical evaluation of mobility, energy, physical activity, and function (12, 13) as follows: CFS 1 (Very Fit) indicates that the patient is robust, active, energetic, highly motivated, exercises regularly, and exceptionally fit for their age; CFS 2 (Fit) indicates absence of active disease and that the patient is less fit compared with patients in category CFS 1; CFS 3 (Managing Well) indicates that disease symptoms are better managed compared with patients in category CFS 4; CFS 4 (Living with Very Mild Frailty) indicates that, although independent, these individuals commonly report having slowed down and/or having symptoms of disease; CFS 5 (Living with Mild Frailty) indicates patients with limited dependence on others for instrumental activities of daily living; CFS 6 (Living with Moderate Frailty) indicates that patients need help with both instrumental and non-instrumental activities of daily living; CFS 7 (Living with Severe Frailty) indicates that patients are completely dependent on others for all personal care; CFS 8 (Living with Very Severe Frailty) indicates that patients are completely dependent and are approaching the end of life; and CFS 9 (Terminally ill) indicates that patients have a life expectancy of <6 months. Nurses in the surgery department who were blinded to clinical and pathological data evaluated CFS scores on the basis of clinical information upon admission for surgery. Frailty was defined as a CFS score of ≥4 in this study.
Statistical analysis. The unpaired Mann-Whitney U-test or Chi-squared test was used to compare clinico-pathological variables between the two groups. Overall survival and disease-free survival curves were generated using the Kaplan–Meier method, and differences were compared using the log-rank test. A multivariate analysis was carried out based on the Cox proportional hazards regression model. For all statistical analyses, a p-value of <0.05 was considered statistically significant. All statistical analyses were performed using statistical software (JMP 8.0.1., SAS Campus Drive, Cary, NC, USA).
Results
Clinicopathological characteristics according to frailty status. According to the frailty score, twenty-nine of 87 patients with CRLM (33.3%) were deemed frail at liver resection. Patients were divided into frailty (n=29) and non-frailty (n = 58) groups, and relevant clinicopathological features were compared between the two groups. The median CFS score was 4 (range=4-7) in the frailty group and 2 (range=2-3) in the non-frailty group, respectively. Table I shows the clinicopathological characteristics of patients according to frailty status. There were no significant differences in sex or liver function between the two groups. The frailty group was significantly older than the non-frailty group (p<0.01). The median age of the frailty group was 78 years (range=56-90 years) and that of the non-frailty group was 65 years (range=43-92 years). The number of co-morbidities (diabetes and hypertension) was significantly higher in the frailty group compared with the non-frailty group. Regarding immuno–nutritional status, CRP concentration and CAR were significantly higher in the frailty group compared with the non-frailty group (p<0.05). Albumin concentrations tended to be lower in the frailty group compared with the non-frailty group (p=0.08). Regarding tumor factors, CEA concentration was significantly higher in the frailty group compared with the non-frailty group (p<0.05). There were no significant differences in other tumor factors between the two groups. Regarding perioperative factors, although there was no significant difference, the frailty group had a longer postoperative stay compared with the non-frailty group (p=0.12).
Clinicopathological characteristics according to frailty status.
Overall and disease-free survival rates. Figure 1A shows that the overall survival rate after hepatectomy in patients with CRLM was significantly worse in the frailty group compared with the non-frailty group (p<0.01). Overall, 3-year survival rates in the frailty and non-frailty groups were 63.9% and 89.1%, respectively. Figure 1B shows that the cancer-specific survival rate after hepatectomy in patients with CRLM was significantly worse in the frailty group compared with the non-frailty group (p<0.01). Cancer-specific 3-year survival rates in the frailty and non-frailty groups were 69.3% and 91.0%, respectively. In the frailty group, 15 patients died; 12 patients were cancer-related deaths and 3 patients were unrelated deaths. Among these three patients, one patient died due to pneumonia and two due to natural death. In the non-frailty group, 16 patients died; 14 patients were cancer-related deaths and two patients were unrelated deaths. Table II shows that the univariate analysis of cancer-specific survival detected CEA concentration (≥20 ng/ml) and frailty as prognostic factors. According to the multivariate analysis, only frailty was an independent prognostic factor (p=0.0477).
Overall and cancer-specific survival rate of patients with metastatic liver cancer. A: Overall survival rate of patients with metastatic liver cancer. B: Cancer-specific survival rate of patients with metastatic liver cancer. Red line, frailty group; Blue line, non-frailty group.
Results of univariate and multivariate analyses of cancer-specific survival.
Figure 2 shows that disease-free survival tended to be worse in the frailty group compared with the non-frailty group, although this difference was not significant (p=0.054). Three-year disease-free survival rates in the frailty and non-frailty groups were 22.7% and 38.6%, respectively. Table III shows that the univariate analysis of disease-free survival detected tumor number (multiple), CEA concentration (≥20 ng/ml), and liver resection (metachronous) as prognostic factors. According to the multivariate analysis, frailty tended to be an independent prognostic factor (p=0.0734).
Disease-free survival of patients with metastatic liver cancer. Red line, frailty group; Blue line, non-frailty group.
Results of univariate and multivariate analyses of disease-free survival.
Treatment after recurrence. Figure 3 shows additional therapy for recurrent patients after liver resection. Fifty-eight patients (66.7%) relapsed after the first liver resection. Twenty-one of 58 recurrent patients were allocated to the frailty group. After recurrence, 10 patients in the frailty group (47.6%) and 21 patients in the non-frailty group (56.8%) underwent repeat surgery. The rate of repeat surgery was not significantly different between the two groups. Chemotherapy was performed in 6 patients (28.6%) in the frailty group and 25 patients (67.6%) in the non-frailty group, and chemotherapy was significantly more frequently performed after hepatectomy in the non-frailty group compared with the frailty group (p<0.01). The indication of repeat surgery was determined by the type of recurrence rather than patient status, but chemotherapy was determined by patient status. Of the 21 patients in the frailty group who relapsed, 19 patients excluding 2 patients who received both of chemotherapy and repeat surgery were divided into 3 groups; chemotherapy group, surgery group, and no-treatment group. Figure 4 shows that the cancer-specific survival rate was significantly worse in the no-treatment group compared with the chemotherapy and surgery group (p<0.01).
Additional therapy after recurrence. A: Comparison of the number of patients who underwent surgery after recurrence. B: Comparison of the number of patients who underwent chemotherapy after recurrence.
Cancer-specific survival rate of the frailty group with recurrence. Red line, surgery group; Blue line, no-treatment group; Green line, chemotherapy group.
Discussion
This study showed that frailty is a better indicator of the long-term outcome of patients with CRLM undergoing hepatectomy compared with NLR and mGPS, which have been considered useful to date. Frailty is widely discussed in medical and surgical disciplines and is increasingly recognized as a useful predictor of healthcare outcomes as well as age. Frailty screening has been proposed as a potential method for preoperative risk classification in a number of conditions (20). Studies in the field of geriatrics have reported that frailty is associated with dysfunction, hospitalization, and death (27, 28). It is also reported that frailty is a prognostic factor in patients with cirrhosis, and should thus be considered one of many criteria for assessing patients’ suitability for liver transplantation (29, 30). There are many methods to assess vulnerabilities, and sarcopenia is one marker of patient vulnerability. Voron et al. (31). reported that sarcopenia was an independent prognostic factor for mortality after hepatectomy for hepatocellular carcinoma and could be used to evaluate patient eligibility before surgery. However, to assess sarcopenia, measurement of muscle mass by computed tomography and evaluation using complicated calculations are necessary (32). CFS can easily be used to assess the general condition and frailty of patients at first visit. The Kihon checklist recently reported by Tanaka et al. (5) is categorized in detail by 25 items. The difference from CFS is that it contains many scales such as depressed and withdrawal risk. CFS is an evaluation method that emphasizes motor function, and is considered to be a more objective evaluation method than the Kihon checklist.
Frailty affects cancer-specific survival in patients with breast cancer (33), gastric cancer (34), pancreatic cancer (35), and hepatocellular carcinoma (19). In this study, we examined the relationship between frailty and prognosis in patients of all ages with CRLM undergoing hepatectomy. CRLM is common not only in the elderly but also in young people. In fact, in this study as well, 12 of 29 cases in the frailty group and 45 of 58 cases in the non-frailty group were under 75 years of age. Therefore, it was considered that it would be more clinically useful to conduct a study of patients of all ages. This study is the first to assess the effect of frailty on long-term outcomes for patients with CRLM undergoing hepatectomy. Frailty was an independent prognostic factor in overall survival. In terms of disease-free survival, frailty tended to be a poor prognostic factor, although there was no significant difference. This result was thought to be due to the fact that additional therapy after recurrence, especially chemotherapy, was frequently performed in the non-frailty group. Therefore, there was a significant difference in overall survival even though there was no significant difference in disease-free survival. According to previous reports, frailty is associated with a high risk of chemotherapy-related toxicity and low resistance to treatment, and has been identified as a predictor of postoperative complications, chemotherapy intolerance, disease progression, and death (36, 37). These studies support our results.
Patients classed as frail are associated with significantly higher serum inflammatory parameters compared with healthy individuals. Soysal et al. (38) observed a large increase in CRP and interleukin-6 in patients classed as frail. Nishijima et al. (39) suggested the relationship between frailty and NLR in the elderly with cancer, and Mei et al. (40) reported that preoperative high blood NLR levels could be an indicator of poor prognosis for advanced tumor. Lealdini et al. (41) reported that mGPS correlated significantly with frailty and more advanced clinical stage in solid tumors and Kishiki et al. (42) reported that mGPS was a poor prognostic factor in patients with incurable stage Ⅳ CRC. In this study, there was no correlation between NLR or mGPS and overall survival or disease-free survival, but inflammatory markers, such as NLR and mGPS, may be prognostic factors in patients with CRLM undergoing hepatectomy; however, further investigation is needed to clarify this. Achilli et al. (43) reported that preoperative immunonutrition reduced both surgical site infections and the duration of antibiotic treatment after surgery. They encouraged the preoperative nutrition therapy in patients with frail colorectal cancer. In this study, the frailty group had a lower chemotherapy induction rate after recurrence, and the prognosis for those who could not receive chemotherapy was significantly worse. From this result, it is clear that some therapeutic intervention is required after recurrence. Efforts to improve the frailty, such as nutritional management, may be important to provide additional treatment after recurrence. Further research is needed to determine the criteria for preoperative nutrition therapy, however preoperative nutrition therapy may become important for frail patients of CRLM. Recently, laparoscopic hepatectomy for CRLM has also been widely performed, and Tajiri et al. (44) reported simultaneous laparoscopic resection of primary colorectal cancer and its liver metastatic lesions. Minimally invasive surgery may be required for frail patients.
In conclusion, the overall and cancer-specific post-hepatectomy survival rate of patients with CRLM was significantly worse in the frailty group compared with the non-frailty group. Frailty can predict the prognosis of patients with CRLM undergoing hepatectomy, suggesting the importance of preoperative nutritional therapy.
Footnotes
Authors’ Contributions
K.T. designed the study, performed data analysis and wrote the manuscript. Y.M. and M.S. designed the study and approved the manuscript finally. K.M., S.Y., Y.S., M.N., and T.I. performed data interpretation and analysis. All Authors read and approved the final manuscript.
Conflicts of Interest
The Authors declare no conflicts of interest related to this study.
- Received April 21, 2021.
- Revision received July 25, 2021.
- Accepted July 26, 2021.
- Copyright © 2021 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.