Serum hemoglobin and albumin levels serve as prognostic predictors after adrenal metastasectomy Authors: Nan Ma, Xiaobo He, Qiwei Nei, Zhenhua Liu, Xianda Chen, Hui Chang, Kai Yao, Shengjie Guo Published: 2024 ABSTRACT AbstractPurposeAdrenal metastasectomy is a common local treatment for adrenal metastases, with a lack of serum prognostic factors. Here, we identified the prognostic significance of preoperative hemoglobin and albumin levels in patients undergoing adrenal metastasectomy.Materials and MethodsData from 93 patients who underwent adrenal metastasectomy were assessed retrospectively. Receiver operating characteristic (ROC) curve analysis was used to determine the optimal cutoff values of hemoglobin and albumin for survival. Overall survival (OS) and disease-free survival (DFS) were evaluated using the Kaplan–Meier method, thereafter Cox regression models and subgroup analyses were applied to adjust confounding factors. A risk stratification developed by joint use of serum hemoglobin and albumin levels was also tested.ResultsOptimal cut-off points were 130.5 g/L and 44.8 g/L for hemoglobin and albumin, respectively. Multivariate Cox regression analysis identified decreased hemoglobin (HR [95% CI]: 0.41 [0.18—0.91], P = 0.029) and albumin (HR [95% CI]: 0.12 [0.02—0.88], P = 0.038) levels as independent factors for poorer OS. Patient with both decreased hemoglobin and albumin levels had the worst OS (P = 0.001) and DFS (P = 0.001) than other risk groups. Subgroup analyses proved decreased hemoglobin and albumin levels predicted poorer OS independent of cancer types.ConclusionsPreoperative hemoglobin and albumin levels may serve as prognostic predictors after adrenal metastasectomy. A prognostic model that combines hemoglobin and albumin can improve accuracy in predicting patient outcomes and be easily implemented in clinical practice.Keywords: Adrenal metastasectomy, Hemoglobin, Albumin, Prognosis Introduction The adrenal gland is one of common metastatic sites for malignant tumors [1]. With prolonged survival of tumor patients, the rate of adrenal metastases (AM) also tends to increase, and about 10–27% of cancer patients are found with adrenal metastases in autopsies [2]. Studies revealed the presence of adrenal metastases in 14% of the patients with colorectal cancer, 42% with lung cancer, and 58% with breast cancer [3]. Systemic treatment strategies, including immunotherapy, targeted therapies, and chemotherapy have been the primary therapeutic approaches for patients with adrenal metastases. However, these treatments often encounter challenges related to the development of resistance [4, 5].Therefore, intervention by local treatment is commonly adapted and might prolong patient survival. Adrenal metastasectomy, especially minimally invasive surgery, is a preferred modality fby urologists, with the advantages of favorable disease control, rapid recovery, and shortened interruption of pharmacotherapy [3, 6, 7]. The one-year and five-years disease-free survivals (DFSs) of adrenal metastasectomy were 54% and 31%, and the one-year and five-year overall survival (OSs) were 83% and 43%, respectively [8]. As the adrenal metastasectomy is inherently a palliative intervention, patients undergoing this procedure often exhibit reduced tolerance for surgery. Consequently, strict adherence to surgical indications becomes imperative in determining the appropriateness of the intervention. Currently, patients with isolated adrenal metastases, well-controlled systemic metastases and progressive adrenal metastases who physically tolerate surgery are considered to be the beneficiaries [7, 9]. Studies have shown that tumor size and lymph node metastasis are poor prognostic factors for adrenal metastasectomy [8]. Since most studies are limited to pathologic features, they hardly meet the clinical needs for prognostic prediction. The prognostic value of patients' physical status factors remains to be elucidated, of which serum albumin and hemoglobin levels was already known as prognostic predictors for multiple malignancies. Studies have shown that the prognosis of tumors is not only related to the characteristics of the tumor itself, but also to the nutritional status and systemic inflammation of patients [10, 11]. Malnutrition was revealed to be associated with decreased immunity, which increases the incidence of postoperative complications and mortality, leading to poor prognosis for patients after surgery [12, 13]. Hemoglobin and albumin are the two most important indicators of nutritional status. Low hemoglobin levels can lead to tumor hypoxia, which can accelerate tumor growth and promote tumor cell angiogenesis [14]. Meanwhile, low serum albumin levels can impair cellular immune function and lead to worse prognosis in cancer patients [15].It was found that low preoperative serum albumin and hemoglobin levels were strongly associated with a poor prognosis for malignant tumors [16]. However, no similar studies have been reported for patients undergoing adrenal metastasectomy. In this study, we retrospectively analyzed the prognostic indicators of patients undergoing adrenal metastasectomy. In addition to clinicopathological features of primary and metastatic tumors, hemoglobin and albumin indexes were the first to be analyzed for correlation with patients’ prognosis, proving an insight for future screening of high-risk patients. Methods Patients This single-center retrospective study included data from 93 consecutive AM patients who underwent adrenalectomy at our center between January 2013 and April 2022. The inclusion criteria were as follows: (i) diagnosed as AM pathologically; (ii) accessible to complete clinical data and preoperative blood samplings for hemoglobin and albumin levels; (iii) primary tumor have received standard drug therapy; (iv) effective and accurate follow-up. The exclusion criteria were as follows: (1) direct adrenal involvement from local tumor invasion; (2) R1 or R2 resection. All the clinicopathological data was retrieved from the electronic patient records of our center. Informed consent was waived because of the retrospective nature of the study. Oligometastatic disease was defined as 1–4 sites of disease. For external validation of the proposed prognostic model, we subsequently selected 24 patients who had AMs from lung cancers and underwent adrenal metastasectomy between May 2022 and December 2023, based on the same criteria aforementioned. Follow-up and study outcome Patients were monitored through both telephone contact. Data collected during follow-up included details on adjuvant therapy, current living status, disease progression, and sites of metastasis. The final date of follow-up was recorded on June 1, 2022. The primary endpoint of this study was overall survival (OS) [8], defined as the time from the date of surgery to either the date of last follow-up or the date of death, whichever occurred first. The secondary endpoint was disease-free survival (DFS) [8], which was defined as the time from surgery to the occurrence of any of the following events: last follow-up, recurrence of adrenal lesions, or death. Statistical analysis Continuous variables were reported as means with standard deviations, whereas categorical variables were presented as frequencies and percentages [17]. Receiver Operating Characteristic (ROC) [18] curve analysis was employed to identify the optimal cutoff values for hemoglobin and albumin levels. The optimal cutoff point, determined by the most significant difference in survival via the log-rank test across cohorts, was based on overall survival status. Survival outcomes, specifically OS and DFS, were assessed using Kaplan–Meier survival curves and analyzed with the log-rank test. Cox proportional hazards models were utilized to explore factors influencing survival outcomes. Variables that reached statistical significance (P < 0.05) in the univariate analysis were subsequently included in multivariate Cox regression analyses to assess their independent effects on survival [19]. All statistical analyses were conducted using SPSS software version 26.0 (IBM, Armonk, NY) and R version 4.2.1. All tests performed were two-sided, with a significance threshold set at P < 0.05. Results Patient characteristics A total of 93 patients were enrolled in our study. Thirty-two (34.4%) patients had oligometastatic disease and the median tumor size of 3.2 cm (IQR: 2.1–4.8 cm). The most common primary malignancy was lung cancer (43, 46.2%), followed by liver cancer (17, 18.3%), gastrointestinal tumors (11, 11.8%), renal cell carcinoma (7, 7.5%), etc. Fifty-eight patients (62.4%) received postoperative systemic therapy (chemotherapy, radiotherapy, targeted, immunotherapy). (Table 1). Table 1Baseline characteristics of patients from this studyCharacteristicsN, Median or MeanSex (%)male72 (77.4)female21 (22.6)Age [IQR], yr56.00 [50.00, 63.00]Primary tumor (%)lung cancer43 (46.2)liver cancer17 (18.3)gastrointestinal tumors11 (11.8%)renal cell carcinoma7 (7.5)pancreatic cancer1 (1.1%)ovarian cancer1 (1.1%)oral squamous carcinoma1 (1.1%)mediastinal yolk sac tumor1 (1.1%)sarcoma2 (2.2%)osteosarcoma1 (1.1%)melanoma1 (1.1%)Primary resection (%)Non-R013 (14.0)R080 (86.0)Lymph node invasion of the primary tumor(%)no50 (53.8)yes43 (46.2)Vascular infiltration of the primary tumor(%)no45 (77.6)yes13 (22.4)Adjuvant systemic therapy of primary tumor(%)yes35 (37.6)no58 (62.4)Timing of metastasis with respect to primary tumorSynchronous (< 6 months)20 (21.5)Metachronous (> 6 months)73 (78.5)Oligometastatic disease (%)32 (34.4)Size of adrenal tumor (%) < 4.5 cm (%)66 (71.0) ≥ 4.5 cm (%)27 (29.0)Operative approach (%)Laparoscopic78 (83.9)Open15 (16.1)Extent of surgery (%)adrenal gland78 (83.9)adrenal gland plus other organs15 (16.1)Postoperative therapy (%)no40 (43.0)yes53 (57.0)DFS (median [IQR]),months7.00 [2.00, 19.00]OS (median [IQR]),months20.00 [11.00, 31.00]IQR inter-quatile range, OS overall survival, DFS disease-free survival Baseline characteristics of patients from this study IQR inter-quatile range, OS overall survival, DFS disease-free survival Cutoff value of preoperative serum hemoglobin and albumin We generated a ROC curve utilizing data from 93 cases to assess the prognostic efficacy of multiple parameters in predicting survival. The likelihood ratio was calculated by examining the specificity and sensitivity values acquired from the analysis. Based on the ROC curves, we observed that only hemoglobin and albumin had C-indexes (0.646 and 0.643) larger than 0.6 (Fig. 1). The optimal cut-off points of 130.5 g/L (specificity, 0.683; sensitivity, 0.586) and 44.8 g/L (specificity, 0.349; sensitivity, 0.931) for hemoglobin and albumin, respectively. ROC curves of multiple parameters for predicting 1-year mortality. HGB, hemoglobin. ALB, albumin. WBC, white blood cell. PLT, platelet. FGB, fibrinogen. CYSC, cystatin C Patient survival The median follow-up after adrenal metastasectomy was 20 months (IQR: 6–41). The median OS was 20 months (IQR: 11–31), the 1-year OS was 98.6%, and 5-year OS was 68.7%. The median DFS was 7 months (IQR: 3–18 months), the 1-year DFS was 92.5% and 5-year DFS was 46.5%. Univariate Cox analysis revealed that HGB (HR [95% CI]: 0.32 [0.15–0.67], P = 0.003) and ALB (HR [95% CI]: 0.09(0.01–0.69), P = 0.02) were associated with OS (Table 2); while vascular infiltration of the primary tumor [HR (95% CI) 2.15 (1.02–4.54), P = 0.044], and concentration of HGB (HR [95% CI]: 0.42 [0.25–0.71], P = 0.001) and ALB (HR [95% CI]: 0.48 [0.23–0.99], P = 0.047) were associated with DFS (Table 3). Multivariate analysis revealed that low HGB (HR [95% CI]: 0.41 [0.18—0.91], P = 0.029) and low ALB (HR [95% CI]: 0.12 [0.02—0.88], P = 0.038) levels were independent factors associated with poorer OS. Table 2Univariate and multivariate Cox regression analyses of baseline predictors associated with OSVariablesHR comparisonUV HR (95% CI)UV p valueMV HR (95% CI)MV p valueSexmale vs. female1.43(0.44–2.73)0.836Age2.04(0.94–4.43)0.07Lunglung vs. non-lung0.82(0.39–1.72)0.603Primary resectionyes vs. no1.38(0.42–4.57)0.599Lymph node invasion of the primary tumoryes vs. no0.8(0.38–1.68)0.56Vascular infiltration of the primary tumoryes vs. no2.03(0.76–5.43)0.159Postoperative therapy of primary tumoryes vs. no1.05(0.49–2.27)0.9Oligometastatic at presentationyes vs. no1.43(0.66–3.11)0.363Surgery aimradical operation vs. palliative surgery0.61(0.25–1.5)0.281Postoperative therapy of AMyes vs. no0.63(0.3–1.32)0.22HGB > 130.5 g/L vs. < 130.5 g/L0.32(0.15–0.67)0.0030.41(0.18—0.91)0.029ALB > 44.8 g/L vs. < 44.8 g/L0.09(0.01–0.69)0.020.12(0.02—0.88)0.038HR hazard ratio, CI confidence interval, MV multivariate, UV univariate Univariate and multivariate Cox regression analyses of baseline predictors associated with OS HR hazard ratio, CI confidence interval, MV multivariate, UV univariate Table 3Univariate and multivariate Cox regression analyses of baseline predictors associated with DFSVariablesOR comparisonUV HR (95% CI)UV p valueMV HR (95% CI)MV p valueSexmale vs. female1.03 (0.55–1.93)0.917Age1.12 (0.66–1.9)0.679Lunglung vs. non-lung0.91 (0.54–1.54)0.726Primary resectionyes vs. no0.75 (0.35–1.59)0.452Lymph node invasion of the primary tumoryes vs. no1.15 (0.68–1.94)0.614Vascular infiltration of the primary tumoryes vs. no2.15 (1.02–4.54)0.0442.55 (1.17–5.58)0.019Postoperative therapy of primary tumoryes vs. no1.76 (0.98–3.15)0.057Oligometastatic at presentationyes vs. no1.53 (0.89–2.64)0.124Surgery aimradical operation vs. palliative surgery0.66 (0.32–1.36)0.263Postoperative therapy of AMyes vs. no1.23 (0.71–2.14)0.469HGB > 130.5 g/L vs. < 130.5 g/L0.42 (0.25–0.71)0.0010.34 (0.15- 0.76)0.009ALB > 44.8 g/L vs. < 44.8 g/L0.48 (0.23–0.99)0.0470.84 (0.32–2.26)0.736HR hazard ratio, CI confidence interval, MV multivariate, UV univariate Univariate and multivariate Cox regression analyses of baseline predictors associated with DFS HR hazard ratio, CI confidence interval, MV multivariate, UV univariate Development of a model jointed by hemoglobin and albumin Patients were stratified into three risk groups according to preoperative serum hemoglobin and albumin levels (high risk: hemoglobin ≤ 130.5 g/L and albumin ≤ 44.8 g/L: intermediate risk: hemoglobin ≤ 130.5 g/L and albumin > 44.8 g/L or hemoglobin > 130.5 g/L and albumin ≤ 44.8 g/L; low risk: hemoglobin > 130.5 g/L and albumin > 44.8 g/L). The high-risk group had poorest OS (high-risk vs. intermediate-risk vs. low-risk, mean OS time, 37.11 vs. 49.45 vs. 73.41 months, P = 0.001) and DFS (high-risk vs. intermediate-risk vs. low-risk, mean DFS time (9.93 vs. 26.52 vs. 28.75 months, P = 0.001) than the other two groups (Fig. 2). Kaplan–Meier curve and log-rank analysis were used to analyze prognostic factors of patients undergoing adrenal metastasis surgery. A OS of preoperative HGB levels. B DFS of preoperative HGB levels. C OS of preoperative ALB levels. D DFS of preoperative ALB levels. E OS of different risk stratifications (ALB-HGB). F DFS of different risk stratifications (ALB-HGB). DFS disease-free survival, OS overall survival, HGB hemoglobin, ALB albumin Subgroup analyses of cancer types For AMs derived from lung cancers, OS (high-risk vs. intermediate-risk vs. low-risk, mean OS time, 23.05 vs. 58.39 vs. 67.20 months, P < 0.001) and DFS (high-risk vs. intermediate-risk vs. low-risk, mean DFS time, 11.11 vs. 43.76 vs. 39.62 months, P = 0.014) were the worst in the high-risk group. (Fig. 3A and B). However, for AMs derived from non-lung cancers, we observed no significant difference in DFS (P = 0.260) nor OS (P = 0.280) between the three groups (Fig. 3C and D). Subgroup analyses of cancer types. A OS of different risk stratifications for AMs of lung cancers. B DFS of different risk stratifications for AMs of lung cancers. C OS of different risk stratifications for AMs of non-lung cancers. D DFS of different risk stratifications for AMs of non-lung cancers. DFS disease-free survival, OS overall survival, HGB hemoglobin, ALB albumin External validation We collected 24 newly diagnosed AM cases derived from lung cancers to validate our findings. Median follow-up time was 13.1 months. Consistent with the training cohort, OS (high-risk vs. intermediate-risk vs. low-risk, mean OS time, 13.4 vs. 19.9 vs. 22.8 months, P < 0.001) and DFS (high-risk vs. intermediate-risk vs. low-risk, mean DFS time, 6.6 vs. 7.8 vs. 15.7 months, P = 0.014) were the worst in the high-risk group (Fig. 4). Kaplan–Meier curves of OS (A) and DFS (B) of risk groups of the external validation cohort Discussion In this study, by examining the prognostic indicators after adrenal metastasectomy, we found that preoperative HGB (HR 0.41, P = 0.029) and ALB (HR 0.12, P = 0.038) were poorly associated with OS. We also found that primary tumor vascular infiltration (HR 2.55, P = 0.019) and HGB levels (HR 0.34, P = 0.009) were the independent predictors for DFS. Further studies showed that a model based on preoperative HGB and ABL levels well predicted OS after adrenal metastasectomy, and subgroup analyses proved its performance in most patient types. To our knowledge, this is the first study to investigate the prognostic value of preoperative HGB and ALB in patients undergoing adrenal metastasectomy, providing a reference for clinical practice. Adrenal metastasectomy is one of the most common modalities for the local treatment of AMs, which has been proven to confer significant survival benefit to selected patients, either in open or laparoscopic surgery [1].With the development of radiotherapy, stereotactic body radiotherapy (SBRT) plays an important role in making patients intolerable to surgery, which not only improves tumor symptoms but also has a synergistic effect with immunotherapy [20]. With the advantages of less damage, faster recovery, effective short-term local control and improved survival, ablation can be an alternative therapy for patients with poor physical condition [21]. Although interventional embolization of AMs can precisely occlude the tumor vessels, operations are challenging because the adrenal gland receives blood from three arteries. The return of embolic agents to normal vessels may lead to ischemia in non-target organs [22]. Predictors for the efficacy of adrenal metastasectomy currently focus on the type of primary tumor and the status of the metastases. Studies have shown that significant vascular invasion of the primary tumor may be associated with systemic multiple metastases, increasing the risk of distant metastases after local treatment, and is an independent predictor of poor prognosis [8, 23]. Indeed, the primary tumors in the present study tend to be vascular ones. This study also showed that patients with vascular infiltration of the primary tumor have an increased probability of distant recurrent metastases. Therefore, resection of adrenal metastases is insufficient for cancer management and systemic therapy is needed to better control the disease and improve survival. The most common source of AMs is lung cancer, and studies have shown that adrenal metastasectomy for lung cancer is more effective than for patients with non-lung cancer, mainly because of greater drug selectivity and efficacy in advanced lung cancer [8, 24]. However, our study showed that the survival difference between the primary tumor non-lung cancer and lung cancer groups is not significant, partly because the non-lung cancer group is dominated by gastrointestinal tumors and liver cancer, which was also benefited by more selective drugs and better efficacy. Adrenal metastasectomy may confer therapeutic efficiency and survival prognosis to more malignant types in addition to lung cancer [8]. Some studies have also shown that the presence of concomitant metastases from other sites is associated with the prognosis of adrenal metastasectomy, yet the present study didn’t observe such a phenomenon, consistent with the study of Ashley E et al [25]. This is partly due to differences in the composition of the primary tumor, the number of surgical cases, and other differences. Most of the current studies have focused on features of primary and metastatic lesions of AMs, and few molecular indicators have been studied to predict the prognosis of AMs. Numerous studies have attempted to excavate serum markers to predict cancer patients’ prognosis. Previous studies have found that low preoperative HGB and ALB are associated with shorter survival in patients with gastric, renal, and pancreatic cancers [26].The present study also found that low hemoglobin and albumin were associated with worse DFS and OS in patients with AMs. Anemia with low hemoglobin can cause hypoxia, which may accelerate tumor angiogenesis, promote tumor progression, and induce tumor resistance to radiotherapy and chemotherapy as well [14, 27]. Under conditions of hypoxia, the expression of hypoxia-inducible factor-1 was upregulated, which activates gene products that promote cancer progression, thus increasing the risk of tumor invasion and metastasis [28]. A recent study evaluated the outcome of 367 patients with soft tissue sarcoma and found that low preoperative hemoglobin levels were significantly associated with decreased cancer-specific survival and OS [29]. Studies from our center have reported that preoperative anemia was an independent predictor of outcome in gastric cancer [30]. In addition, malnutrition and inflammation can inhibit albumin synthesis; thus, serum albumin is also an independent predictor of poor prognosis in malignant tumors [31]. In the present study, we found that low albumin levels in oncologic disease may be secondary to a strong systemic inflammatory response syndrome, which increases the synthesis of acute-phase proteins and thus competitively inhibited the synthesis of albumin. Long-term malnutrition may lead to protein depletion in organisms, immune resistance, and even early death [32]. This study is the first to report the clinical utility of preoperative HGB and ALB in predicting the prognosis of adrenal metastasectomy. Low HGB and ALB were negatively associated with DFS and OS in patients undergoing adrenal metastasectomy. When preoperative HGB and ALB were risk stratified, patients in the high-risk group had lower OS (P = 0.0013) and DFS (P = 0.0014) than those in the intermediate and low-risk groups. While the current indications for surgical treatment of adrenal metastases are routinely judged based on tumor factors, this study provides new predictive models for high-risk patients that are suitable for less invasive options besides surgery, such as SBRT radiotherapy or ablation, which may lead to better overall survival. To adjust confounding effect of cancer types, we performed subgroup analyses in which a stable predictive effect was achieved in the lung cancer group rather than the non-lung cancer group, probably because of the extreme mixture of cancer types of the latter. Therefore, we selected newly diagnosed AM cases of lung cancers for external validation, which proved the prognostic significance of our model. Despite of it, its performance in other cancer types still requires validation. Chen et al. first described this score system to predict the prognosis of gastric cancer. Later studies adapted in esophageal squamous cell carcinoma, renal cell carcinoma, bladder cancer, prostate cancer, lung cancer, colorectal cancer, and pancreatic cancer [33]. Systemic inflammation stimulates angiogenesis, immunosuppression, and the formation of supporting microenvironments that can promote the initiation, progression and metastasis of tumor cells [26].Therefore, preoperative stratification of HGB and ALB has a strong prognostic predictive value for predicting adrenal metastases especially in patients with lung cancer as the primary tumor. The shortcomings of this study are mainly its retrospective nature, insufficient sample size, and lack of prognostic analysis in patients undergoing resection for non-adrenal metastases; therefore, we are looking forward to prospective and multi-center studies to give greater credibility to preoperative HGB and ALB in predicting prognostic indicators for adrenal metastasectomy. Conclusion Preoperative hemoglobin and albumin levels were independent predictors of prognosis in patients undergoing adrenal metastasectomy, and patients in the high-risk stratification had the worst prognosis; therefore, patients undergoing adrenal metastasectomy can be better screened for better survival according to HGB and ALB stratification. Author contributions S.G and K.Y designed and conducted the study; N.M, X.H and Z.L analyzed the data statistically. N.M, X.H, Q.N and Z.L wrote the manuscript. X.C, H.C and S.G critically reviewed the manuscript. All the authors have read and approved the manuscript. Funding There is no funding or donating support for this work. Data availability The data that support the findings of this study are available from the corresponding author, [Shengjie Guo], upon reasonable request. Declarations Ethics approval and consent to participate This retrospective observational study has gained approval (approval number: B2022‐247‐01) from the ethics committee of the Sun Yat-sen University Cancer Center (SYSUCC). In accordance with the committee's regulations, the requirement for informed consent was waived due to the study's retrospective nature. Additionally, all patient information was handled with strict confidentiality to ensure privacy and compliance with ethical standards. Competing interests The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Nan Ma, Xiaobo He, Qiwei Nei and Zhenhua Liu have authors contributed equally to this work. Contributor Information Kai Yao, Email: yaokai@sysucc.org.cn. Shengjie Guo, Email: guoshj@sysucc.org.cn. References Associated Data Data Availability Statement The data that support the findings of this study are available from the corresponding author, [Shengjie Guo], upon reasonable request.