Metabolic syndrome and its components predict the biochemical recurrence and adverse pathological features of patients following radical prostatectomy: a propensity score matching study

Prostate cancer (PCa) is the most common malignancy cancer of the urinary system and the significant cause of cancer-related deaths in men worldwide. There were an estimated 268,490 new cases of PCa and 34,500 PCa related deaths in 2022 in the United States [1]. In contrast, although the number of newly diagnosed PCa patients in China has been increasing in recent years, the incidence and mortality of PCa are still significantly lower than those in western countries [2, 3]. Apart from genetic factors including ethnic origin and family history, lifestyle‐related factors such as eating habits and sedentariness might also play a significant role in the obvious geographical disparity in PCa risk [4].

Radical prostatectomy (RP) has become the standard treatment for eligible patients due to its superior cancer control and survival benefits [5]. Although most patients are disease-free following RP, nearly 30% of patients continue to experience biochemical recurrence (BCR) during follow-up [6, 7]. Patients with BCR exhibit an extremely worse prognosis due to its association with progression to distant metastases and cancer-specific mortality [8]. Therefore, the assessment of reliable prognostic predictors of BCR after RP is clinically important for guiding clinical decision-making and patient counseling. To date, several traditional clinicopathological factors, such as preoperative prostate‐specific antigen (PSA) levels, Gleason score, tumor stage, surgical margin status, lymph node invasion, extracapsular extension (ECE) and seminal vesicle invasion (SVI) have been identified as prognostic factors for BCR after RP [9]. Although these factors are commonly used to predict BCR‐free survival (BCRFS) after RP, they are unsatisfactory and limited due to their irreversibility. Therefore, more reliable and reversible prognostic factors are required to allow for improvements in disease outcome. In addition, given the significant impact of adverse pathological features on BCR, exploring the potential risk factors of these adverse features can also provide guidance for the preoperative treatment options and effective postoperative management of RP.

Over the past few decades, metabolic syndrome (MetS) has become a prevalent global major health issue, which has attracted extensive attention [10]. MetS are metabolic abnormalities resulting from sedentary lifestyle and excessive diet in a genetically predisposed individual [11], which is characterized by a constellation of metabolic disturbances including abdominal obesity, hypertension, hyperglycemia, hypertriglyceridemia, and low high-density lipoprotein cholesterol (HDL-C) [10]. There is mounting clinical and epidemiologic evidence suggests that MetS is strongly associated with increased risk of PCa incidence [12], cancer progression [13] and poor prognosis [14]. However, regard as the potential association between MetS and BCR or adverse pathological features after RP, the current studies conducted to so far are inconsistent to draw a definitive conclusion. For instance, some studies reported that the presence of MetS was associated with an increased risk of BCR, higher Gleason score, pT3–4 diseases and lymph node involvement after RP [15, 16]. In contrast, another study did not find any significant association between MetS at the time of diagnosis and the risk of BCR or clinicopathological features of PCa after RP [17]. These discrepancies may be explained by the large existence of confounders, the wide heterogeneity of the criteria used for MetS evaluation and ethnicities of the populations.

Therefore, given the inconsistency of existing evidence, the study was aimed to investigate the predictive value of MetS and its components in BCR and adverse pathological features of patients with PCa treated with RP by eliminating potential confounding factors using propensity score matching (PSM), and to provide additional evidence for the current study on the correlation between oncological outcomes after RP and MetS.

To the best of our knowledge, this is the first study to comprehensively explore whether MetS or its components influence BCR as well as adverse pathological features of patients following RP by using PSM. Balancing the differences of baseline characteristics between the patients in the MetS and non-MetS groups and eliminating potential confounding factors by using the method of PSM is an advantage of our study. We found that the presence of MetS were significantly associated with worse BCRFS in Kaplan–Meier analysis, while hyperglycemia and hypertriglyceridemia were identified as independent prognostic factors for BCRFS by multivariate Cox analysis adjusted for other clinicopathological factors. BMI was also found to be significantly associated with worse BCRFS in population with age ≥ 70 and with NOCD in the subgroup analysis. In addition, we performed further analyses for exploring the association between MetS and its components and adverse pathological features, the results revealed that hypertriglyceridemia was the only component independently associated with NOCD (≥ pT3a), ECE and upgrading in propensity matched patients.

We observed that the prevalence of MetS in RP population is 25.9% in our study, which is consistent with previous studies that reported incidence rate ranging from 18%-30% [23‐25]. The risk of morbidity and mortality associated with PCa varies in different countries, with the highest in Western countries and the lowest in Asian countries [26]. The epidemiological characteristics of MetS are also similar to those of PCa that are manifested by an obviously higher incidence rate in Western countries compared with Asian countries [27], which indicates that MetS plays a significant role in pathogenesis of PCa. In addition, the presence of MetS was supposed to be closely associated with increased incidence, aggressiveness and unfavorable prognosis of PCa as well, while several potential molecular mechanisms and metabolic pathways are well characterized including insulin resistance and comorbid hyperinsulinemia [dysregulation of the insulin-like growth factor (IGF) signaling pathway] [28], pro-inflammatory condition and abnormal adipokines levels [16, 29], and a microenvironment conducive to tumor formation induced by adipose tissue [30, 31]. However, the current knowledge might represent a small part of the biological mechanisms underlying these associations, and further studies are still warranted.

In addition to traditional clinicopathological factors such as preoperative PSA level, tumor stage, GS and positive surgical margin, the results of our study also showed that MetS was significantly associated with worse BCRFS after RP in PCa patients, although the predictive role was not independent. It indicated that there is a higher incidence of BCR after RP among patients with MetS. However, current researches on the association between MetS and BCR after RP remain controversial. Shiota M et al. [15] considered that the feature of MetS is an independent risk factor for BCR after RP among Japanese men. Similarly, Castillejos-Molina R et al. [25] also concluded that MetS is independently associated with the risk of biochemical progression in both OCD and locally advanced PCa. In contrast, neither the result of Xu X et al. [32] and Morlacco A et al. [33] showed any relationship between the presence of MetS and BCR among the cohort in China and Europe, respectively. These discrepancies might result from the wide heterogeneity of MetS definition and ethnicities of the populations. In addition, given the significant impact of adverse pathological features on BCR, quantities of studies have investigated and revealed the close association between MetS and its components and adverse pathological features after RP, despite of the considerable differences between the findings of various studies. A retrospective study consisted of 1016 Chinese patients with PCa who received RP revealed that MetS indicated an increased risk of prostatectomy GS ≥ 8, pT3-4 disease and lymph node involvement [16]. On the contrary, this positive relationship was not found in studies by Beebe-Dimmer JL et al. [34] and Xu X et al. [32]. Furthermore, MetS also represents a significant risk factor for positive surgical margin [23], upgrading and upstaging after RP [35]. However, there was no statistical significance between MetS and any postoperative adverse pathological features after adjusting other clinicopathological variables using multivariate analysis in our study. It could be explained by the fact that MetS is a syndrome consisting of at least three components, and the individual component may exert antagonistic roles (like positive and negative functions cancel each other out), thus ultimately manifesting as a meaningless composite outcome.

Apart from MetS itself, meaningful outcomes were found when focus on individual MetS components alone in further analysis as well. hyperglycemia and hypertriglyceridemia were the only two MetS components both identified as independent risk factors for increased risk of BCR after RP. Our findings on hyperglycemia can be supported by previous other studies with similar conclusions. The result of a case–control study showed that diabetes was significantly associated with an increased likelihood of BCR after RP regardless of metformin use [36]. Wright JL et al. [37] also concluded that glucose levels at the time of PCa diagnosis is an independent predictor of BCR for men undergoing RP for localized disease. This positive association comes in contrast to other researches. Rieken M et al. [38] could not detect a significant association between diabetes mellitus and increased risk of postoperative BCR in a cohort of 6,863 RP patients. More recent studies also failed to establish a significant link between hyperglycemia and BCR [39, 40]. These contradictory findings might be related to the fact that diabetes could interact with PCa cells at different levels. On the one hand, diabetes may have a potential protective effect against the progression of PCa by reducing the activity of IGF‐I and testosterone levels [41, 42]. On the other hand, elevated glucose levels accompanied by hyperinsulinemia represent the underlying mechanism by which the PCa development and BCR could be induced by diabetes [43]. Therefore, further well-controlled clinical researches with large sample sizes are still warranted to provide more evidence regarding the association between hyperglycemia and BCR with corresponding numerous mechanisms.

As for hypertriglyceridemia, the existing data is both scarce and divergent. A meta-analysis integrating 12 articles involving 11,108 patients concluded that there was no significant correlation between hypertriglyceridemia with BCR after RP [44]. Inconsistent with the above results, our study demonstrated that patients with hypertriglyceridemia was significantly associated with worse BCRFS. This was also further confirmed by Kaplan–Meier analysis, which revealed an increased risk of BCR in patients with hypertriglyceridemia. In addition, further analysis also showed that hypertriglyceridemia is independently association high risk of upgrading after RP. The results of Arthur R et al. [45] and Hayashi et al. [46] about that hypertriglyceridemia was positively associated with high-grade PCa (GS ≥ 8), which contribute to support our conclusions indirectly. The underlying mechanisms for this positive risk correlation could be explained by the results of several experimental studies using in vitro models [47]. However, several interesting results in our study can also be observed that hypertriglyceridemia was the protective factor significantly associated reduced risk of NOCD (≥ pT3) and ECE, which is inconsistent with Zheng X et al. [44] who suggested that hypertriglyceridemia was linked with higher risk of pT ≥ T3. Kang M et al. [48] also revealed that preoperative hypertriglyceridemia was significantly associated with a reduced risk of BCR after RP. Serum cholesterol levels can be reduced by cancer cells, which provides a mechanism basis for this positive protective association. Malignant cells grow and proliferate rapidly, which required consuming large amounts of serum cholesterol for the biosynthesis of new cell membrane to meet accelerated metabolic turnover [49]. In this respect, patients with unfavorable tumor phenotype are considered likely to have lower baseline cholesterol due to abundant storage of lipids in their cancer cells [48]. Thus, further studies are required to verify whether there is a significant accumulation of intracellular triglyceride in the surgical specimens of PCa patients with BCR to provide more convincing evidence for our conclusion.

At present, it has been proposed that higher BMI increases the risk of tumor aggressiveness, BCR and cancer-specific mortality of PCa in several studies [50‐52]. In contrast, the primary analysis of our study did not show any association between overweight and/or obesity (BMI ≥ 25) and BCR as well as adverse pathological features in RP patients, which is consistent with most previous studies [53‐55]. The differences may be related to variations in the cutoff values of BMI and treatment selection. Nevertheless, we observed the independent predictive role of BMI in the patients with age ≥ 70 and with NOCD in subgroup analysis, which suggests that BMI might influence the outcomes in a distinct manner from dyslipidemia. On the one hand, obese men more often have lower testosterone levels, low testosterone concentrations are thought to promote the development of aggressive forms of PCa through a carcinogenic inflammatory pathway [52]. As testosterone levels in elderly men decline with age gradually, this unfavorable mechanism of obesity will be further exacerbated. On the other hand, due to technical difficulties of performing surgery, obesity is thought to be associated with a higher risk of positive surgical margins after RP, thereby increasing the likelihood of BCR [56, 57]. When patients accompanied by NOCD, the presence of ECE and/or SVI would inevitably increase the possibility of positive surgical margin and further in the risk of BCR in collaboration with obesity.

Our findings have several clinical implications. First off, our results provide evidence to support the role of MetS and its components (especially hyperglycemia and hypertriglyceridemia) in preoperative risk stratification for BCR and adverse pathological features after RP by using the method of PSM. The findings may provide a research direction about new strategies or approaches for PCa treatment, such as diet, exercise and medications (like statins and metformin) for MetS, which contribute to improve the prognosis of PCa by reversing MetS. Then, the natural history of BCR after RP could be long but variable. The determination of risk assessment factors such as preoperative MetS, hyperglycemia, hypertriglyceridemia and BMI could be conducive to recognize patients with high-risk BCR after RP, and may benefit from aggressive salvage treatment. In addition, accurate preoperative staging is vital for the management strategy of PCa. Thus, urgent improvement of the current situation of a great quantity of tumors are over- or under-staged is needed [58]. Although no association between MetS and adverse pathological features can be found in this study, this information of the significant relationship between hypertriglyceridemia and the features of NOCD and ECE would contribute to predict stage in PCa patients accurately.

There are several limitations of the study that need to be acknowledged. First and foremost, the retrospective design is the primary limitation, which only allowed us to evaluate the temporal link between MetS and BCR, thereby causal inferences are limited, and the potential selection bias can not be ignored due to the retrospective nature as well. Second, this study mainly focused on the Chinese population and we used the diagnostic criteria of MetS that are most suitable for the Chinese population, which might affect the applicability of research results in other ethnic populations. Meanwhile, we adopted BMI rather than waist circumferences to define overweight or obesity while waist circumference may be more closely related to metabolic changes compared with BMI [59], which could lead to misclassification. Last but not least, additional potential confounding factors, such as family history of cancer, dietary habits, physical activities and drug treatment including stains, aspirin or metformin were not evaluated in the study. Data on these factors would be of great important since they are known to play a vital role in the association between MetS and BCR as well as adverse pathological features, which might limit the statistical power of the study.

In conclusion, hyperglycemia and hypertriglyceridemia were the two effective components of MetS that were identified as independent prognostic factors for BCRFS. The independently reduced association of hypertriglyceridemia with NOCD and ECE appear to a survival benefit, while the independently increased risk of upgrading points to a worse prognosis. Hyperglycemia and hypertriglyceridemia may become reliable and reversible predictors in predicting BCR and adverse pathological features of patients following RP that are benefit for helping clinicians improve patient counseling as well as risk-adapted strategies.