Survival benefits of para-aortic lymphadenectomy in colorectal cancer with clinically suspected para-aortic lymph node metastasis: a meta-analysis and systematic review

Colorectal cancer (CRC) is the third most commonly diagnosed cancer and is the second-leading cause of cancer-related deaths worldwide, with an incidence of over 1.9 million new cases and more than 935,000 deaths in 2020 according to the GLOBOCAN 2020 report [1]. Compared to liver and lung metastasis, para-aortic lymph node (PALN) metastasis is a particularly rare pattern of distant metastasis with an occurrence rate of less than 2% [2]. Once PALN metastasis occurs, it may lead to an early recurrence after surgery and extremely worse survival in patients with pancreatic cancer, biliary cancer, or cervix cancer [3‐5]. Surgical excision of primary and metastatic lesions is still considered to be the most effective way to cure CRC with distant metastasis. Imaging data are reproducible and effective methods for routine clinical diagnosis of PALN metastasis [6]. However, the best treatment for patients with CRC and clinically suspected PALN metastasis is still controversial due to the different definitions in the past.

The 2017 AJCC 8^th Edition TNM staging classified PALN metastasis in patients with CRC as stage M1 or distant metastasis, rather than grouping it with regional lymph nodes [6]. PALN metastasis may occur in the form of an oligometastatic state [7], which provides an opportunity for patients to receive radical para-aortic lymphadenectomy (PALND). So far, many studies have pointed out the survival benefit of PALND for clinically diagnosed PALN metastasis in patients with CRC. However, due to the small sample data and no prospective studies, the evidence level is not convincing enough.

The era of precision medicine warrants the application of a more personalized approach for the treatment of patients with CRC and clinically suspected PALN metastasis. Since PALN is close to important blood vessels, surgery is somewhat dangerous and difficult. The question of which type of patients receiving PALND would benefit from survival also needs to be answered. In this study, we aimed to perform a meta-analysis of survival outcomes and prognostic factors in patients with CRC and clinically suspected PALN metastasis undergoing PALND.

PALN metastases were defined as histologically confirmed retroperitoneal lymph nodes metastases rather than local tumor recurrence. The latter refers to tumor cell recurrence at a local site after surgical removal of the primary tumor, without lymph node involvement [24]. According to the classification system of the Japanese Society of Clinical Oncology, the retroperitoneal region, in which PALN metastases occur, is classified as A (supra-renal vessels) and B (infra-renal vessels) [28]. The upper boundary of the B region is the renal vein, and the lower boundary is the iliac bifurcation, surrounding the abdominal aorta and inferior vena cava. The A retroperitoneal area with a starting point slightly higher than the renal vein is unconventionally included in the anatomical range of PALN because of the difficulty of surgical resection and the low possibility of complete resection [2, 7].

Lymph nodes, a type of normal tissue structure, can be imaged and measured regardless of whether they are involved in metastasis. While biopsies are invasive and difficult to replicate, computed tomography, magnetic resonance imaging, and positron emission tomography are reproducible, straightforward, and practically noninvasive methods for measuring the size of lymph nodes. Although the literature included in this study must have clear radiological evidence supporting PALN metastasis as described above, the definition of measurable and identifiable target metastatic lymph nodes in terms of computed tomography is controversial. Some institutions measure the long axis diameter of lymph nodes to check for enlargement. For example, in the study by Sahara et al., the diameter of the long axis of lymph used lymph node long axis diameter ≥10 mm as the inclusion criterion [18]. Other groups have used the short axis for diameter measurement; however, they also contain varying lengths, such as 5, 8, and 10 mm [17, 19, 22, 29, 30]. There was still no consensus on the selection of lymph node diameter line, perhaps due to multiple medical equipment and different judgments of imaging technology scientists. Lymph node diameter measurement was also commonly used to evaluate the effect of chemotherapy or radiotherapy, namely the treatment response and post-resection recurrence by imaging. Therefore, patients with CRC should be routinely followed up every 3 months for the first 2 years and every 6 months thereafter [31].

There were two ways of time classification: one is the operative time of primary tumor and PALN mentioned in our study, and the other is the time of PALN metastasis after the diagnosis of primary tumor [11]. They should not be confused. According to our subgroup analysis results, the survival effect achieved by PALND was independent of the surgery time for primary tumor and clinically suspected PALN metastases. Increasing articles emphasized the classification of PALN metastasis time. Gagniere et al. manifested that the OS was not affected by PALN metastasis time (HR = 2.83, 95% CI: 0.61–13.08, P = 0.18) [2]. A retrospective analysis by Ichikawa et al. also showed that the 3-year RFS of 28 patients with PALND was not affected by the time of PALN metastasis (HR = 0.792, 95% CI: 0.66–3.58, P = 0.301) [23]. In another small sample study, Arimoto et al. reported that the 3-year OSs for simultaneous (n = 9) and metachronous (n = 5) PALN metastases were 40 and 100%, respectively [29]. Due to the small sample size and the inclusion of patients with other distant metastatic lesions, the conclusions of these three studies were not statistically significant. Choi et al. included 24 participants with pathologically positive PALN metastasis, but without other distant metastasis, indicating that patients with metachronous metastasis (n = 5) had a longer median survival time than patients with simultaneous metastasis (n = 19) (median OS: 61 months (95% CI: 50–71) and 29 months (95% CI: 1–57), P = 0.227) [11]. However, the results were not statistically significant. We were unable to analyze in detail whether there was a difference in survival between patients with simultaneous and metachronous PALN metastasis according to the time of PALN metastasis after the diagnosis of primary tumor because few studies have provided accurate time limits to distinguish these two groups of patients.

Radical surgical resection of stage I–III CRC is still the mainstay of treatment, which is associated with a 5-year OS ranging from 50 to 94% [32]. Although there have been several studies on retroperitoneal lymphadenectomy, the choice between PALND and adjuvant therapy remains uncertain. Some studies had reported no significant survival benefit from extensive lymphadenectomy [33, 34]. Based on the 20 studies we included, the 5-year OSs for patients with CRC undergoing PALND ranged from 0 to 70.3%, and the 3-year OSs ranged from 33.15 to 93.90%. Eight articles presented the 5-year OSs for participants who underwent PALND and those who did not, and pooled results suggested a survival benefit for patients who received PLAND (OR = 3.73, 95% CI: 2.05–6.78). Additionally, none of the patients included in these eight articles had other extra-retroperitoneal metastasis before surgery. Only 6/16 patients in the study by Ogura et al. and 3/13 in the study by Kim et al. did not confirm pathological PALN metastasis [12, 14]. To some extent, our analysis indicated that para-aortic lymphadenectomy rather than chemotherapy alone was beneficial to the survival of patients with CRC and clinically suspected PALN metastasis. The ideal margin of retroperitoneal lymph node resection should be negative, and our results were consistent (OR = 5.26, 95% CI: 2.02–13.69) (Table 2). Laparotomy and endoscopy are the alternative surgical approaches; however, few studies have compared their operative difficulty, duration, blood loss, and survival outcome. Furthermore, a recent case-cohort analysis found no difference in overall survival between endoscopic and open approaches (HR = 0.941, 95% CI: 0.571–1.831, P = 0.101) [10]. From the renal vessel to the iliac vessel bifurcation, PALND was conducted along the abdominal aorta. Because there are so many important vascular pathways nearby, PALND is a highly challenging surgical procedure and requires a more experienced general surgeon. The popularity of endoscopic surgery has undoubtedly increased the difficulty of lymph node dissection and also brought the risk of surgical complications to some extent. However, endoscopic approach had no significant survival benefit compared to open approach. Complications after PALND were reported in 18 included literatures, and the incidence of complications in the PALND group ranged from 8.00 to 42.90%. In terms of the number of complications, the most common complications in PALND group were intestinal obstruction (50 cases), followed by incision infection (44 cases), anastomotic leakage (28 cases), urinary retention (25 cases), pneumonia (21 cases), urinary tract infection (18 cases, abdominal abscess (18 cases), chylous leakage (16 cases), and abdominal hemorrhage (7 cases). Rare complications include atelectasis, venous embolism, ureter or bladder damage, and so on (Supplementary table 1). In addition, according to our result, PALND had no effect on the incidence of complications (OR = 0.97, 95% CI: 0.46–2.08).

Not all patients with suspicious preoperative imaging results of PALN have pathologically postoperative positive lymph nodes. After performing PALND on 33 patients with CRC who exhibited signs of PALN metastasis on preoperative radiologic examination, Lee et al. found that only 14 patients were confirmed as pathologically positive PALN metastasis [35]. In other words, the pathological findings of PALN metastasis had a 42.42 percent likelihood of agreeing with the radiological findings. The advantage of surgical resection over imaging detection is that it can provide comprehensive diagnostic pathological information to guide treatment in terms of genes. However, metastasectomy is not always feasible, especially if metastatic lymph nodes have invaded important blood vessels or organs, or if the patient's physical state prevents them from undergoing complex surgery. Except for metastasectomy, salvage adjuvant chemotherapy may be another option for patients with advanced CRC. Yeo et al. showed that radical chemotherapy is an effective salvage treatment for retroperitoneal lymph node metastasis in CRC, with a 5-year OS of approximately 36.4% [36]. Currently, National Comprehensive Cancer Network guidelines recommend the use of capecitabine-base and 5-fluorouracil-base as the most commonly used first-line chemotherapy regimens for CRC. If the disease progresses and distant metastasis to para-aortic lymph nodes occurs, targeted therapies may also be considered. Bevacizumab targeting vascular endothelial growth factor ligand and cetuximab targeting epidermal growth factor receptor are key standard agents for improving survival outcomes in patients with metastatic CRC [37].

Due to the small number of patients undergoing PALN metastasectomy and the limited literature on this topic, more prospective large multicenter randomized trials are urgently needed to confirm the merits of extended lymphadenectomy. However, the use of surgical resection, chemotherapy, or local radiation therapy can be incorporated into a multimodal strategy for the treatment of these patients.

To sum up, PALND had a survival benefit for patients with CRC and clinically diagnosed para-aortic lymph node metastasis, in which pathologically negative PALN, complete resection, less metastatic PALN and medium-high histological differentiation were protective factors for survival and the presence of preoperative metastasis elsewhere was a risk factor for recurrence-free survival.