Insufficient lymph node assessment in gastric adenocarcinoma

This study aimed to investigate the sufficient (≥ 16) lymph node assessment in 449 patients with gastric adenocarcinoma and literature review. Four hundred and forty-nine patients with pathologically confirmed locoregional invasive gastric adenocarcinoma from 2004 to 2013 were included. A standard surgical resection was performed for all the patients with (n = 16) or without (n = 433) neoadjuvant treatment. In this study, 301 men and 148 women with a median age of 58 (range 21–88) years were included. The median total numbers of examined lymph nodes were 9 (range 0–55). Ninety-five patients (21.2%) had adequate (≥ 16) lymph node examination, and 70 patients (15.6%) had no examined lymph nodes. In univariate analysis, total or near total gastrectomy (P <  0.001), advanced node stage (P < 0.001), primary tumor size > 6 cm (P < 0.001), and the presence of perineural invasion (P = 0.039) were associated with more average number of examined lymph nodes. On multivariate analysis, node stage (P < 0.001) and type of surgery (P = 0.008) were independent predictive factors. In this study, approximately one in five patients with gastric adenocarcinoma had sufficient lymph node assessment. More studies are suggested for identifying a true inadequate lymph node dissection from insufficient lymph node assessment.

1. In this study, the median total numbers of examined lymph nodes were 9 (range 0-55). 2. Ninety-five patients (21.2%) had adequate (≥ 16) lymph nodes examination, and 70 patients (15.6%) had no examined lymph nodes. 3. Total or near total gastrectomy, advanced primary tumor and node stage, tumor size > 6 cm, and the presence of perineural invasion were associated with more median number of examined lymph nodes. 4. Advanced node stage and total or near total gastrectomy were independent predictive factors for adequate lymph node assessment.

Background
Gastric cancer remains the fourth most frequent cancer and the second leading cause of cancer death worldwide. Despite a decline in incidence of gastric cancer in the western countries, it is still a major malignant disease [1]. In most countries, this malignancy present at late stage due to undefined risk factors and non-specific symptoms. Surgery is the mainstay of curative treatment of gastric cancer. In gastric cancer surgery, no residual tumor (R0) resection is an ultimate goal; however, there has been strong argue regarding the degree of lymph node (LN) dissection. This argument involves sufficient surgical and pathological staging and satisfactory adjuvant therapy. Generally, limited LN dissection (D1) involves perigastric LNs surrounded by 3 cm from the primary tumor, extended LN dissection (D2) extends the dissection outside D1 to include LNs surrounding the hepatic and splenic arteries, and superextended LN dissection (D3) further includes LNs in the retropancreatic, paraaortic, and the root of the mesocolon LNs. Further LN dissection such as D2 resection may potentially offer more precise pathologic staging, enhanced regional tumor control, and potential survival improvement. All these issues, however, need to be proven, because the results of randomized clinical trials have failed to confirm a clear disease control and survival benefit to date [2].
In the fifth edition of the American Joint Committee on Cancer (AJCC) Tumor, Node, Metastasis (TNM) classification for gastric cancer, the number of positive nodes was considered as the base of LN classification. This classification showed superiority over the prior classification system in terms of LN stage as a prognostic factor. However, as it is expected, the more LN evaluated, nodal metastases detection is more likely. The number of nodal examined depends not only on the extent of LN dissection, but also on the lymph node retrieval. Therefore, when the number of examined LNs is insufficient for diagnosis, the nodal stage may be underestimated, which is so-called stage migration [3].
The 7th edition of the AJCC staging manual (2010) revised the nodal classification system such that N1 = 1-2 positive LNs; N2 = 3-6 positive LNs; N3a = 7-15 positive LNs; and N3b > 15 LNs. As such, the AJCC now recommends that at least 16 LNs be assessed per patient. The staging changes attempt to minimize the impact of surgical dissection on gastric cancer staging and to improve the prognostic ability of N-staging compared to that in the 5th/6th editions. Unfortunately, despite the changes to simplify staging, the number of LNs assessed in each gastric cancer case varies, and in many cases, the number reported per specimen is less than current recommendations [3].
The present study aimed to investigate the sufficient (≥ 16) lymph node evaluation in patients with resected gastric adenocarcinoma, and review of the literature.

Methods
In this retrospective study, a chart review was performed on 449 patients with resected locoregional invasive gastric adenocarcinoma who were treated and followed up at a referral academic hospital, between 2004 and 2013. A minimum sample size required for the study was calculated based on the value of standard deviation of the mean total number of assessed LN (SD = 11.02) in previous study by Zhao [4]. Accordingly, a minimum of 117 patients were estimated for a precision of 2%. Exclusion criteria in this study were in situ or metastatic tumor, pathologies other than adenocarcinoma, and unresectable or inoperable gastric cancer. Additionally, patients who had been treated with palliative surgery were excluded. The patients' cancers were reclassified according to the 8th edition of the AJCC. Four hundred and thirtythree patients underwent primary surgery, and the remaining 16 cases received neoadjuvant chemotherapy and/or chemoradiation before curative surgical resection. Initial investigation involved a history taking and physical examination, upper GI endoscopy, laboratory test, and chest, abdominal, and pelvic computed tomography (CT) scans.
All statistical analysis was performed using IBM SPSS 22. The median, as well as the percentage of patients with sufficient (≥ 16) and insufficient (< 16) evaluated LN, was initially calculated. All potential clinical (age, sex, tumor location, type of surgery) and pathological (histologic form, tumor stage, node stage, tumor grade, lymphovascular invasion, perineural invasion, tumor size, surgical margin status, and the total and adequate assessed LNs) variables were analyzed. The outcome variable were measured based on the median and adequate (≥ 16) total number of assessed lymph node. The impact of all potential variables on the median and adequate (≥ 16) total number of assessed lymph node was analyzed using non-parametric (Mann-Whitney U test and Kruskal-Wallis test) tests respectively. Initially, a univariate analysis for the dependent variable adequate (≥ 16) LN assessment was performed. In the final step, all significant factors were included in a stepwise multivariate logistic analysis. All statistical tests were twoended, and p values less than 0.05 were considered statistically significant.

Results
In this study, 301 men and 148 women with a median age of 60 (range 21-88) years were included. Two hundred and twenty-one patients were less than 60 years old, and 228 patients were older than or equal to 60 years old. The distribution of histologic form (P < 0.001), tumor grade (P = 0.027), and tumor size (P = 0.016) was significantly different among tumor location (Table 1). Accordingly, diffuse gastric involvement tended to be presented with larger tumor size, to have higher rate of poorly differentiated tumor and diffuse histologic type. The median total numbers of evaluated lymph nodes were 9 (range 0-55). Only 95 patients (21.2%) had equal or more than 16 lymph nodes evaluation, and 70 patients (15.6%) had no any lymph nodes for evaluation. Relative distribution of total lymph node evaluation has been illustrated in Fig. 1. Two hundred and sixty-two patients (58.3%) were node positive. Association of potential variables on median total lymph node count has been illustrated in Table 2. Additionally, an association was found between positive node and lymphatic vascular invasion (P < 0.001), the presence of perineural invasion (P < 0.001), and advanced T stages (P < 0.001). In univariate analysis, type of gastrectomy (P < 0.001, node stage (P < 0.001), primary tumor size (P < 0.001), and perineural invasion (P = 0.039) were significant variables (Table 3). In multivariate analysis using logistic regression method, N3 node stage (P < 0.001, OR = 5.907, CI = 3.462-10.081) and total or  (Table 4).

Discussion
Gastric cancer remains a major health problem not only in developing, but also in developed countries. Surgical treatment including complete resection of primary tumor and regional LN dissection plays an essential role in treating these patients. Theoretically, an insufficient LN dissection increases the risk of potential microscopic and gross residual tumor cell, higher rate of recurrent disease, and poorer prognosis [5]. It is believed that by increasing the number of surgically dissected and pathologically harvested LNs, surgical and pathological staging will be more accurate. Subsequently, this can potentially enhance locoregional tumor control and improves oncologic outcomes following gastric cancer surgery. Accordingly, a minimum number of 16-LN evaluation was recommended to achieve precise staging [6]. Despite many years after changing in gastric cancer staging, only 33% of patients had an adequate lymph node assessment. The occurrence of gastric cancer in patients younger than 40 years old is uncommon; however, after which, its incidence increases progressively. In the current study, the patients with a median age of 60 years old were younger than that of the results of previous reports in which the average median age of 27,214 patients in 15 report series was 65.5 (range 54-71) years old [1,4,[6][7][8][9][10][11][12][13][14][15][16][17][18] (Table 5). During recent decades, the frequency of primary tumor location has been changed in favor of proximal gastric cancers. At present, proximal gastric cancers including gastroesophageal junction lesions are diagnosed more frequently than in the past. Nevertheless, most gastric cancers still originate from distal stomach. The largest percentage of gastric cancers still arise within the antrum or distal stomach [19]. In this study, most of the lesions were in the distal of stomach (50.5%). In major reported series, the average total number of lymph node evaluated was 27.7 (range 8.4-35.3) for 7643 patients in 8 studies [4,6,[8][9][10][11][12]20]. In this research, the average total number of evaluated lymph node was 11.1. Furthermore, in the current study, 58% of all patients had stage III which was higher than that of major reported series in which this value was 31.2% (range 15.5%-64%) for 22,994 patients in 10 series [2, 7-10, 13-16, 18, 20]. In gastric cancer, insufficient lymph node assessment is a common finding in the literature. By analyzing data of 15 series including 27,942 patients, 52.2% (range 17.6-94.2%) of all patients had adequate (≥ 16) lymph node evaluation [1, 2, 4, 6-9, 13-18, 21, 22] (Table 5). In the present study, only 21.2% of patients had a sufficient lymph node assessment which seems to be much lower than that of average value in the literature.
The cause of insufficient lymph node assessment is multifactorial. A cooperation between the surgeon for doing optimal lymph node dissection and pathologist for doing sufficient lymph node assessment is required [23]. However, according to multiple studies, a variety of factors can affect both parties, including ethic background such as body mass index (that affect the surgeon's ability to perform an adequate lymphadenectomy), age (younger patients are more likely to have an adequate LN assessment), and region in which the surgery is performed [21]. In this study, age and sex have no significant relation with examined lymph nodes. Regarding the tumor and treatment factors, the average number of evaluated lymph node was associated significantly with primary site, node stage, tumor size, perineural invasion, type of surgery, and neoadjuvant treatment. Higher lymph node   examined was yield in total or near total gastrectomy, stage N3, tumor size > 6 cm, and presence of perineural invasion. The type of gastric surgery and extent of LN dissection remains an important factor in the adequacy of lymph node sampling and survival outcomes. Therefore, the surgeon can be considered as one of the most independent predictive factor in gastric cancer for achieving adequate lymph node sampling, R0 resection and locoregional disease control [24]. In this study, total gastrectomy was associated with more average number of examined lymph nodes compared to those with subtotal gastrectomy. In the current study, due to large number of surgeons, pathologists, and hospitals, to investigate any association between adequate lymph node assessment and the surgeon, pathologists, and hospitals was not possible.    [19]. Many reports showed a major role for surgical volume and experience of the surgeons and pathologist on lymph node retrieval in gastric cancer. High volume of surgery such as total and near total gastrectomy provides more removal and examination of regional LNs. In addition, in cases with higher T stage, gross LN involvement is more frequent which facilitates more LN detection and dissection by the surgeon and more LN retrieval by pathologist. Likewise, neoadjuvant therapies such as chemotherapy and chemoradiation can potentially shrinkage and disappear gross involved LNs and subsequently decrease the number of LN detection, dissection, and examination in gastric cancer. Conversely, in patients with early-stage gastric cancer, low-volume limited surgery may be associated with inadequate LN staging in these patients [17,25].
In Iran, many factors including the paucity of experienced surgical oncologist and pathologist and particularly infrequent extended LN dissection may contribute to insufficient LN assessment in gastric cancer.
The limitation of the current study were retrospective chart review of the patients' medical records, unknown operative details regarding surgical approach and LN location, non-uniform pathologic reports, and relatively small sample size of a single institution data based.

Conclusion
This study indicates that only one fifth of patients with gastric adenocarcinoma underwent sufficient lymph nodes assessment in Shiraz, Iran. As well, in this research, about 15% of the patients had no lymph nodes for assessment. Multicenter studies with larger sample size are suggested to confirm these results and to identify a true insufficient lymph node dissection from insufficient lymph node detection.