Storli K, Lindboe CF, Kristoffersen C, Kleiven K, Søndenaa K. Lymph node harvest in colon cancer specimens depends on tumour factors, patients and doctors, but foremost on specimen handling, APMIS 2010; 119: 127–34.
There are good indications that the number of lymph nodes found in the specimen after resections for colon cancer somehow has a bearing on prognosis. Many factors have been reported in the literature to influence lymph node retrieval. We wanted to assess these closer with special focus on the pathology handling process in our own practice. A range of international literature was reviewed to study what has been found to influence lymph node harvest. A questionnaire was sent to 13 renowned national and international institutions to explore their handling of the colon cancer specimens to obtain a histological diagnosis. A retrospective, hospital audit was undertaken to examine if the number of lymph nodes and staging after examinations of the specimens varied between individual pathologists. In the literature, tumour and patient characteristics, as well as the surgeon and the pathologist, are found to be influential, but it is difficult to ascertain which ones are truly essential. Fat solvents were found by several to increase the lymph node yield, although some also opposed this finding. Our questionnaire showed some variations in the routines of each Department. A junior pathologist was more likely to inspect the specimen first hand and not more than half employed specific lymph node detection strategies while three of 13 did not seek a minimum number of lymph nodes. Still every department had implemented a standard procedure for such examinations. The internal audit showed without doubt that the devotion of the pathologist secured significantly more lymph nodes from the specimen and this may also have detected more stage III cancers. Several tumour and individual patient characteristics, surgical approach and specimen handling may influence lymph node yield and theoretically, TNM staging. Our investigation specifically suggests that tissue handling by pathologists may be a prominent factor in lymph node harvest from colon cancer specimens.
Colon cancer is a prominent cause of death in the Western world. However, the prognosis has been surpassed by that of rectal cancer (1) after the surgical principle of total mesorectal excision (TME) has improved rectal cancer outcome (2). Recently, however, some centres have focused on colon cancer and shown improved results both for local recurrence and for survival (3). Their approach has been based on a radical principle that includes respect for the mesocolic planes in concert with that of TME for rectal cancer. Accordingly, this has been called complete mesocolic excision (CME) (3). In short, this includes following the mesocolic fascia, or Told’s, plane with removal of apical lymph nodes and a central tie of the feeding vessel, also called a D3 resection (or N3 node removal) in Japanese nomenclature (3, 4). Although it has been shown by one proponent of this technique that a high cut-off level of the number of lymph nodes may in addition indicate a better outcome, it remains obscure if the number of lymph nodes just mirrors radical, anatomical surgery with removal of more tissue or if some, as yet not fully identified, oncogene process is reduced or eliminated by removal of more lymph nodes.
Nevertheless, it has been reported that the number of lymph nodes harvested influenced prognosis even for TNM stage II (5, 6). In this setting, a minimum number of 12 nodes have been proposed as an ideal standard for lymph node harvest in colon cancer specimens (7) even though 18 or higher figures may be important (8). Under these circumstances, it is appropriate to examine if the lymph node harvest is affected by other factors too, such as patient and tumour characteristics, technical issues, the pathologist’s tissue handling and interpretation (9–11). However, there seems to be a lack of implementation of a standardized pathology procedure or awareness of specimen handling (9).
Our aim was to examine factors that may influence the harvest of lymph nodes in colon cancer specimens based on a literature search, a report of the procedure at 13 different international hospitals, and an audit of individual pathologists in a single national institution.
Materials and Methods
A specially designed questionnaire (by author CFL) about the routine treatment of colon cancer specimens was sent to 12 pathologists and one surgeon in 13 hospitals (Norway, n = 3; rest of Scandinavia, n = 3; the Netherlands, n = 1; UK, n = 1; Germany, n = 1; USA, n = 4).
Pathology quality control
Pathology records at one local hospital in Norway were studied retrospectively for 12 years from 1995 to 2006 to assess investigations carried out for colon cancer. The following criteria were recorded: gender and age, presence of adenocarcinoma, tumour location, colon specimen length, number of all lymph nodes and nodes with metastases.
A limited search was done in PubMed for relevant literature from 1943 to 2010. Only papers in the English literature were assessed.
It is well known that colon cancer has a poorer prognosis when it has metastasized to local lymph nodes (3, 12, 13). Consequently, it is an accepted principle that all tissue containing lymph node metastases should be removed in an effort to improve outcome (4). However, there are indications that removal of as many as possible negative nodes is also important (3, 7, 8). This should even improve the prognosis for TNM stage II tumours (5, 6, 13–15), although the oncological principle behind this remains unclear.
It can be seen from our international practice query that there is no complete agreement as to how the specimen should be handled to obtain as many lymph nodes as possible. A crucial point in aiming for a minimum number of lymph nodes, as is generally recommended, was only done by 10 of the 13 who answered this question. A minority had routine inspection of the specimen by a senior pathologist, and particular techniques for finding as many lymph nodes as possible were only used by no more than half of the Departments of Pathology. There was also a lack of uniform agreement on what should be regarded as a lymph node or discontinuous tumour growth. Still, all confirmed having implemented a standard procedure.
Our retrospective audit showed a great variation in the performance of individual pathologists during a 12-year period during which time, the routine work done by surgeons and pathologists remained unchanged. The recognition of such variations, however, later led to implementation of national standards of surgical treatment and registration of colon cancer from 2007. There were significant differences between two of the pathologists and the other eleven in our audit. Not only did two of the pathologists find more lymph nodes, the number of TNM stage III tumours was also greater. Our results are supported by a large US study that indicated that the two main factors behind a large lymph node yield are the surgeon and the pathologist (16). However, some oppose this and suggest that tumour factors are more important (17, 18). Our result may support the contention that identification of more lymph nodes is needed to detect metastases and a primary focus of histological examination is not to miss a tumour that should have been upstaged. Some feel that the true demonstration of stage III tumours can only occur following scrutiny of as many lymph nodes as possible (19), but a recent report contradicted this (20). Stage was not affected as long as the specimens contained at least seven lymph nodes for examination. Results from our audit would support that. However, the proportion of specimens with more than 11 lymph nodes varied from 21% to 44%. These figures are not acceptable, but it is noticeable that the number of patients still was at the same level as that reported by others (3). It is very unusual that the share of stage III patients exceeds 45% in study cohorts.
Several reports have shown that the pathologist holds the key to tissue examination (10, 14, 21). Specialized pathologists will more often find in excess of 12 lymph nodes in the specimen (9). Different departments may show variation in number of lymph nodes perhaps because of different procedures and implementation of guidelines (10, 14, 22, 23). The contribution of specialists and technicians alike has been shown to be decisive (21, 24). This suggests that dedication and meticulous work are important apart from solid experience or training. It is important to remember that tissue handling requires special skill and knowledge of how the lymph nodes can be detected (25, 26). Smaller lymph nodes may be missed by palpation and it is helpful to slice the mesentery at short intervals to inspect and detect lymph nodes that may be present.
Our literature search confirmed that the individual patient, the (open) surgical technique, pathology handling, tumour or specimen characteristics, and hospital volume, all seem to have an impact. The only issue that was not found to matter was the laparoscopic technique. Whether it is just as good or just as poor remains unclear. At least one study (27) did not report the number of lymph nodes harvested. From a pathologist’s perspective, only a couple of studies focused on the total number of lymph nodes in colon cancer specimens (10, 12).
The most important technical factor for harvesting more lymph nodes from the specimen is, from our experience and supported by and large in the literature, the use of fat solvents or lymph node detecting fluids, including the dedication and skill of the pathologist. True, some have not found this to be important, but all other factors being constant, this additional treatment of the specimen is capable of producing more lymph nodes in the mesentery when the fat is eliminated or the lymph nodes are clearly marked as shown in a randomized trial (28). The crucial issue is, of course, if this is only relevant in rare cases when skip lesions occur after jumping the nearest (N1) station (29). Upstaging has the potential to improve outcome by administration of adjuvant chemotherapy.
Many studies have shown that optimal fixation of the specimen is of paramount importance for the detection of lymph nodes. Specially designed fixatives, like GEWF (glacial acetic acid-ethanol-water-formaldehyde) solution, produce more nodes than the traditional formalin fixation (30–32), and fat clearance methods are probably superior to all other methods in this respect (33–38). However, the latter methods are time-consuming, toxic, and too expensive for routine use.
The essential question concerning tumour characteristics is really: why is it important to remove as many lymph nodes as possible? The burden of positive lymph nodes (6), the distance from the tumour by affected nodes, closeness to the resection line in the mesentery (Told’s plane) (3), apical nodes (D3/N3 nodes) and affected ‘retrograde’ nodes – have all been implicated in local recurrence and spread of tumour cells. Hohenberger et al. (3) found a cut-off value of 28 lymph nodes to be important. Is it because there are minimal oncological changes (micrometastases) in these apical or N3 nodes and will removal of these nodes halt tumour progress? Nevertheless, Prandi et al. (13) did not find any significance in an excess of 12 lymph nodes. It has been shown that removal of apical metastatic nodes will improve prognosis (3, 4). In contrast, a recent study found that the tumour burden or lymph node ratio was more important and survival remained unaffected by radical removal of nodes in these cases (39). However, the same authors recommended that as many negative nodes as possible in stage II tumours should be removed, but the oncological principle that should justify this approach remained obscure (39).
An increased chance of lymph node metastasis and death from disease is seen when tumour cells are observed in vascular structures (40, 41). The significance of tumour deposits (TD) in the mesentery remains unclear and controversial even though this finding has been incorporated in the 7th edition of UICC (41). It has the potential ability to upstage tumours in up to 8% by its influence on pN with its consequences for prognosis estimation. As pointed out in two recent publications (59, 60), this has been the cause of much concern among European pathologists, oncologists and surgeons, particularly because the definition has varied from edition to edition including both size more or less than 3 mm, and contour. Tumour deposits are present in 3–25% (41). A lump definition of TD increases the chance of inter-observer variability (41, 60). In addition, this affects pT with a choice of either T2 or T3 as shown by Nagtegaal and Quirke (60). The 5th edition was published in 1997, the 6th in 2002 and the last version in 2009 (62). A large US report did not take this change over time into account and thus raised criticism from two prominent European pathologists (60, 61). That a uniform and widely accepted classification system is essential for pathology reports and outcome comparisons goes without saying, and it has been suggested that quality assurance of the TNM classification should be implemented along the lines of clinical trials (61–63).
Until further studies explore these hypotheses with relation to patient survival, it seems prudent to accept that a minimum number of lymph nodes should be examined to discover effectively lymph node metastases in the specimen for correct staging and at least act as a surrogate marker for radical surgery that is presently the gold standard.
Apart from the amount of tissue removed, it has been shown that the use of solutions like GEWF, and the dedication and competence of the pathologist are crucial in harvesting as many lymph nodes as possible from the specimen. These factors may probably override the importance of individual patient and tumour characteristics. A standardized pathology procedure in a dedicated service therefore seems essential if patient outcome should be assessed based on optimal tissue handling. Surgeons should in turn make an effort to remove a necessary number of lymph nodes. This will help surgeons and pathologists alike to report outcome more comprehensively after surgery.
We are grateful to all the pathologists who answered our questionnaire and thus contributed to the study. The study was supported by grants from the Western Norway Hospital Trust, Haraldsplass Deaconal Hospital, and the University of Bergen.