| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
SUPPLEMENT |
From the Department of Anatomy, Sapporo Medical University School of Medicine, Sapporo, Japan.
Correspondence: Address correspondence and reprint requests to: Gen Murakami, MD, PhD, Department of Anatomy, Sapporo Medical University School of Medicine, South 1, West 17, Chuo-ku, Sapporo 060–8556 Japan; Fax: 81-11-618-4288; E-mail: chisa{at}sapmed.ac.jp
ABSTRACT
Gaps of the superficial cortex of the lymph node provide intranodal shunts that are more often the cause of skip metastasis than are collateral vessels. Examination of lymph nodes from cadavers of elderly subjects often revealed cortical gaps, especially in specific three-dimensional assembled cords; these cortical gaps were readily seen in para-aortic and pelvic nodes. This architecture seemed to be more appropriate for a systemic immune response than a local defense. Evidence of poorly developed cortices, anthracosis, and hyalinization also suggested impaired nodal function. We suspect that this histologic heterogeneity, perhaps a result of aging, affects the nodal trapping of colorimetric/isotopic tracers and metastatic cancer cells. This may have implications for lymphatic mapping of the sentinel lymph node in elderly patients with early-stage cancer.
Key Words: Aging • Cortex • Gaps • Lymph nodes • Skip metastasis
Our group has demonstrated an anatomical basis of skip metastasis.1 These metastases have been attributed to tumor-cell transfer via collateral vessels and/or a thick vascular trunk that bypasses one or more nodes of the lymphatic basin (Fig. 1A). Although some surgeons may believe that skip metastasis limits the application of the sentinel node (SN) procedure, we think that the SN concept accommodates physiologic variations in individual patients, including the possibility of skip metastasis. However, current SN techniques do not consider lymphatic vessels that drain directly to the thoracic duct or venous angle from the esophagus, pancreas, and thyroid gland (Fig. 1B).2,3 In this case, there is no intermediate nodal relay and therefore no SN. The use of esophageal endoscopy in combination with submucosal lymphangiography may allow modification of the SN procedure to identify direct lymphatic drainage vessels from the esophagus. In our experience the submucosal plexus is consistently dilated to over 100 µm in diameter in cases of direct drainage.4
|
|
MATERIALS AND METHODS
A total of 419 lymph nodes (40 cervical, 14 axillary, 154 thoracic, 205 abdominal and pelvic, and 6 inguinal) from 30 sites were dissected from 27 donated cadavers (12 Japanese males and 15 Japanese females, aged 72 to 95 years at death). One node was dissected from each site. During dissections, the efferent vessels of each node were labeled by carbon particles if possible. These cadavers carried no macroscopically visible tumors, and according to their death certificates, the causes of death did not include neoplasms. All nodal specimens were fixed with 10% formalin solution.
A total of 150 nodes were dissected from 10 separate sites in 15 other donated cadavers (9 white American males and 6 white American females, aged 68 to 93 years at death) and used as a comparison group. These 150 nodes were a gift from the late Professor Sandy C. Marks, Jr., of the Department of Cell Biology, Massachusetts State University Medical School.
After routine procedures to prepare paraffin-embedded blocks and complete serial sections (10-µm thickness), we performed hematoxylin and eosin (H&E) staining and immunohistochemistry for CD3 (DAKO, Glostrup, Denmark), CD20 (BioGenex, San Ramon, CA), and CD68 (DAKO). Morphometry of cortical areas was assessed with the aid of Photoshop (Adobe, San Jose, CA) and the National Institutes of Health (NIH) image program.
RESULTS
Specific Architecture of Small Nodes
Of the 419 nodes from the Japanese group, 12 nodes less than 3 mm in maximum diameter were the largest nodes encountered in the infrapyloric, retropancreatic, and pericolic regions and along the internal iliac artery. These nodes had a centrally located cortex. The subcapsular sinus was separated from the medullary sinus and identified outside the marginally located medullary sinus (Fig. 2B).
Nearly Empty and/or Ruined Nodes
In 18 mediastinal and intrapulmonary nodes and 12 paracolic nodes (Fig. 2C), the superficial cortex was thinner than the maximum diameter of a follicle. Gaps of the cortex were evident in these nodes. The paracortex area was often difficult to identify.
Fatty tissue infiltration into the medullary sinus was observed in 5 cervical nodes, 3 left gastric artery nodes, and 1 inguinal node. However, in contrast to nodes from elderly white Americans, fatty change rarely seemed to involve most of the node. Instead of nodes, we sometimes identified a capsule-bound bag filled with fat in the lower abdominal para-aortic region in elderly Japanese.
Hyalinization was another cause of a "ruined" node.12 The mediastinal and intrapulmonary nodes (72 of 154) and, notably, pelvic nodes consistently exhibited hyalinization, although the extent of this hyalinization varied. Mediastinal hyalinization occurred in the medullary sinus and was composed of thin collagen fibrils (40 to 50 nm in diameter, according to transmission electron microscopy). In contrast, hyalinization of the pelvic nodes consisted of very thick (100 to 150 nm) fibrils that branched diffusely along the perivascular area in the cortex (Fig. 3A). This type of hyalinization was sometimes seen in the abdominal and axillary nodes but was more restricted in its distribution within the node. Nodal specimens from elderly white Americans also often had extensive hyalinization in these regions. Hyalinization in mediastinal nodes was usually restricted to 20% of the entire node, but eight nodes at or around the tracheal bifurcation had large amounts of hyalinization that occupied more than 50% of the nodal volume. Twelve of the 46 pelvic nodes from elderly Japanese demonstrated hyalinization of 20% to 40% of the cortical volume.
|
Primary and Secondary Follicles
Primary follicles were almost always identifiable, even in small or nearly empty nodes. Their diameter ranged from 150 to 250 µm (mean, 189.0 µm). Large primary follicles were found in the pelvic nodes, whereas small follicles were seen in the gastric cardia and infrapyloric areas.
Multiple secondary follicles were seen in 25 of 154 mediastinal nodes and 12 of 252 abdominal and pelvic nodes. All 12 abdominal/pelvic nodes with secondary follicles were from 2 cadavers; medical histories for these 2 cases were not available. The secondary follicle was usually larger than the primary follicle, and its maximum diameter was 200 to 370 µm (mean, 263 µm).
Proportional Volume of the Cortex
The cortex and medullary cords, i.e., lymphocyte accumulation areas, usually occupied 70 to 80% of the node. Variations between individuals were evident, and we found region-specific differences (Fig. 4). The pulmonary hilar nodes and intrapulmonary nodes usually had a small cortex (almost 30%), whereas cervical, axillary, and upper mediastinal nodes tended to have a large cortex. Of note, the cortices of the cervical nodes were significantly (P < .01) larger in specimens from elderly white Americans (64.7 ± 9.4%) than elderly Japanese (55.9 ± 20.4%).
|
|
|
The present study demonstrated that the cortical shield along the afferent side of lymph nodes was commonly incomplete because of gaps in the thin superficial cortex, especially in thoracic and abdominal visceral nodes. Thus, an intranodal shunt for lymph flow (Fig. 2A) seemed to be frequent in visceral nodes from the elderly. It is likely that small, nearly empty, and/or "ruined" nodes have lost cortical functions, such as trapping of antigens. Severe anthracosis also seemed to impair trapping activities in the cortex and sinus. As a result, particles and/or antigens should be able to pass easily through modified visceral nodes.
Region-specific differences in the proportion of cortical volume seemed to be a critical indicator of nodal immune capacity. The intrapulmonary and pulmonary hilar nodes, which usually displayed severe anthracosis, had the smallest cortices of all nodes. Therefore, their protective capacity against lung cancers would be different from and perhaps less effective than that of the submandibular node against lingual cancer, because the submandibular nodes have the largest cortical volume in the body. Lymph nodes from axillary, cervical, and inguinal areas almost always had large cortical volume, whereas mediastinal, abdominal, and pelvic nodes were characterized by small cortices, with significant variations between individuals and unique architecture.
The three-dimensional tetrapod-like cortical architecture in the para-aortic nodes, seen as an island-like pattern two-dimensionally, seemed to allow free communication between the subcapsular and marginal sinuses throughout the node. This architecture may correspond to a cluster of small nodules, each of which is surrounded by a common sinus.13 This finding argues against functional units that are centered on the opening of an afferent lymphatic10; instead, in the island pattern, activities of the follicular dendritic cells may depend on any afferent route due to the common sinus. This unique architecture is likely to have greatly increased cortical surface areas in which various critical interactions occur between cortex and sinus components. A cluster formation of T and B lymphocytes and dendritic cells is one of the critical steps for cancer immunity.14 In contrast to the separated T- and B-cell areas in most experimental animals, such a cluster appears to form easily in the island-like nodal architecture because of the intermingling of lymphocyte areas and because of increased cortical surface areas. However, these factors also can accelerate inactivation of nodal T lymphocytes by the primary cancer.15 Because of the unique architecture with a common sinus, para-aortic node metastasis may indicate an increased likelihood that the cancer has already become a systemic disease. Suggested active cell-cell interactions in the island pattern or para-aortic nodes seem to be suitable for accelerating a systemic immune response rather than providing a local barrier. The para-aortic nodes may be specialized as the last gate along lymphatic drainage routes of the abdominal and pelvic viscera.
Why do clinical pathologists ignore cortical gaps or island-like nodal architecture? We hypothesize that, in cancer patients, even in the elderly, the histologic heterogeneity has already changed to a simple homogeneous morphology with numerous and/or large secondary follicles in the well-developed cortex. However, the SN procedure is essentially useful for finding nodal metastasis in the very early stages of cancer. Tracer particles as well as cancer cells are likely to migrate to nodes characterized by age-related architectural changes. Histologic heterogeneity may modulate homing of either the tracer or cancer cell. Trapping activities for tracers seem to be unequal among the multiple nodes within the lymphatic basin, even in proximal nodes. Consequently, region-specific modification or correction may be required to optimize application of SN mapping, especially for elderly patients with early-stage cancer.
FOOTNOTES
Histologic heterogeneity, perhaps a result of aging, may affect the nodal trapping of colorimetric/isotopic tracers and metastatic cancer cells. This may have implications for lymphatic mapping of the sentinel lymph node in elderly patients with early-stage cancer.
Received for publication November 4, 2003. Accepted for publication November 13, 2003.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
