Concepts about the spread of cancer, specifically of melanoma and of breast carcinoma, were spawned as a result of the work of Handley³ and Halsted⁴ near the beginning of the 20th century. Those two surgeons, the first an Englishman and the second an American, believed that cancer grew centrifugally from the site of the primary neoplasm, then moved in continuous fashion along lymphatics, and traveled subsequently in continuity along them to nodes. They did not believe that single neoplastic cells or groups of them actually entered the lumen of lymphatic (or blood) vessels whence they were transported via lymph to regional nodes and beyond them.

Thirty years later, the ideas of Halstead and Handley were challenged by Grey,⁹⁵ who had studied the relationship between lymphatics and the spread of cancer. He injected thorotrast into tissue within which lodged a malignant neoplasm, among those being carcinoma of the breast, the bowel, the skin, and the thyroid. Thorotrast filled the lymphatics and Grey was then able to study the relationship between lymphatics and neoplastic cells in sections of skin that housed the malignant neoplasm. In a few instances, neoplastic cells were seen in the lumen of lymphatics. In no instance, however, did Grey visualize a neoplasm growing as an uninterrupted mass from the site of the primary to regional lymphatics, as Halsted and Handley had insisted really happened. The observations led Grey to these conclusions: "The mode of spread of cancer to lymph glands is generally by lymphatic emboli of cancer cells carried along the lymph stream," and "Cancer cells generally do not remain for any length of time within the lumen of a lymphatic vessel."

Grey came to think that once cells of a cancer enter lymphatics they are swept rapidly to regional nodes, where they proliferate and grow. In time, the concept that a malignant neoplasm grew in "more or less interrupted or quite uninterrupted way" came to be understood as being wrong. Lymphoscintigraphy confirms the impression that a metastasis from melanoma spreads principally by way of lymphatic channels to regional nodes, which is the site where metastatic melanoma often manifests itself first, but is not the only site remote from the primary neoplasm where cells of melanoma are found.

Also in time it became axiomatic that melanoma progressed in a stepwise manner from its earliest development in the skin to its causing death by metastases to vital organs. In 1953, Meyer and Gumport⁸ wrote about melanoma coming to be present in nodes by virtue of metastasis and they did that in these words: " . . . this metastatic melanoma may be a focus and source from which malignant cells may spread to other nodes, or may enter the blood stream and cause distant metastases. These metastases, of course, will seal the fate of the patient."

Balch, et al.,³⁰ in 1996, made known their idea about the behavior of melanoma in regard to metastasis thus: "Our hypotheses was that melanomas metastasize first to regional lymph nodes and then to distant sites . . . " They reiterated that hypothesis four years later³¹ and explained it in this way: "The window of time for which early surgical intervention halts the further dissemination of melanoma metastases from regional to distant sites is estimated at 16 months, which is the average relapse time for patients who had wide excision only, and whose original nodal metastases subsequently evolved into clinical detectable disease." That estimate, namely, approximately 16 months from the time a metastasis of melanoma reaches a node to the node becoming apparent clinically, is not substantiated in the article by Balch and coworkers. Pack, et al.,⁷ in 1945, wrote this about the matter: " . . . we have shown that an average time of fifteen months elapses from the time the primary melanoma is recognized before metastasis in inguinal nodes are apparent." Although those latter authors averred that that figure derived from a series of patients with melanoma at Memorial Hospital in New York City, they do not cite results of a specific study as the basis for it. Balch, et al. may have taken the estimate of "15 months" from Pack, et al., an estimate that also was unverified. As has been the case for so much of what they have written about melanoma, Balch and collaborators not only were incorrect, but had engaged in mere supposition without producing any compelling data in support of it.

In the old-fashioned, conventional view of Balch and coworkers, once cells of a melanoma are released into lymphatics, they are carried forthwith to regional nodes, are retained there, proliferate there for a time, and, eventually, are released into the systemic circulation, whence they are disseminated widely. That hypothesis dovetails with the assumption that cells of melanoma are carried along with lymph to the first node in the path of a particular lymphatic, that node being the sentinel node, where they proliferate and then are released to secondary nodes before being spread widely. The fact that patients with a negative sentinel node as judged by conventional microscopy almost always have negative secondary nodes has been cited by Reintgen, et al.⁹⁶ and others as evidence that when melanoma metastasizes, the process is orderly, sequential, and, more or less predictable.

The principle of END also was predicated on the notion that there is a period of time during which cure of a patient can be accomplished by complete removal of the melanoma and of the nodes that harbor metastasis from it. END was thought to be beneficial because it was believed to be capable of achieving cure. That misapprehension arose from the fact that the procedure was carried out mainly on patients who did not have a metastasis of melanoma and, therefore, were cured by wide excision of the primary neoplasm, the END being redundant. It has been estimated that only 20% of patients considered eligible for END actually had metastasis of melanoma at the time they were deemed eligible.

Advocates of SNB were devotees, too, of the notion that metastasis of melanoma progresses in an orderly fashion. They argued that SNB permits identification of patients whose metastasis of melanoma was "localized," thereby enabling the nodal basin that housed the metastasis to be eliminated. That operation is performed on a select set of patients whose nodes are not involved clinically, and for that reason, the operation qualifies as "END." END has been shown, however, to have no benefit to patients and, that being the case, a SNB performed prior to an operation that, in essence, is an END cannot be of any benefit either.

In 1968, Cameron,¹⁰ in writing about the spread of melanoma and the mode whereby it metastasized, observed that (1) "spread within the lymphatics is by embolism and also by permeation" and (2) "Melanoma may gain access to the veins and at an early stage travel by the blood stream, causing widespread dissemination."

Nodes are not traps or dams, but filters; they have been shown to allow passage of fluid, air, erythrocytes, lymphocytes, bacteria and intact neoplastic cells into the efferent lymphatics and into the rest of the vascular system.^{34,37,38,97,98,99,100} Slingluff, et al.,²⁹ in 1994, made this statement about the issue: "It is likely that, as in breast cancer and other malignancies, the appearance of tumor in the regional nodes may be a marker of aggressive disease than a cause of subsequent systemic dissemination of disease."

Experience with mapping of lymphatics and with SNB has revealed that vital dyes and radioactive tracers pass through nodes and disseminate. Vital dye may pass through the nodes of a basin, which may preclude the identification correctly of a SN. The rate of this movement is sufficiently rapid that repeated injections of vital dyes into the dermis, for example, as often as every 20 minutes,⁴⁵ is requisite. Radioactive tracers may take 24 to 48 hours to disseminate, yet after just four hours it becomes difficult to distinguish the SN from secondary nodes. Morton and Chan,⁸⁷ in 2000, asserted that 20–30 minutes after intradermal injection of radioactive tracer, the SN is the only "hot" node; after that, radioactive material accumulates in the secondary nodes. In some instances, vital dye is not visualized in the SN, yet radioactive material still becomes concentrated in that node. The dye seems to disappear, but, in reality, it must go somewhere! Either it never gained access to lymphatics or it took a course different from that predicted for it.

Evidence that metastases are not held up in nodes for a significant period of time is derived from research on SNs. Metastasis of melanoma may manifest itself in a patient whose SN did not contain any cells of melanoma, no matter how rigorous the examination of sections of tissue stained by hematoxylin and eosin and by immunohistochemical stains. The metastasis may make itself known (1) in the same basin from which the SN was harvested, (2) nearby the site of the primary melanoma, i.e., a "satellite" metastasis, (3) en route to regional nodes i.e., an "in transit" metastasis, or (4) beyond the regional nodes, i.e., a "distant" metastasis. Molecular techniques confirm the randomness inherent in metastasis by demonstrating, using antigens particular to them, the presence of cells of melanoma in nodes, including sentinel ones, in peripheral blood, and in bone marrow. That these findings are important is evidenced by finding a metastasis in some of those patients in whom no cells of melanoma were found in SNs or in nodes harvested from the basin. A SN that does not appear to contain cells of melanoma after examination by conventional microscopy conveys no information about metastases elsewhere, whereas molecular biologic techniques could, conceivably, be sensitive enough to do that.

Patients with a positive SN and negative secondary nodes, and those with a positive SN and positive secondary nodes have metastases of melanoma that, in time, will become apparent clinically, irrespective of whether those patients have undergone END or received immunotherapy. A SN that contains cells of melanoma is a sure sign of widespread metastases, even if secondary nodes are negative for melanoma. No benefit accrues from attempting to predict the "extent of the melanoma" by the number of nodes involved or the number of cells of melanoma in a node. Those are mere exercises bereft of value. These scenarios (SNs and secondary nodes with/without melanoma) act only as vehicles for speculation about when metastases in an individual patient will make themselves known clinically; it may be days or decades, but, if the patient does not die of another cause, those metastasis will become manifest in time; the only uncertainty is when precisely.

Articles^50,51,78 telling of patients with metastasis of melanoma proven by SNB and in whom END was performed often are misleading because they indicate that patients do well following those surgeries when, in fact, the period of time of follow up is insufficient to permit any accurate judgment to be made about the effectiveness of those procedures. An international prospective randomized study, the Multicenter Selective Lymphadenectomy Trial, with Morton as the principal investigator and the John Wayne Cancer Institute as the organizing center,⁷⁰ currently is being conducted and no doubt will reveal that those operations, namely, a SNB with dissection of the entire basin if the SN contains metastasis of melanoma, have no efficacy at all because once cells of a melanoma enter lymphatics, they are likely to be swept to and through nodes or bypass them completely, either by way of lymphatic channels that anastomose or by entering the blood vascular system directly. For that reason the various terms chosen to convey progressively worse prognosis for metastasis of melanoma, to wit, "satellite," "in transit," "regional," and "distant" are deceptive and wholly without merit. A metastasis is a metastasis, with all that that implies in regard to prognosis, to wit, being grim. In fact, a "satellite" metastasis is a sure indicator of "distant" metastases.

That lymph nodes do not halt malignant cells in their tracks has been demonstrated unequivocally by studies performed recently on carcinoma of the breast³⁶ in which cells of the carcinoma were observed in bone marrow by conventional microscopy, even when regional nodes were negative for carcinoma. This fact, together with molecular studies carried out on tissue of patients with melanoma in whom metastases were not detectable by conventional techniques, indicates that once cells of a malignant neoplasm are released into the circulation, either lymphatic or blood vascular, or both, they disseminate rapidly and widely, and do not advance in stepwise fashion from node to node. When, exactly, metastases will manifest themselves clinically, however, cannot be gauged with any degree of confidence.

Almost without exception, articles that deal with the utility and validity scientifically of SNB or END for patients with melanoma refer to metastases as "recurrences." According to Webster's New Universal Unabridged Dictionary,¹⁰¹ the word "recur" means "to occur again as an event, experience, etc." Patients with so-called recurrence of melanoma, do not, in actuality have a recurrence, but rather a metastasis which was not detected clinically. Those patients certainly were not cured by SNB and END, and moreover, neither of those procedures did anything to limit the extent of the disease. They merely provided evidence of whether or not a particular node harbored neoplastic cells, and that judgment was compromised by limitations imposed by examination of tissue by conventional microscopy.

In brief, the word "recurrence" is used loosely. It refers to a metastasis that was there already and only just now became apparent clinically. Moreover, the very same word "recurrent" is employed universally by surgeons and pathologists, as well as dermatologists, for persistence of melanoma at the local site by virtue of incomplete excision of it. There is no relationship between persistence locally and metastasis widely; prognosis is not necessarily grim for persistence of melanoma (it may be wholly in situ), whereas prognosis is grave for metastasis of melanoma. In short, "recurrent" and "recurrence" have no place in the parlance of behavior of melanoma.

All of this is made even more dizzying by Balch and acolytes of him. Balch has written unabashedly about the matter thus: " A local recurrence is defined as any tumor that occurs within 5 cm of the scar of a previously excised melanoma. "In actuality, he includes both persistence locally and metastasis locally (a satellite metastasis sets within 5 cm. of the primary melanoma) under the rubric of "local recurrence." (C.M. Blach, et al. Cutaneous melanoma. Second Edition, 1992. J.B. Lippincott Company, Philadelphia, p. 288). If those who established the rules blur boundaries so badly, how can those physicians' charged soberly with responsibility for management of patients with melanoma be expected to play by them as they endeavor to carry out their task competently.