In 1925, Kreibich, working in the German Clinic for Dermatology in Prague, told of a 20-year-old man who presented himself with an asymptomatic condition that, over the course of nine months, had come to affect the buttocks, thighs, arms, abdomen, and head.⁴⁴ He described the lesions as being red to violaceous plaques with a yellowish hue; some were typified by prominence of follicles in a manner reminiscent of comedones. A plug of horny material, along with some sticky fluid, could be expressed readily from ostia of those follicles. Other lesions showed wrinkling of their surface in a fashion said to be like the crinkling of cigarette paper, that condition being superficial atrophy (Figure 3A ).

In sections of tissue of biopsy specimens taken from the patient, Kreibich observed the following (translated from the German): "Mucin is present in abundance within the follicular epithelium where an infiltrate of inflammatory cells is noticeable. Mucin and round cells of the lymphocytic type, along with an occasional plasma cell, also are present around the follicles. Hair shafts are degenerated and cyst-like dilations of follicles are topped by parakeratotic comedones." (Fig. 3B ). Kreibich noted, too, that round cells in infiltrates present within follicular epithelium were larger than those around blood vessels of the superficial plexus. In addition to making astute observations, he reviewed literature pertinent to the finding histopathologically of deposits of mucin in skin, but could find no report in which changes similar to those in his patient were recorded.

Kreibich did recognize, however, that his patient had presented himself clinically in a manner very much like that of a disease known in those days as "érythrodermie pityriasique en plaques disséminée" or "Brocq's disease," which later came to be called "parapsoriasis en plaque" and which nowadays is regarded as one expression of a patch of mycosis fungoides.

In 1939, Lehner and Szodoray wrote of their experience with a 20-year-old man who had had extensive disease of the skin of the trunk for what was claimed to be two months.⁴⁵ The lesions were oval patches and plaques punctuated by individual follicular papules, many of which had become confluent (Fig. 4A ). The patient complained about pruritus, but apart from that was without other symptoms and was in good health.

Findings in sections of tissue of a biopsy specimen taken from this patient were catalogued by Lehner and Szodoray thus (translated from the German): "Marked changes are present in the follicles, whereas the epidermis is not involved in the process. Infundibula are dilated and filled with keratotic material. The epithelium of the root sheath is destroyed; keratinocytes are edematous and show vacuolar alteration. Cysts filled with metachromatic fibrillar substance have formed within the epithelium, in which many leukocytes are visualizable, too. Lymphocytes and eosinophils are situated in the vicinity of follicles. Sweat glands seem to be normal." (Fig. 4B and C ).

Lehner and Szodoray acknowledged that they did not know how, exactly, to classify the disease in their patient, but for them the considerations were scleroderma, scleredema, myxedema, and keratosis follicularis squamosa Dohi. They understood, full well, however, that the changes in their patient, clinically and histopathologically, differed from those in all the conditions they had included as possibilities in differential diagnosis. The coworkers called attention to the fact that the disease in their patient was very similar to the one in the 20-year-old man who had been a patient of Kreibich and concluded that both patients had the same disease, one that they regarded as being singular and not like any other disease yet described. They made no comment about Kreibich's suggestion that the malady belonged in the spectrum of "parapsoriasis," although the oval patches pictured by them on the trunk of the patient they reported on seem to us to be consonant with that general category of disease, to wit, mycosis fungoides.

When H. Pinkus, in 1957, transmitted his observations of six patients with what he called alopecia mucinosa, he was completely unaware of the articles that had appeared decades before in German by Kreibich and by Lehner and Szodoray.¹⁵ He thought, therefore, that the disease in his patients marked by peculiar histopathologic findings had not been described previously and he gave the name alopecia mucinosa to it in order to emphasize what for him were the two most striking aspects, namely, alopecia clinically and mucin in hair follicles histopathologically (Figs. 5A –D ). The plaques, as Pinkus characterized them, were red, only slightly elevated, sometimes scaly, and typified further by the presence of dilated ostia of follicles, some of which were plugged by horny material and marked by absence of hair. The lesions were localized mainly to the head and neck, but in some patients were distributed over the trunk and extremities; in most instances, the patient had been cognizant of the condition for several weeks prior to a dermatologist having been consulted, the presenting complaint usually being loss of hair within the plaques.

According to Pinkus, the findings histopathologically in alopecia mucinosa were a dermal infiltrate of lymphocytes, eosinophils, and histiocytes (the latter mononuclear cells sometimes being so numerous that they conveyed an impression of granulomatous inflammation), prominent spongiosis in the epithelium of folliculosebaceous units (the cells in the sebaceous glands having become separated from one another and, in the process, assuming a stellate shape), and mucin in what was claimed to be the external root sheath (the acid mucopolysaccharide there being said to be so abundant that it resulted in formation of cystic spaces). Pinkus stated unambiguously that he was unaware of any disease similar to the one he was bringing to the attention of colleagues. Nowhere in his article did Pinkus mention mycosis fungoides.

In 1957, Pinkus presented his observations about alopecia mucinosa at the International Congress of Dermatology that convened in Stockholm where Braun-Falco, from the audience, commented that he, himself, had made similar observations on patients of his own. Shortly thereafter, Braun-Falco published those findings in the Dermatologische Wochenschrift under the title "mucophanerosis intrafollicularis et seboglandularis," that designation being derived, in part, from the Greek, "phaneros" (meaning "visualizable").⁴⁶ By that term, Braun-Falco attempted to advance his hypothesis that mucin comes into being as a consequence of keratocytes having undergone necrosis. He remarked that findings histopathologically in the skin of his four patients were nearly identical to those reported on by Pinkus. Lesions in two of the patients of Braun-Falco were restricted to the face (Fig. 6A ). In one patient, a widespread condition of longstanding had been diagnosed clinically and histopathologically as mycosis fungoides. In the fourth patient, red plaques at different anatomic sites for 30 years had been diagnosed as "eosinophilic reticulosis" (no explanation being given for what that diagnosis was meant to communicate, but the implication of it seems to be lymphoma).

Braun-Falco held that there were two kinds of what he called mucophanerosis intrafollicularis et seboglandularis: a primary or idiopathic type (Fig. 6B ) that corresponded to that which had been described by Pinkus, and a secondary or symptomatic type (Fig. 6C ) that was associated consistently with a particular dermatosis, the examplar of which was mycosis fungoides.

Braun-Falco reviewed the work of Kreibich and of Lehner and Szodoray, assigned Kreibich's patient to his second symptomatic group, and placed Lehner and Szodoray's patient in the primary idiopathic group, even though the lesions in both patients were widespread and very similar to each other. In the discussion of his findings, Braun-Falco referred to a report by Korting that had appeared a few months before publication of his own article — and in the same journal. Korting had recounted the peculiar course of a 52-year-old man who had mycosis fungoides for eight years, at which time, in addition to patches and plaques typical of that disease, he developed plaques dotted by numerous follicular papules.⁴⁷ The lesions, situated on the scalp and eyebrows, were alopecic. This is how Korting described the findings histopathologically (translated from the German): "Follicles are dilated and surrounded by an infiltrate of lymphoreticular cells and numerous eosinophils and mast cells. Metachromatic mucoid substance is present in marked amount within the follicles. Focally, the epithelium of follicles is destroyed completely, the follicle is ruptured, and a foreign body reaction is apparent." (Fig. 7 ).

Braun-Falco considered the patient of Korting to qualify for his own secondary or symptomatic category of "mucophanerosis intrafollicularis et seboglandularis" and he proceeded to criticize Korting for not having endeavored to give meaning to the findings he, Korting, had written about. Korting, in fact, had interpreted the changes as being those of one manifestation of mycosis fungoides. The legend to Korting's photomicrograph states that explicitly as follows (translated from the German): "Mycosis fungoides with mucoid degeneration of follicular epithelium."

In 1958, W. Johnson, at the Armed Forces Institute of Pathology, reviewed findings, contained in sections of eight biopsy specimens, that were typical histopathologically of alopecia mucinosa.⁴⁸ Most of the patients had presented themselves clinically at first with but a solitary lesion and only later did the lesions become widespread. In the majority of patients, the disease was typified by progression slowly and regression equally slowly over a period of years. Although many of the patients were thought by Johnson to show features reminiscent clinically of large plaque parapsoriasis, in none of them did he encounter changes that he could identify as those of mycosis fungoides histopathologically. It was not surprising, therefore, that Johnson concluded that "alopecia mucinosa as a clinical and pathologic entity is primarily a change in the epithelial cells within the pilosebaceous follicle, which results in the accumulation of mucin, the inflammation in the corium being secondary." For him, the "recurrent nature" of the condition seemed to be indicative of a metabolic process like that of myxedema.

In 1959, Kopf and Steagall shared their experiences with a 36-year-old man who had alopecic patches and plaques on the face, chin, and scalp (Fig. 8A ). In addition to features typical histopathologically of alopecia mucinosa, as that condition had been characterized by Pinkus, the authors observed hyperkeratosis, acanthosis, and inflammatory cells scattered both in the dermis and in the epidermis, where, in the latter site, they were said to destroy the basal layer (Fig. 8B ).⁴⁹ For them the diagnosis was alopecia mucinosa; they seemed unimpressed by involvement of the epidermis by lymphocytes.

Neither large plaque parapsoriasis nor mycosis fungoides was mentioned by Kopf and Steagall as possibilities diagnostically, only seborrheic dermatitis and lupus erythematosus having been raised as considerations clinically.

In the same year, 1959, Jablonska and collaborators reported on a patient who had skin lesions in which abundant mucin was thought by them to be restricted to follicular epithelium.¹⁶ The patient presented himself with infiltrated plaques that were slightly erythematous, those on the face and scalp being alopecic (Figs. 9A –B ).

A sticky fluid could be expressed easily from the follicles. Despite a variety of efforts therapeutically, the disease continued to progress, albeit slowly. By virtue to their study of that single patient and from their review of the reports by Lehner and Szodoray, Pinkus, and Braun Falco, Jablonska and coworkers concluded that the disease named alopecia mucinosa by Pinkus and mucophanerosis intrafollicularis et seboglandularis by Braun-Falco was the one borne by their patient (Fig. 9C ). They were critical of the designations given to the condition by both Pinkus and Braun-Falco, and proposed instead the title "mucinosis follicularis" ("follicular mucinosis"). In their experience, lesions of the condition at sites other than the scalp did not show signs, consistently, of alopecia. Moreover, Braun-Falco's term "mucophanerosis intrafollicularis et seboglandularis" was deemed by them to be too abstruse to be useful. They agreed, however, that the disease, irrespective of the name given to it, was highly distinctive, but not related to mycosis fungoides.

In 1960, Haber published his observations about alopecia mucinosa as it presented itself in six patients, all of whom had slightly erythematous and alopecic patches, plaques, and nodules marked histopathologically by mucin in infundibular and follicular epithelium and by dilated infundibula filled with a keratotic plug. In the majority of those lesions, he noted infiltration by round cells, as well as spongiosis, acanthosis, and parakeratosis.⁵⁰ This led Haber to conceive of alopecia mucinosa as a form of "follicular eczema" that came into being either independently or in the course of a chronic dermatosis, such as seborrheic dermatitis, atopic dermatitis, and a reticulosis that could be benign or malignant.

One of Haber's pictures (Fig. 10A ) shows a lesion typical clinically of mycosis fungoides. His photomicrographs exhibit a range of findings characteristic of follicular mucinosis in so-called alopecia mucinosa, Fig. 10B displaying copious mucin in follicular epithelium and Fig. 10C plugs of corneocytes that not only stuff dilated infundibula, but protrude far above the surface of the skin in the manner of a caricature of keratosis pilaris.

Haber commented that in Fig. 10D , in which mucin is prominent in infundibula, the "infiltration in the cutis suggested localized lymphoma." Nevertheless, he thought that "alopecia mucinosa . . . is a . . . form of eczema of hair follicles," a notion he reinforced by stating that "In follicular mucinosis, there is spongiosis which leads to vesicle formaion, as in eczema." Haber, like everyone who had written about the subject before him, garbled it by mixing up terms, as exemplified by the synonymy found in his statement, "alopecia mucinosa, or better still, follicular mucinosis."

In 1961, Findlay and Loewenthal told of their own experience with six patients whose lesions possessed findings of alopecia mucinosa in company with what they called eosinophil granuloma, lichen spinulosus, and chancriform lesions (Figs. 11 A–C).⁵¹ The coworkers also reviewed all reports published until that time about mucin in follicles.

In a Table, the collaborators listed all patients who, to their knowledge, had been diagnosed until then as having alopecia mucinosa. In that Table they included the differential diagnoses that had been offered clinically for the lesions in those patients, the most common of those being parapsoriasis en plaque, reticulo-granuloma eosinophilicum, nummular eczema, eczematid, mycosis fungoides, and eosinophilic reticulosis. The authors declared that "alopecia mucinosa has features of premycosis and mycosis fungoides without so far as we know developing the irreversible changes expected in those conditions" and that alopecia mucinosa is "potentially a reticulo-granuloma in its own right with a benign course."

Findlay and Loewenthal mentioned, too, that "alopecia mucinosa may be associated with many non-specific eczematoid changes" and that "the clinical alopecia may be minimal in cases where there is at present no other diagnosis possible than alopecia mucinosa." (Fig. 11D ). By believing alopecia mucinosa to be a "reticulo-granuloma in its own right," the authors implied that they disagreed with the idea of Braun-Falco, and of others later, that that condition could be divided neatly into primary (idiopathic) and secondary (symptomatic) types. By invoking the word "potentially" in regard to "reticulo-granuloma" and by enunciating the phrase "associated with many non-specific eczematoid changes," however, Findley and Loewenthal not only failed to establish solidly the actual nature of alopecia mucinosa, they contributed further to confusion about it.

An article, in 1962, by Kim and Winkelmann, communicates, once again, how vast was the muddle that results from imprecise use of terms by those who had engaged the subjects of follicular mucinosis and alopecia mucinosa.⁵² Among the 10 patients of Kim and Winkelmann was one they pictured with lesions typical clinically of large plaque parapsoriasis (Fig. 12A ) and another who bore frank tumors of mycosis fungoides (Figs. 12B–C .). Yet because of findings histopathologically of follicular mucinosis, the authors insisted that the diagnosis was not mycosis fungoides, but alopecia mucinosa (Fig. 12D ).

Parenthetically, Kim and Winkelmann opined that when mycosis fungoides is accompanied by follicular mucinosis, the follicular involvement is secondary to the primary lymphomatous process.

As another parenthesis, the coworkers recounted having observed a boy who had developed groups of follicular papules on the arms and legs, lesions that were found histopathologically to exhibit follicular hyperkeratosis and mucinosis. On the basis of findings met with later in lymph nodes, the boy was diagnosed as having Hodgkin's disease; he died of systemic lymphoma the same year. Curiously, Kim and Winkelmann thought the presence of follicular mucinosis in this youngster with unquestionable lymphoma was mere coincidence. In a discussion that followed presentation of a paper of Plotnick in 1962,¹⁷ it turned out that one of the patients reported on by Kim and Winkelmann, and diagnosed at first as having alopecia mucinosa, developed what later was diagnosed as undubitable mycosis fungoides. New lesions had continued to appear, old ones had continued to progress, and new biopsies continued to be performed until, at last, a diagnosis of mycosis fungoides became undeniable. The patient of Plotnick (Figs. 13A-C ) had had a course very similar to that of the patient of Kim and Winkelmann; lesions that were widespread evolved from plaques to nodules and it was just a matter of time before a diagnosis of mycosis fungoides became obvious.¹⁷

In the years that followed, many articles given to the subject of alopecia mucinosa and of follicular mucinosis, and of their significance vis-à-vis risk for developing lymphoma, had in common a particular theme as follows: a patient with typical mycosis fungoides clinically but follicular mucinosis histopathologically, that is, presence of mucin in infundibula (and often in sebaceous lobules), was diagnosed as having alopecia mucinosa, despite the fact that lymphocytes peppered the epidermis in a manner typical of mycosis fungoides.^53–59 Some who wrote about that seeming disconnect recognized the illogic of what appeared to be patches and plaques of mycosis fungoides clinically interpreted to be alopecia mucinosa/follicular mucinosis histopathologically, the latter propelling them headlong to a final diagnosis of idiopathic alopecia mucinosa. Hyman, in 1961, commented at a meeting about a patient with widespread patches and plaques like those of mycosis fungoides but with changes histopathologically of follicular mucinosis, those latter findings preventing the presenters, as well as all the other participants at that particular session, from issuing a definitive diagnosis of lymphoma.¹⁸ In closing remarks of the discussion that followed the presentation, Hyman advocated the concept that mucin in what he regarded to be follicular epithelium was merely a pattern histopathologically and not a distinctive disease sui generis.

In 1964, Pinkus presented attributes morphologically of four patients whose lesions of alopecia mucinosa he thought had progressed to lymphoma, that malignant neoplastic process usually being mycosis fungoides. One of the four was the same patient Plotnick had published about previously.⁶⁰ The clinical lesions were erythematous or brownish plaques that became infiltrated increasingly, eventually becoming tumors, some of which underwent ulceration. Sections of tissue from the biopsy specimen obtained from the patient who also had been reported on by Plotnick did not seem to contain any abnormal lymphocytes, but such cells were observed, and in large number, in sections from succeeding biopsy specimens. Pinkus remarked that in some instances of alopecia mucinosa the uppermost part of the follicle (which actually was infundibular epidermis, Pinkus being of the belief that the infundibulum represented the uppermost part of the hair follicle) and even the perifollicular epidermis might be engaged in the process. He stated, too, that "the inflammatory infiltrate, although mild in some cases, usually has a peculiar character that approaches granulomatous inflammation on one hand, and proliferative disease of the reticuloendothelial system on the other." Last Pinkus averred that in the context of the finding of follicular mucinosis, only three possibilities for differential diagnosis clinically existed, those being: "(1) idiopathic alopecia mucinosa that may occur (a) in children and young adults, or (b) in persons over 40 years, (2) lymphoblastoma with (secondary?) follicular mucinosis, and (3) alopecia mucinosa transforming into lymphoblastoma." He continued to insist that the term "alopecia mucinosa" was preferrable to "follicular mucinosis" for characterizing what he conceived to be a distinctive inflammatory condition that had to be distinguished from "the histologic phenomenon of follicular mucinosis which may also be found with other connotation." That last statement, typical of so many others of others about the subject under discussion here, did not help clarify matters. Pinkus did not make clear what, precisely, the changes were in sections of tissue that allowed a diagnosis of follicular mucinosis to be made histopathologically, nor did he provide any criteria clinically that enabled a diagnosis of "idiopathic alopecia mucinosa" to be rendered with confidence. In the view of Pinkus, alopecia mucinosa in some patients was typified by an "inflammatory stage" that, for reasons unknown, transformed into a malignant neoplastic process, his speculation being that because those patients were older and their immune system defective, they were unable to hold the inflammatory process in check and that was the explanation for why it proceeded to become neoplastic. Unfortunately, no pictures of lesions clinically were shown in the article of Pinkus.

After the publication of Pinkus, the number of reports about patients who at first had been diagnosed as having alopecia mucinosa and later were found to have mycosis fungoides and about patients who, from the outset, seemed to have alopecia mucinosa in conjunction with mycosis fungoides began to increase greatly.^61–74 More and more, it became accepted that the primary/idiopathic form of alopecia mucinosa might present itself as patches, follicular papules arranged in a group, plaques, nodules, and, at times, even necrotic masses,^62,75–82 all of which were indistinguishable clinically from lesions of doubtless mycosis fungoides (Fig. 14A –C ). Astonishingly, none of the lesions, even the necrotic tumors that were accompanied histopathologically by signs of follicular mucinosis, were diagnosed as being authentic mycosis fungoides; the diagnosis invariably was either alopecia mucinosa or follicular mucinosis.

The categorization of alopecia mucinosa into two types, as proposed by Braun-Falco (primary/idiopathic or secondary/symptomatic), or into three types, as suggested by Pinkus (idiopathic alopecia mucinosa, lymphoblastoma with follicular mucinosis, or alopecia mucinosa transforming into lymphoblastoma), not only was accepted without resistance, but was repeated, again and again, by other authors. Still, differentiation between and among those "types" proved to be extremely difficult because no criteria for distinguishing among them ever had been set forth convincingly. It became obvious that one type of alopecia mucinosa that looked, more or less, the same, clinically and histopathologically from patient to patient, "transformed" eventually into lymphoma, either into mycosis fungoides^83–88 or another kind.^89–95 Although some students of the subject were struck by the dynamic of the "process of transformation," it sometimes occurring rapidly⁹⁶ and sometimes slowly over the course of as many as 20 years,⁷¹ none proposed that what was being called alopecia mucinosa truly was, from the very beginning, mycosis fungoides.

Curious it is, too, that an association of alopecia mucinosa with lymphoma came to be recognized for sure in young adults and even in children,^97–99 but Pinkus remained adamant that alopecia mucinosa always was a self-limited inflammatory disorder of younger persons.

It was Tappeiner, in 1967,¹⁰⁰ who proposed that the fundamental process pathophysiologically of so-called primary/idiopathic and of secondary/symptomatic follicular mucinosis might be one and the same, but, in the end, he decided that the process likely was the result of a localized deficiency in perfusion of blood and not a consequence of malignant neoplasia. A similar vantage was offered in 1969 by Emmerson, who reported on findings of a major study undertaken by him on 47 patients who had been diagnosed as having follicular mucinosis (Figs. 15A–E ). He provided follow-up data on 40 of them.¹⁰¹ In 10 patients, the disease persisted for years and did not evolve to certifiable mycosis fungoides. In eight patients, follicular mucinosis was accompanied from the outset by a "reticulosis," four of those patients later dying of lymphoma. Emmerson stressed that in all the patients a diagnosis of lymphoma had been made prior to, or at the same time of, diagnosis histopathologically of follicular mucinosis and that he was unable to observe anything that signalled a transition from follicular mucinosis to mycosis fungoides.

Emmerson was induced, therefore, to conclude that there was no "direct temporal relationship" between the course of the lymphoma and the development of follicular mucinosis. This is how he put it: "the two processes were occuring alongside of one another, and that the combined picture [as found in some patients] resulted from the merging of the two pathological processes rather than any more direct relationship." Although Emmerson implied that follicular mucinosis (he used that term as a synonym for alopecia mucinosa) when it developed idiopathically might be the very same disease as when it appeared in patients with lymphoma, he excluded any possibility of a relationship between follicular mucinosis and lymphoma. Emmerson regarded the presence, together, of follicular mucinosis and mycosis fungoides as mere coincidence, stating as he did that "it seems very unlikely that the 2 processes [follicular mucinosis and overt reticulosis] are closely related or interdependent" and cautioning that failure "to regard the two as entirely separate" would abet confusion about a subject that already had risen to a level of incomprehensibility. Confounding, too, is the statement by Emmerson that although malignant cells may be present also in the damaged follicles, their presence there does not imply that the changes in the follicles have resulted from an invasion of them by malignant cells (Figs. 15F–G ).

To Emmerson's credit, however, was his statement about changes histopathologically in sections of tissue "read out" originally as "alopecia mucinosa" or "follicular mucinosis" in patients who, years later, had a biopsy that produced a specimen from which sections showed signs, incontrovertibly, of mycosis fungoides (Figs. 15H –J ). Emmerson pondered the matter thus: "It would be of interest to know whether reappraisal of the original sections might in fact show changes of reticulosis which had been previously overlooked."

Of particular interest in the ever more intriguing saga of alopecia mucinosa was a patient reported on in 1969 by Felman and colleagues. Those coworkers had the opportunity to witness an autopsy of a 49-year-old man who had been diagnosed as having alopecia mucinosa (Fig. 16 ). The patient had presented himself with what seemed to be a pruritic papular eruption that continued to progress despite application topically of corticosteroids.¹⁰²

Findings histopathologically were interpreted as being those of a nonspecific dermatitis. A year later, the patient developed plaques that showed changes by conventional microscopy of follicular mucinosis in association with a dense infiltrate of rather uniform histiocytes, as well as eosinophils. Shortly thereafter, the man developed nodules and tumors, some of which ulcerated. Radiation therapy was begun and some lesions began to wane, but, after a few months, the tumors reappeared, hepatosplenomegaly developed, and the patient died. At examination post mortem, lymphoma was found to be systemic, involving the skin, lung, small bowel, and bone marrow. From the vantage of the authors, the lymphoma, which they said had occurred "in association" with alopecia mucinosa, "originated" in the skin. Similar was the situation of a patient of Grosshans et al. who had been diagnosed as having follicular mucinosis and who died shortly afterward from the effects of systemic lymphoma. For those French colleagues, follicular mucinosis was a paraneoplastic process.¹⁰³

In summary, when, in 1925, Kreibich reported on a patient with widespread plaques, some of which were made up largely of follicular papules, he noted similarities clinically between those lesions and ones of large plaque parapsoriasis.⁴⁴ Kreibich was the first to record the finding of mucin in infundibular and follicular epithelium. Lehner and Szodoray knew of the report of Kreibich and when, in 1939, they published their observations about the same subject, they acknowledged their debt to him.⁴⁵ H. Pinkus, however, did not review scrupulously the literature pertinent to alopecia in general and to alopecia typified by mucin in follicles in particular so that it was inevitable that when, in 1957, he wrote about "alopecia mucinosa," he made no mention of the articles by Kreibich and by Lehner and Szodoray.¹⁵ Moreover, Pinkus did not appreciate the similarities clinically between what he called alopecia mucinosa and what surely were patches and plaques of mycosis fungoides. Only years later, after he, himself, had seen patients who had evidences concurrently of mycosis fungoides clinically and of follicular mucinosis histopathologically, did he acknowledge that there might be a relationship between what he called alopecia mucinosa and mycosis fungoides. Braun-Falco mentioned, correctly, that some patients with alopecia mucinosa did, indeed, have mycosis fungoides,⁴⁶ but he separated the former condition into primary (idiopathic) and secondary (symptomatic) types, which, for almost half a century, became an insuperable impediment to comprehension of the subject. The new name proposed by him for the condition known already as alopecia mucinosa, to wit, "mucophanerosis intrafollicularis et seboglandularis," not only failed to clarify the essential character of the condition, but it became yet another curiosity in a lexicon of dermatology already burdened by imponderable phrases. Jablonska and coworkers were right to criticize Braun-Falco for muddying matters, but they went wrong when they proposed that the term follicular mucinosis be used as a synonym for alopecia mucinosa.¹⁶ Pinkus and other students of the subject agreed with Braun-Falco on the merit of dividing alopecia mucinosa into primary and secondary types, an idea that is flawed irreparably because everything is secondary to something and what is called primary cutaneous lymphoma almost certainly is secondary to a fundamentally systemic disease. In the more than four decades that have ensued since Pinkus, Braun-Falco, and Jablonska et al. voiced their opinions about the same subject, no author has rejected outright classifying alopecia mucinosa as primary and secondary. Neither did Johnson who observed a striking similarity clinically between lesions of the purported primary type of alopecia mucinosa and those of parapsoriasis,⁴⁸ implying by the use of that latter imprecise term a relationship of alopecia mucinosa to mycosis fungoides. And neither did Kopf and Steagall, and many others, who took note of lymphocytes in the epidermis in conjunction with the changes of follicular mucinosis,⁴⁹ connect those findings to mycosis fungoides.

Findlay and Loewenthal,⁵¹ Tappeiner,⁶⁵ and, later, Emmerson,¹⁰¹ for different reasons, were the only ones who thought alopecia mucinosa was a single condition consequent to the effects of a single pathologic process, but all of them fell victim to the ideas of Pinkus and of Braun-Falco to the effect that there were two basic types of alopecia mucinosa, namely, one inflammatory and the other neoplastic (lymphomatous). The vast majority of other authors who wrote about the matter continued to interpret the association of alopecia mucinosa in patients with mycosis fungoides as being pure coincidence, a precursor of lymphoma (alopecia mucinosa), or authentic lymphoma (usually mycosis fungoides). Some considered alopecia mucinosa to be a paraneoplastic process.

The notion of Haber, enunciated in 1961,⁵⁰ that follicular mucinosis could be found in sections of tissue from specimens extracted from a variety of diseases led Hyman, in 1962, to articulate the concept that follicular mucinosis simply was a finding histopathologically and not a specific disease.¹⁸ That view was shared by Pinkus, who, also in 1962, admitted that follicular mucinosis could be encountered in a several conditions, among those being cutaneous lymphoma. That recognition prompted him to propose that "the term alopecia mucinosa should be used for the idiopathic clinical disease," whereas follicular mucinosis (mucinosis follicularis) should be restricted to "the histologic picture that may be found as a primary change in alopecia mucinosa or as a symptomatic feature in certain cases of mycosis fungoides."¹⁷ Although Pinkus in these statements telegraphed an advance in his thinking beyond that which he had expressed in his seminal article in 1957, it is evident from the sentence just quoted that he still had not integrated effectively information then available about the subject. Failure to grasp the essence of things in a unified way is indicated by Pinkus' reference, repeatedly, to phrases such as "the idiopathic clinical disease," "the histologic picture of [mucinosis follicularis] that may be found as a primary change in alopecia mucinosa," and follicular mucinosis that can be found "as a symptomatic feature in certain cases of mycosis fungoides."

In 1978, H. Wolff, Kinney, and Ackerman lent support to the idea that a distinction should be made between the skin disease "alopecia mucinosa" and the histopathologic finding "follicular mucinosis." They did that in a report on an incidental finding of follicular mucinosis noted by them in a lesion of angiolymphoid hyperplasia.¹⁰⁴ They advocated the proposition that deposits of mucin in what they believed to be follicular epithelium was just one of several very different and distinctive patterns histopathologically of cutaneous epithelium in general. In the first edition of his text titled Histologic Diagnosis of Inflammatory Skin Diseases published that same year, Ackerman championed the same argument, to wit, that "follicular mucinosis is but one of many examples of morphologic expression of altered epithelial metabolism in the skin" and that "confusion would be avoided if the term follicular mucinosis was [sic] used only histologically."³⁵

Seven years later, in 1985, Hempstead and Ackerman made a clear distinction between follicular mucinosis, a pattern of epithelium recognizable histopathologically, on one hand and alopecia mucinosa, a disease of the skin, on the other.¹⁹ They maintained that follicular mucinosis should be employed only as a description for a particular finding histopathologically (what now is appreciated to be mucin in the infundibulum, the follicle below it, and the sebaceous gland), that phenomenon being analogous to other distinctive patterns of cutaneous epithelium histopathologically, in particular, epidermolytic hyperkeratosis, acantholytic dyskeratosis, cornoid lamellation, and pale-cell acanthosis (Figs. 17A –F ).

Hempstead and Ackerman thought, incorrectly, that what was called alopecia mucinosa was an inflammatory disease associated with follicular mucinosis and not related in any way to lymphoma, whereas what was termed follicular mucinosis was a finding histopathologically that might be encountered episodically in mycosis fungoides, in a variety of other conditions, and as an epiphenomenon (Figs. 17G–H ).

In spite of these efforts to clarify terminology in regard to follicular mucinosis and alopecia mucinosa, those two terms continued to be used interchangeably by virtually every author^{73,78,105–114} (Figs. 18A –C ). In 1979, Wilkinson and coauthors brought further confusion to the subject by introducing the designation "follicular mucinosis (lymphoma)"¹¹⁵ for histopathologic findings in plaques, situated on the head and neck, made up of follicular papules. The diagnosis of lymphoma was predicated on atypical mononuclear cells in sections of tissue from a biopsy specimen taken from one of those lesions.

Terminology about the matters under discussion here was rendered yet more unfathomable in an article by Truhan and Roenigk in 1986 in which the authors set out to elucidate the issue of "the cutaneous mucinoses."¹¹⁶ Throughout their piece, Truhan and Roenigk used follicular mucinosis as a synonym for alopecia mucinosa and included it (follicular mucinosis/alopecia mucinosa) along with myxedema, lichen myxedematosus, reticular erythematous mucinosis, scleredema, cutaneous focal mucinosis, and myxoid cyst among the "primary mucinoses," by which they meant a primary "metabolic" process and not a secondary "catabolic" one. Nowhere, however, did they inform what they intended by the words "metabolic" and "catabolic," and nowhere did they make a distinction between "epithelial mucin," such as that generated by keratocytes in "follicular mucinosis," and "connective tissue mucin," like that manufactured by fibrocytes in scleredema. Despite a trenchant critique by Alajlan and Ackerman, published in 2000, of the notion of cutaneous mucinoses,¹¹⁷ that term and concept continues to be employed unbridled.¹¹⁸

This patient presented himself with two lesions on the face which, according to Pinkus, indicated benignancy, but, in reality, this patient had, and has, mycosis fungoides.⁷⁶

By the 1970s, some authors did endeavor to establish criteria for distinguishing "primary" from "secondary" alopecia mucinosa. For example, J. Civatte et al., in 1973, took note of hypertrophy of folliculosebaceous units affected by follicular mucinosis.¹¹⁹ Those coworkers thought that the hypertrophy was more marked in so-called primary follicular mucinosis and, for that reason, proposed that the finding of hypertrophy allowed the primary type of alopecia mucinosa to be differentiated from the secondary type. Others who attempted to determine the source of the mucin in follicular mucinosis, such as did Langner, Jablonska, and Darzynkiewitcz in 1969, could observe no difference between the primary and secondary manifestations in regard to the mechanism whereby mucin came to be produced (which, in their view, resulted from degeneration of keratocytes, rather than from production of mucin by them).¹²⁰ Interestingly, Zambal, in 1970, drew attention to the fact that no matter whether the expression of follicular mucinosis was "primary" or "secondary," the amount of mucin in folliculosebaceous units varied extraordinarily, even in sections of biopsy specimens removed from the very same lesion said to be that of alopecia mucinosa.¹²¹ Studies by electron micoscropy failed to enlighten about where mucin was produced in lesions of follicular mucinosis in patients who had, or did not have, mycosis fungoides.^{110,122–124} Nonetheless, each of the investigators persisted in the claim that the mucin in follicular mucinosis seemed to originate from cells of follicular epithelium, especially those of the outer sheath.

In 1983, Lancer and coworkers undertook what they characterized as "a detailed morphologic and immunopathologic study" of follicular mucinosis.¹²⁵ They examined sections of tissue of biopsy specimens taken from two patients, both of whom had been diagnosed by them as having the "clinically benign form" of follicular mucinosis, a diagnosis they utilized as a synonym for alopecia mucinosa. Lancer et al. found "degenerative changes" in the outer sheath and in sebaceous glands, always in company with an infiltrate of lymphocytes that they said looked "morphologically benign." Their studies by electron micoscropy and direct immunofluorescence added nothing of substance to the understanding of the condition under surveillance here, although immunoperoxidase techniques did reveal the infiltrate to consist of T lymphocytes mostly, thereby prompting Lancer et al. to conclude that "cell-mediated immune mechanisms may play a role in the pathogenesis" of alopecia mucinosa.

In 1985, Nickoloff and Wood made the assumption that an infiltrate confined to follicular, perifollicular, and perivascular zones in the absence of Pautrier's "abscesses," plasma cells, and eosinophils enabled "benign" alopecia mucinosa to be distinguished from "malignant" alopecia mucinosa, that is, the type associated with lymphoma.¹²⁶ In 1988, Gibson and colleagues came to a different position based on the results of a study of 59 patients who had been diagnosed histopathologically as having follicular mucinosis, 19 of whom had mycosis fungoides and two Hodgkin's disease.¹²⁷ Gibson et al. were not able to identify any finding (lymphocytes, eosinophils, exocytosis of lymphocytes, hyperplasia of the epidermis, and deposits of mucin) that enabled follicular mucinosis affiliated with lymphoma to be distinguished from follicular mucinosis unrelated to lymphoma. These observations forced the coworkers to admit that "no single clinical or histopathologic observation predicts which patients with follicular mucinosis will have a benign course." In another article devoted to this general subject, Gibson wrote about his experience with follicular mucinosis in children and emphasized that, in that setting, too, the condition may persist in an inocuous way for years or it may be joined by signs of lymphoma, Hodgkin's disease especially.¹²⁸ In this regard, the authors contended that it mattered not a whit whether follicular mucinosis had presented itself first as a solitary plaque or as widespread lesions. In 1991, Mehregan and collaborators told of a review of 33 patients with follicular mucinosis, nine of whom had mycosis fungoides and three of whom had lymphoproliferative disorders of another kind, such as Hodgkin's disease and chronic lymphatic leukemia.¹²⁹ They agreed fundamentally with the conclusion of Gibson et al. to the effect that concerning follicular mucinosis in adults, no criteria histopathologically permitted a distinction to be made between lesions of follicular mucinosis in patients with lymphoma and those in patients who did not have lymphoma.

Ackerman et al., in one volume (the second edition of Volume II) of four dedicated to the issue of differential diagnosis in dermatopathology, set forth criteria for differentiation histopathologically of alopecia mucinosa from mycosis fungoides in which follicular mucinosis was accompanying.¹⁰ The authors regarded alopecia mucinosa as an inflammatory disease unrelated to mycosis fungoides, claiming that absence of cerebriform nuclei of lymphocytes, absence of lymphocytes within the epidermis, presence of a dense infiltrate of lymphocytes, and involvement of nearly every follicle in a section of tissue were findings indicative of alopecia mucinosa, rather than of mycosis fungoides accompanied by follicular mucinosis. The clinical pictures and photomicrographs presented by them show hardly any difference between what the authors perceived to be two unrelated diseases, to wit, alopecia mucinosa (an inflammatory process) and mycosis fungoides (a lymphomatous process).

Recently, new techniques have been utilized to investigate the issue of "follicular mucinosis/alopecia mucinosa" and its relationship to cutaneous lymphoma. For example, in 1991, Zelickson published his analysis of gene rearrangement of T-cell receptors in cutaneous lymphoma. The results of his study indicated that not every patient with mycosis fungoides or Sezary's syndrome had clonal rearrangement of T-cell receptors, and, moreover, in conditions thought to be benign or premalignant, as follicular mucinosis was conceived by him to be, clonal rearrangement of T-cell receptors could be detected.¹³⁰ In 1993, Weinberg, on the basis of his review of articles about diagnosis using techniques of molecular pathology for skin lesions of what he called "primary" follicular mucinosis, concluded that evidence of clonal rearrangement of T-cell receptors was in favor of the condition being a lymphoproliferative disorder.³⁸ In 1997, Willemze and coworkers published what they called the "EORTC Classification for Primary Cutaneous Lymphomas," in which they categorized lymphomas based on study of 626 patients. Although they made no reference to alopecia mucinosa having any relationship whatsoever to mycosis fungoides, they nonetheless listed mycosis fungoides with follicular mucinosis as a specific variant of mycosis fungoides. According to them, that variant was: "characterized by the presence of folliculotropic infiltrates with sparing of the epidermis, mucinous degeneration of hair follicles and preferential involvement of the head and neck area." No criteria were set forth for enabling distinction to be made by conventional microscopy between alopecia mucinosa and mycosis fungoides with follicular mucinosis, but features that allowed that particular variant to be distinguished from mycosis fungoides, unmodified, were said to be "folliculotropism instead of epidermotropism." The proponents claimed that clonal T-cell receptor gene rearrangement was found in most examples of mycosis fungoides with follicular mucinosis, but not in alopecia mucinosa.¹³¹ Jackow, in 1997, told of having encountered follicular mucinosis together with expansion of oligoclonal T-cell receptors, inferring therefrom that clonal T-cell expansion alone might not be sufficient to permit a particular disease to be classified as benign or malignant.¹³² Wittenberg, one year later, reported on follicular mucinosis in four young adults whose course was protraced; clonal rearrangement of the T-cell receptor was demonstrable in two of those patients, but not in the other two.¹³³ Curiously, those four patients presented themselves not with patches, as described originally by Pinkus, but with "acneiform" lesions on the face that were reminiscent of acne vulgaris or of rosacea. Nevertheless, the authors interpreted the changes, clinically and histopathologically, to be those of the same disease that had been described by Pinkus in 1957. That in these patients infundibular mucinosis might be a finding unrelated to lymphoma was not given any credence by the authors.

Hodak, in 2000, communicated the results of her study of seven patients diagnosed as having unilesional mycosis fungoides, one of whom had findings histopathologically of both mycosis fungoides and follicular mucinosis in the same section of tissue.¹³⁴ She concluded that unilesional mycosis fungoides is a variant of cutaneous T-cell lymphoma. Although only a single lesion was present at the time Hodak's report came to be published, attributes catalogued by her clinically, histopathologically, and immunophenotypically were indistinguishable from those in the same condition when it was widespread.

In sum, wrong terminology contributed mightily to confusion ongoing about the matters of follicular mucinosis and alopecia mucinosa. Even as of this writing, the terms alopecia mucinosa and follicular mucinosis are used interchangeably and synonymously in nearly every textbook of dermatology, general pathology, and dermatopathology, despite sustained efforts at denunciation of that practice and refutation of reasons marshalled on behalf of engaging in it.^19,35,104 No repeatable, dependable criteria have been established for enabling clinical features and histopathologic findings in patients with "primary" alopecia mucinosa to be distinguished from those with "secondary" alopecia mucinosa, not electron micoscropy, not immunofluorescence, not immunohistochemistry, and not employment of molecular techniques like T-cell receptor gene rearrangement. Moreover, many articles devoted to the subject suffer from lack of precise clinico-pathological correlation, and that is why patients in whom infundibular mucinosis is an incidental finding in diseases as disparate as allergic contact dermatitis, rosacea, and so-called eosinophilic folliculitis are grouped with patients who have unquestionable mycosis fungoides under the same rubric "follicular mucinosis," that designation being descriptive of a distinctive pattern of epithelium and not of a particular disease.

By 1983, 16 years after publication of the article in which he had introduced the term alopecia mucinosa, Pinkus was in receipt of more data concerning the very same patients who had been diagnosed as having that condition by him originally.¹³⁵ At that later time he acknowledged that in several patients lesions of alopecia mucinosa not only had become widespread, but plaques and tumors of sure mycosis fungoides had become manifest. Pinkus now advised that alopecia mucinosa should be "referred to as a potential first stage of mycosis fungoides," that assessment being offered by him irrespective of the fact that he, himself, had experience with only three of the many patients with alopecia mucinosa reported on whose lesions had seemed to progress to undoubted mycosis fungoides. Pinkus went on to question whether alopecia mucinosa is "one disease or several" and stated that for him it remained an "unanswerable question." Contrary to the opinion of Pinkus, however, was that of Winkelmann, who, shortly afterward, in 1984, in a chapter in a volume concerning "Controversies in Dermatology," edited by Ervin Epstein, Sr., wrote that "alopecia mucinosa or follicular mucinosis has a definitive relationship to mycosis fungoides" and that it is "so indicative of mycosis fungoides in the adult that it is considered often as a skin manifestation of mycosis fungoides."¹³⁶

Kanno and colleagues, in 1983, reviewed the entire Japanese literature devoted to follicular mucinosis and contributed to it two patients of their own whose lesions, according to them, had developed into lymphoma.¹³⁷ Reassessment of sections from the first biopsy specimen of one patient, diagnosed at first as having benign follicular mucinosis, disclosed findings typical of follicular mucinosis but in conjunction with a "few atypical lymphocytes." The review included assessment of 64 patients with follicular mucinosis, in six of whom signs of lymphoma were accompanying. Four of those six patients had been diagnosed at the beginning as having benign follicular mucinosis, but within three years each of them had developed unquestionable lymphoma. In spite of the finding of atypical lymphocytes in "early" lesions of follicular mucinosis, Kanno et al. concluded that although follicular mucinosis may be a "forerunner" of malignant disease, it is not in itself a malignant disease.

Sentis, in 1988, transmitted his experience with two patients who were diagnosed initially as having alopecia mucinosa in the form of a facial solitary plaque that sported attributes of follicular mucinosis histopathologically.¹³⁸ Within a few years, both patients developed dramatic signs of mycosis fungoides, that is, widespread lesions clinically and changes histopathologically that were stereotypical of that particular lymphoma (Figs. 20A–C ).

The same patient developed on the trunk patches and plaques (C) that are reminiscent of lesions as they were characterized by Kreibich, Lehner and Szodoray, and Kim and Winkelmann.¹³⁸

When sections of tissue from the first biopsy specimen of one patient were studied retrospectively, some lymphocytyes with large, hyperchromatic, cerebriform nuclei became apparent (Figs. 20D–E ). The authors interpreted to be progression of a benign condition to a malignant lymphoma, yet they illustrated that advance with clinical photographs that show, in sequence, a plaque on the forehead that enlarged continuously over time (Fig. 20B ), becoming more and more exophytic.

In the 1980's, the notion was propagated that some patients with Sézary's syndrome had follicular mucinosis as the only change apparent histopathologically in their skin. As an example of that thesis, Fairis et al., in 1987, reported on a patient with follicular mucinosis who presented himself clinically with an erythroderma¹³⁹ that did not, in their judgment, merit a diagnosis of Sézary's syndrome because no changes of mycosis fungoides could be detected by conventional microscopy, only findings of follicular mucinosis. They did, however, observe in the perpheral blood of the patient some "abnormally shaped activated T-cells." That same year, Rivers told of having seen follicular mucinosis in the erythroderma of Sézary's syndrome, the diagnosis of that syndrome having been made by him on the basis of findings clinically and hematologically, namely, an erythroderma that was pruritic and Sézary cells that were present in the blood.¹⁴⁰ Yet again, changes histopathologically in this patient were said not to be diagnostic of mycosis fungoides, sections of tissue from an alopecic plaque on the scalp showing only follicular mucinosis. LeBoit and collaborators, in 1988, recounted a similar situation in which a 12-year-old girl had an erythroderma and circulating Sézary cells.¹⁴¹ In the skin, the findings histopathologically were those of follicular mucinosis. In sections of tissue cut from a biopsy specimen removed from the erythroderma, no abnormal lymphocytes or any involvement of the epidermis by lymphocytes was noticeable. Studies using immunohistochemical techniques revealed that the infiltrate within and around follicles consisted of T cells mostly. A clonal pattern of rearrangement of the T-cell receptor beta gene was interpreted by the authors as confirmation of their diagnosis of a T-cell lymphoma. They were not sure, however, whether their patient truly had Sézary's syndrome or "another unique lymphoproliferative disorder."

Another peculiar finding was brought to attention in 1987 by Berger, who reported on a patient with findings histopathologically of follicular mucinosis and, in the same section of tissue, signs of hyperplasia of eccrine units.¹⁴² A similar condition had been recorded in 1984 by Vakilzadeh and Brocker,¹⁴³ who termed it "syringolymphoid hyperplasia with alopecia." They told of a man who, on a leg, had a hairless patch studded with reddish-brown papules and that, in addition, exhibited follicular hyperkeratosis and anhidrosis. Sections from a biopsy specimen of skin showed infiltrates of lymphocytes in the dermis and around hyperplastic sweat glands and ducts. In 1992, Burg and Schmoeckel¹⁴⁴ commented on a patient with a solitary lesion characterized histopathologically by hypertrophy of eccrine ducts and a lymphoid infiltrate of T helper phenotype, a condition for which they proposed the term "syringotropic variant of mycosis fungoides," a concept that came to be shared by others.¹⁴⁵ Tomaszewski et al., in 1994, wrote of a patient who, for more than 13 years, had an alopecic plaque on the scalp in which were noted syringolymphoid hyperplasia, an infiltrate of lymphocytes around eccrine units and follicules, and follicular mucinosis.¹⁴⁶ Clonal rearrangement of T-cell receptor genes was detected in the plaque, leading the coworkers to propose that so-called syringolymphoid hyperplasia, in actuality, is a syringotropic form of mycosis fungoides. In 1999, Tannous et al. remarked on changes in the skin of a patient with a cutaneous T-cell lymphoma that expressed itself as patches and plaques typical of mycosis fungoides (Fig. 21A ).¹⁴⁷ Sections of tissue showed findings diagnostic of mycosis fungoides, as well as changes of follicular mucinosis and hyperplasia of eccrine glands and ducts, those latter structures being surrounded by an infiltrate of atypical lymphocytes (Figs. 21B–C ).

In 2000, Rongioletti and Smoller published the results of a study of sections of tissue from biopsy specimens taken from 104 patients with mycosis fungoides, their purpose in that undertaking being to assess involvement of adnexal structures by lymphocytes.¹⁴⁸ They found eccrine glands and ducts to be infiltrated by lymphocytes in lesions of 30 % of those patients and hyperplasia of eccrine units in 5 % of them. Follicles were affected in 57 % of patients and follicular mucinosis was present in 8 % of them. The authors averred that infiltration of adnexal structures by lymphocytes might be a "useful diagnostic tool" for diagnosis of mycosis fungoides, especially when epidermotropism by lymphocytes is slight.

Confusion reigns not only about primary and secondary alopecia mucinosa and the role played by follicular mucinosis in them, but also about the relationship between follicular mucinosis and alopecia mucinosa on one hand and "folliculotropic mycosis fungoides" on the other. That infiltrates of lymphocytes in mycosis fungoides may involve hair follicles is not a new revelation; Kim, while a visiting fellow (of Ackerman) in the dermatopathology unit at New York University Medical Center in 1985, that year wrote a Letter to the Editor (also Ackerman) of the American Journal of Dermatopathology in which he told of having observed numerous atypical lymphocytes in the epithelium of hair follicles and no such lymphocytes in surface epidermis.¹⁴⁹ In 1993, Lacour commented on a patient whose alopecic patch was characterized histopathologically by lymphocytes with hyperchromatic cerebriform nuclei, folliculotropism by dense infiltrates of mononuclear cells that spared surface epidermis, Pautrier's "microabscesses" in infundibular epithelium, comedo-like plugging of infundibula, and folliculotropism of lymphocytes.¹⁵⁰ The author asserted that some patients with mycosis fungoides purported to have alopecia mucinosa clinically had no mucin in follicles of those lesions. Goldenhersh et al., in 1994, introduced the term follicular mycosis fungoides for what they noted in a single patient, namely, widespread plaques of mycosis fungoides, some of which were dotted by follicular papules.¹⁵¹ The authors paid heed not only to perifollicular infiltrates of lymphocytes, but to folliculotropism of lymphocytes, thereby motivating them to stress how different were the findings they themselves had encountered from those of follicular mucinosis affiliated with mycosis fungoides.

In 1997, Gilliam et al. memorialized their experience with a patient at the tumor stage of mycosis fungoides who had striking changes of follicular mucinosis histopathologically.¹⁵² Strangely enough, the authors diagnosed their patient as having "folliculotropic mycosis fungoides . . . presenting as dissecting cellulitis of the scalp," and they went on to claim priority for having been the first to report on that phenomenon. When Pereyo et al., the same year, shared the results of their study of three patients with what they called follicular mycosis fungoides, they stated that the disease is a "rare variant" of cutaneous T-cell lymphoma, even though long before their article had been published it was known that infundibula, and even follicular epithelium, may be affected by infiltrates of mycosis fungoides.¹⁵³ The authors underlined the importance of distinguishing follicular mycosis fungoides from follicular mucinosis, that task being accomplished, in their estimation, by scrutinizing follicular epithelium for the presence or absence of mucin. Pereyo and cohorts admitted that, in fact, such distinction clinically could be extremely difficult, the reason being that both conditions present themselves usually as follicular papules and alopecic plaques.

Klemke, in 1999, told of findings in a patient with "follicular mycosis fungoides" who presented himself with patches interrupted by comedo-like keratotic plugs, as well as with plaques, nodules, and alopecia of the scalp.¹⁵⁴ A perifollicular and folliculotropic infiltrate of pleomorphic lymphocytes, large Pautrier's "abscesses" in follicular epithelium, and infundibular hyperkeratosis were recognizable readily, but follicular mucinosis could not be detected, a situation that had been met with by others previously.^155–159

In sum, the concept of alopecia mucinosa as an inflammatory disease that presents itself first as patches and plaques that, over time, may or may not progress to nodules and tumors is untenable in the context of all the evidence now available (Figs. 22A–B ). Also futile is the attempt to distinguish alopecia mucinosa from mycosis fungoides with follicular mucinosis according to age of the patient, as Pinkus insisted could be done, young patients being claimed by him to have a self limited benign course (alopecia mucinosa), whereas adults were at risk of progressing to lymphoma (mycosis fungoides with follicular mucinosis). That thesis was given the lie as more and more young persons with alopecia mucinosa were noted to develop over time obvious lymphoma. Puzzling, as well, was the supposed "association" of follicular mucinosis histopathologically with Sézary's syndrome in which no evidence of mycosis fungoides could be found by conventional micoscropy. Some authors even went so far as to exclude a diagnosis of Sézary's syndrome in erythrodermic patients whose peripheral blood contained atypical lymphocytes simply by virtue of the finding of follicular mucinosis histopathologically, that latter change propelling them to a diagnosis of follicular mucinosis rather than of mycosis fungoides. "Syringolymphoid hyperplasia with alopecia" has been noted in association with changes of follicular mucinosis. Hyperplasia of eccrine glands and ducts, especially the latter, together with clonal T cells within both components of eccrine units has been interpreted as "syringotropism" of neoplastic lymphocytes of mycosis fungoides, an assessment that we deem to be reasonable. It has yet to be stated straightforwardly and unambigously, however, that the condition called alopecia mucinosa is an expression of mycosis fungoides and that follicular mucinosis in these patients comes into being by effects of neoplastic lymphocytes on epithelial elements in the skin (the infundibular epidermis, sebaceous lobules, and the follicle all being affected) in a manner similar to that of syringotropism of lymphocytes and subsequent hyperplasia of eccrine structures. It is particularly enigmatic that never has that statement been made, directly and unequivocally, when it is known that clonal rearrangement of T-cell receptors was detectable in patients who had been diagnosed as having alopecia mucinosa. The term follicular mycosis fungoides is advantagous only in the sense that it acknowledges that cutaneous T-cell lymphoma may be affiliated not only with epidermotropism, but with folliculotropism (as well as sebaceotropism and eccrotropism).

Assertions about alopecia mucinosa and mycosis fungoides, even in the very most recent publications, are contradictory. Flaig et al., in 2001, addressed the matter of follicular mycosis fungoides in nine patients,¹⁶⁰ in all of whom they recognized a folliculotropic infiltrate of lymphocytes mostly, those cells exhibiting nuclear atypia (Figs. 23A–B ). The degree of epidermotropism varied considerably, even when biopsy specimens were taken from different lesions in a single patient. The same seemed to be true for the degree of follicular mucinosis present in different lesions of a single patient. The authors suggested that "follicular mycosis fungoides and follicular mucinosis-associated mycosis fungoides belong to the spectrum of mycosis fungoides with a predilection of [sic] hair follicles" and that "the term follicular mucinosis-associated mycosis fungoides should be replaced by follicular mycosis fungoides." What the authors claimed to be involvement of follicles by lymphocytes of mycosis fungoides, however, was involvement mostly of infundibular epidermis.

Haller et al., also in 2001, reported on a patient whom they had diagnosed as having syringolymphoid hyperplasia with alopecia and anhidrosis.¹⁶¹ They interpreted the basic disease to be "a syringotropic variant of mucinosis follicularis" and "a facultative precursor lesion of mycosis fungoides."

In 2002, Van Doorn and coworkers wrote about 51 patients with follicular mycosis fungoides, the prognosis of which they found to be worse than that of patients with classic mycosis fungoides.¹⁶² The lesions were present on the head and neck, and consisted of follicular papules, some of which were acneiform, alopecic, and associated with "mucinorrhea." Severe pruritus was said by them to be an oft-experienced symptom. The clinical photographs pictured show changes highly reminiscent of what is claimed to be the stereotypical presentation of alopecia mucinosa (Figure 24A ).

When scoured by conventional microscopy, the sections of tissue were said to be characterized by perifollicular infiltrates of lymphocytes and by folliculotropism of lymphocytes, but not by epidermotropism (Fig. 24B ). Eosinophils and plasma cells were apparent in the infiltrate. Follicular mucinosis was identifiable in lesions from all 51 patients. In synchrony with the conclusion of Flaig et al., the authors advised that it was not possible to distinguish between follicular mycosis fungoides with follicular mucinosis and follicular mycosis fungoides without of follicular mucinosis. That being the case, they urged that both should be called follicular mycosis fungoides.

Campanati, shortly after the publication of van Doorn et al., wrote about a patient with lesions marked by both folliculotropism of atypical lymphocytes and follicular mucinosis. They contended that that phenomenon (follicular mycosis fungoides in company with follicular mucinosis) was rare, in contrast to the position of Flaig et al. and van Doorn et al. who had found that that combination of findings was common (Figs. 25A–B ).¹⁶³

Also in contrast to Flaig et al. and van Doorn et al., Monopoli and coworkers, as recently as April 2003, on the basis of their observations in two patients, proposed that follicular mycosis fungoides devoid of follicular mucinosis be separated from that manifestation of mycosis fungoides that presents itself with follicular mucinosis as determined histopathologically.¹⁶⁴

Later in 2002, Cerroni et al. reported on a study of 44 patients with follicular mucinosis, 28 of whom (68%) had signs, clinically and histopathologically, of T-cell lymphoma.¹⁶⁵ No criteria, by inspection grossly or examination microscopically, enabled them to distinguish the patients with lymphoma from those who did not have lymphoma. Not even rearrangement of T-cell receptor gamma chain, which was found in about 50% of both groups, namely, the one with proven lymphoma and the one in which no lymphoma could be proven, was decisive in regard to that differentiation. The coworkers came to conclude that so-called idiopathic follicular mucinosis might be a localized form of T-cell lymphoma rather than a distinctive inflammatory disease (Figs. 26A–F ). They proposed that idiopathic follicular mucinosis "belongs to the variant forms of mycosis fungoides that show a prolonged, nonaggressive clinical course," akin to that of the localized variant of mycosis fungoides known as pagetoid reticulosis (Woringer-Kolopp disease).¹⁶⁵

Cerroni and associates made a worthy effort at untying the knots of this Gordian conundrum by stating that "one must question whether idiopathic and lymphoma-associated follicular mucinosis are unrelated diseases . . . or rather different names for a single disease . . ." Nevertheless, they continued to make a distinction between a manifestation of mycosis fungoides associated with follicular mucinosis in which lesions are widespread and an expression in which the disease presents itself in "localized" fashion, they claiming that that latter variant was associated consistently with a favorable prognosis.

Contrary to the assessments of Cerroni et al., however, were those of Brown et al., who, in 2002, too, communicated their experience with seven patients who for several years had had follicular mucinosis.¹⁶⁶ All of those patients had more than one lesion, in three of them the lesions being widespread. Findings histopathologically in lesions from all seven patients were said to be typical of follicular mucinosis. Rearrangement of T cells clonally was detected in five of the seven patients. The authors insisted that none of the patients had any sign clinically of cutaneous T-cell lymphoma and, therefore, they were led to conclude that primary follicular mucinosis is a "clonal disorder with limited or 'benign' cutaneous manifestations." A photograph of lesions in one of their patients shows patches of what surely is mycosis fungoides (Figs. 27A–B ).

The findings of both sets of authors, Cerroni et al. and Brown et al., prompted letters to be written to the editor,^167–169 but despite the colloquy that ensued, no agreement was reached in regard to what, precisely, was the essential character of alopecia mucinosa.

As recently as June 2003, Lee and coworkers published their findings histopathologically in Ofuji's disease (eosinophilic pustular folliculitis). In 7 of 13 Taiwanese patients with Ofuji's disease, deposits of mucin were discernable in infundibulosebaceous epithelium, thereby posing a problem in differential diagnosis histopathologically of what they called alopecia mucinosa. The authors averred that the amount of mucin seemed to be less impressive in Ofuji's disease than in alopecia mucinosa, and eosinophils were more striking in Ofuji's disease, some of those granulocytes even forming collections within epithelial structures. Lee et al. conceived of alopecia mucinosa as consisting of at least two different conditions, namely, a neoplastic process (mycosis fungoides) and an inflammatory disease. Nonetheless, they acknowledged the interpretation of Cerroni et al. that both of those conditions might represent different manifestations of the very same process, to wit, mycosis fungoides.¹⁷⁰

In brief, Flaig et al. and van Doorn et al. agreed with each other that no distinction need be made between follicular mucinosis associated with mycosis fungoides on one hand and follicular mycosis fungoides with or without follicular mucinosis on the other. Van Doorn, however, conceived of follicular mycosis fungoides as being associated consistently with a poor prognosis.

Cerroni et al. thought of "follicular mucinosis" (which for them is synonymous with "alopecia mucinosa") as a "localized cutaneous T-cell lymphoma," rather than as a disease unrelated wholly to lymphoma. Although this interpretation acknowledges, correctly, how flawed is the concept of a type of alopecia mucinosa that is an inflammatory disease, it implies that there are different "variants" of mycosis fungoides and that each of them is associated consistently with a particular prognosis. Cerroni and collaborators did not emphasize strongly enough that so-called alopecia mucinosa presents itself often as a widespread condition, rather than as a single patch, and that in this regard alopecia mucinosa is more comparable to large plaque parapsoriasis than to pagetoid reticulosis, a condition that almost always is localized to the vicinity of the foot, and they did not stress sufficiently that even when presenting itself as a single lesion, alopecia mucinosa not uncommonly manifests itself as a papule, plaque, or nodule, rather than as a patch. Their conclusion that alopecia mucinosa is an "'indolent' CTCL [cutaneous T-cell lymphoma]" contributes to sustaining the distinction, made erroneously for decades, between a "clinically benign" disease, namely, alopecia mucinosa, and a "clinically malignant" one, to wit, mycosis fungoides associated with follicular mucinosis.

Very similar to the conclusion of Cerroni et al. was that, more than 40 years before, of Findlay and Lowenthal who averred that alopecia mucinosa is "a reticulo-granuloma in its own right with a benign course." Although Cerroni and associates included the article by Findlay and Lowenthal in their bibliography, they did not acknowledge the forethinking of those forebears.⁵¹ Also to be recalled are the prescient words of Winkelmann in 1984 to that effect that "alopecia mucinosa or follicular mucinosis . . . [is] so indicative of mycosis fungoides in the adult that it is considered often as a skin manifestation of mycosis fungoides."¹³⁶ Winkelmann's perceptions were not cited by Cerroni et al.

Findlay and Lowenthal, Winkelmann, and Cerroni et al. never stated, without equivocation, that alopecia mucinosa, whether it presents itself as a single lesion or as many patches, plaques, or nodules, is mycosis fungoides from the outset and that that single process is responsible for the many different guises of the condition, clinically and histopathologically.

The finding of Lee et al. that infundibulosebaceous mucinosis often is present concurrently in eosinophilic folliculitis (both Ofuji's disease and that phenomenon affiliated with infection by HIV) therefore poses a challenge in differentiation histopathologically from alopecia mucinosa. For decades, various inflammatory diseases such as allergic contact dermatitis, rosacea, and eosinophilic folliculitis affiliated episodically with epithelial mucinosis present as a mere incidental finding, have been grouped indiscriminately as "idiopathic alopecia mucinosa" along with mycosis fungoides in which epithelial mucinosis also was accompanying. Had clinico-pathologic correlation been undertaken scrupulously of the condition in each of these individuals, it would have become apparent that so-called follicular mucinosis simply was a denominator in common of very different pathologic processes, the two in which the finding occurred most often being "eosinophilic folliculitis" on one hand and mycosis fungoides on the other.

It should be mentioned in passing that prognosis of a particular condition is not related necessarily to attributes of it morphologically, and, moreover, prognosis with any degree of accuracy for an individual patient who has patches of mycosis fungoides is not possible on the basis of assessment merely of clinical features and/or histopathologic findings. As but one example of those principles, follicular mycosis fungoides (with and without follicular mucinosis) has been said to have a poor prognosis,¹⁶³ whereas localized alopecia mucinosa has been stated to have a good prognosis.¹⁶⁶ If both conditions are regarded as manifestations of a single pathological process, namely, mycosis fungoides that happens to affect infundibular epidermis especially, the prognosis of that condition is likely to be no different fundamentally than mycosis fungoides which effects surface epidermis preponderantly. The clinical course of alopecia mucinosa is as uncertain as that of small or large plaque parapsoriasis, and of all the other manifestations of mycosis fungoides, except for nodules and tumors, particularly ulcerated ones, for which prognosis is poor uniformly. In actuality, the vast majority of patients with mycosis fungoides have patches and only slightly elevated plaques for a lifetime, never developing nodules and tumors; they do not die from the effects of the lymphoma but from another unrelated malady. In contrast, an individual patient with mycosis fungoides, in a rather short period of time, may generate nodules and tumors in the skin and manifest signs of involvement of lymph nodes and other organs (Figs. 28A–B ). It is accurate to state, therefore, that prediction of the course of mycosis fungoides in an individual person, either from features clinically of flat or slightly elevated lesions or from findings histopathologically in such lesions, while easy to make, cannot be offered with any degree of certainty. It is true that when in slightly raised plaques there are within the epidermis large collections of numerous abnormal lymphocytes, in dermal papillae arrangement compactly of numerous abnormal lymphocytes, and lymphocytes in the epidermis decidedly larger than those in the dermis, as well as individual lymphocytes that are strikingly atypical, the lymphomatous process known as mycosis fungoides can be inferred to be accelerating. No judgment with surety, however, can be made about the eventual course of mycosis fungoides when all lesions still are patches or extraordinarily subtle plaques.