Over the last 50-60 years, medical genetics has been the epitome of scientific disciplines being translated into everyday life, both professionally and in the population at large. Sometimes this progress is in terms of specific biochemical reactions and at other times in broader, more general ways. Presented here is one account of the most common human genetic disorder, neurofibromatosis type 1 (NF1), exemplifying such a general translation. A careful consideration of the various types of the disorder’s mass lesions translates pathogenetic schemata and improved patient care. There are four major points of emphasis. First, that the mast cell is common to all types of NF1 neurofibromas. Second, the three primary types of NF1 neurofibromas have very different natural histories. Third, elements of those natural histories put an emphasis on the potential role of trauma in NF1 neurofibroma pathogenesis. Fourth, the NF1 neurofibroma differences cannot be accounted for by the NF1 gene locus itself, requiring emphasis on how the NF1 gene is “put into practice”, that is, its praxitype.
Keywords: Neurofibromatosis type 1, neurofibroma, nerve sheath, Schwann cell, mast cell, trauma
Translational genetics and genomics is the literal translation of highly technical, esoteric scientific genetic jargon and associated data into relatively simpler and more vernacular words, phrases and processes that afford clinicians and patients relatively new diagnostic and treatment options. For example, up through the first half of the 20th century, chromosomes could hardly be considered familiar or ordinary for most people. On the other hand, there was some modest familiarity with a particular type of mental retardation disorder, the Down syndrome. Then, in 1959, Jerome Lejeune identified the cause, namely an extra chromosome 21. The affected person thus had three copies of chromosome 21, otherwise referred to as Trisomy 21. All of a sudden, chromosomes were poised to become incorporated into household jargon. In the 15 or so years that followed, geneticists dedicated intense energy to the technology of identifying and characterizing human chromosomes and defining their various roles in human morbidity, including disorders affecting the embryo and fetus as well as the newborn infant and child.
One additional critical element of technology changed everything: the ability to culture cells from amniotic fluid, the liquid surrounding and cushioning the fetus in the womb! Such a technique afforded the opportunity to diagnose Trisomy 21 (and other chromosome abnormalities) in the first trimester, in turn affording a termination of the pregnancy to avoid the birth of a mentally retarded child. To that technology, two more societal developments laid the groundwork for routine prenatal diagnosis and consequent termination of pregnancy to be legal and casually feasible. One was the “Roe v. Wade” Supreme Court decision establishing first-trimester abortion a legal right for the mother. The second was the coincidental availability of reliable overnight delivery of the amniotic fluid specimen from where it was obtained to the diagnostic laboratory, courtesy of Federal Express and similar commercial enterprises. As the first Medical Genetics Fellow at the Massachusetts General Hospital (1968-1969 and 1972), I was personally involved in the early stages of facilitating the translation of the underlying technology to daily, ordinary household activities. This was and is an example par excellence of translational genetics, although now it is at a much more complicated and laborious level, often focused on degenerative neurological disorders and various types of cancer.
The establishment of a journal dedicated to translational genetics and genomics is timely and uniformly welcomed four-fold by the data-generating scientists, by the application-minded clinicians, by the specifically-benefited patients and their families, and by a health-conscious society as a whole. I am pleased to be part of the amplification of this modern paradigm and I wish to celebrate my involvement by focusing on the most common human single-gene-mutation disorder, neurofibromatosis type 1 (NF1). I suggest herein that the translation of laboratory and clinical research on the mast cell (MC) and MC-stabilizer pharmacology to the treatment of NF1 patients with readily available, inexpensive and safe drugs represents a milestone. Although genetic details have not been paramount in the early stages of this work, there is already a major trend to an increasing focus on genetic details per se. For example, comparative and anthropological genetics, epigenetics (imprinting, microRNAs) and embellishment of the praxitype concept will be at the leading edges of translating basic research into cogent and compelling treatment of humanity’s most common single-gene disorder, NF1.
NF1 and mast cells
NF1 is a complex and fascinating disorder. There are many reasons for the fascination, not the least of which is the nature of its sentinel feature, the neurofibroma. I have lived that fascination for over 40 years. In the late 1970s, based on cogent clinical data, both those personally derived and those previously published, I suggested that the MC was a key to NF1 neurofibroma pathogenesis. Multiple NF1 mouse models soon established that MCs were necessary but not sufficient for the development of the syndrome’s neurofibromas.[2-4] In addition, Lloyd’s group have documented that the diplosufficient (DS) MC is as effective in this capacity as the Nf1+/- MC. That is, it is the MC native characteristics, not those derived from Nf1 haploinsufficiency (HI), that account for MC complicity in the development and progression of NF1 neurofibromas.
In the meantime, I initiated a series of therapeutic protocols demonstrating that blocking MC degranulation could ameliorate neurofibroma-associated symptoms (e.g. itching, pain, tenderness) and even interfere with NF1 neurofibroma development and progression. The positive results and limitations of treating NF1 patients with the MC stabilizer, ketotifen (Zaditen®, Sandoz Pharmaceuticals), were reported in the Archives of Dermatology in 1987 and 1993, and in a book chapter in 1990. For several patients, in addition to symptom relief, the rate of growth cutaneous, subcutaneous and plexiform neurofibromas appeared to decrease and/or the rate of appearance of new cutaneous neurofibromas decreased. While there were never any refutations of the data, likewise there seemed to be little, if any, enthusiasm for building on or exploiting this approach to neurofibroma treatment. One exception was an enthusiastic 1989 publication from Germany emphasizing similar findings. I had expected these cumulative data to spark an approach to neurofibromas that focused on arrest and preemption, but, alas, the model relying on post hoc reversal or removal prevailed as the prototypic approach.
Above and beyond an unequivocal decrease in NF1 neurofibroma-associated symptoms, particularly itching, pain and tenderness, documentation of ketotifen’s ability to minimize NF1 cutaneous neurofibroma growth was only published in December 2015. I would also suggest that this patient’s right ankle diffuse plexiform neurofibroma also benefited from the ketotifen treatment, but there are those who would attribute its remarkably small size after 33 years of treatment merely to chance or coincidence. Among my reasons for enthusiasm are both a modest number of similar results among NF1 patients who were essentially “self-treated”, and the documented efficacy of ketotifen in treating other MC-dependent disorders, including and especially systemic mastocytosis.[11-13]
Ketotifen is a cycloheptathiophene that has two primary functions: preventing or blocking MC degranulation and directly antagonizing histamine. It has been used worldwide to treat a variety of allergic disorders, including asthma, sinusitis and urticaria,[15-17] mastocytosis,[11-13,18] and a variety of parasitic infestations. It has also been used in ameliorating scleroderma and other fibrosing conditions, including keloids.[20-23] As an extension of the latter, it has been successfully used to enhance or temper post-surgical scarring or wound-healing in general.[24,25] Even a casual familiarity with the sum of ketotifen’s salutary influences in MC-related disorders would suggest a potential benefit of the drug in the development and progression of NF1 MC-dependent neurofibromas. In turn, the nature and magnitude of the documented positive effects of ketotifen thus far in NF1 can and do serve as an impetus for further exploring and refining such treatment.
As detailed in the aforementioned 2015 article describing 30 years of age-appropriate ketotifen treatment from the age of four months, blocking MC degranulation appears to have arrested the progression of NF1 cutaneous neurofibromas at their earliest stages, mostly at the earliest nascent/latent stage (see below). At 30 years of age, his entire skin surface was essentially free of any ordinary later stage NF1 cutaneous neurofibromas, although there were several areas of arrested early (flat) neurofibromas. They were very localized slight elevations that required tangential illumination for visualization. While arrest of one or several cutaneous neurofibromas might logically be consigned to chance, the latter is hardly an explanation for well over a square meter of at-risk human skin.
Ketotifen’s effects in other conditions are consistent with emphasis on arrest and preemption. In terms of its most widespread use, the treatment of asthma, the primary goals are avoidance of “attacks” and less intense symptoms during an acute episode. Moreover, ketotifen treatment prior to exposure to a variety of microbial, parasitic or biochemical challenges precludes or mollifies the development of the intrusive gastrointestinal symptoms[26,27] and their underlying inflammation,[19,28-31] including gout. On the other hand, ketotifen treatment of established allergic rhinitis or mastocytosis can effectively reverse symptoms and other consequences of excessive or inopportune MC degranulation. For example, a series of IgE allergenic rhinitis patients not only realized symptomatic relief with ketotifen treatment, the drug also reversed biochemical consequences of excess MC degranulation. And in a case of flagrant mastocytosis with associated hypovitaminosis D, the MC contribution to the latter was obviated as the mastocytosis responded to the ketotifen. This case suggests that the low serum calcidiol (vitamin D2) often seen in NF1[34,35] may be corrected with ketotifen treatment.
Respecting a variety of issues, including selection biases and placebo effects, reliance on self-treatment data is fraught with logical and logistical hazards. Nonetheless, data derived from cohorts of self-treated patients may provide useful ancillary insights. As considered recently,[36,37] I have interacted with a cadre of NF1 persons whose use of ketotifen was based on self-determination in cooperation with a thoughtful physician. Among those persons are ones who used ketotifen over three years or more and who described the benefits with surprising consistency from one person to another. The epitome of such a person is the subject of the 2015 article considered above. The three relatively consistent items of feedback are as follows: (1) a major decrease in neurofibroma-associated itching, pain and tenderness; (2) a decrease in the rate of appearance of new cutaneous neurofibromas and/or a slowing in the rate of progression of a visible plexiform neurofibroma; and (3) an improvement in an overall sense of well-being. In one NF1 patient, these changes were accompanied by a decrease in the number/density of MCs in biopsied cutaneous neurofibromas. In addition, I have had comparable feedback from physicians outside the US who regularly prescribe ketotifen for their NF1 patients.
Based on my own observations and an immense international literature, there are no major or serious adverse side-effects associated with oral ketotifen treatment over long periods of time. The one definite side-effect is occasional unwanted weight gain, almost always a matter of “a few” pounds or kilograms,[38,39] though there are also data to suggest ketotifen can actually contribute to avoidance of an obese habitus.[40-42] Moreover, almost all the relevant data are from countries other than the United States, in turn reflecting the absence of USA Food and Drug Administration (FDA) approval. Withholding of FDA approval in the 1980s reflected details about efficacy in treating asthma, as opposed to concerns about safety. Toxicity and adverse effects were not an issue.
In the NF1 ketotifen treatment results published in 1987 and 1993, the primary endpoints were decreased localized neurofibroma-associated symptoms and decreased growth dynamics (rate of appearance of cutaneous neurofibromas and/or slowed enlargement of any neurofibroma). However, during the study period, another potential endpoint was identified. For several ketotifen-treated NF1 patients who coincidentally underwent surgery involving a plexiform neurofibroma, the surgeon involved contacted me later to declare that intra-operative hemorrhage was much less than anticipated based on previous experiences. The surgeon usually queried something like “what did you do to the patient so that he/she didn’t bleed”? I published these anecdotal data in the 1990 book chapter noted earlier and have since frequently broadcast the associated potential for MC stabilizers to minimize one of the most important hazards linked to NF1 neurofibroma surgery: intra-operative hemorrhage. This surgical complication inter alia immediately jeopardizes the patient’s life and risks otherwise avoidable blood transfusions and premature cessation of the surgery.
Among the reasons for pressing the matter presently and elsewhere are two interrelated considerations. One is the immediate health of the NF1 patient. A second is the establishment of a compelling, readily quantifiable general benefit of ketotifen treatment. Once such a benefit is acknowledged, expansion of the benefits of NF1 MC stabilization will likely follow quickly. Please know that I structured this last sentence very carefully, focusing on MC stabilization in general, not merely or solely as a benefit of a particular stabilizer, ketotifen. There are a number of other already-established MC stabilizing agents with American FDA approval. In addition, quite different pharmacological strategies are emerging, including and especially, antibodies to siglec molecules, particularly Siglec-8, in humans.[43-45] (Siglec is the acronym for sialic acid binding immunoglobulin-like lectins.)
In the above considerations, I referred to MC stabilization as a strategy directed at two related, but ultimately distinct pathologies: on the one hand cancer or tumors in the strict sense, and wounds (e.g. scars) on the other hand. The MC contributes to both types of processes and can be a neoplastic cell itself, to wit, in mastocytosis. Ketotifen and MC blockers in general seem to be treatment considerations in both types of lesions. Thus, I ask, when we treat NF1 neurofibromas are we treating wounds or tumors? To address this matter, let me consider NF1 neurofibromas themselves.
NF1 neurofibroma pathogenesis
When is a mass lesion a tumor? Is a keloid scar in some sense a tumor? In terms of etymology, the word, “tumor”, simply specifies a “swelling” or a tissue enlargement therefrom, but in the 21st century the designation, “tumor”, requires greater perspicacity. Key tumor definition criteria consider rate of growth, genetic changes (mutations, chromosome rearrangements) and metabolic changes (e.g. Warburg effect). I suggest that Recklinologists have been a bit too casual in denoting NF1 neurofibromas as tumors, such that we have largely ignored stages or phases of NF1 neurofibroma development that are more assiduously considered to represent wounds. This oversight ignores the fact that aberrant NF1 wound-healing processes are part of neurofibroma development, based both on clinical grounds[8,37] and the use of mouse models. I have consistently considered that NF1 neurofibromas originate as wounds, specifically, Schwann cells (SC)-enriched granulation tissue. It is here that MC stabilizers have an extraordinary potential-to prevent the transition from wound-healing to frank neoplasia, from wound to tumor. This is translational genetics in broad, general terms. And, there’s more.
In almost all recent publications dealing with NF1 neurofibromas, the NF1 gene is introduced as a “tumor suppressor gene”, reflecting the well-established association of NF1 diploinsufficiency (DI) with malignant transformation, particularly the transformation of a neurofibroma to a sarcoma, most often a neurofibrosarcoma. But, it is more complicated than that. Without question, some NF1 neurofibromas are DI, yet they are not malignancies, nor, perhaps, even be tumors sensu strictu. Does the NF1 DI necessarily make these neurofibromas tumors? No! Consider that NF1 DI can also be present both in NF1 some café-au-lait spots[46-48] and in elements of the NF1 skeletal aberration known as pseudarthrosis.[49,50] Yet neither are tumors in any ordinary sense. Thus, for the early-stage uncharacterized NF1 neurofibroma, I consistently refer to them more neutrally and simply as “mass lesions”.
The development of NF1 DI in NF1 neurofibromas establishes a very important general principle about these mass lesions: there is an evolution, a progression of NF1 neurofibromas from one stage or phase to another. I have suggested that there are actually eight defined stages/phases of an NF1 neurofibroma, the initial (first) stage being axon/SC disruption and the immediate local response thereto, and the final (eighth) stage being the sarcoma [Tables 1 and 2]. During the intervening six stages there is a series of transitions, one of which involves the conversion from NF1 HI to DI. As of yet, there are no additional histopathologic or intrinsic behavioral criteria to document a change from “wound” to “tumor”.
Table 1: Neurofibromatosis type 1 neurofibroma phases of initiation and maturation-overviewClick here to view
Table 2: Neurofibromatosis type 1 neurofibroma phases of initiation and maturation-detailsClick here to view
In the early 1980s, a trio of NF1 patients critically influenced me to explore and eventually embrace crush trauma as a trigger for NF1 cutaneous neurofibroma development.
The first was a middle-aged woman who pointed out to me an 8-9 mm sessile cutaneous neurofibroma on the volar surface of her right forearm as she declared. “This is where I accidentally closed the car door on my arm. The next day it itched intensely and I knew I was going to develop a neurofibroma there”.
The second person was a middle-aged automobile mechanic who spent many hours daily lying on back working on the underside of automobiles and trucks. He described how a relatively small pedunculated neurofibroma was constantly subjected to the crushing weight of his body during his workday. He was very clear that he assumed he caused this mass lesion constantly to increase in size by virtue of his crushing it against the floor supporting his back. This soccer-ball-sized neurofibroma was remarkable both for its size and the multiple hypopigmented abrasion scars punctuating the skin overlying the mass’ posterior surface.
The third person was a middle-aged nurse who directed my attention to two large (5-6 cm) cutaneous neurofibromas. The one on her right shoulder she declared was “where my horse bit me”. The other, on the lateral surface of her right buttock, was “where my horse kicked me”.
All three persons volunteered local intense itching as part of the symptomatology associated with the initiation and/or increasing size of the mass lesions.
These and numerous similar, though less striking, NF1 patient vignettes influenced me to consider both crush trauma as a trigger for NF1 cutaneous neurofibroma development and the MC as a contributor to that development. Decades later, the investigations of Lloyd’s group documented that trauma (scalpel incision) incited neurofibroma development at the incision site in an Nf1+/- mouse model and that the contributing MCs could be either Nf1+/- (HI) or Nf1+/+ (DS).
Respecting these data, two logical conclusions are suggested: (1) the canonical schema holds that, distinct from the trauma, the neurofibroma arises as a direct and immediate result of the conversion from NF1 HI to DI and that the trauma is either coincidental or a matter of ad hoc aggravation; and (2) the new derivative schema holds that the trauma initiates the neurofibroma and that the conversion from NF1 HI to DI develops subsequently, the pre-conversion neurofibroma acting as an incubator of sorts, accumulating and/or maturing NF1 HI SCs that eventually become NF1 DI. If the latter schema obtains, thereby is an opportunity to avoid the NF1 DI that sets the stage for malignant transformation. In other words, arresting the neurofibroma in the early wound phase(s) might preclude further progression from wound phases into tumor phases, including the sarcoma eventuality. A rigorous evaluation of the data favors the wound-to-tumor scenario.
Additional current canonical models focus on interaction of the NF1 gene product, neurofibromin (Nfn), with the oncoprotein, Ras. That is, while the Ras-Nfn interaction takes place at the plasma membrane (PM), the subsequent (consequent) pathogenic phenomena occur away from the PM in the cytoplasmic milieu via Ras-phosphorylated Raf. The latter constitutively up-regulates the MAPK-ERK1/2 regulatory pathway, effectively establishing Ras-GTP overdrive (RTO). Totally ignored in this schema is the potential for the interaction of mutant Nfn with Ras actually to pervert the PM, such that the PM perversion underlies at least some of the adverse impact of NF1 mutations. The pioneering work of Bloomfield et al.,[52,53] abetted by the work of Zhang et al., focusing on the amoeba, dictiostylium, documents rather nicely that mutant NF1 actually perturbs the PM. In turn, this PM perturbation directly and immediately accounts for elements of the amoeba’s mutant phenotype. This set of considerations fits well with many of the aberrant behaviors of NF1 mutant SCs in general and more specifically with the requirement of properly placed PM processes by “terminal Schwann cells” (TSC) in nerve repair after crush trauma[55-58] (see below).
In other words, a key element of NF1 pathogenesis may be distortion of PM processes in the TSC such that, rather than timely repair of damaged cutaneous sensory nerve endings, the NF1 mutant TSC cannot contribute to ordinary repair and-instead-the result is a cutaneous neurofibroma. Other lines of evidence complement this approach, including and especially the role of dual leucine zipper kinase (DLK) in nerve repair and its reliance therefor on a normally functioning MAPK/ERK regulatory process.[59-63] In the presence of NF1-mutant-based RTO, DLK is primed to confound the already perturbed PM functions of the TSC. That is, while the presence of an NF1 HI germinal mutation affords normal embryologic development of sensory nerves, these same NF1 HI nerves simply cannot repair properly after trauma, particularly crush trauma.
Types of NF1 neurofibromas
The classification and naming of NF1 neurofibromas is way overdue. The reasons for the delay are not entirely clear, but it probably boils down to who, or which group, is sufficiently authoritative. I think it is the purview and responsibility of the loosely defined cadre of “Recklinologists” who have devoted all or most of their professional careers to the neurofibromatoses and NF1 in particular. As a member of this cadre, I have already made my recommendations in 2007 in the online journal, Neurosurgery Focus and reiterated them in 2016. I discounted purely clinical and purely pathological perspectives as being the “only” disciplines having the wherewithal for this task. I have focused on additional developmental and anatomical vantage points as well as the traditional clinical, pathological and radiologic/imaging considerations.
I start with the microanatomical characterization of the nerve sheath and its three major elements, deriving the endoneurial neurofibroma (Endonf), the perineurial neurofibroma (Perinf) and the epineurial neurofibroma (Epinf).[37,64] From there I considered the mammalian organism topography, the pathologist’s sensibilities, the rapidly evolving technology of radiology/imaging, developmental perspectives and clinical acumen. I also relied heavily on the thoughtfulness of Pierre Masson, who characterized two major NF1 neurofibroma groupings, namely, “encapsulated” neurofibromas and “diffuse” neurofibromas. Two articles, for which I was a coauthor, made effective use of Masson’s approach, affording the authors several cogent insights thereby.[66,67] Indeed, the “encapsulated” neurofibromas correspond to Perinfs and the “diffuse” neurofibromas correspond to Endonfs and Epinfs. More recently, recognition of the key role of the TSC has allowed a more sophisticated approach to subtypes of NF1 Endonfs.
Endoneurial neurofibromas at one point were considered synonymous with the vernacular designation of “cutaneous neurofibroma”. Presently, however, considering the details of the central role the TSC’s similar behaviors in the lanceolate nerve endings of the skin and the penicillate nerve endings of the neuromuscular junction (NMJ), I suggest two types of Endonf, namely the traditional cutaneous neurofibroma and the NMJ neurofibroma.
When we speak of trauma we have to remind ourselves of the various types of trauma, including mechanical, electromagnetic, metabolic, hypoxic and toxic. Then, with specific regard to NF1, we have to consider what is traumatized. For starters, I am focusing on the axon/SC relationship that is typical of sensory nerves. As well, ultimately the ganglia and plexuses associated with these nerves likely should be considered. Simplifying things somewhat, let us here focus on NF1 cutaneous neurofibromas and the termini of sensory nerves in the skin. Extensive investigations have documented that these termini are characterized by a specific class of SC that literally caps the nerve ending, though occasionally a short (nanometer-micrometer length) tongue of naked axon protrudes. This special SC is specifically known as the terminal SC[55,68,69] or perisynaptic SC,[70-74] either or both designated by the acronym, TSC. The just-cited sophisticated and elegant publications have documented that disruption of a sensory nerve’s axon/SC relationship causes the TSC to orchestrate the repair processes. Briefly, the TSC generates multiple PM processes, at least one of which forms a literal bridge with an intact sensory nerve terminus. In the skin, the intact nerve uses that bridge to guide an axonal branch to restructure and thereby repair the damaged nerve. A similar schema effects repair of nerve-crush injury at the NMJ.
Multiple confluent NMJ Endonfs might initially suggest an Epinf. One potentially distinguishing feature may be the lesion’s size at the surface of the skin, anticipating that the NMJ lesions will be smaller and more circumscribed than the Epinf. Also, given the role of NF1 somatic mutations in the progression of NF1 neurofibromas, one might expect distinctive NF1 second hits in each of the NMJ lesions, while a larger Epinf involving the skin is expected to have only one distinctive NF1 second hit. (It is possible that a very large Epinf is actually not a “single” lesion, but the coalescence of one or more adjacent lesions.) In addition, the Epinf involving the skin often manifests overlying hyperpigmentation and/or hirsutism, either diffusely throughout the lesion or at the edges, where hairs are like bristles (frequently seen in Asian or Hispanic NF1 patients).
As indicated below, epineurial neurofibromas are congenital lesions, ultimately deriving from an identifiable nerve or nerve root. In contrast, postnatal crush trauma disrupts the NMJ TSC/Axon relationship such that, in the presence of NF1 HI, alternative to “normal” repair, a localized neurofibroma develops. Growth and progression is local, distinctively below the skin, enlarging on and in the involved muscles. This appearance on neuroimaging is, again, distinctive, although in early 2017 there are relatively few reliable clinical and neuroimaging data.
Although all NF1 cutaneous neurofibromas derive from skin nerve termini, they are distinctive in having a variety of morphologies as a function of each lesion’s chronological progression. In addition, there are potential contributions from skin-located adnexae (e.g. hair follicles, sebaceous glands) and stem/progenitor cells. That is, NF1 cutaneous neurofibromas (like other NF1 neurofibromas) are heterogeneous. In any event, an NF1 cutaneous neurofibroma has five sequential configurations: Nascent/Latent, Flat, Sessile, Globular, and Peduncular.
The nascent/latent cutaneous neurofibroma [Figure 1] is not apparent by inspection or palpation of the skin, but might be appreciated by proper imaging techniques, such as high-resolution ultrasound. The contributory cells may be dispersed or clustered.
Figure 1: The nascent (A)/latent (B) neurofibromatosis type 1 cutaneous neurofibroma (Endonf)Click here to view
The flat cutaneous neurofibroma [Figure 2] is visible at the skin surface, distinguished primarily by its slightly raised surface; the surface may appear thinner and somewhat pale compared to adjacent uninvolved skin; alternatively, there may be some increased pigmentation and/or peripheral coarse hairs (bristles); for larger/older lesions, pushing down on them elicits the “button-holing” phenomenon. The size range is from about 0.5 mm to about 12 mm.
Figure 2: The flat neurofibromatosis type 1 cutaneous neurofibroma (Endonf)Click here to view
The sessile cutaneous neurofibroma [Figure 3] is definitely raised compared to the adjacent skin and there is usually an apparent apex. The height of the apex may be as much as 8-10 mm. The surface and texture are essentially as for the flat lesion, but for pale skin erythema may be apparent and for others there may be some hyperpigmentation, as well as the bristle hirsutism noted above. Button-holing may be elicited. The circumference at the base is usually round and measures from about 1 mm to 10-12 mm. It is not usually possible to palpate neurofibroma elements below the surface of the skin. Ultrasound and magnetic resonance imaging (MRI) analyses may reveal such sub-surface elements.
Figure 3: The sessile neurofibromatosis type 1 cutaneous neurofibroma (Endonf)Click here to view
The globular cutaneous neurofibroma [Figure 4] represents progression of the sessile cutaneous neurofibroma. The base at the surface of the skin may be 20-30 mm in diameter, with a comparable height and a globular shape, reflecting a larger circumference some distance above the base. Moving the portion above the skin surface indicates the presence of a portion of the lesion below the skin surface, though there is no distinctive stalk joining/separating the portions above and below the surface of the adjoining skin.
Figure 4: The globular neurofibromatosis type 1 cutaneous neurofibroma (Endonf)Click here to view
The peduncular cutaneous neurofibroma [Figure 5] is a progression of the globular cutaneous neurofibroma. A distinct stalk separates/joins the portions above and below the skin surface. The stalk is usually several mm long and 1-3 mm in diameter. (Rarely, the stalk diameter may be measured in cm as opposed to mm.) A key issue is that substantial amounts of the cutaneous neurofibroma are below the surface of the skin and the nature of its connection to the superficial portion has never been investigated. The diameter of the superficial portion is usually between 5 and 25 mm, with a comparable mass below the surface of the skin. The surface skin of the lesion usually is not distinctive.
Figure 5: The peduncular neurofibromatosis type 1 cutaneous neurofibroma (Endonf)Click here to view
Two variant types of NF1 cutaneous neurofibromas must also be acknowledged. Superficially, these variants seem to confound, if not contradict the definitions for the five types of cutaneous neurofibroma. The first is the “combined post-traumatic neurofibroma”. Virtually always there is a previous episode of crush trauma, such as a horse-bite or a horse’s kick (described above). The size at the base may be up to 8 cm or so and the elevation from the skin surface may be as high as several cm. The overlying skin is somewhat thinner and pale compared to the surrounding skin surface. Itching/pruritus may be prominent, especially during the initial and early phases. The element of “combination” derives from the almost certain involvement of intact nerve elements (including perineurial components) in addition to the cutaneous nerve termini that define the cutaneous neurofibroma. The second NF1 cutaneous neurofibroma variant is the “endoneurial diffuse neurofibroma” characterized by a large expanse of skin (e.g. the entire surface of the posterior thorax) with a palpable difference from adjacent “normal” skin and some degree of hyperpigmentation. As appreciated by appropriate MRI studies, involvement is limited to the skin, without apparent involvement of distinct nerves and their identifiable branches.
As a reminder, specifically excluded from the NF1 cutaneous neurofibroma category are the cutaneous components of Epinfs and the subcutaneous variety of the Perinf, visible and palpable under the skin. While Endonfs may share elements with Perinfs and Epinfs, ultimately they are all sufficiently different to warrant consideration of distinctive therapies.
Perineurial neurofibromas have, at least initially and in the early stages, an intact perineurium. In later stages, the perineurial sheath may be disrupted and fragmented, obscuring that originally it was intact and functional as a distinct anatomical boundary. In any event, Perinfs account for what are ordinarily referred to as subcutaneous neurofibromas and nodular plexiform neurofibromas. These two subgroups are distinguished merely as a function of duration and size. Perinfs are located more or less anywhere along the length of a sensory nerve, from the nerve root to its more terminal portions. Subcutaneous neurofibromas are visible and palpable under the surface of the skin, ranging in size from about 2 mm to 5-6 cm in diameter. When the overlying skin is moved it glides smoothly over the lesion. Subcutaneous neurofibromas may be solitary or multiple, and when closely adjacent to each to each other over many centimeters, they correspond to von Recklinghausen’s “Rankenneurom”. When present over some substantial length, the Rankenneurome presents a nodular plexiform neurofibroma. In either case, they are easily separated from non-nervous tissue adjacent to the involved nerve, for example, by blunt dissection at surgery or during an autopsy. This distinct mass is firm and nodular and may be the source of local pain and tenderness for the NF1 patient. By plain radiography, computed tomography (CT) scan, MRI and positron emission tomography (PET) scanning they are utterly distinct from surrounding tissue. Primarily as a function of location and size they can be the cause of sensory, motor and autonomic deficits.
Perinfs of both types are at substantial risk for sarcoma transformation. For NF1 patients manifesting an NF1 whole gene deletion as the disorder’s etiology, Perinfs have an early onset and widespread distribution, including and especially at the proximal dorsal nerve roots. A key issue is the presumption that the intact perineurium affords an intact blood-nerve barrier, perhaps influencing the efficacy of various pharmacologic agents. Figure 66 of the Harkin and Reed pathology treatise depicts a typical Perinf (nodular plexiform neurofibroma. That figure’s label is simply “plexiform neurofibroma”. Confounding matters is the nearby Figure 64 of the Harkin and Reed treatise, which figure is also labeled simply, “plexiform neurofibroma”. as though the specimens in the two figures were/are the “same”! Rather, Figure 64 represents an Epinf. Even a casual perusal of these two figures declares fundamental lesional differences. While they share the same high propensity for malignant transformation and have some histopathology overlap, they are very different features of NF1.
Epineurial neurofibromas are so named because its sole outer boundary is the epineurium of the nerve sheath (at least in the lesion’s earliest stages) and the fact that it has no intact perineurium.[37,64] Endonfs, by definition, do not have an epineurial or perineurial boundary at any stage, and Perinfs, by definition, have by a perineurial boundary. In addition to the propensity of growing to a very large size, an Epinf manifests two sentinel features: (1) its ability to invade adjacent tissues if and when the epineurial membrane is breached or fragmented; and (2) late stage metamorphosis of its cellular constituents to adipocytes (unpublished observations). And there are other less dramatic differences such as the pattern of MC distribution: the MCs are diffusely interstitial in the Epinf and are circumferential in the Perinf.
Another historical confounding factor for understanding the Epinf is its frequent involvement of the skin. When flat at the skin surface (in the lesion’s early stages), some clinicians confuse it with a large café-au-lait spot. And when it involves the skin more intensely, some clinicians incorrectly construe the lesion to be a large cutaneous neurofibroma. While Endonfs and Perinfs characteristically initiate post-partum, the Epinf is basically a congenital lesion. Based on the work of Gamble and others,[80-82] I have concluded the congenital origin of the Epinf can be understood as follows. During the sensory nerve’s early development, the axon elongates centrifugally, followed by the SCs in various patterns. The perineurial sheath temporally and spatially follows the SCs. At first the perineurium is incomplete, with numerous gaps and lacunae, which may persist for two weeks or more. If an NF1 neurofibroma initiates from embryologic/fetal endoneurial precursors while the perineurial lacunae are present, perineurial components are present only as scattered perineurial fibroblasts and the lesion’s peripheral boundary is necessarily the epineurium.
Moreover, there appear to be two sets of growth dynamics. The earlier first phase involves globular budding within the epineurial sheath, as depicted in the Harkin and Reed Figure 64, cited earlier. The cumulative increase in total surface area (TSA) is thereby substantial, suggesting that growth monitoring of Epinfs by TSA might be preferable to the usual volumetric calculations.[83-85] At some point, the Epinf epineurium is breached and fragmented such that further Epinf growth proceeds by infiltrating and replacing (not merely displacing) adjacent tissue and, in subsequent stages, there is the fatty metamorphosis alluded to above. Separation of the Epinf from adjacent structures utilizing blunt dissection is impossible. MRI and CT scanning (with or without PET) depict indistinct perimeters and there may be density heterogeneity reflecting a number of variables, including the incorporation of adjacent tissues. Often there is erosion of adjacent boney structures. Moreover, Epinfs are utterly distinct in their ability to become massive, weighing hundreds of kilograms, even hundreds of pounds. On the other hand, some Epinfs become quiescent after periods of rapid growth, a significant confounding factor in therapeutic drug trials. Moreover, the propensity for Epinfs to undergo sarcomatous degeneration has one remarkable exception. Epinfs involving the trigeminal ganglion and/or its three nerve branches undergo malignant transformation only very rarely, if at all. Although the reasons for this discrepancy is uncertain, facial skin temperature may be a critical factor.
Epinf skin involvement varies from none, through merely nonspecific protrusion, to frank corruption, often involving hyperpigmentation and various types of hirsutism. The latter ranges from diffuse hairiness to short, thick and dark bristles at the edges. Epinf skin involvement is distinguished from that of the “endoneurial diffuse neurofibroma” variant on the basis of the latter’s having no involvement of tissues below the dermis. Perinfs, including nodular plexiform neurofibromas, do not alter the nature or integrity of overlying skin. Endonfs, as cutaneous neurofibromas, by definition always involve the skin.
At the least, these comparisons make it absolutely clear that there two distinct types of plexiform neurofibromas consistent with the pathologist’s use of the term, “plexiform neurofibroma”. One is the Epinf, concordant with both Masson’s “diffuse neurofibroma” and the vernacular designation, “diffuse plexiform neurofibroma”. The other is a Perinf specified as a “nodular plexiform neurofibroma”. This latter designation should not be confused with the designation of a “nodular neurofibroma” as specified Widemann and Dombi and their collaborators referring to a portion of an atypical neurofibroma or STEP lesions (Suspicious Tumor Enlargement and PET scan positive). Given the many differences in the two types of plexiform neurofibromas, it would not be surprising that they respond differently to various treatment strategies.
NF1 atypical neurofibromas (or STEP Lesions, as noted above) are not in the same pathogenetic category as the three nerve-sheath-defined NF1 neurofibromas. Rather, STEP Lesions are derivatives of Epinfs and Perinfs and as such are NF1 pathogenetic consequences, not NF1 pathogenetic features in the strict sense. NF1 atypical neurofibromas are particularly important as they often represent a transitional phase from a benign lesion to a malignant sarcoma, which is an NF1 complication. Such malignancy is most often a neurofibrosarcoma, with other types including liposarcoma, rhabdoid sarcoma or malignant fibrous histiocytoma.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
There is no patient involved.
This review is waived for ethical approval.
- 1. Riccardi VM. NF1 and the praxitype. JSM Genet Genomics 2015;2:1006.
- 2. Yang FC, Ingram DA, Chen S, Hingtgen CM, Ratner N, Monk KR. Clegg T, White H, Mead L, Wenning MJ, Williams DA, Kapur R, Atkinson SJ, Clapp DW. Neurofibromin-deficient Schwann cells secrete a potent migratory stimulus for Nf1+/- mast cells. J Clin Invest 2003;112:1851-61.
- 3. Yang FC, Chen S, Clegg T, Li X, Morgan T, Estwick SA, Yuan J, Khalaf W, Burgin S, Travers J, Parada LF, Ingram DA, Clapp DW. Nf1+/- mast cells induce neurofibroma like phenotypes through secreted TGF-beta signaling. Hum Molecular Genet 2006;15:2421-37.
- 4. Staser K, Yang FC, Clapp DW. Mast cells and the neurofibroma microenvironment. Blood 2010;116:157-64.
- 5. Ribeiro S, Napoli I, White IJ, Parrinello S, Flanagan AM, Suter U, Parada LF, Lloyd AC. Injury signals cooperate with Nf1 loss to relieve the tumor-suppressive environment of adult peripheral nerve. Cell Rep 2013;5:1-11.
- 6. Riccardi VM. Mast cell stabilization to decrease neurofibroma growth.Preliminary experience with ketotifen. Arch Dermatol 1987;123:1011-6.
- 7. Riccardi VM. A controlled multiphase trial of ketotifen to minimize neurofibroma-associated pain and itching. Arch Dermatol 1993;129:577-81.
- 8. Riccardi VM. The potential role of trauma and mast cells in the pathogenesis of neurofibromas. In: Ishibashi Y, Hori Y, editors. Tuberous Sclerosis and Neurofibromatosis: Epidemiology, Pathophysiology, Biology and Management. Amsterdam: Elsevier; 1990. p. 167-90.
- 9. Haustein UF. Ketotifen inhibits urticaria and tumor progression in neurofibromatosis. Dermatol Monatsschr 1989;175:581-4. (in German)
- 10. Riccardi VM. Ketotifen suppression of NF1 neurofibroma growth over 30 years. Am J Med Genet A 2015;167:1570-7.
- 11. Póvoa P, Ducla-Soares J, Fernandes A, Palma-Carlos AG. A case of systemic mastocytosis; therapeutic efficacy of ketotifen. J Intern Med 1991;229:475-7.
- 12. Graves L 3rd, Stechschulte DJ, Morris DC, Lukert BP. Inhibition of mediator release in systemic mastocytosis is associated with reversal of bone changes. J Bone Miner Res 1990;5:1113-9.
- 13. Kurosawa M, Amano H, Kanbe N, Akimoto S, Takeuchi Y, Yamashita T, Hashimoto T, Kurimoto F, Miyachi Y. Heterogeneity of mast cells in mastocytosis and inhibitory effect of ketotifen and ranitidine on indolent systemic mastocytosis. J Allergy Clin Immunol 1997;100:S25-32.
- 14. Martin U, Baggiolini M. Dissociation between the anti-anaphylactic and the anti-histaminic actions of ketotifen. Naunyn Schmiedebergs Arch Pharmacol 1981;316:186-9.
- 15. Edge JA, Osborne JP. Terbutaline and ketotifen in cold urticaria in a child. J R Soc Med 1989;82:439-40.
- 16. Huston DP, Bressler RB, Kaliner MA, Sowell LK, Baylor MW. Prevention of mast cell degranulation by ketotifen in patients with physical urticarias. Ann Internal Med 1986;104:507-10.
- 17. Iikura Y, Naspitz CK, Mikawa H, Talaricoficho S, Baba M, Sole D, Nishima S. Prevention of asthma by ketotifen in infants with atopic dermatitis. Ann Allergy 1992;68:233-6.
- 18. Ting S. Ketotifen and systemic mastocytosis. J Allergy Clin Immunol 1990;85:818.
- 19. Serna H, Porras M, Vergara P. Mast cell stabilizer ketotifen [4-(1-methyl-4-piperidylidene)-4h-benzo[4,5]cyclohepta[1,2-b]thiophen-10(9H)-one fumarate] prevents mucosal mast cell hyperplasia and intestinal dysmotility in experimental Trichinella spiralis inflammation in the rat. J Pharmacol Exp Ther 2006;319:1104-11.
- 20. Gruber BL, Kaufman LD. Ketotifen-induced remission in progressive early diffuse scleroderma: evidence for the role of mast cells in disease pathogenesis. Am J Med 1990;89:392-5.
- 21. Gallant-Behm CL, Hildebrand KA, Hart DA. The mast cell stabilizer ketotifen prevents development of excessive skin wound contraction and fibrosis in red Duroc pigs. Wound Repair Regen 2008;16:226-33.
- 22. Monument MJ, Hart DA, Befus AD, Salo PT, Zhang M, Hildebrand KA. The mast cell stabilizer ketotifen reduces joint capsule fibrosis in a rabbit model of post-traumatic joint contractures. Inflamm Res 2012;61:285-92.
- 23. Monument MJ, Hart DA, Befus AD, Salo PT, Zhang M, Hildebrand KA. The mast cell stabilizer ketotifen fumarate lessens contracture severity and myofibroblast hyperplasia: a study of a rabbit model of posttraumatic joint contractures. J Bone Joint Surg Am 2010;92:1468-77.
- 24. Oskeritzian CA. Mast cells and wound healing. Adv Wound Care (New Rochelle) 2012;1:23-8.
- 25. Ng MF. The role of mast cells in wound healing. Int Wound J 2010;7:55-61.
- 26. Karmeli F, Eliakim R, Okon E, Rachmilewitz D. Gastric mucosal damage by ethanol is mediated by substance P and prevented by ketotifen, a mast cell stabilizer. Gastroenterology 1991;100:1206-16.
- 27. Eliakim R, Karmeli F, Okon E, Rachmilewitz D. Ketotifen effectively prevents mucosal damage in experimental colitis. Gut 1992;33:1498-503.
- 28. Eliakim R, Karmeli F, Rachmilewitz D. Ketotifen--Old drug, new indication: reduction of gastric mucosal injury. Scand J Gastroentero 1993;28:202-4.
- 29. Pothoulakis C, Karmeli F, Kelly CP, Eliakim R, Joshi MA, O’Keane CJ, Castagliuolo I, LaMont JT, Rachmilewitz D. Ketotifen inhibits Clostridium difficile toxin A-induced enteritis in rat ileum. Gastroenterology 1993;105:701-7.
- 30. Klooker TK, Braak B, Koopman KE, Welting O, Wouters MM, van der Heide S, Schemann M, Bischoff SC, van den Wijngaard RM, Boeckxstaens GE. The mast cell stabiliser ketotifen decreases visceral hypersensitivity and improves intestinal symptoms in patients with irritable bowel syndrome. Gut 2010;59:1213-21.
- 31. Karmeli F, Eliakim R, Okon E, Rachmilewitz D. Ketotifen and nitroxides decrease capsaicin-augmented ethanol-induced gastric damage in rats. Dig Dis Sci 1995;40:1140-6.
- 32. Hsu DZ, Chu PY, Chen SJ, Liu MY. Mast cell stabilizer ketotifen inhibits gouty inflammation in rats. Am J Ther 2016;23:e1009-15.
- 33. Talaat MA, Inaam PK, Mohammed MH, Ibrahim TE. The histological and histochemical effects of ketotifen in allergic rhinitis. J Asthma 1991;28:117-28.
- 34. Lammert M, Friedman JM, Roth HJ, Friedrich RE, Kluwe L, Atkins D, Schooler T, Mautner VF. Vitamin D deficiency associated with number of neurofibromas in neurofibromatosis 1. J Med Genet 2006;43:810-3.
- 35. Stevenson DA, Viskochil DH, Carey JC, Sheng X, Murray M, Moyer-Mileur L, Shelton J, Roberts WL, Bunker AM, Hanson H, Bauer S, D’Astous JL. Pediatric 25-hydroxyvitamin D concentrations in neurofibromatosis type 1. J Pediat ndocrinol Metab 2011;24:169-74.
- 36. Riccardi VM. Current utilization of mast cell stabilizers for preemptive treatment of NF1 neurofibromas. Neuro Open J 2015;2:67-73.
- 37. Riccardi VM. NF1 clinical elements and the NF1 neurofibroma burden. J J Neurol Neurosci 2016;3:025.
- 38. Mylona-Karayanni C, Hadziargurou D, Liapi-Adamidou G, Anagnostakis I, Sinaniotis C, Saxoni-Papageorgiou F. Effect of ketotifen on childhood asthma: a double-blind study. J Asthma 1990;27:87-93.
- 39. Asnaashari S, Delazar A, Habibi B, Vasfi R, Nahar L, Hamedeyazdan S, Sarker SD. Essential oil from Citrus aurantifolia prevents ketotifen-induced weight-gain in mice. Phytother Res 2010;24:1893-7.
- 40. Lee MN, Ye C, Villani AC, Raj T, Li W, Eisenhaure TM, Imboywa SH, Chipendo PI, Ran FA, Slowikowski K, Ward LD, Raddassi K, McCabe C, Lee MH, Frohlich IY, Hafler DA, Kellis M, Raychaudhuri S, Zhang F, Stranger BE, Benoist CO, De Jager PL, Regev A, Hacohen N. Common genetic variants modulate pathogen-sensing responses in human dendritic cells. Science 2014;343:1246980.
- 41. Shi MA, Shi GP. Different roles of mast cells in obesity and diabetes: lessons from experimental animals and humans. Front Immunol 2012;3:7.
- 42. Habibi AB, Vaez H, Imankhah T, Hamidi S. Impact of caffeine on weight changes due to ketotifen administration. Adv Pharm Bull 2014;4:83-9.
- 43. O’Reilly MK, Paulson JC. Siglecs as targets for therapy in immune-cell-mediated disease. Trends Pharmacol Sci 2009;30:240-8.
- 44. Bochner BS. Siglec-8 on human eosinophils and mast cells, and Siglec-F on murine eosinophils, are functionally related inhibitory receptors. Clin Exp Allergy 2009;39:317-24.
- 45. Bochner BS. “Siglec”ting the allergic response for therapeutic targeting. Glycobiology 2016;26:546-52.
- 46. Stark M, Assum G, Kaufmann D, Kehrer H, Krone W. Analysis of segregation and expression of an identified mutation at the neurofibromatosis type 1 locus. Hum Genet 1992;90:356-9.
- 47. Malhotra R, Ratner N. Localization of neurofibromin to keratinocytes and melanocytes in developing rat and human skin. J Invest Dermatol 1994;102:812-8.
- 48. De Schepper S, Maertens O, Callens T, Naeyaert JM, Lambert J, Messiaen L. Somatic mutation analysis in NF1 cafe au lait spots reveals two NF1 hits in the melanocytes. J Invest Dermatol 2008;128:1050-3.
- 49. Stevenson DA, Zhou H, Ashrafi S, Messiaen LM, Carey JC, D’Astous JL, Santora SD, Viskochil DH. Double inactivation of NF1 in tibial pseudarthrosis. Am J Hum Genet 2006;79:143-8.
- 50. Lee SM, Choi IH, Lee DY, Lee HR, Park MS, Yoo WJ, Chung CY, Cho TJ. Is double inactivation of the Nf1 gene responsible for the development of congenital pseudarthrosis of the tibia associated with NF1? J Orthop Res 2012;30:1535-40.
- 51. Riccardi VM. Neurofibromatosis: Phenotype, Natural History and Pathogenesis. Baltimore: Johns Hopkins University Press; 1992.
- 52. Bloomfield G, Traynor D, Sander SP, Veltman DM, Pachebat JA, Kay RR. Neurofibromin controls macropinocytosis and phagocytosis in Dictyostelium. Elife 2015;4:e04940.
- 53. Bloomfield G, Kay RR. Uses and abuses of macropinocytosis. J Cell Sci 2016;129:2697-705.
- 54. Zhang S, Charest PG, Firtel RA. Spatiotemporal regulation of Ras activity provides directional sensing. Curr Biol 2008;18:1587-93.
- 55. Li L, Ginty DD. The structure and organization of lanceolate mechanosensory complexes at mouse hair follicles. Elife 2014;3:e01901.
- 56. Munger BL, Renehan WE. Degeneration and regeneration of peripheral nerve in the rat trigeminal system: III. Abnormal sensory reinnervation of rat guard hairs following nerve transection and crush. J Comp Neurol 1989;283:169-76.
- 57. Cauna N. Fine morphological characteristics and microtopography of the free nerve endings of the human digital skin. Anat Rec 1980;198:643-56.
- 58. Son YJ, Thompson WJ. Nerve sprouting in muscle is induced and guided by processes extended by Schwann cells. Neuron 1995;14:133-41.
- 59. Hammarlund M, Nix P, Hauth L, Jorgensen EM, Bastiani M. Axon regeneration requires a conserved MAP kinase pathway. Science 2009;323:802-6.
- 60. Valakh V, Frey E, Babetto E, Walker LJ, DiAntonio A. Cytoskeletal disruption activates the DLK/JNK pathway, which promotes axonal regeneration and mimics a preconditioning injury. Neurobiol Dis 2015;77:13-25.
- 61. Frey E, Valakh V, Karney-Grobe S, Shi Y, Milbrandt J, DiAntonio A. An in vitro assay to study induction of the regenerative state in sensory neurons. Exp Neurol 2015;263:350-63.
- 62. Nix P, Hisamoto N, Matsumoto K, Bastiani M. Axon regeneration requires coordinate activation of p38 and JNK MAPK pathways. Proc National Acad Sci U S A 2011;108:10738-43.
- 63. Chen M, Geoffroy CG, Wong HN, Tress O, Nguyen MT, Holzman LB, Jin Y, Zheng B. Leucine Zipper-bearing Kinase promotes axon growth in mammalian central nervous system neurons. Sci Rep 2016;6:31482.
- 64. Riccardi VM. The genetic predisposition to and histogenesis of neurofibromas and neurofibrosarcoma in neurofibromatosis type 1. Neurosurg Focus 2007;22:E3.
- 65. Masson P. Human tumors: Histology, Diagnosis, Technique. Detroit: Wayne State University Press; 1970.
- 66. Tucker T, Riccardi VM, Brown C, Fee J, Sutcliffe M, Vielkind J, Wechsler J, Wolkenstein P, Friedman JM. S100B and neurofibromin immunostaining and X-inactivation patterns of laser microdissected cells indicate a multicellular origin of some NF1-associated neurofibromas. J Neurosci Res 2011;89:1451-60.
- 67. Tucker T, Riccardi VM, Sutcliffe M, Vielkind J, Wechsler J, Wolkenstein P, Friedman JM. Different patterns of mast cells distinguish diffuse from encapsulated neurofibromas in patients with neurofibromatosis 1. J Histochem Cytochem 2011;59:584-90.
- 68. Cauna N. The free penicillate nerve endings of the human hairy skin. J Anat 1973;115:277-88.
- 69. Son YJ, Trachtenberg JT, Thompson WJ. Schwann cells induce and guide sprouting and reinnervation of neuromuscular junctions. Trends Neurosci 1996;19:280-5.
- 70. Auld DS, Robitaille R. Perisynaptic Schwann cells at the neuromuscular junction: nerve- and activity-dependent contributions to synaptic efficacy, plasticity, and reinnervation. Neuroscientist 2003;9:144-57.
- 71. Son YJ, Thompson WJ. Schwann cell processes guide regeneration of peripheral axons. Neuron 1995;14:125-32.
- 72. O’Malley JP, Waran MT, Balice-Gordon RJ. In vivo observations of terminal Schwann cells at normal, denervated, and reinnervated mouse neuromuscular junctions. J Neurobiol 1999;38:270-86.
- 73. Carrasco DI, Seburn KL, Pinter MJ. Altered terminal Schwann cell morphology precedes denervation in SOD1 mice. Exp Neurol 2016;275:172-81.
- 74. Rimer M. Neuregulins at the neuromuscular synapse: past, present, and future. J Neurosci Res 2007;85:1827-33.
- 75. Ball NJ, Kho GT. Melanocytic nevi are associated with neurofibromas in neurofibromatosis, type I, but not sporadic neurofibromas: a study of 226 cases. J Cutan Pathol 2005;32:523-32.
- 76. Von Recklinghausen F. Uber die Multiplen Fibrome der Haut und ihre Beziehung zu Multiplen Neuromen. Berlin: August Hirschwald; 1882.
- 77. Harkin JC, Reed RJ. Tumors and lesions of neurofibromatosis and other types of neurocutaneous phakometosis. Atlas of tumor pathology second series. Washington DC: Armed Forces Institute of Pathology; 1969. p. 67-106.
- 78. Riccardi VM. Neurofibromatosis type 1 is a disorder of dysplasia: the importance of distinguishing features, consequences, and complications. Birth Defects Res A Clin Mol Teratol 2010;88:9-14.
- 79. Riccardi VM. Von Recklinghausen neurofibromatosis. N Engl J Med 1981;305:1617-27.
- 80. Eames RA, Gamble HJ. Schwann cell relationships in normal human cutaneous nerves. J Anat 1970;106:417-35.
- 81. Gamble HJ, Fenwick RG, Allsopp G. Electron microscope observations on the changing relationships between unmyelinated axons and Schwann cells in human fetal nerves. J Anat 1978;127:363-78.
- 82. Gamble HJ. Further electron microscope studies of human foetal peripheral nerves. J Anat 1966;100:487-502.
- 83. Shahar KH, Solaiyappan M, Bluemke DA. Quantitative differentiation of breast lesions based on three-dimensional morphology from magnetic resonance imaging. J Comput Assist Tomogr 2002;26:1047-53.
- 84. Ghosh SS, Kakunoori S, Augustinack J, Nieto-Castanon A, Kovelman I, Gaab N, Christodoulou JA, Triantafyllou C, Gabrieli JD, Fischl B. Evaluating the validity of volume-based and surface-based brain image registration for developmental cognitive neuroscience studies in children 4 to 11 years of age. Neuroimage 2010;53:85-93.
- 85. Yuan F, Leunig M, Berk DA, Jain RK. Microvascular permeability of albumin, vascular surface area, and vascular volume measured in human adenocarcinoma LS174T using dorsal chamber in SCID mice. Microvasc Res 1993;45:269-89.
- 86. Kaufmann D, Tinschert S, Algermissen B. Is the distribution of dermal neurofibromas in neurofibromatosis type 1 (NF1) related to the pattern of the skin surface temperature? Eur J Dermatol 2001;11:521-6.
- 87. Meany H, Dombi E, Reynolds J, Whatley M, Kurwa A, Tsokos M, Salzer W, Gillespie A, Baldwin A, Derdak J, Widemann B. 18-fluorodeoxyglucose-positron emission tomography (FDG-PET) evaluation of nodular lesions in patients with neurofibromatosis type 1 and plexiform neurofibromas (PN) or malignant peripheral nerve sheath tumors (MPNST). Pediatr Blood Cancer 2012;60:59-64.