Untitled Document


Wladislaw Ellis


During investigation of patients with neurogenic thoracic outlet syndrome, a syndrome that produces significant disability and can dramatically improve with decompressive surgery but one without consistent objective findings, our group found that relatively small structures, when resected, produced relief. These structures were presumed to be fibrotic (scar tissue), but when examined histologically and biochemically included significant amounts of growing neural tissue producing pro-inflammatory cytokines. This led directly to our development of the concept we termed  "innervated fibrosis" (1). This concept, that apparently passive fibrotic tissue can contain metabolically active neural tissue causing pain, organ dysfunction and further innervated fibrosis, has been found to apply to  other disorders, including endometriosis, painful keloids, interstitial cystitis, post myocardial infarction scarring, and chronic appendicitis among others. It is possible that this concept could explain some or all of these disorders suggesting treatment strategies for them as well.

Thoracic Outlet Syndrome- What is the lesion?

Neurogenic thoracic outlet syndrome (NTOS) is an upper extremity pain syndrome thought to be caused by compressive irritation of the brachial plexus at the scalene triangle. Unfortunately, there are no absolute, concrete, "objective" criteria for diagnosis and no measurable, consistent findings on examination or imaging that prove its existence or predict success of treatment to any significant degree (2).

Because of this, our group attempted to determine what factors correlated with success after surgery by means of intraoperative infrared imaging. This technique, intraoperative thermography, allows the almost immediate (within seconds) visual identification of the irritation or relief of a specific  peripheral nerve or trunk during surgery by imaging the more distal skin surface., Thus, the removal of any offensive structures is immediately apparent.  Not surprisingly, such structures turned out to be mainly compressive, actively deforming the large nerve trunks with hypertrophied or fibrotic scalene muscles or fibrous bands. However, we also noticed small (1/2-2cm.) adherent fibrotic tags which were consistently present. When these were resected, we found that better results were seen more consistently than with any other of the offending lesions (3).

Interestingly, we have seen the same thing during spinal surgery where the removal of dural, adherent, fibrotic tags in cases of severe arachnoiditis normalized peripheral nerve symptoms and function (4). These findings (small, scarred  lesions in NTOS and arachnoiditis which, when resected, were associated with disproportionate benefit) seemed hard to understand as most would suggest that fibrosis is not innervated. Fascinatingly, however, histology of these fibrotic tags showed a rich cellular structure in addition to the macroscopic fibrosis. Mature and immature nerves, largely unmyelinated, coursed throughout the tissue samples. Mast cells, fibroblasts, and neutrophils were present. Were the fibrotic tags sites of "diffusable dysfunction" neurogenically producing cytokines, growth factors and possibly transcription factors in a paracrine fashion causing reversible (largely) damage but no cell death (necrosis, apoptosis)? Similar inflammatory residues are widely accepted as driving a wide variety of diseases(5). As a result, we came to local neural dysfunction (driving inflammation and growth) as the dominant lesion in NTOS.

Innervated Fibrosis- a different kind of pathology

We are proposing a parenchymal disorder driven by the products of neurogenic inflammation in local tissue neural nets and occasioned by the continued lack of maturation and consequent continued hypersensitivity, immaturity, of those same neural nets. Symptoms are largely a function of the specific tissue the innervated fibrosis (IF) resides in. Importantly, we propose that the continued hypersensitivity is the result of phenotypic changes in local sensory networks driven by unknown transcription factors (probably more than one) that de-differentiate or trans-differentiate the affected neural nets into mechano- and chemo-sensitive phenotypes providing the essential substratum for continued pathology. This occurs by virtue of their continuing to initiate inflammation and growth. As a result, otherwise normal mechanical or biochemical perturbations give rise to focal inflammation, pain,  dysfunction and the disorder spreads.

Evidence for this hypothesis consists of our work with NTOS, mentioned above, clinical findings and correlations, and consistent histological and immunochemical abnormalities across many varied disorders. Let's list what we know, proceed sequentially through the suspect disorders and then sum up.

We know that the fibrosis characteristic of many different chronic, painful disorders is, in fact innervated and capable of initiating local inflammation.  We know this by virtue of specific immune staining for both immaturity (GAP-43, varied others)  and the secretion of pro-inflammatory neuropeptides (SubP, CGRP, BK, NGF for a partial and evolving list)  in those nerves and the isolation of these secreted inflammatory compounds on a cellular level (citations given by disease, below).

We also know what they have in common clinically: chronic pain punctuated by flares for little apparent reason; lesion hypersensitivy to a wide variety of sensory stimuli; chronicity and intransigence to treatment (steroid insensitive). Besides etiology, it is these factors that are largely unexplained and continue to make  these disorders contentious. IF offers an alternative by restricting the pathology to existing neural nets and the inflammatory symptoms and growth they can produce.

Other disorders showing evidence of IF

Endometriosis, thought to affect up to ¼ of the young women in the United States, disables with focal inflammation of endometrial-like fibrotic tissues ectopically embedded in, usually, peritoneum or pelvis.These lesions are mechanosensitive, as demonstrated intraoperatively, and richly innervated with immature GAP43+ myelinated and unmyelinated fibers(6). They release pro-inflammatory peptides which cascade cytokines and growth factors with NGF being prominent. They also activate the neuro-immune axis, stimulating the high concentration of resident and sensitized mast cells. Mast cells feature prominently as a final biochemical effector-- they can selectively release a large number of vaso- and neuro active substances or degranulate totally(7). So far, when looked for, mast cells have been consistently involved in IF. When activated, mast cells can be neuronally discharged and cause pain, swelling and systemic symptoms such as malaise/dysphoria. These and other symptoms, depending  on the organs adhered to, follow the proliferating, fibrotic foci of endometrial-like tissue with its resident active nerve and mast cell effectors. Total eradication of these fibrotic, inflammatory foci even surgically is rare. Etiology continues being contentious as are many treatment approaches(8). Importantly, heparin delivered locally produces time limited (2wks.) improvement(9). Heparin is well recognized for its ability to sequester and inactivate many physiologically active molecules, including the cytokines and growth factors produced by IF (10).

Entrapment syndromes constitute another large patient population suffering from progressive pain and dysfunction with evidence for IF. Symptoms, at least initially, are localizable to discrete nerve groupings at, for example, the carpal tunnel, ulnar sulcus, or supraclavicularly at the thoracic outlet where, when looked for, there is fibrosis and local inflammation about the nerves in question(11). Looking further, we determined that this innervated fibrosis was pro-inflammatory and causing symptoms best explained by the unusual high concentration of new, immature, neurites secreting pro-inflammatory (+SubP, +CGRP, +NGF) substances(12). Immature neurites are very mechano- and chemo-sensitive as is easily seen in examining carefully any incisional wound margin. The same thing happens in and about nerve trunks with IF. Another group, looking at carpal tunnel syndrome has shown pro-inflammatory neurite growth with fibrosis, explaining symptoms similarly.  They showed that neurite growth was initiated by stretch sensitive Schwann cells which de-differentiate, dysregulate myelin associated glycoprotein (MAG) which then releases new neurite formation and the accompanying sensitivity. Because of this they termed entrapments to be "Schwannopathies", but with no determination as to what sensitizes or de-differentiates Schwann/Remak cells (13). Heparin has been found to consistently  alleviate symptoms in otherwise refractory cases of nerve entrapments, underscoring the importance of the neurogenically induced  cytokines and growth factors(14). Erythropoietin (Epo), recognized as a neuroprotectant and as a transcription factor inducing neural and Schwann cell differentiation, is possibly even more effective in controlling symptoms (15). This underscores the importance of the immature state.

Hypertrophic painful scars or keloids turn out to be highly innervated. Many of the neurons stain for GAP43, SubP, CGRP as evidence for continued neurogenic inflammation (16). Activated mast cells in conjunction with monocytes and extensive fibroblasts create a proliferating and innervated matrix. Prurigo nodularis shows similar symptoms and histology albeit with increased activated mast cells and much less fibrosis (17).

Post myocardial infarction scarring shows not only dense myofibroblast proliferation but also Schwann cell and new neurite ingrowth (GAP43+), neural stem cell marker nestin as well as NGF (18). The new innervation is thought to contribute arhythmogenicity although local inflammatory effects remain an open question (19).

Discogenic low back pain has been shown by multiple groups to result from extensive neurite  (GAP43+, NGF+) ingrowth into the degraded disc (20). Symptoms can resolve dramatically with the intra-discal injection of methylene blue, a known neurotoxin, emphasizing the primacy of the ingrown and immature nerves (21). This new nerve growth is accompanied by various degrees of fibrosis and immune activation (increased activated mast cells, endothelial leakage), seen also in tendinoses which are also responsible for much pain and disability. (22)

Arachnoiditis, thought to be largely due to instrumentation (anesthesia, pain control) has been reported to have free nerve endings in the often extensive adhesive fibrosis (23). Further histology seems to be lacking but predicting the existence of mechano and chemo-sensitive immature neurites initiating pro-inflammatory cascades seems appropriate. Cytokines, growth factors and increased immune (mast) cells and fibroblasts are likely. Indeed, nerve driven localized inflammatory dural foci offer a reasonable explanation of events.

Interstitial cystitis is yet another disorder with multiple contending etiologies. Symptoms are often severe, progressive, and difficult to treat. Increased submucosal nerves staining for NGF, CGRP, and SubP intimately co-exist with mast cells which are often degranulated on electron microscopy, all proliferating in an increasingly fibrotic extra cellular matrix (ECM) (24). These submucosal nerves are immature and sensitized to respond to minimal distension with proinflammatory peptides. This neurogenic inflammation forms a vicious cycle of increasing pain, bladder dysfunction, dyspareunia and malaise. Again, the self-perpetuation of new neurites, inflammation and fibrosis, restricted , in this case, to the ECM, produces and then spreads symptoms and dysfunction. The perpetuated neural immaturity and inflammation seem key to the disease process. Heparin will often be effective when instilled into the bladder, showing the importance of neurogenically induced cytokines and growth factors (25).

Peritoneal adhesions show significant innervation with +NGF, +SubP, +CGRP and neurites in close approximation to mast cells (26). These innervated fibrotic tags and adhesions show varied degrees of immune infiltration and inflammatory changes, provoking pain when manipulated intraoperatively, and spread with symptom progression (27). Treatment is difficult, at best, given that re-operation often provokes more IF. Anecdotal use of heparin has improved symptoms.

Chronic appendicitis, pancreatitis, and hepatitis all show increased mucosal nerve densities in addition to the obvious fibrotic changes (28,29,30). As with the previous examples of putative IF, low grade, focal neurogenic inflammation in an expanding and innervated fibrotic matrix, largely unresponsive to treatment and mechano- and chemo-sensitive speaks to its likely inclusion in the spectrum of disorders that feature IF prominently.

Asthma is a disease characterized by perpetuated chronic inflammation and airway remodeling. The pulmonary stroma shows increase fibrosis as well as increased neurites delivering NGF, SubP, CGRP, VIP, and varied other growth factors to a sensitized pulmonary tree (31). This sets the stage for substances that are otherwise innocuous to provoke bronchoconstriction, plasma extravasation, and the continued cycle of symptomatology and dysfunction, organ (lung) based, showing steroid sensitivity short but not long term. Chronic, continued focal  inflammation in the context of growing, increasingly sensitized nerves capable of inflammation and  bronchoconstriction, as well as poor treatment response point at IF.

A number of other disorders with typically contested etiologies, chronic pain, changes in the ECM, and evidence for neurogenic inflammation come to mind. These include migraine, complex regional pain syndrome, fibromyalgia, temporo-mandibular joint syndrome, chronic cough, adiposis dolorosa, and even chronic fatigue syndrome (32,33,34,35,36,37,38). There are, probably, others.


We have seen that abnormal, immature, neurites, secreting pro-inflammatory factors in a progressively fibrotic matrix, are present and active in a wide range of disorders. Although seen largely  as an effect of the disorder, this innervated fibrosis is, in our estimation, causal in producing the disorder's symptoms and dysfunction. This has been shown by direct manipulation of the fibrotic tags with the patient awake or by thermographic imaging when anesthetized. Neurogenic inflammation has been found when looked for with symptoms abating when neuropeptides were inhibited. Again, when looked for, pro-inflammatory peptides have been found consistently and associated both in time and place with immature neurites and symptomatic tissues. Symptom spread corresponds with new tisssue involvement. All of this, given the demonstrated morbidity attendant to cytokines and other pro-inflammatory factors, argues for neuronally triggered disorders where symptoms reflect the organ affected (39).

One of the more troublesome aspects, clinically, of all of these disorders is their chronicity. Our own conclusion is that the abnormally sensitized nets stay that way because the neurites don't differentiate into the mature  and much less mechano- or chemo-sensitive phenotype. As a result the initial inflammatory changes in the ECM largely remain or enlarge with repeated episodes of local inflammation and cellular neogenesis. This is certainly speculative, but is buttressed by the immuno-staining evidence of neural immaturity, continued response to agents neutralizing neuropeptides and growth factors,  and a positive response to neural differentiation.

Fundamentally, this approach allows for entirely new interventions for many, if not all of these disorders. The most productive would be the identification of which (highly likely) environmental factors and/or genetic predispositions are responsible for sensitizing susceptible individuals. This is also the most problematic but advances in genetic and tissue engineering, and micro-analysis (proteomics, genomics, complex systems) offer hope.

Next, inactivating cytokines, growth factors, cell adhesion molecules, and complex lipids involved in various non-COX mediated inflammatory cascades has shown promise and needs development. TNF alpha inhibitors, heparin, octreotide all shown  benefit if used appropriately. More cytokine inhibitors are in the pipeline (40).

BDNF and NGF are obvious candidates for inhibition. The recent lack of benefit from anti-NGF treatment of a painful peripheral neuropathy might reflect a problem common to many of these disorders- that of systemic toxicity before the local concentration of the medication reaches therapeutic levels. We largely by-pass this problem by delivering small volumes and concentrations (but enough for benefit) by injection. DMSO can be useful in transporting many substances to some depth subcutaneously.

Vasospasm is a common problem, especially in patients with nerve entrapments, often becoming a frank Raynaud's phenomenon and contributing to morbidity. Relieving vasospasm relieves symptoms. Fasudil, a Rho-kinase inhibitor shows exceptional promise (41).

Lastly, finding more terminally differentiating factors besides Epo for what might be a disease of developmental regulation should be exciting and productive. If the neurons, Schwann cells, and mast cells could be made to function responsibly, we might have a cure.

In summary, we propose that developmental dysregulation (dedifferentiation and stasis) of tissue neural networks, making them increasingly sensitive to otherwise normal mechanical and chemical stimuli, producing frequent, punctuated episodes of inflammation and pain results in diseases characterized by tissue specific symptoms. Chronicity and spread are conjectured, plausibly, to be the result of continued dedifferentiation/developmental arrest at an immature and very sensitive phenotype,a form of aberrant wound healing and likely subtle small fiber polyneuropathy. Verification or falsification are clearly in order.

June 2012


1) Cheng S, Ellis W, Stoney R. Thoracic Outlet Syndrome: Supraclavicular Approach, in Rob and Smith's Operative Surgery/ Vascular Surgery; 5th ed. Editors Jamieson C, Yao J: 475-482, 1994.

2) Cheng S, Reilly K, Ellis W, Stoney R. Neurogenic Thoracic Outlet Decompression: Rationale for Sparing the First Rib. Cardiovasc Surg, 3 (8):617-623, 1995.

3) Ellis W, Cheng S. Intraoperative Thermographic Monitoring During Thoracic Outlet Decompressive Surgery. Vasc & Endovasc Surg, 37(4):253-257, 2003.

4) Ellis W, Morris J, Hosobuchi Y. Intraoperative Thermography in Spinal Disease. Clinical Proc J Academy Neuromuscular Thermography, 1:60-62, Oct 1988.

5) Richardson J, Vasko M. Ceelular mechanisms of Neurogenic Inflammation. J Pharm & Exp Therapeutics, 302(3): 839-45, 2002.

6) Bokor A, et al. Density of  small diameter sensory nerve fibers in endometrium: a semiinvasive diagnostic test for minimal to moderate endometriosis. Human Reproduction, 24(12):3025-3032, 2009.

7) Tariverdian N, et al. Neuro-endorcrine-immune disequilibrium and endometriosis: an interdisciplinary approach. Semin Immunopathol, 29(2):193-210, 2007.

8) de Ziegler D, Borghese B, Chapson,C. Endometriosis and infertility: Pathophysiology and management. Lancet, 376:730-737, 2010.

9)  Anecdotal, personal experience with a small number of patients.

10) Engelberg H. Heparin and the Prevention of Atherosclerosis Basic Research and Clinical Application. New York-Wiley- Liss, pp 39-139, 1990.

11) McCall T, et al. Treatment of Recurrent Peripheral Nerve Entrapment Problems- the role of scar formation and its possible treatment. Neurosurg Clinics of No Am, 12(2):329-339, 2001.

12) Ellis W. Thoracic-Outlet Syndrome as a Disorder of Neurogenic Inflammation. Vasc & Endovasc Surg, 40(3):251-254, 2006.

13) Gupta R, Rummler L, Steward O. Understanding the biology of compressive neuropathies. Clin Orto & Related Res, 436:251-260, 2005.

14) Ellis W. Heparin alleviates pain in nerve entrapments. Amer J Pain Mgmt, 13(2):54-59, 2003.

15) Ellis W. Erythropoietin as a novel analgesic for neuropathic pain. Abstracts 12th World Congress on Pain, Control # 08-A-1245-IASP, Glasgow, Scotland, 8/17-22 2008.

16) Crowe R, et al. Neuropeptide containing nerves in painful hypertrophic scar tissue. British J Derm, 130(4):444-452, 1994.

17) Haas S, Straub R, Stander S. Low density of sympathetic fibers relative to Substance P positive nerve fibers in lesional skin of chronic pruritis and prurigerm Sci, 50(3):193-197, 2010.

18) Drapeau J, et al. Nestin Expressing Neural Stem Cells Identified in Scar Following Myocardial Infarction. J Cell Physiol, 204:51-62, 2005.

19) Oh Y, etal. Arrhythmogenic Gene Changes and Nerve Sprouting after Acute Myocardial Infarction in Mice. Korean Circulation J, 37:399-407, 2007.

20) Fremont A. The cellular pathobiology of the degenerate intervertebral disc and discogenic low back pain. Rheumatology, 48:5-10, 2009.

21) Peng B, etal. A randomized placebo controlled trial of intradiscal methylene blue injection for the treatment of chronic discogenic low back pain. Pain, 149:123-129, 2010.

22) Fredberg U, Stengaard-Pederson K. Chronic tendinopathy  tissue pathology, pain mechanisms, and etiology with special focus on inflammation. Scanf J Med Sci Sports, 18:3-15, 2008.

23) Richardson J. Obstetric epidural and chronic adhesive arachnoiditis. Br J Anesth, 92(5):765-767,2004.

24) Vizzard M. Alterations in neuropeptide expression in lumbosacral bladder pathways following chronic cystitis. J Chem Neuroanat, 21(2):125-136, 2001.

25) ChenD, Kuo H. Treatment of lower urinary tract symptoms and overactive bladder- focus on bladder sensory innervation. Tzu Chi Med J, 18(5):333-344, 2006.

26) Herrick S, et al. Human peritoneal adhesions are highly cellular, innervated and vascularized. J Pathol, 192:67-72, 2000.

27) Sulaiman H, et al. Presence and distribution of sensory nerve fibers in human peritoneal adhesions. Ann Surg, 234(2):256-261, 2001.

28) SebastianoP, et al. Neuroimmune appendicitis. Lancet, 354:461-466,1999.

29) Sebastiano P, et al. Chronic Pancreatitis: the perspective of pain generated by neuro-immune interaction. Gut, 52(6):907-911, 2003.

30) Matsunaga Y, Kawasaki H, Terda T. Stromal Mast Cells and Nerve Fibers in Various Chronic Liver Diseases; Relevance to Hepatic Fibrosis. Am J Gastroent, 94(7):1923-1932, 1999.

31) Groneberg D, et al. Neurogenic mechanisms in bronchial inflammatory disease. Allergy, 59(11):1139-1152, 2004.

32) Waeber C, Moskowitz M. Migraine as an inflammatory disorder. Neurology, 64(supp2):809-815, 2005.

33) Watkins L, Maier S. Beyond Neurons: Evidence that Immune and Glial Cells Contribute to Pathological Pain States. Physiol Rev, 82:981-1011, 2002.

34) Baraniuk J, Zheng Y. Relationshibs among rhinitis, fibromyalgia, and chronic fatigue. Allergy and Asthma Proceedings, 31(3):169-178, 2010.

35) Field L. Temporomandibular Syndrome: The Great Imboster. The Pain Practitioner, 20(2):49-51, 2010.

36) Widdicombe J. Neurophysiology of the cough reflex. Eur Respir J, 8:1193-1202, 1995.

37) Herbst K, Asare-Bediako S. Adiposis Dolorosa is more then Painful Fat. The Endocrinologist, 17:326-334, 2007.

38) Bruno R, Creange S, Frick N. Parallels between post-polio fatigue and chronic fatigue syndrome: a common pathophysiology? Am J Med, 105(3A):66-73, 19998.

39) Black P. Stress and the inflammatory response: a review of neurogenic inflammation. Brain Behav Immun, 16(6):622-653, 2002.

40) Ellis W. Octreotide, a small peptide, alleviates burning pain and hyperesthesia: a preliminary study. The Pain Clinic, 3(4):239-242, 1990.

41) Boyce-Rustay J, et al. Characterization of Fasudil in Preclinical Models of Pain. The J of Pain, 11(10):941-949, 2010.