Murray Kornfeld, American College of Chest Physician, and Sarcoidosis: a Historical Footnote^*: 2004 Murray Kornfeld Memorial Founders Lecture FREE TO VIEW

Emerson wrote, “A great institution is the lengthened shadow of one man.” If this is so, the American College of Chest Physicians is the lengthened shadow of Murray Kornfeld. Son of a successful New York businessman, Murray was born in 1893 in the city’s East Side. Tuberculosis afflicted him in his adolescent years, but the disease went into natural remission. Soon after, he joined the South American division of an exporting firm. At the age of 30, tuberculosis came back and Murray was admitted to America’s “magic mountain,” the sanatorium at Saranac Lake founded by another tuberculosis patient, Edward Livingston Trudeau.¹

In 1876, Dr. Edward Livingston Trudeau (1848–1915), ailing from consumption, came to Saranac Lake for the hospitable climate that was conducive to healing. His health improved, and in 1884 he started Adirondack Cottage Sanatorium that became the Trudeau Sanatorium after his death. The hospital attracted patients with tuberculosis and other types of chest diseases, including bronchiectasis, chronic bronchitis, lung abscess, and lung fibrosis. Dr. Trudeau employed the open-air treatment of the disease and organized the first laboratory for the study of tuberculosis. In that era, it was not uncommon for sarcoidosis patients to be admitted to tuberculosis sanatoriums because of the clinical similarities between the two diseases.

Many years before Murray’s admission to the sanatorium, Robert Louis Stevenson stayed at the Saranac Lake Hospital (1887–1889). He had shortness of breath and cough but no weight loss, fever, or other systemic symptoms. Repeated sputum examinations by Dr. Trudeau failed to reveal any tubercle bacilli. It is probable that Stevenson had sarcoidosis. In 1894, at the young age of 44 years, while in Samoa, a slow heart rate developed and he died suddenly. Did the creator of Dr. Jekyll and Mr. Hyde die of myocardial sarcoidosis?² There were, during Murray Kornfeld’s long stay at the sanatorium, patients with sarcoidosis masquerading as tuberculosis. These patients, however, remained undetected for sarcoidosis was not yet discovered in America.

Murray Kornfeld was curious and energetic. The medical and social aspects of tuberculosis fascinated him. He was keen on developing new approaches to teach tuberculosis to medical practitioners and the general public. In 1924, he started a monthly journal called Woodland Whispers. He was able to persuade the townspeople and local merchants to underwrite its costs. Later, he invited Dr. Charles Hendricks, a chest physician with an interest in writing, to edit a journal on chest diseases. In early 1935, the first issue of Diseases of the Chest, the predecessor of present CHEST, appeared. In the later part of that year, in Albuquerque, NM, the Federation of American Sanatoria was created, which was renamed the American College of Chest Physicians in June 1937.

In the earlier decades of the 20th century, sarcoidosis was not a significant illness; its existence was shrouded in the veil of tuberculosis. The disease was not in any of the editions of William Osler’s Textbook of Medicine or in the medical tomes by Drs. Austin Flint and George Pepper.

Nevertheless, I believe that the case of Hannah W., a 14-year-old African-American girl admitted to the Johns Hopkins Medical Center, published in the American Journal of Medical Sciences, reported by Osler, had sarcoidosis. The patient had bilateral parotid and lacrimal enlargement, lung disease, and pleural involvement. At autopsy, no evidence of tuberculosis was found.³

In 1937, the first significant publication of the disease that hitherto had not gained much attention in America appeared; it was a 75-page review from Johns Hopkins Hospital by Longcope,⁴ which introduced sarcoidosis to the American medical audience. This event took place almost 6 decades after the first recorded and illustrated case of sarcoidosis by Jonathan Hutchinson, a London dermatologist, who published the case of John W, in Illustrations of Clinical Surgery in 1877, just a year after Dr. Trudeau opened the doors of his institution, under the title of “Case of livid papillary psoriasis.“⁵ In 1895, Dr. Hutchinson presented a 64-year-old woman to the Dermatological Society of London⁶: When Dr. Hutchinson suggested skin biopsy for histologic proof, Mrs. Mortimer promptly disappeared.

I have to describe a form of skin disease which has, I believe, escaped especial recognition. It may not improbably be a tuberculous affection and one of the lupus families, but if so it differs widely from all other forms of lupus, both in features and its course. The multiplicity of the patches; their occurrence in groups; their bilateral symmetry; and the absence of all tendencies to ulcerate or form crusts are the features that separate the malady from lupus vulgaris… I prefer to recognize it by the name of one of its subjects, as Mortimer’s malady.

In 1897, Caesar Boeck presented to the Medical Society of Christiania a 34-year-old police constable with “multiple benign sarkoid of the skin” and drew attention to its similarity to the Mortimer malady. Unlike Mrs. Mortimer, Dr. Hutchinson’s patient, Dr. Boeck’s patient agreed to undergo a skin biopsy. The histologic picture showed sharp, well-defined foci of epithelioid cells with some giant cells permeating the corium. Caesar Boeck became the first physician to describe histology of sarcoid granuloma.⁷ With the advent of chest radiography in 1915, two Swedish physicians, Jorgen Schaumann, a dermatologist, and Sven Lofgren, a general practitioner, defined the typical features of hilar and mediastinal node enlargement and linked them with noncaseating granulomas.

Although Dr. Hutchinson’s teachings of sarcoidosis had not reached America, he remained the most sought-after teacher and physician in Britain and parts of Europe. Dr. Hutchinson’s contemporary, Dr. Arthur Conan Doyle, creator of the fictional character Sherlock Holmes, made a skin disease—most likely cutaneous sarcoidosis—the basic ingredient of the plot of the Adventures of the Blanched Soldier.⁸ Whether or not Dr. Doyle ever discussed patients with Dr. Hutchinson remains unclear, but one can safely surmise that the two most likely had met and exchanged ideas during the London Medical Society Meetings, where Dr. Hutchinson was a member and regular presenter and discussant. Dr. Doyle practiced at 2 Devonshire Street, just a few steps away from 10 Chandos Street, home of the Medical Society of London. In a letter to The Lancet, Dr. Doyle described a family in which three members had systemic illness resembling sarcoidosis.⁹ This most likely was the first allusion to familial sarcoidosis. The familial occurrence of sarcoidosis is now established. Recent studies from the United Kingdom and the United States and A Case Control Etiologic Study of Sarcoidosis show that the relative risk of sarcoidosis developing ranges from 36 to 73 if a first-degree or a second-degree family member has sarcoidosis.¹⁰

In 1954, in India, during my first year of pathology class, I was told that the sarcoidosis granuloma first described by Dr. Boeck was an example of perimeter defense mounted by the immune system against invading antigens or organisms. Tuberculosis then, as now, a widespread cause of granulomatous inflammation, was an excellent example of the phenomenon. However, granulomas were also seen in fungal disorders, tropical parasites (schistosomiasis and leishmaniasis), syphilis, brucellosis, and foreign body reactions. Sarcoid granuloma, we were told, was different: it was noncaseating and in it no organisms were to be seen, no matter how hard we tried in our laboratory. Because of clinical and radiologic similarities, however, sarcoidosis was then considered “atypical” tuberculosis.¹¹ Sarcoidosis, particularly in India and in Africa, is also confused with leprosy; in tuberculoid leprosy, the granuloma is noncaseating and indistinguishable from a sarcoid granuloma.¹² During the middle of the last century, most of the poor countries had no economic resources to sustain radiograph departments and functional biochemical and pathology laboratories. The differentiation between sarcoidosis and leprosy rested mainly on clinical examination. If a patient with noncaseating granulomas in a skin biopsy had anesthetic skin patches on the body, it was leprosy; if not, it was sarcoidosis. We used to carry a hatpin in our coat lapel to examine for the sensory loss! If the patient had palpable supraclavicular nerves or thickened ulnar nerve, the diagnosis of leprosy was clinched.

Mankiewicz and Van Wallbeck¹³ proposed that sarcoidosis resulted from tissue invasion by phage-infected mycobacteria in individuals who lacked the capacity to form phage-neutralizing antibodies. Kozmin-Sokolow and Kostina,¹⁴ however, found mycobacteriophage in only 2 of 40 samples from 22 patients with sarcoidosis and in 1 of 26 samples from 9 patients with tuberculosis. Chapman and Speight¹⁵ showed that 90% of their patients with sarcoidosis had significant titers of antibodies to atypical mycobacteria. Reid and Wolinski¹⁶ found no evidence of precipitating antibodies to atypical mycobacteria in their 29 sarcoidosis patients. In an exhaustive study, Bowman and associates¹⁷ were unable to find any acid-fast bacteria, fungi, or Mycoplasma from sarcoid granulomas. Milman and associates¹⁸ subjected lymph node biopsy specimens from 15 patients with sarcoidosis to prolonged culture (12 months) looking for mycobacteria. No evidence of Mycobacterium tuberculosis and paratuberculosis was found.

Many investigators have used polymerase chain reaction (PCR) technology to detect mycobacterial DNA in clinical specimens from patients with sarcoidosis; only half of the studies showed mycobacterial DNA. Vokurka and colleagues,¹⁹ using highly sensitive PCR technology, which it can detect as few as one copy of the genome, were unable to identify any mycobacterial DNA in patients with sarcoidosis. Brown and associates,²⁰ however, recovered cell wall-deficient mycobacteria with equal frequency from sarcoidosis patients as well as control subjects.

Hosoda and colleagues²¹ analyzed a series of health surveillance data in a Japanese work population of 460,000 employees. The work population underwent radiography and tuberculin testing annually. Hilar adenopathy was observed as a common marker of the two diseases. Primary tuberculosis showed unilateral hilar adenopathy, whereas sarcoidosis displayed bilateral hilar adenopathy. The incidence of tuberculosis and sarcoidosis showed a reversed relationship; in a community, the disappearance of unilateral hilar adenopathy caused by tuberculosis was followed by appearance of bilateral hilar adenopathy due to sarcoidosis. In both diseases, hilar adenopathy subsided within a few years; however, resolution of sarcoid hilar adenopathy was delayed if accompanied by extrapulmonary lesions. Age distribution of sarcoidosis incidence showed a bimodal curve but that of tuberculosis had monomodal curve increasing with advanced age. Sarcoidosis had a higher standardized prevalence ratio in northern Japan, while tuberculosis is higher in southern Japan. The authors²¹ found no epidemiologic evidence of any causal relationship between tuberculosis and sarcoidosis.

Propionibacterium acnes has been reported extensively. Abe and colleagues²² found that lymph nodes from 31 of 40 sarcoidosis patients (77%) and 38 samples of nonsarcoidosis tissue (21.1%) also revealed P acnes in culture. The difference was significant (p < 0.001). Ishige et al²³ used PCR analysis and observed that 15 of 15 sarcoidosis tissue samples were positive for propionibacterial (P acnes or Propionibacterium granulosum) RNA; only 3 were positive for mycobacterial DNA. However, 2 of 15 tissue samples from patients with tuberculosis were positive for propionibacteria and 15 of 15 samples were positive for mycobacteria.²³ In an international study,²⁴ propionibacterial DNA was detected in all but 2 of 108 sarcoidosis lymph node samples compared with M tuberculosis DNA, which was found in 0 to 9% of the samples. Propionibacterial DNA was also found in 0 to 60% of M tuberculosis and control samples.²⁴

The diagnostic test that occupies an exalted place in the annals of differential diagnoses of sarcoidosis is the Kveim-Siltzbach skin test, first described by Williams and Nickerson.²⁵ The antigen was prepared by suspending skin tissue in saline solution, sterilized, and then injected into four patients with sarcoidosis. Within 24 h appeared a firm nodule that was still present after a week. Dr. Kveim, an Oslo dermatologist, added that the papule consisted of granulomatous reaction and asserted that the test was useful to differentiate sarcoidosis from tuberculosis.²⁶ Louis Siltzbach popularized the test and carried extensive clinical studies.

Kveim test antigen is thought to contain antigen/antigens that induce an immune-mediated granulomatous reaction. To date, no causative antigen has been identified in the Kveim suspension. A proteomic approach to identify unique proteins in sarcoidosis tissue might be helpful.

The 1960s and 1970s were the “golden years” for lung function investigation related to sarcoidosis. Nils Svanborg systematically studied cardiopulmonary functions and published the monograph that has been surpassed in merit and pragmatic and academic value.²⁷ The workers at Albert Einstein Medical College Hospital demonstrated that even in stage I disease, airway resistance and diffusing capacity may be severely impaired and cause symptoms, particularly the dry, persistent cough.²⁸

Immunologic alterations have long been described in sarcoidosis. Hypergammaglobulinemia and the presence of cutaneous anergy were recognized early in the history of the disease. However, the widespread use of bronchoscopy and bronchoalveolar fluid studies that began in the mid 1970s led to recognition of increased lymphocytes in BAL and the concept of “active alveolitis.” Ronald Crystal et al²⁹ popularized bronchoscopy during the 1980s. In early sarcoidosis, a large number of CD4 cells appear at the site of inflammation. The cytokine profile is T-helper type 1. The activated cells express increased interleukin (IL)-2 receptors and the spontaneous release of IL-2 by these cells. There is associated macrophage activation, and IL-18 and interferon-γ production.²⁹

Soon after, monoclonal antibody evaluation of lymphocytes quickly established the typical patterns of sarcoid lung involvement. The events that lead to spontaneous resolution of the granuloma formation include an influx of CD8+ cells. IL-10 helps resolution, as it suppresses the inflammatory response. However, the cytokines associated with chronic disease are IL-8, IL-12, and tumor necrosis factor (TNF). Patients with chronic disease have raised IL-8 levels. An increase in IL-8 is associated with an increase of transforming growth factor, a cytokine that promotes fibrosis.³⁰ The role of BAL in diagnosing sarcoidosis remains unsettled.

Louis E. Siltzbach worked at the Mount Sinai Hospital, NY, and so did Burrell M. Crohn. Dr. Siltzbach spent all his life studying sarcoidosis in its infinite details. Dr. Crohn did the same for the disease that bears his name. For years, we have compared and contrasted sarcoidosis with Crohn disease because of the histologic similarity. No causal connection has ever been established between the two disorders, and no one knows the cause.³¹ Both granulomatous disorders can cause hypercalcemia. At Los Angeles County, University of Southern California Medical Center, John Adams and colleagues³² demonstrated that activated alveolar macrophages and granulomas produced 1–25-dihydroxy-vitamin D that caused hypercalcemia by increasing calcium absorption by the gut. Thus, granulomas of sarcoidosis and Crohn disease not only look similar, they have also have similar biochemical properties.

About 20 years ago, the first report of isolation of Mycobacterium avium subspecies paratuberculosis (MAP) from patients with Crohn disease appeared.³³ Recently, Saleh Naser reported the detection of MAP DNA in circulation of patients with inflammatory bowel disease, and viable MAP organisms were cultured in the patients with Crohn disease.³⁴

There is strong evidence that Crohn disease is a complex genetic disease with an environmental factor. Genome-wide scans have found many chromosomal region links with Crohn disease, including the human leukocyte antigen (HLA) region of chromosome 6, cytokine gene clusters on chromosome 5, and the region for CARD 15/NOD2 gene on chromosome 16. The gene product of CARD 15 appears to be an intramolecular pattern-recognition molecule for bacterial peptidoglycans. The CARD 15 mutations in Crohn disease are loss-of-function mutations. Homozygous loss-of-function mutations are associated with high relative risk of Crohn disease in humans and in cattle with Johnne disease.³⁵

Genetic susceptibility to sarcoidosis involves several genes; as yet no single gene responsible for the disease has been found. The major histocompatibility complex (MHC) region on chromosome 6 continues to be an area of interest. The genes that encode the HLA portion of the MHC gene play an important role in determining the risk and clinical course of sarcoidosis; HLA DQB*0602 is transmitted more often than expected to affected offspring and appears to associated with progression of the disease, whereas HLA DQB*1201 is found half as often as expected. HLA DR-11, -12, -14, -15, and -17 seem to confer susceptibility to the disease, and HLA DR-1, DR-4, and HLA DQ*0202 seem to be protective. The TNF complex located near the MHC class II loci on the chromosome 6 appears to play an important part in determining severity of disease together with MHC class I alleles. The − 308 promotor polymorphism of TNF-α is associated with Lofgren syndrome, a form of sarcoidosis with good prognosis.³⁶

In 1960, James,³⁷ in the Second International Congress on Sarcoidosis, introduced the concept of acute and chronic sarcoidosis. Acute disease is characterized by hilar adenopathy and erythema nodosum. The combination reflects good genes and occurs frequently in white patients. Erythema nodosum is not seen in Japanese patients. Lupus pernio, however, represents the chronic, progressive disease with fibrosis and bone lesions. The diagnosis of sarcoidosis is best made when the clinical or radiologic picture is enhanced by the demonstration of noncaseating granulomas and exclusion of tuberculosis and other infections. Biopsy is almost always needed to exclude other causes of granulomatous diseases.³⁸³⁹

In sarcoidosis, clinical presentation, organ involvement, course, and response to treatment vary with the genetic makeup of the host.^40414243 There are many drugs used for controlling sarcoidosis, but none cure the disease.⁴⁴ In selected cases, organ transplantation (lungs, heart, and liver) is successful. The disease, however, returns and attacks the newly transplanted organ.^45464748

Sarcoidosis keeps its distance. It stubbornly refuses to reveal its identity.⁴⁹ We persevere and keep in mind what another Bostonian had so perceptively observed: Robert Frost

We dance around in a ring and suppose, but the secret sits in the middle and knows.