Coexistence of sarcoidosis and malignancy.

Abstract: In 1974, Brincker and Wilbek noted a statistically significant increase of malignant tumors among sarcoid patients and speculated that immunologic deficiencies stemming from sarcoidosis may predispose such patients to malignancy. In this article, we describe three additional cases of sarcoidosis and coexisting malignancy. The temporal presentation, radiographic features, and histologic findings are discussed in detail. The evidence for and against a proposed link between sarcoid and malignancy is reviewed.

Key Words: Hodgkin's disease, lymphoma, sarcoidosis

**********

The coexistence of sarcoidosis and malignancy has been noted for nearly three decades. In 1974, Brincker and Wilbek (1) noted that malignancy occurred in 48 of 2,544 patients enrolled in the Danish sarcoid registry from 1962 to 1971. On the basis of an expected age- and sex-specific incidence of only 33, Brincker and Wilbek concluded that there was an excess number of malignancies in the sarcoid patients. When analyzed further, this higher incidence could be attributed to an increased number of lung cancers and malignant lymphomas. Since then, multiple investigators have reported various observational and epidemiologic studies that suggest a link between sarcoidosis and malignancy. Moreover, the proposed link has been extended to a wide variety of lymphoreticular and solid organ neoplasms. (2-5) In this article, we report three additional cases of malignancy that occurred in the setting of biopsy-proved sarcoidosis.

Case Reports

Patient 1

A 32-year-old black man was evaluated in the pulmonary clinic for an abnormal chest film. The film showed bilateral hilar and right paratracheal adenopathy consistent with Stage 1 sarcoidosis and was unchanged from a chest x-ray (CXR) obtained 3 years previously in 1990 (Fig. 1). Laboratory evaluation was remarkable for an elevation of serum angiotensin-converting enzyme level of 64 U (normal range, 8 52 U/L). A transbronchial biopsy revealed mononuclear inflammation and focal fibrosis but no clearly defined granuloma. He remained asymptomatic until 1996, at which time he noted fatigue, night sweats, and unintentional weight loss. A chest film showed a markedly enlarged right paratracheal node complex with bilateral widening of the superior mediastinum (Fig. 2). the hilar adenopathy remained unchanged. A punch biopsy of a skin rash along the dorsum of the neck revealed multiple well-defined granulomas (Fig. 3). He was given a 2-week course of corticosteroids but showed no interval resolution of his mediastinal adenopathy. A cervical mediastinal exploration with lymph node excision was obtained revealing Hodgkin's disease (HD), nodular sclerosing subtype (Fig. 4). He received six courses of alternating ABVD/ MOPP combination chemotherapy--course 1, Adriamycin (57 mg), bleomycin (23 U), Velban (14 mg), and DTIC (848 mg); course 2, Mustargen (13.6 mg), vincristine (2.0 mg), procarbazine (225 mg), and prednisone (90 mg)--in addition to mantle radiation to the mediastinum (4,140 rad). During a follow-up gallium scan, uptake was noted in both mediastinal nodes and the left scapula. Because there was no other evidence of recurrent HD, he underwent an excisional bone biopsy. The bone biopsy specimen contained marrow elements and multiple noncaseating granulomas without evidence of HD (not shown). Four years later, he remained in clinical remission with stable anterior and subcarinal mediastinal adenopathy by chest computed tomographic (CT) scanning.

[FIGURE 1-4 OMITTED]

Both sarcoid and HD are documented histologically. Noncaseating granulomas are occasionally found in HD and other malignancies as a nonspecific reaction (see Discussion section). In this case, however, bilateral hilar adenopathy preceded the diagnosis of HD by 3 years. After treatment and remission of HD, mediastinal and hilar adenopathy have remained stable for over 4 years. This radiographic history along with the presence of multiple noncaseating granulomas at two distinct tissue sites is most consistent with sarcoidosis.

Patient 2

A 38-year-old black man presented to the pulmonary clinic with a 9-year history of exertional dyspnea, which began in 1991 shortly after he returned from assignment in Saudi Arabia. For the same period, he had noted a sporadic cough that on occasion was productive of colored phlegm. In the past 4 months, he has noted progressive abdominal distention without associated pain, changes in bowel habit, or jaundice. Physical examination disclosed the presence of ascites without organomegaly. CXR and chest CT scan were remarkable for hilar and paratracheal adenopathy along with perihilar consolidations (not shown). The patient underwent bronchoscopy and paracentesis. Bronchoalveolar lavage disclosed many macrophages and rare multinucleated giant cells but no cytologic evidence of malignancy. Transbronchial biopsies showed multiple ill-defined noncaseating granulomas consistent with pulmonary sarcoidosis (Fig. 5). Special stains and cultures were negative for acid-fast and fungal organisms. The peritoneal fluid contained malignant cells consistent with an adenocarcinoma (Fig. 6). One week later, an endoscopy biopsy specimen from the gastric fundus was positive for moderately differentiated adenocarcinoma with focal signet ring features (not shown). Due to the advanced nature of his malignancy, the patient was treated with supportive nutritional care (percutaneous endoscopic gastrostomy tube) and died 4 months later in a hospice center. Permission for autopsy was not given.

[FIGURE 5-6 OMITTED]

Again, both sarcoid and malignancy are documented histologically. The 9-year history of cough and dyspnea along with the CT findings are consistent with Stage 2 pulmonary sarcoidosis. Bronchoalveolar lavage and transbronchial biopsy both supported a diagnosis of sarcoid and showed no evidence of carcinoma. The coexistence of sarcoid and gastric carcinoma is rare. Some authors have speculated that choronic sarcoid of the gastric mucosal lining may predispose to gastric malignancy. (2)

Patient 3

A 56-year-old man originally presented in July 2000 with right-sided "pulling" flank pain. He underwent abdominal and pelvic CT scanning that disclosed a large right retroperitoneal mass with adjacent spread to the right renal hilum. A chest film disclosed a right-sided pleural effusion and prominent hila bilaterally. An excisional lymph node biopsy from the left inguinal region showed non-Hodgkin's lymphoma (NHL) of the large cell subtype (Fig. 7A). After a negative bone marrow biopsy and aspiration, he underwent eight cycles of CHOP chemotherapy (Cytoxan [1,650 mg], vincristine [2 mg] Adriamycin [110 mg], and prednisone [100 mg]) from August until December. Afterward, a surveillance positron emission tomographic scan disclosed uptake in the mediastinum. A chest CT scan obtained in February 2000 disclosed right hilar and subcarinal adenopathy (Fig. 8). A bronchoscopically directed transbronchial needle (Wang) aspiration revealed a mixed population of lymphoid cells without clear evidence of malignancy. Central mediastinal exploration with lymph node excision disclosed scattered noncaseating granulomas without evidence of malignancy (Fig. 7B). Special stains were negative for acid-fast and fungal organisms. At his most recent follow-up appointment, he remained in remission nearly 1 year after his last treatment.

[FIGURE 7-8 OMITTED]

In clinical practice, radiographic abnormalities in cancer patients are often presumed to represent metastasis. However, in this case, lymph node excision clearly showed noncaseating granulomas consistent with mediastinal sarcoid during a time of clinical remission from NHL. Retrospectively, the prominent hila noted on CXR at presentation may have represented coexistent sarcoidosis.

Discussion

The proposed link between sarcoidosis and malignancy has remained controversial ever since Brincker and Wilbek's (1) original investigation nearly three decades ago. Evidence favoring an association between sarcoidosis and malignancy is based on case reporting and statistical analysis from cross-registry studies.

After Brincker and Wilbek's (1) study in 1973, many investigators began to report a wide variety of lymphoreticular neoplasms among sarcoid patients. Finke et al (3) reported a case of malignant lymphoplasmacytoid lymphoma associated with a monoclonal band on electrophoresis in a patient with prior histologically confirmed sarcoidosis. Karakantza et al (4) reported five cases of lymphoreticular neoplasms in patients with a biopsy-proved sarcoid. On the basis of cumulative experience, the term sarcoid-lymphoma syndrome emerged. The typical features of this syndrome are the development of lymphoreticular neoplasms, especially Hodgkin's lymphoma, in a patient with a preexisting diagnosis of sarcoidosis. Typically, the patient is older than average for the sarcoid population (median, 45 years of age) and has had chronically active pulmonary sarcoidosis for a period of several years. Patient 1 in this article represents an example of the sarcoid-lymphoma syndrome.

In Brincker and Wilbek's (1) original investigation, an excess of lung malignancies was also noted. Since then, other investigators have noted the development of solid tumors in patients with biopsy-proved sarcoidosis. Most of these reports have involved cases of breast or lung cancer, but other solid tumors including testicular, ovarian, uterine, and gastric cancer have been noted as well. (5) In 1990, Suen et al (6) reported six cases of patients with biopsy-proved malignancy that preceded the diagnosis of sarcoidosis and used the term malignancy-sarcoidosis syndrome to describe their clinical experience. Four of their six cases involved lymphoproliferative diseases (HD, NHL, and T-cell lymphoma), and all had biopsy-proved sarcoidosis that was diagnosed an average of 9 months after the development of malignancy.

To avoid the observer bias inherent with case reporting, epidemiologic surveys have been used to examine the relationship between sarcoid and malignancy. Yamaguchi et al (7) examined the incidence of malignancy in a nationwide prevalence survey of sarcoid patients in Japan. Excess deaths were analyzed by calculating standardized mortality ratios. Excess lung cancer and uterine cancer deaths were noted (standardized mortality ratios of 3.26 and 8.7, respectively). Askling et al (8) studied two cohorts of sarcoid patients, one consisting of 474 consecutively diagnosed sarcoid patients from Uppsala County in Sweden and a second, larger group of 8,541 patients identified from a national inpatient sarcoid registry. Over a 30-year period, both groups were cross-referenced to nationwide cancer and death registries. Standardized incidence ratios were determined for a variety of malignancies and were increased in both cohorts for lung cancer, skin cancer, and melanoma. The risk of lymphomas was increased in the inpatient cohort as well.

As with case reporting, there are problems inherent with epidemiologic studies, which can confound any link between sarcoid and malignancy. Misclassification is perhaps the most significant. Misclassification occurs when either sarcoid or malignancy is misdiagnosed. For example, a mass-like lesion on CXR in a patient with sarcoidosis may be attributed to malignancy when, in fact, it is due to sarcoid. Likewise, mediastinal adenopathy in a patient with known malignancy is often assumed to be cancer when, in fact, it could represent thoracic sarcoid. To make matters more confusing, the histologic hallmark of sarcoid, noncaseating granulomas, can be seen in lymph nodes and other tissues as a nonspecific reaction to malignancy and has been termed the sarcoid reaction. (9) The importance of misclassification was stressed by Seersholm et al, (10) who noted misclassification of 3 of 36 malignancies in an epidemiologic survey of 254 sarcoid patients. After taking these into account, they found no link between sarcoid and lymphoma, although a link between sarcoid and other malignancies could not be disproved. In an ef]brt to circumvent some of these problems, Reich et al (11) used "clinical linkage criteria" to assess the true relationship between malignancy and sarcoidosis. Their linkage criteria included close temporal relationship between sarcoid and malignancy and the presence of local sarcoid reactions that generalize over time. Reich et al proposed that sarcoidosis might represent an immunologic reaction to dispersal of tumor antigen. The complex linkage criteria proposed by Reich et al have thus far not been uniformly accepted in establishing linkage between sarcoid and malignancy.

To date, the link between sarcoid and malignancy is unproved, and some authors continue to doubt whether any association truly exists. However, clinicians should be aware of the potential risk of malignancy in sarcoid patients. Until more definitive data are available, sarcoid patients who develop new lesions or adenopathy should be considered for histologic biopsy to rule out coexistent neoplasm. Furthermore, in patients with established malignancy and newly discovered radiographic findings, the possibility of sarcoidosis should be considered.

Key Points

* Physicians should be aware of the possible link between sarcoid and malignancy.

* Sarcoid patients with new or enlarging lesions should be considered for biopsy to rule out coexisting malignancy.

* Sarcoid lesions of the lung and mediastinum can coexist in patients with previously documented malignancy.

Acknowledgment

We extend special thanks to Rami Choi for detailed assistance with manuscript preparation.

References

(1.) Brincker H, Wilbek E. The incidence of malignant tumours in patients with respiratory sarcoidosis. Br J Cancer 1974; 29:247-251.

(2.) Roth D, West B, Madison J, et al. Gastric carcinoma in a patient with sarcoidosis of the gastrointestinal tract. Am J Gastroenterol 1994;89: 1589-1591.

(3.) Finke R, Lydtin H, Prechtel K. Sarcoidosis and immunocytoma. Am J Med 1986; 80:939-942.

(4.) Karakantza M, Matures E. MacLennan K, et al. Association between sarcoidosis and lymphoma revisited. J Clin Pathol 1996; 49:208-212.

(5.) Brincker H. Solid tumors preceding or following sarcoidosis. Med Pediatr Oncol 1987; 15:82-88.

(6.) Suen JS, Forse MS, Hyland RH, et al. The malignancy-sarcoidosis syndrome. Chest 1990; 98:1300-1302.

(7.) Yamaguchi M, Odaka M, Hosoda Y, et al. Excess death of lung cancer among sarcoidosis patients. Sarcoidosis 1991; 8:51-55.

(8.) Askling J, Grunewald J, Eklund A, et al. Increased risk for cancer following sarcoidosis. Am J Respir Crit Care Med 1999; 160:1668-1672.

(9.) Brincker H. Sarcoid reactions in malignant tumours. Cancer Treat Rev 1986; 13:147-156.

(10.) Seersholm N, Vestbo J, Viskum K. Risk of malignant neoplasms in patients with pulmonary sarcoidosis. Thorax 1997; 52:892-894.

(11.) Reich JM, Mullooly JP, Johnson RE. Linkage analysis of malignancy-associated sarcoidosis. Chest 1995; 107:605-613.

LTC (ret) William E. Caras, MD, FCCP, COL (ret) Thomas Dillard, MD, FCCP, LTC Thomas Baker, MD, and COL (ret) Jerry Pluss, MD, FCCP

From the Department of Medicine, Pulmonary/Critical Care Service, and the Department of Pathology, Madigan Army Medical Center, Tacoma, WA.

The views expressed herein are those of the authors and do not reflect the official views of the U.S. Army or the U.S. Department of Defense. No financial support has received by any of the authors in support of this publication. The authors do not have a proprietary interest in any drug, device, or equipment mentioned in this article.

Reprint requests to LTC (ret) William E. Caras. MD, 2618 Thornhill Road, Puyallup, WA 98374.

Accepted February 17, 2003.

Copyright [c] 2003 by The Southern Medical Association 0038-4348/03/9609-0918