Reply to P. B. et al., to Fahim and Rosewarne, and to Reich

From the Authors:

The American Thoracic Society (ATS) Sarcoidosis Diagnosis and Detection Guideline Committee relishes the engagement of the sarcoidosis community and would like to sincerely thank correspondents of these three letters for their care of patients with sarcoidosis, contributions to sarcoidosis research, and raising these excellent discussion points.

Notably, ATS guideline development follows strict Grading of Recommendations, Assessment, Development and Evaluations (GRADE) methodology to ensure that the highest level of evidence is reviewed and synthesized. As a corollary to this rigorous method, a limited number of key questions with highest clinical impact to the field of sarcoidosis and for which sufficient supportive data were available were selected for a comprehensive search and review of the literature. Therefore, a number of compelling questions, such as the utility of cervical lymph node biopsy, were not selected for review. In this regard, we are indebted to the ATS methodologists and ATS librarian who searched and reviewed thousands of publications for this guideline.

We certainly appreciate the suggestion to consider cervical lymph node biopsy in patients suspected of having sarcoidosis as proposed by Fahim and Rosewarne. Lower procedure risk and cost to the patient as well as potentially lower risk of infectious exposure to health care providers are important aspects of this approach. Addressing cervical lymph node biopsy for the diagnosis of sarcoidosis is an excellent suggestion for future guidelines. In the interim, we encourage the accumulation of data from additional studies to provide a more robust evidence base to evaluate the procedure in the context of a guideline.

The committee further notes that the ATS standard for guideline rigor includes tightly framing each research question using patient/intervention/comparator/outcome (PICO) methodology. Thus, the recommendation for endobronchial ultrasound–guided transbronchial needle aspiration over mediastinoscopy did not include review of all possible alternative methods of endoscopic ultrasound technique. We appreciate and agree with the benefits raised by P. B. and colleagues of potentially improved procedural tolerance of esophageal endoscopic approaches such as endoscopic ultrasound with fine needle aspiration (EUS-FNA) compared with bronchoscopic approaches. We note that three of the studies selected for the question about mediastinal/hilar lymph node sampling versus no sampling featured EUS-FNA studies (1–3). Compared with endobronchial ultrasound, the number of thoracic lymph nodes that can be sampled via an esophageal approach is more limited. In addition, the majority of interventional pulmonologists do not routinely perform EUS-FNA or endoscopic ultrasound using echobronchoscope. For these reasons, we did not feel it reasonable to include these modalities in our recommendation on endoscopic approaches to biopsy in comparison to traditional mediastinoscopy. Advances in the evidence base for the safety and efficacy of these varied approaches could be addressed in future guidelines.

We also appreciate Dr. Reich’s contributions to the field of sarcoidosis, including the nuanced risk and cost analysis in his letter to the editor favoring no biopsy of asymptomatic bilateral hilar lymphadenopathy (ABHL) based on his original meta-analysis of the sarcoidosis literature of the 20th century, in which higher risk mediastinoscopy of mediastinal lymph nodes present at the time of ABHL was the biopsy technique (4). The challenge for clinicians and patients remains that not a single clinical study has been published in the contemporary literature on the narrowly defined topic of ABHL in the absence of mediastinal lymphadenopathy. Thus, the committee considered 16 studies that enrolled a total of 425 patients with asymptomatic radiographic stage I sarcoidosis, comprising patients who have ABHL and who may also have asymptomatic mediastinal lymph node enlargement. These studies capture center-to-center variation in practice and patient populations as well as a range of techniques to sample lymph nodes. They adequately document the existence of alternative diagnoses that mimic stage I sarcoidosis. The rate of alternative diagnoses (8 of 425, or 1.9%) is low, yet it is not trivial, and the alternative diagnoses of tuberculosis (TB) and lymphoma are of particular import to patients, who were represented by a patient advocate on the clinical practice guideline committee. In addition, the presumptive diagnosis of sarcoidosis engenders lifelong screening and clinical follow-up for emergence of multisystemic disease, prompting significant concern for some patients and thus reduction in long-term quality of life and increased cost to patients. Thus, the clinical practice guidelines committee did not make a recommendation for or against biopsy in the setting of ABHL, and our recommendations conform with those of Dr. Reich’s based on the premise that close clinical follow-up is the key to safely recommending a nonbiopsy approach. However, we also recognize that some patients have limited access to health care because of socioeconomic healthcare disparities and the decentralized nature of our American health system, such that close follow-up may not be realistic in all cases.

In sum, we have found an overall paucity of high-quality evidence for or against biopsy of stage I sarcoidosis, with an overall detection rate of 85% for sarcoidosis, 11% nondiagnostic, and 2% alternative diagnosis including TB and lymphoma, using the full range of contemporary biopsy techniques (5). The present guideline is rooted in a decision by our multidisciplinary committee to recommend that an individualized patient/healthcare provider choice regarding biopsy can be made. Key factors guiding this personalized decision include the risk profile of the preferred/available biopsy technique; patient preference; the community prevalence of alternative infectious disease such as TB, histoplasmosis, or coccidioidomycosis; the patient’s cancer and occupational/exposure history; and the ability to closely follow for development of symptoms if biopsy is not performed. We acknowledge in research statements throughout this guideline that more high-quality data is sorely needed to guide these important decisions. We encourage members of the sarcoidosis research community to address these questions in future publications to inform future sarcoidosis clinical practice guidelines.