Secondary Logo

Journal Logo

Should Conjunctival Biopsy Be Performed for Definitive Diagnosis of Sarcoidosis?

Jirawuthiworavong, Guy, MD, MA; Rubinstein, Daniel, MD; Turbin, Roger, E., MD

Section Editor(s): Lee, Andrew G. MD; Van Stavern, Gregory MD

Journal of Neuro-Ophthalmology: June 2018 - Volume 38 - Issue 2 - p 217–222
doi: 10.1097/WNO.0000000000000600
Point Counter-Point

Southern California Kaiser Permanente Medical Group (GJ), La Palma, California; and Department of Ophthalmology (DR, RET), Institute of Ophthalmology and Visual Science, Rutgers New Jersey Medical School, Newark, New Jersey.

Address correspondence to Roger E. Turbin, MD, Institute of Ophthalmology and Visual Science, Rutgers New Jersey Medical School, 90 Bergen Street, DOC 6100, Newark, NJ 07101-1763; E-mail:

The authors report no conflicts of interest.

Sarcoidosis is a systemic, autoimmune disease with a variety of presentations and manifestations, including pulmonary, neurologic, hepatic, and joint disease. Ocular involvement has been reported in up to 80% of cases (1) and is noted at the time of presentation in approximately one-fifth of patients (2). Neurosarcoid occurs when sarcoid involves the central nervous system and may affect the meninges cranial nerves, brain parenchyma, and spinal cord. The Zajicek criteria have been proposed to stratify the diagnosis of neurosarcoid into “definite, probable, and possible” (3). Diagnostic criteria of sarcoid and neurosarcoid include histologic identification of noncaseating granuloma, supportive laboratory and imaging studies, and compatible clinical course. Despite advancing imaging techniques and greater understanding of immunologic abnormalities detected in patients with sarcoid, the diagnosis can be elusive and may remain uncertain if not histologically confirmed by biopsy. Trends favoring biopsy vary from specialty to specialty; in particular, utilization of conjunctival biopsy varies widely among sarcoidosis experts. We invited 2 of our colleagues to weigh in on the pros and cons of conjunctival biopsy in the diagnosis and management of sarcoidosis as it pertains to neuro-ophthalmology.

Back to Top | Article Outline

Pro: Patients With Probable Sarcoidosis of the Eye or the Central Nervous System Should Be Considered for Conjunctival Biopsy: Guy Jirawuthiworavong, MD, MA

Sarcoidosis is a granulomatous disease that affects multiple organ systems. Neurosarcoidosis occurs in approximately 5%–15% of patients. A multitude of neuro-ophthalmic disorders can occur as a result of neurosarcoidosis (4). In addition, up to 3% of neurosarcoid patients may present without symptoms. Other conditions such as tuberculosis, syphilis, fungal infections, and demyelinating disease can mimic the manifestations of neurosarcoidosis.

The prognosis of neurosarcoidosis can be devastating. Permanent morbidity and even death can occur without treatment with steroids or steroid-sparing agents. Complicating matters further, the side effects of treatment are not trivial. Oral and intravenous steroids can cause a multitude of unwanted systemic effects that are well known. These steroid side effects can be prevented in some patients using steroid-sparing agents. However, steroid-sparing agents, including anti-metabolites and biologics, have their own risks for morbidity. These immune modulators and immunosuppresants can increase susceptibility to recurrent viral and bacterial infections, and can increase the risk for patients to acquire lymphoma and other cancers. In my own practice over the last 10 years, I can think of at least 2 patients who died of complications of steroids or steroid-sparing agents. One had been on infliximab and acquired a fatal pulmonary coccidiomycosis infection, and the other was on long-term prednisone, which ultimately caused her demise because of weight gain, inactivity, and eventually pneumonia.

The importance of histological confirmation of the diagnosis of sarcoidosis cannot be taken lightly. There are multiple diagnostic criteria based on the organ system affected, including the central nervous system (Table 1). These diagnostic criteria group patients according to their clinical presentation of signs, symptoms, laboratory, and radiological findings (3–5). As a result, sarcoidosis can be considered as possible, probable, or definite. Definite sarcoidosis can only be diagnosed if there is histological confirmation. Typically, the biopsy of the particular organ of interest, will show the classic, noncaseating granuloma that is associated with sarcoidosis.



Of the different biopsy options to confirm sarcoidosis (e.g., transbronchial, mediastinoscopy, meningeal, skin), conjunctival biopsy is one of the easiest to perform. This minimally invasive, quick procedure can be completed in the office and requires only topical anesthetic. The patient can easily appreciate advantages of a conjunctival biopsy, including low risk, low cost, and its outpatient nature. The inferior conjunctival cul-de sac is commonly used for tissue sampling. The biopsy of follicles, nodules, or fibrous scarring can be performed (6–8) or it can be performed “blindly” (9) (Fig. 1). A sample of tissue can be as small as 5 × 3 mm, and specimens can be biopsied bilaterally to increase yield. The overall yield for a confirmative conjunctival biopsy ranges from 24% to 75% (10–14).

FIG. 1

FIG. 1

In a study by Leavitt and Campbell (15), the cost-effectiveness was analyzed comparing conjunctival biopsy with other more commonly used biopsy methods for the diagnosis of sarcoidosis. Transbronchial biopsy was twice as expensive as conjunctival biopsy, and mediastinoscopy with biopsy cost almost 7 times more.

Considering the gravity of the morbidity and mortality that is associated with sarcoidosis as well as the diagnostic dilemma it poses, an attempt for a definitive diagnosis with biopsy needs to be initiated by the clinician. I strongly recommend the procedure as it is:

  1. Simple and minimally invasive
  2. Cost effective
  3. Confirmatory in a moderate number of patients
  4. Helpful in avoiding unnecessary diagnostic testing
  5. Evidence to justify risks of steroids and steroid-sparing agents
Back to Top | Article Outline

Con: Alternative Methods of Diagnosis Should Be Considered for Patients With Probable Sarcoidosis of the Eye or Central Nervous System: Roger E. Turbin, MD

Here we go again… Revisiting an old idea that dates back to the 1950s with no definitive answer. Should we perform a biopsy? Where should we perform a biopsy? Should we perform a biopsy in clinically nonlesional tissue? To these questions I'll answer, “yes, we should perform a biopsy in most patients because otherwise the risk of a misdiagnosis and inappropriate treatment remains high in many patients.” To the second question of where to perform a biopsy, I'll reference Willie Sutton's response to the question as to why he robbed banks, answering “because that is where the money is.” To the last question about nonlesional tissue biopsy, I'll answer tongue in cheek, “sometimes, if we ain't got nothing better.”

In 1995, Crick, Hoyle and Mather (10) described their technique for performing office-based conjunctival biopsy as a “simple method… of diagnostic value in sarcoidosis.” They performed biopsy of a “follicle containing” 2-mm sections of fornix or bulbar conjunctiva anesthetized with 4% cocaine solution, with a 50-section histologic serial analysis of the specimen. They found lesions consistent with conjunctival sarcoid in 10 of their 28 biopsies (36%), in 3 of 10 patients (30%) defined as generalized sarcoid without ocular symptoms (30%), in 4 of 6 patients (67%) with generalized sarcoid with ocular manifestations of sarcoid (67%), and 7 of 16 patients (44%) with generalized sarcoid with or without ocular signs. The authors covered 4 points of discussion, including biopsy specimen size, presence of clinical abnormality such as follicles, histologic analytic technique, and diagnostic value/safety compared with other biopsy sites. Each of these ideas remains germane in today's clinical practice.

Since then, experts have opined both for and against conjunctival biopsy, suggested a variety of clinical risk factors in patient populations that are likely to lead to the highest conjunctival biopsy yields, and described surgical and histologic methods to increase the sensitivity of the conjunctival biopsy. For instance, Spaide and Ward (14) described highest yields from conjunctival biopsies in patients with “conjunctival follicles, ocular abnormalities consistent with sarcoid, and X-ray diagnosed pulmonary infiltrates.” Biopsies typically 10–12 mm in length were positive in 67% of the patients with conjunctival follicles and 31% without conjunctival follicles, supporting a yield of nearly one-third of patients undergoing “blind” nonlesional biopsy. Bui and colleagues (16) emphasized the importance of performing a “directed” biopsy and showed slight increase in diagnostic yield from 43% to 63% if the surgeon requested multiplane histologic sectioning. Chung et al (6) advocated bilateral blind biopsies with multiple samples taken from each eye to increase the diagnostic yield. Leavitt and Campbell (15) found 71% of the Mayo Clinic's (see additional discussion below) conjunctival bilateral biopsies to be positive on only 1 side, suggesting the importance of bilateral biopsy to increase the diagnostic yield. In their series, the yield was highest in patients with more advanced clinical disease.

Of interest are 2 articles from the Mayo Clinic that reported contrasting results. In 1998, Leavitt and Campbell (15) described their experience with 41 consecutive patients over a 3-year period who had proven sarcoid. They cited a 51% yield from “blind” conjunctival biopsies and recommended performing a biopsy strip of 8–10 mm of inferior fornix conjunctiva. The authors believed the yield to be similar to mediastinal biopsy at one-tenth of the cost. However, in 2015, Pichler and colleagues (17) published a follow-up study, citing 440 consecutive patients over a 6-year period who underwent blind conjunctival biopsy. These patients were suspected of having neurosarcoid, each having evidence of inflammatory neurologic disease, which was not a selection criterion of the earlier study. By contrast, the latter report produced a yield of 3% having positive conjunctival biopsies. Ultimately the diagnosis of neurosarcoidosis was established in 60 patients (14%) by the Zacijek criteria (definite, 15: probable 22: possible 23).

I believe that we can agree that the optimal source of tissue for analysis remains a subject of debate. The debate highlights one of the fundamental weaknesses of a “blind” diagnostic method: random sampling without clinical localization has the propensity for missing patchy or skip-type lesions through sampling error. Surprise, surprise… it seems that selecting areas that appear clinically abnormal, with respect to conjunctiva or other tissue, improves diagnostic yield. It is interesting, however, that biopsy of nonspecific but lesional areas of conjunctiva including scar tissue or follicles increases the yield and is, therefore, preferable to a true-blind biopsy of nonlesional conjunctiva.

Other anatomic sites, in addition to the traditional mediastinal biopsy, may be targeted to make a tissue diagnosis of sarcoidosis. The lacrimal gland is a frequent consideration, especially in the context of an abnormal gallium, or positron emission tomography (PET) scan. A report from our institution demonstrated that both subjective visual scoring (SVS) as well as region of interest quantitative analysis of gallium scan findings aid in determining a biopsy strategy for a patient suspected of sarcoidosis (18). Subjective visual scoring (Fig. 2) rates the uptake of the gallium tracer in the lacrimal gland as mild, moderate, or intense. This correlated with a positive to lacrimal gland biopsy in 0%, 44%, and 50% of patients, respectively. A simple mathematical score comparing the tracer uptake in the lacrimal gland to the background signal, termed the lacrimal uptake ratio (LUR) (Fig. 3) was calculated and produced a nearly linear correlation to lacrimal gland biopsy positivity. When this signal-to-noise ratio was greater than 6, approximately 75% of biopsy specimens were positive and the rate approached 100% at an LUR >8.

FIG. 2

FIG. 2

FIG. 3

FIG. 3

So I quote Mr. Sutton in my resident clinic frequently, and direct biopsy to the area of highest systemic inflammation, because “that's where the money is.” This corresponds to areas of highest uptake on gallium scan as assessed by SVS or the LUR ratio. We have generalized this principle to uptake intensity on PET scans. We biopsy the areas of highest glandular, parenchymal, or nodal uptake, taking into account relative risk of each particular biopsy site and patient preference. If the parotid gland has particularly high uptake, we may substitute a minor salivary gland biopsy of the lower lip (Fig. 4) as an initial “safe” biopsy to avoid the risk of iatrogenic damage to the facial nerve as branches traverse the parotid gland. Minor salivary biopsy is performed as an office-based procedure under local anesthesia, or can be added to increase tissue sampling when performing other procedures under anesthesia.

FIG. 4

FIG. 4

Given its diverse manifestations, the diagnosis of sarcoidosis, and especially neurosarcoidosis, presents a challenge to the clinician. Rather than using a 1-size-fits-all approach, such as blind biopsy of commonly involved tissues, we have, at our disposal, tools capable of tailoring the diagnostic workup to the individual patient. Chief among these tools is an advanced interpretation of imaging, including gallium or PET scans. Our patients only stand to benefit from this approach by way of fewer invasive procedures, a shortened time to diagnosis and treatment, and fewer health care dollars spent on low-yield procedures.

Back to Top | Article Outline

Rebuttal: Dr. Jirawuthiworavong

I agree with Dr. Turbin that most patients with sarcoidosis of the eye or central nervous system will require a biopsy. As for the location and the role of blind biopsies, the conjunctival blind biopsy still should not be overlooked. It has been cited in multiple studies and has extremely low morbidity.

Dr. Turbin's research has shown that the use of the LUR can increase lacrimal gland biopsy positivity for diagnosing sarcoidosis. However, to calculate the LUR, an expensive gallium scan is required. I am not necessarily opposed to a gallium scan for certain patients with systemic sarcoidosis, but a gallium scan should not be an initial diagnostic study as it is very costly.

Most ophthalmologists are not comfortable performing a lacrimal gland biopsy and would have to refer the patient to a subspecialist for the procedure. By contrast, the conjunctival biopsy is a simple, quick procedure that is easy to perform. One can increase the yield of blind conjunctival biopsies by performing them bilaterally and in patients with elevated serum angiotensin-converting enzyme (ACE) and/or a chest imaging study showing hilar adenopathy (17). I would advocate that if initially the conjunctival biopsy is negative, Dr. Turbin's LUR protocol would be the next step in the work up.

As gallium scan LUR is novel, it needs to be verified with additional studies. Its availability likely is limited to only a few academic centers or health maintenance organizations. Furthermore, many clinicians may have abandoned gallium scans for PET scans because PET has shown less acquisition time, higher interobserver agreement, and lower radiation exposure (19). Research also needs to be conducted on the usefulness of LUR quantification using PET.

There is no question that nonlesional, blind biopsies of the conjunctiva can diagnose sarcoidosis. This has been proven by multiple studies. The report by Pichler et al (17) from the Mayo Clinic documented a 3% positivity rate for the diagnosis of neurosarcoid, which is much lower than that of the previous studies. However, the yield may have been much higher if one takes in account that 125 of the 400 biopsied patients were on immunosuppresants within 1 week of the biopsy. Immunosuppression may have caused the biopsy result to be negative. It must also be noted that 23 patients did not have an updated medication list, so we do not know if they also were taking immunosuppressive medications at the time of biopsy. Nevertheless, the positive predictive value was 69.2%, which is a reasonable percentage for the uncommon condition of neurosarcoidosis.

Here is my diagnostic approach for neurosarcoid:

  1. Exclude patients without signs and symptoms of neurosarcoidosis
  2. Serum angiotensin-converting enzyme level
  3. Chest CT
  4. MRI of the brain and spine


  1. Bilateral conjunctival biopsy
  2. If possible, no immunosuppressants 2 weeks prior
  3. Special stains to exclude infectious etiologies

If NEGATIVE conjunctival biopsy


  1. Gallium or PET scan (LUR quantitation)
  2. Directed biopsy of tissue of interest (e.g., Lacrimal gland, parotid gland, hilar lymph nodes)
Back to Top | Article Outline

Rebuttal: Dr. Turbin

Dr. Jirawuthiworavong's assertion that a secure diagnosis of sarcoidosis before the initiation of high-risk therapy is well taken and also is consistent with our clinical practice. There are cases, however, that we treat as “presumed sarcoid or neurosarcoid” despite negative biopsy. We agree that conjunctival biopsy is a relatively straightforward, safe, and convenient procedure in the appropriate context. It is particularly appealing in the settings of targeting obvious conjunctival nodules or other abnormalities, patients considered to be very high anesthesia risks, or patients seeking immediate biopsy before imaging. However, our opinions differ on the cost-effectiveness of conjunctival biopsy and it's utility as an initial diagnostic procedure. I would also argue that “blind” minor salivary biopsy has similar benefits to conjunctival biopsy, is easy and safe to perform in an office setting, and also has the potential to diagnoses other diseases such as Sjogren syndrome or, occasionally, lymphoma. The strongest evidence against the utility of the “blind” conjunctival biopsy is the publication from the Mayo Clinic by Pichler et al (17). In addition to the low yield of positive biopsies in patients suspected of neurosarcoid, approximately 29% of the conjunctive biopsies consistent with sarcoid were likely false-positive biopsies because these patients were found to have other diagnoses to explain their neurologic disorders. I would conclude by saying that a blind conjunctival biopsy may ultimately provide a diagnosis, but the high rate of false positives and negatives limits its utility, especially in patients suspected of having neurosarcoid.

Back to Top | Article Outline


Neurosarcoidosis is frequently in the differential diagnosis for patients presenting to a neuro-ophthalmologist. The authors provide strong arguments for pursuing a tissue diagnosis to provide better guidance for management and prognosis. The most recent evidence suggests that “blind” conjunctival biopsy is a low yield procedure, but the low yield may be offset by its low cost and low morbidity. A careful history and physical examination may provide a more targeted approach, with biopsy of a lesional site, which would likely increase yield. Conjunctival biopsy may still have a role in the diagnostic work up of select patients suspected of having neurosarcoid.

Back to Top | Article Outline


1. Bonfioli AA, Orefice F. Sarcoidosis. Sem Ophthalmol. 2005;20:177–182.
2. Heiligenhaus A, Wefelmeyer D, Wefelmeyer E, Rosel M, Schrenk M. The eye as a common site for the early clinical manifestation of sarcoidosis. Ophthalmic Res. 2011;46:9–12.
3. Zajicek JP, Scolding NJ, Foster O, Rovaris J, Evanson J, Moseley IF, Scadding JW, Thompson EJ, Charmour V, Prither DH, McDonald WI, Mitchell D. Central nervous system sarcoidosis–diagnosis and management. QJM. 1999;92:103–117.
4. Mijajlovic M, Mirkovic M, Mihailovic-Vucinic V, Aleksic V, Covickovic-Sternic N. Neurosarcoidosis: two case reports with multiple cranial nerve involvement and review of the literature. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2014;158:662–667.
5. Herbort CP, Rao NA, Mochizuki M. International criteria for the diagnosis of ocular sarcoidosis: results of the first International Workshop on Ocular Sarcoidosis (IWOS). Ocul Immunol Inflamm. 2009;17:160–169.
6. Chung YM, Lin YC, Huang DF, Hwang DK, Ho DM. Conjunctival biopsy in sarcoidosis. J Chin Med Assoc. 2006;69:472–477.
7. Birnbaum AD, Ojh FS, Chakrabarti A, Tessler HH, Goldstein DA. Clinical features and diagnostic evaluation of biopsy-proven ocular sarcoidosis. Arch Ophthalmol. 2011;129:409–413.
8. Bornstein JS, Frank MI, Radner DB. Conjunctival biopsy in the diagnosis of sarcoidosis. N Engl J Med. 1962;267:60–64.
9. Nichols CW, Eagle RC Jr, Yanoff M, Menocal NG. Conjunctival biopsy as an aid in the evaluation of the patient with suspected sarcoidosis. Ophthalmology. 1980;87:287–291.
10. Crick R, Hoyle C, Mather G. Conjunctival biopsy in sarcoidosis. Br Med J. 1955;2:1180–1181.
11. Niitu Y, Horikawa M, Suetake T, Hasegawa S, Komatsu S. Comparison of clinical and laboratory findings of intrathoracic sarcoidosis between children and adults. Ann NY Acad Sci. 1976;278:532–538.
12. Merritt JC, Lipper SL, Peiffer RL, Hale LM. Conjunctival biopsy in sarcoidosis. J Natl Med Assoc. 1980;72:347–349.
13. Solomon DA, Horn BR, Byrd RB, Lorfel RS, Griggs GA. The diagnosis of sarcoidosis by conjunctival biopsy. Chest. 1978;74:271–273.
14. Spaide RF, Ward DL. Conjunctival biopsy in the diagnosis of sarcoidosis. Br J Ophthalmol. 1990;74:469–471.
15. Leavitt JA, Campbell RJ. Cost-effectiveness in the diagnosis of sarcoidosis: the conjunctival biopsy. Eye. 1998;12:959–962.
16. Bui KM, Garcia-Gonzalez JM, Patel SS, Lin AY, Edward DP, Goldstein DA. Directed conjunctival biopsy and impact of histologic sectioning methodology on the diagnosis of ocular sarcoidosis. J Ophthalmic Inflamm Infec. 2014;4:8.
17. Pichler M, Flanagam E, Aksamit A, Leavitt J, Salomao D, Keegan M. Conjunctival biopsy to diagnosis neurosarcoidosis in patients with inflammatory nervous system disease unknown etiology. Neurol Clin Pract. 2015;5:216–223.
18. Tannen BL, Kolomeyer AM, Turbin RE, Frohman L, Langer PD, Oh C, Ghesani NV, Zuckier LS, Chu DS. Lacrimal gland uptake of (67) Ga-gallium citrate correlates with biopsy results in patients with suspected sarcoidosis. Ocul Immunol Inflamm. 2014;22:15–22.
19. Mostard RL, Van Kroonenbergh MJ, Drent M. The role of the PET scan in the management of sarcoidosis. Curr Opin Pulm Med. 2013;19:538–534.
© 2018 by North American Neuro-Ophthalmology Society