Safety and usefulness of minor salivary gland biopsy: Retrospective analysis of 502 procedures performed at a single center




To analyze the safety of our biopsy technique and the effectiveness of minor salivary gland biopsy (MSGB) for the diagnosis of Sjögren's syndrome (SS) and amyloidosis.


We conducted a retrospective analysis of 452 patients with suspected SS and 50 with suspected amyloidosis and negative periumbilical fat aspiration analysis who underwent MSGB at a single center. Diagnostic evaluation for SS included Schirmer's test, unstimulated whole salivary flow, detection of antinuclear antibodies and anti-SSA/SSB, erythrocyte sedimentation rate, C-reactive protein, IgM rheumatoid factor, and serology for hepatitis C virus. For all biopsy samples, a cumulative focus score on multilevel sections was calculated. SS was diagnosed according to American-European Consensus Group (AECG) criteria. Histologic evaluation for amyloidosis was performed using Congo red staining and polarized-light microscopy. Adverse events were recorded on a questionnaire immediately after the procedure and 7 days, 14 days, and 6 months thereafter.


Sixty-four patients (12.7%) reported transient adverse events: 40 paresthesias lasting <7 days, 17 paresthesias lasting <14 days, 27 cases of local swelling, and 8 external hematoma. One patient has had local paresthesia for 2 years. A total of 498 (99.2%) samples provided adequate material for histologic analysis. Of 452 patients evaluated for SS, 378 were finally evaluated. Ninety-three patients (24.5%) had a cumulative focus score ≥1, and 87 (94.5%) of 93 satisfied the AECG criteria. Classification of SS was possible for 124 (32.8%) of 378 patients. In 51 (41%) of 124, MSGB was essential to reach the number of criteria needed for classification. Of 50 patients evaluated for amyloidosis, 10 (20%) had positive Congo red staining.


MSGB is a simple, safe, and reliable tool for the diagnosis of SS and amyloidosis, and therefore is suitable for more extensive application.


Minor salivary gland biopsy (MSGB) is a simple procedure that is currently extensively used in internal medicine and rheumatology for the diagnosis of Sjögren's syndrome (SS) and other infiltrative diseases, such as amyloidosis (1), hemochromatosis (2), and sarcoidosis (3). The original technique was introduced in 1966 (4, 5) for the diagnosis of SS and remained the prerogative of surgeons for many years. Until now, a variety of methods have been described, each accompanied by various inconveniences such as inadequate sampling or excessively large incision with an increased incidence of side effects (6–9). A minimally invasive procedure, which can be carried out by nonsurgical physicians, has proven to provide sufficient material for analysis with low risk of complications (10, 11).

MSGB is regarded as mandatory in selected cases (e.g., sicca syndrome and negative anti-Ro or anti-La autoantibodies) for the classification of SS following the 2002 publication of the revised version of the European criteria from the American-European Consensus Group (AECG) (12). The group asserted that MSGB with a focus score ≥1 (defined as a number of inflammatory infiltrates of at least 50 cells in 4 mm2 of gland surface unit), together with a specific serologic profile, was a major parameter (12).

How valuable MSGB is for the diagnosis of SS is still the object of debate. This is due to the fact that, although the methodology of sampling and examining MSGB has been standardized (13, 14), the reproducibility of the focus score assessment according to the method described by Greenspan et al (13) and Daniels (9) is still low (14, 15). However, it has been recently reported that the evaluation of a cumulative focus score (cFS) on multilevel slides can improve the diagnostic performance of the AECG criteria (16).

Moreover, the reliability of MSGB in the diagnosis of amyloidosis has been reported in many studies, confirming the high sensitivity and specificity of this tool for this purpose (1, 17–19). Despite the fact that the histologic evaluation of MSGB is simple and easily reproducible, given the definite histologic aspect, periumbilical fat aspiration is still the most commonly used procedure for screening for amyloidosis in many centers.

In this study, we retrospectively analyzed the clinical and laboratory findings of all patients seen at our clinic over an 8-year period who underwent minimally invasive MSGB as part of an evaluation for SS or for suspected amyloidosis with negative or nondiagnostic periumbilical fat aspiration analysis. The primary aim of our study was to analyze the safety of our technique; the secondary aim was to evaluate the efficacy of MSGB in diagnostic routine application.



Between August 1998 and March 2006, we performed 502 consecutive minimally invasive MSGBs as part of an evaluation for primary SS (n = 452) and systemic amyloidosis (n = 50). Patients recruited from our clinic were suspected of having SS if they had ocular and/or oral dryness for >3 months or if they showed positive anti-SSA and/or negative SSB autoantibodies without symptoms. Patients with established diagnosis of connective tissue disease were excluded. We recorded the clinical information of each patient, including demographic features, chief symptoms, and possible historical, physical, or laboratory findings that might be attributed to SS or that might represent exclusion criteria according to the AECG (12). We also obtained a detailed drug history to determine drugs that may cause sicca syndrome and immunosuppressive drugs that may affect the result of the biopsy.

The complete evaluation included the following: 1) performance of Schirmer's test without anesthesia using standardized tear test strips (a value ≤5 mm in 5 minutes was considered positive); 2) evaluation of unstimulated whole salivary flow, considering a flow ≤1.5 ml in 15 minutes as positive; 3) serologic studies, including detection of antinuclear antibodies using indirect immunofluorescence procedure on a Hep-2 substrate (Hep 2000, Immunoconcept, Sacramento, CA), anti-SSA and anti-SSB autoantibodies using enzyme-linked immunosorbent assay (Diastat, Bouty, Milan, Italy), and rheumatoid factor using immunonephelometry (Dade Behring, Newark, DE); 4) serologic screening test for hepatitis C virus (HCV) followed, if positive, by immunoblot analysis; and 5) evaluation of erythrocyte sedimentation rate and C-reactive protein level. SS was diagnosed according to the 2002 AECG criteria (12).

Patients who underwent MSGB for the detection of amyloid deposits were recruited from our hospital's amyloidosis center. These patients were suspected of having amyloidosis if they had monoclonal gammopathy and symptoms/signs indicating kidney, heart, skin, or neurologic involvement (i.e., chronic renal failure, cardiac symptoms with echocardiographic findings indicating cardiomyopathy, autonomic neuropathy) but negative or inconclusive results from periumbilical fat aspiration.

All participants gave written informed consent before the bioptic procedure according to our local ethics committee.

Biopsy technique.

All MSGBs were performed by 2 rheumatologists from our unit (OE and RC) following the evaluation of coagulation parameters. The salivary glands were located by palpation of the lower lip; after disinfection with hydrogen peroxide and the infiltration of a drop of anesthetic (adrenaline plus bupivacaine) near one of the glands, a small incision (2–3 mm) was made with a scalpel (blade 3) on the inner surface of the lower lip. The gland (sometimes more than 1) bulged from the wound so that it could be easily pressed out and removed with biopsy forceps (Figure 1). The incision was sutured with reabsorbable silk and an ice pack was applied for several minutes. In our study, this was sufficient to provide adequate glandular lobules for the histopathologic study (4–5 lobules).

Figure 1.

Minimally invasive minor salivary gland biopsy image showing the small incision required to collect the glands.

Safety assessment.

Adverse events were recorded immediately after the procedure and 7 days, 14 days, and 6 months thereafter using a questionnaire. The assessment was made verbally at 0 days and 7 days with the patients in our outpatient clinic; at 14 days and 6 months the questionnaire was administered via telephone (Table 1).

Table 1. Questionnaire used to assess safety 0 days, 7 days, 14 days, and 6 months after the procedure
After the biopsy was performed, did you have any bleeding?
Did you have any anesthesia/paresthesia of the lip?
Did you have any external hematoma?
Did you have any local swelling of the lip?
Did you have any local pain?
Did you suffer from any other complaints after the biopsy was performed?
If so, please describe them.

Histopathologic study.

Study for SS.

The procedure has been described in a previous report (16). Briefly, samples were fixed in formalin, processed, and embedded in paraffin according to standardized laboratory methods. Sections measuring 4 μm were obtained. The sections were then cut again into 2 additional levels at 200-μm intervals and stained with hematoxylin and eosin. All samples were analyzed by the same pathologist (PM), blinded to clinical and laboratory data, who recorded for each patient the number of glands, the sample surface area, and the presence of alterations suggestive of nonspecific sialadenitis. Lobules with acinar atrophy and diffuse fibrosis were excluded. The area of the biopsy sections was assessed with video-assisted morphometric software capable of measuring the area of delineated surfaces (ImageDB System, Casti Imaging, Cazzago di Pianiga, Italy). Scoring of the lymphocytic infiltrates was performed on each section, according to Greenspan et al (13) and Daniels (9). For all samples, a cFS was calculated as previously described (16). A cFS ≥1 per 4 mm2 was considered abnormal. The pathologist reevaluated each sample in a blinded manner almost 20 days after the first evaluation.

Study for amyloidosis.

The samples collected were fixed in formalin and processed as described, stained with Congo red–saturated alcoholic solution, and examined by polarized-light microscopy.



All patients answered the questionnaire immediately after the procedure as well as 7 and 14 days after the biopsy was performed; 488 (97.2%) of 502 patients answered after 6 months.

The procedure was well tolerated in all cases: no major adverse events were observed. Sixty-four patients (12.7%) reported transient adverse events (Table 2). No significant differences were noted between patients who underwent MSGB for suspected SS versus suspected amyloidosis. Only 1 patient (0.2%), who underwent biopsy in 2004, still has local paresthesia.

Table 2. List of the transient adverse effects referred by 64 (12.7%) of 502 patients*
 Suspected SS (n = 452)Suspected amyloidosis (n = 50)
  • *

    Values are the number (percentage). SS = Sjögren's syndrome.

Transient adverse effects58 (11.5)6 (12)
Paresthesia <7 days37 (8.2)3 (6)
Paresthesia >7 and <14 days16 (3.5)1 (2)
External hematoma <7 days7 (1.5)1 (2)
Local swelling <7 days24 (5.3)3 (6)
Other (granulomas <1 mm, bleeding, internal scarrings <2 mm)5 (0.6)0 (0)


Four (1%) of 435 samples collected for the diagnosis of SS did not provide adequate material for histologic analysis. All samples collected for the diagnosis of amyloidosis were suitable for histologic analysis.

Diagnosis of SS.

We performed 452 MSGBs as part of the evaluation for SS; complete data are available for 435 patients (35 men and 400 women). Results are summarized in Table 3.

Table 3. Clinical and serologic features of 435 patients evaluated for Sjögren's syndrome*
  • *

    Values are the number (percentage) unless otherwise indicated. UCTD = undifferentiated connective tissue disease; UWSF = unstimulated whole salivary flow; ANA = antinuclear antibody; HCV = hepatitis C virus; cFS = cumulative focus score; MSGB = minor salivary gland biopsy.

Male/female, no.35/400
Age, mean (range) years56 (18–87)
Duration of symptoms, mean years3.5
Previous diagnosis of UCTD, no.50
 Anti-Ro positive, no.31
Xerostomia384 (88)
Xerophthalmia381 (87.5)
Parotid swelling41 (9.4)
Raynaud's phenomenon46 (10.5)
Arthralgias186 (42.5)
Fibromyalgia98 (22.5)
Sensory neuropathy22 (5)
Positive Schirmer's test, no./total no. (%)249/422 (59)
Reduced UWSF228 (52)
Rheumatoid factor >15 IU, no./total no. (%)184/410 (44)
ANA ≥1:80257 (59)
Positive Ro/SSA164 (37.4)
Positive La/SSB53 (12.5)
Positive HCV53 (12.5)
cFS ≥1 at MSGB93 (24.5)

The mean age of the patients was 56 years (range 18–87 years). A total of 392 patients were investigated for primary SS and 43 for secondary SS. Fifty patients were followed according to a previous diagnosis of undifferentiated connective tissue disease (UCTD); of these, 31 had persistent positive anti-Ro/anti-La autoantibodies. The mean duration of symptoms was 3.5 years (range 6 months to 20 years).

Fifty-three patients (12.6%; 51 women and 2 men) were excluded due to positive HCV tests; 4 (1%) were excluded for inadequate sampling. In 1 patient, the evaluation of MSGB evidenced a low-grade B cell lymphoma. Ninety-three patients (24.5%) had a cFS ≥1 upon histopathologic evaluation (Figure 2); of these, 87 (94.5%) satisfied the AECG criteria set (12). Classification was possible for 124 (32.8%) of 378 patients. In 51 (41%) of 124 patients, MSGB was essential in order to reach the number of criteria needed for the classification of SS. The reevaluation of the biopsies performed by the same pathologist after a minimum of 20 days gave the same results. The clinical and serologic characteristics of patients diagnosed with SS are summarized in Table 4.

Figure 2.

Example of minor salivary gland biopsy samples showing significantly different focus scores at different cutting levels. Level 1 (a) showed 2 lymphocytic foci (arrowheads) with a focus score of 1.4 (consistent for Sjögren's syndrome [SS]). On level 2 (b), where only 1 focus was present (arrowhead), a focus score of 0.8 was calculated (not consistent for SS). Level 3 (c) showed no lymphocytes. The cumulative focus score was 0.9, thus the final diagnosis was not consistent for SS. Magnification × 10.

Table 4. Clinical and serologic features of patients with final diagnosis of Sjögren's syndrome*
  • *

    Values are the number (percentage) unless otherwise indicated. See Table 3 for definitions.

Total number124
Male/female, no.12/112
Age, mean (range) years55 (18–80)
Xerostomia120 (96.7)
Xerophthalmia115 (92.7)
Positive Schirmer's test113 (91.2)
Reduced UWSF67 (54)
Rheumatoid factor >15 IU85 (68.5)
ANA ≥1:80113 (91.1)
Positive Ro/SSA100 (80.6)
Positive La/SSB34 (29.6)
cFS ≥1 at MSGB86 (69.3)

Among the 50 patients with a previous diagnosis of UCTD, 6 were excluded due to a positive HCV test result, and 21 (47.7%) of 44 achieved the diagnosis of SS; 19 of 21 had previously known positive anti-Ro autoantibodies. Four (1.6%) of 254 patients who did not have the requisites for the diagnosis of SS showed a cFS ≥1 upon MSGB.

Diagnosis of amyloidosis.

From March 2003 to March 2006, 50 MSGBs were performed for suspected amyloidosis. Complete data are available for all patients (31 men and 19 women, mean age 62 years). Examination of periumbilical fat aspirate by Congo red staining was negative in 20 (40%) patients; in 28 (56%) patients, Congo red staining was weakly positive and did not allow a definite diagnosis. In 2 patients (4%), periumbilical fat aspiration could not be performed due to abdominal adherences secondary to previous laparotomy.

Ten patients (20%) had MSGBs with positive Congo red staining. Of these, 4 had negative and 5 had nondiagnostic periumbilical fat aspirate; 1 patient did not undergo periumbilical fat aspiration as explained above.


MSGB is a simple, safe, and reliable tool for the diagnosis and followup of SS and amyloidosis. Extensive application is advisable, although its reliability may be affected by the pathologist's skill (6–9, 20, 21).

The technique introduced in the 1960s has remained the prerogative of surgeons for many years; the collection of adequate samples (from 3–7 glands on average) required larger incisions (2 cm) of normal-appearing mucosa, down to the muscle layer (4–6, 8). The procedure described by Daniels (9) provided enough tissue for examination carrying a low risk of nerve damage. No major adverse effects have been observed; short-term cutaneous ecchymosis, local swelling, pain, and transient numbness have been the most frequent symptoms. In 2002, Friedman et al (10) described a simple and safe sampling technique that could be performed by nonsurgical physicians and required only small incisions (5–7 mm) that could heal spontaneously; the most reported adverse events were local swelling (10%), local infection (4%), short-term local pain (2.5%), and long-term local numbness (1.7%). However, MSGB is still scarcely used as a first-step diagnostic tool, and many authors are still searching for serologic markers (22) or noninvasive imaging procedures (23) that could offer a good prediction of the histologic study and eventually replace it.

In a recent article, Pijpe et al (24) proposed the inclusion of histopathology of the parotid gland in the classification criteria for SS on the basis of its high diagnostic potential and low morbidity compared with labial biopsy in a series of 15 patients with SS. As discussed elsewhere (25), this approach presents some limitations, in particular related to the fact that parotid biopsy requires surgical skills.

Our data show that MSGB can be easily performed, provides sufficient material for histologic studies (inadequate sampling affected <1% of our cases), and is safe: only 12.7% of patients showed transient adverse events, except for 1 woman who still has long-term paresthesia. No major adverse events have been observed, making this technique well accepted by patients and repeatable during the followup.

The high safety of MSGB allows its use in the followup of patients with UCTD who show risk factors for the development of SS (i.e., xerophthalmia and anti-Ro or anti-La autoantibodies) (26). In our study, 21 of 44 patients with a clinical diagnosis of UCTD were diagnosed as having SS after biopsy.

Also, the use of a simple and safe technique, such as MSGB, allows multiple sampling in patients developing symptoms/signs that are highly suggestive of lymphoproliferative diseases (27, 28). It is common knowledge that patients with SS have an increased risk (16–40 times) of developing lymphoma (29, 30). In fact, MSGB enabled the diagnosis of low-grade B cell lymphoma in one of our patients.

A possible concern regarding the application of MSGB in the diagnosis of SS is the standardization of histologic evaluation. With the studies by Chisholm and Mason (6) and Chisholm et al (7) and, later, Daniels (9), histologic evaluation was introduced in different sets of criteria for the classification of SS (31–34).

In 1993, the need for a uniform definition of the disease to be adopted by the scientific community led to the creation of the European Community Study Group on Diagnostic Criteria for Sjögren's Syndrome to validate classification criteria in a multicenter study. MSGB with a focus score ≥1 showed the best balance between sensitivity and specificity (83.5% and 81.8%, respectively) and the highest accuracy among the tests evaluated, and was chosen as an independent criterion (35, 36). The revised international criteria set by the AECG, published in 2002, definitively established the central role of focal sialadenitis in MSGB and the presence of anti-SSA/anti-SSB antibodies in the diagnosis of SS, proposing these as major criteria (12).

Although Greenspan et al (13) and Daniels (9) standardized the methodology in assessing focus score, its reproducibility between different pathologists and at different section levels within the same sample seems to be low (14, 15). In a recently published study, we calculated a cFS on multilevel slides of 120 labial gland biopsy samples obtained from patients with suspected SS. Statistical analysis by receiver operating characteristic curve highlighted that this method increased the specificity of the AECG criteria set by 9.8% with a statistically significant improvement of the diagnostic performance, without affecting sensitivity (16).

With the limitations of retrospective studies, our data show that the routine use of multilevel examination improved the contribution of MSGB to SS diagnosis: up to 69% of all patients with newly diagnosed SS had a cFS ≥1 and >40% of these patients reached the number of criteria sufficient for classification thanks to the histologic data. Only 1.6% of patients who did not have the requisites for the diagnosis of SS had a cFS ≥1 upon MSGB, and this underlines the high specificity of MSGBs when cFS is applied (16).

Since 1989, labial salivary gland biopsy has been widely described for the diagnosis of amyloidosis and has proven to be effective and suitable for routine use (1, 17–19, 37). Despite this evidence, abdominal fat still remains the most commonly biopsied tissue in first-step evaluation of patients with suspected amyloidosis (38). We used MSGB for the diagnosis of amyloidosis in patients with suggestive clinical and/or laboratory findings but who presented negative or uncertain results with periumbilical fat aspiration. In all cases, MSGB was performed without long-term adverse effects and provided adequate material for analysis. In 20% of patients, MSGB permitted diagnosis, avoiding the application of more invasive diagnostic techniques such as renal or myocardial biopsy. Moreover, MSGB could be the first choice in patients who cannot undergo periumbilical fat aspiration due to complications presented by previous abdominal surgery.

We are aware that the study has some limitations. First of all, the study lacked a second blinded pathologist to review all of the biopsy samples, which may weaken the reliability of the histopathologic analysis. To partially overcome this limitation, the same pathologist reviewed each sample in blinded fashion a minimum of 20 days after the first evaluation, confirming the score in all cases. Another possible concern is the lack of objective evaluation of motor and sensory diseases after the performance of the biopsy, as other authors did (24). Finally, the retrospective design of our study may be another limiting factor.

However, despite these concerns, we think that the large size of the sample and the high response rate of the patients may support the conclusion that MSGB is a simple, safe, and reliable tool for the diagnosis and followup of sicca syndrome, anti-Ro–positive UCTD, and amyloidosis. Therefore, it is highly suitable for extensive routine application in internal medicine and in rheumatology.


Dr. Caporali had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Caporali, Bonacci, Montecucco.

Acquisition of data. Caporali, Bonacci, Epis, Bobbio-Pallavicini, Morbini, Montecucco.

Analysis and interpretation of data. Caporali, Bonacci, Morbini, Montecucco.

Manuscript preparation. Caporali, Bonacci.

Statistical analysis. Caporali, Bonacci.