Chapter 29

Classification Criteria

Chiara Baldini, Rosaria Talarico, and Stefano Bombardieri

29.1Introduction

SS is a chronic, systemic inßammatory disorder that mainly affects the salivary and lacrimal glands with focal lymphocytic inÞltration of the exocrine glands, leading to sicca symptoms [1]. Although xerostomia and keratoconjunctivitis sicca are the hallmarks of SS, over the course of disease progression any organ or mucosal surface may be involved. The disease also has an important association with nonHodgkinÕs lymphoma [2, 3]. Thus, pSS is a heterogeneous autoimmune entity that possesses both organ-speciÞc and systemic features and encompasses a wide spectrum of clinical manifestations, serological abnormalities, and shortand long-term complications. The complexity of SS clinical presentation is increased by the fact that SS can occur alone as pSS or in association with other connective tissue diseases, including rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), and systemic sclerosis (SSc), in which case the disorder is considered to be a secondary SS (sSS) variant [4].

Disease heterogeneity is probably the most important reason why multiple classiÞcation criteria sets have cropped up through the years, all proposed by leading

C. Baldini ¥ R. Talarico ¥ S. Bombardieri (*)

Rheumatology Unit, Department of Internal Medicine, University of Pisa, Pisa, Italy

experts in the Þeld, yet none has been widely accepted until 2002 [5]. The key question the different sets of classiÞcation criteria have had to address was whether the term ÒSjšgrenÕs syndromeÓ can only be applied to a rather restricted group of individuals whose exocrinopathy is linked to a systemic, autoimmune basis, to the larger group of patients who share the similar sicca complex symptoms [6, 7]. In 2002, the American-European Consensus Group (AECG) proposed a criteria set that is now considered by most experts in the Þeld to provide a sufÞcient basis for the diagnoses of pSS and sSS [5]. Nevertheless, classiÞcation criteria for SS remain a subject of ongoing debate.

29.2Historical Overview and Sets of Criteria

Four different criteria sets for the deÞnition of SS were presented at the First International Seminar on SS in 1986. The four different sets were known as the Copenhagen [8], Japanese [9], Greek [10], and San Diego criteria [11], respectively. Eleven years earlier, yet another set Ð the San Francisco criteria for SS Ð had been proposed USA [12]. Table 29.1 summarizes their similarities and dissimilarities. All these criteria sets had been presented by leading experts in the Þeld and were routinely used in the leading expertÕs own countries but had not been validated by multicenter studies or by means of standard statistical approaches. The sets focused primarily on assessing the glandular signs and symptoms of the disease, utilizing different procedures with variable levels of sensitivity, speciÞcity and reliability (in many cases still not assessed). The criteria sets appeared to be more or less restrictive depending on their requirements for histological or serologic abnormalities in addition to features of dry eye and dry mouth.

All of the candidate criteria sets except Copenhagen employed the terms ÒprobableÓ and ÒdeÞniteÓ SS. The Copenhagen, Greek, and San Francisco criteria used the terms ÒpSSÓ and ÒsSSÓ [8, 10, 12]. The concept of probable and deÞnite was introduced in the 1958 classiÞcation criteria for rheumatoid arthritis but was viewed as problematic because of issues relating to giving a patient a diagnosis of a ÒprobableÓ disorder [13]. The distinction between pSS and sSS, on the other hand, was justiÞed by the fact that patients with sSS had different genetic, serological, and clinical proÞles with respect to pSS patients.

Some of the proposed criteria sets included both subjective symptoms (e.g., sicca symptoms) and objective Þndings (e.g., positive lip biopsies). In contrast, others relied exclusively on objective Þndings, considering the subjective symptoms to be non-speciÞc. The Copenhagen Criteria and San Francisco Criteria were the most restrictive criteria from this point of view, whereas the Greek criteria and the Japanese were more liberal. The San Diego criteria offered an intermediate view, focusing on objective evidence of sicca symptoms and signs but also including symptomatic xerostomia, provided that it was also proven on clinical examination and conÞrmed by decreases in the basal and stimulated salivary ßow rates.

The choice of diagnostic tests and their ranges has been another topic of debate. For example, the Greek criteria required only one abnormal test for the diagnosis of

Table 29.1 Similarities and dissimilarities of the historical criteria sets for SS: Copenhagen, Japanese, Greek, San Diego and San Francisco criteria

keratoconjunctivitis sicca (KCS). In contrast, the Copenhagen, Japanese, and San Diego criteria required at least two simultaneous tests of ocular dryness (see Table 29.1). The tests most commonly used to diagnose KCS were the Schirmer-I test, the tear break-up time, the slit lamp examination after rose Bengal staining, and the ßuorescein test.

attended by 29 experts from 11 European countries and Israel. The aim of this collaboration was to deÞne and validate simple standardized diagnostic tools for SS and to design a multicenter study to deÞne classiÞcation criteria for SS [20]. The novelty was represented by the fact that previously proposed classiÞcation criteria had generally been formulated by experts on the basis of clinical experience or derived from data coming from single centers. The preliminary European classiÞcation criteria, in contrast, represented the Þrst attempt to create criteria for pSS using the same methodology and statistics that have been used by the American College of Rheumatology for RA: a multicenter study using a generally agreed upon protocol, standardized methodologies, and the collection of data from real patients [13, 20].

The panel of experts included rheumatologists, ophthalmologists, oral pathologists, and epidemiologists and the study protocol was articulated in two parts. The preliminary groundwork was represented by the proposal of a panel of candidate tools for the diagnosis of glandular and extraglandular aspects of the disease. Clear guidelines for their application and interpretation were deÞned and collected in a manual of methods and procedures. Furthermore, the panel of experts decided that for subjective symptoms, a speciÞc questionnaire should be validated on a separate protocol. A simple questionnaire (20 questions: 13 regarding ocular involvement and 7 regarding oral involvement) for dry eyes and dry mouth was therefore validated. Each expert was asked to complete these questionnaires in 15 patients who, in their judgment, had well-deÞned pSS, and also in 15 healthy ageand sex-matched controls.

Data on 480 patients (240 SS and 240 controls) were gathered. Monovariate and multivariate analyses and stepwise multiple regression were used to select those questions and combinations of questions that performed best in classifying patients and controls correctly. A simpliÞed questionnaire consisting of three questions each for dry eyes and dry mouth emerged from this section of the study [20].

For part II, each center recruited 40 patients: ten each with pSS, sSS, other connective tissue disorders (CTDs) without SS, and ten controls. In these study subjects, a limited set of proposed diagnostic tests were validated. These included the Schirmer-I test, rose bengal test, tear break-up time, tear ßuid lactoferrin level, stimulated and unstimulated saliva ßow, biopsy of the minor salivary glands, parotid sialography, and salivary gland scintigraphy. The procedures for each test were described in the protocol. The data from part II were subjected to the same analysis as were those from part I, with the addition of a classiÞcation tree to assist in determining the optimal classiÞcation strategy. From the analysis, the consensus group established a set of four objective criteria for the diagnosis of SS [21]. These four criteria and the two subjective criteria are presented in Table 29.2. The preliminary European criteria were based on any four out of six items, including ocular and oral symptoms (oral and ocular dryness), ocular and oral signs (positive Schirmer-I test, Rose Bengal score, parotid sialography, scintigraphy and unstimulated salivary ßow), immunological parameters, and focal sialoadenitis. For primary SS, the presence of four out of six items had good sensitivity (93.5%) and speciÞcity (94%) [21]. Some exclusion criteria were also added to this classiÞcation set for SS following the recommendations made by Fox et al. [11], namely preexisting lymphoma, acquired immunodeÞ-

Table 29.2 European preliminary criteria for SjšgrenÕs syndrome

I. Ocular symptoms: a positive response to at least one of the following questions:

1.Have you had daily, persistent, troublesome dry eyes for more than 3 months?

2.Do you have a recurrent sensation of sand or gravel in the eyes?

3.Do you use tear substitutes more than 3 times a day

II. Oral symptoms: a positive response to at least one of the following questions:

1.Have you had a daily feeling of dry mouth for more than 3 months?

2.Have you had recurrently or persistently swollen salivary glands as an adult?

3.Do you frequently drink liquids to aid in swallowing dry food?

III. Ocular signs Ð that is, objective evidence of ocular involvement deÞned as a positive result for at least one of the following two tests:

1.SchirmerÕs I test (£5 mm in 5 min)

2.Rose bengal score or other ocular dye score (³4 according to van BijsterveldÕs scoring system)

IV. Histopathology: Focus score ³1 on minor salivary gland biopsy (focus deÞned as an aggregation of at least 50 mononuclear cells; focus score deÞned as the number of foci per 4 mm2 of glandular tissue)

V. Salivary gland involvement: objective evidence of salivary gland involvement deÞned by a positive result for at least one of the following diagnostic tests:

1.Unstimulated whole salivary ßow (<1.5 mL in 15 min)

2.Parotid sialography showing the presence of diffuse sialectasias (punctate, cavitary or destructive pattern), without evidence of obstruction in the major ducts

3.Salivary scintigraphy showing delayed uptake, reduced concentration and/or delayed excretion of tracer

VI. Autoantibodies: presence in the serum of the following autoantibodies:

1.Antibodies to Ro(SSA) or La(SSB) antigens, or both

2.Antinuclear antibodies

3.Rheumatoid factor

Exclusion criteria:

Pre-existing lymphoma

Acquired immunodeÞciency disease (AIDS)

Sarcoidosis

Graft versus host disease

ciency syndrome, sarcoidosis, and graft-versus-host disease. For the diagnosis of sSS, the criterion for any positive serological test was excluded and a consensus emerged from the group that three of Þve items were sufÞcient for diagnosis [21].

In 1996, the criteria set was validated on a total of 278 cases (157 SS patients and 121 non-SS controls) collected from 16 centers in ten countries. The criteria were conÞrmed to have a sensitivity of 97.5% and a speciÞcity of 94.2% [22].

After these validation exercises, the European classiÞcation criteria have been widely received by the scientiÞc community because of their excellent combination of sensitivity and speciÞcity. In fact, when previously proposed criteria [8Ð12] had been used to classify patients with pSS and controls enrolled in the European study, they all demonstrated high speciÞcities (range 97.9Ð100%) but low sensitivities (range 22.9Ð72.2%) [21]. Other advantageous characteristics of the European criteria were that they distinguished between pSS and sSS but avoided the concept

of deÞnite/possible SS. Furthermore, they relied upon unstimulated or basal tests [21, 22] and did not require an invasive procedure such as the minor salivary gland biopsy for diagnosis.

Despite their strong points, the European criteria for the classiÞcation of SS continue to generate extensive discussion. One key point of ongoing debate is that these criteria can be fulÞlled in the absence of either the classic autoantibodies (i.e., antibodies directed against the SSA/Ro or SSB/La antigens) or positive Þndings on labial salivary gland biopsy. Further, the criteria can also be fulÞlled by patients who have sicca symptoms caused by entities other than primary SS. Finally, in a criteria set in which two out of the six items are devoted to subjective complaints, it is difÞcult to classify asymptomatic patients precisely even if the abundance of objective data suggest that a diagnosis of pSS is likely [15, 23].

29.4American-European Criteria: Strengths, Limitations, and Future Proposals

The European Preliminary Criteria represented the Þrst attempt to validate a set of international criteria for pSS [21]. Because of the appropriate criticisms voiced against the European Preliminary Criteria, the SS Foundation proposed that a joint effort be undertaken by the European Study Group on ClassiÞcation Criteria for SS and a group of American experts. The Foundation has organized meetings between the two groups since 1998. A detailed analysis of the European database of the patients and controls collected during the validation phase of the European Criteria was undertaken. A receiver operating characteristic (ROC) curve of the revised criteria was constructed, based on the analysis of 180 cases provided by 16 centers from ten European countries. The patient and control populations included 76 patients affected by pSS, 41 patients with a diagnosis of CTD without SS, and 63 healthy control subjects. The curve was obtained by plotting sensitivity and speciÞcity values for each different combination of positive tests.

Based on these ROC curve analyses, the condition Òpositivity of any four out of the six itemsÓ and the condition Òpositivity of four out of six items with the exclusion of the cases in which both serology and histopathology were negativeÓ showed the same accuracy, 92.7%. However, the second condition had a lower sensitivity (89.5% vs. 97.4%) but a higher speciÞcity (95.2% vs. 89.4%). The presence of any three of the four objective criteria items performed as well a slightly lower accuracy (90.5%) but a speciÞcity of 95.2% and a sensitivity of 84.2%. This combination was, therefore, judged to be reliable, as well. Thus, the American-European Consensus Group introduced the obligatory rule that for a deÞnite diagnosis of SS either the minor salivary gland biopsy or serology had to be positive (see Table 29.3) [5]. Other proposed modiÞcations were suggested to make with item deÞnitions within the European criteria more precise. In particular, it was speciÞed that SchirmerÕs I test should be performed with standardized paper strips in unanesthetized and closed eyes, following the European and the Japanese approaches. Moreover, as rose

Table 29.3 American-European Consensus Group Criteria. Revised international classiÞcation criteria for SjšgrenÕs syndrome

I. Ocular symptoms: a positive response to at least one of the following questions:

1.Have you had daily, persistent, troublesome dry eyes for more than 3 months?

2.Do you have a recurrent sensation of sand or gravel in the eyes?

3.Do you use tear substitutes more than 3 times a day

II. Oral symptoms: a positive response to at least one of the following questions:

1.Have you had a daily feeling of dry mouth for more than 3 months?

2.Have you had recurrently or persistently swollen salivary glands as an adult?

3.Do you frequently drink liquids to aid in swallowing dry food?

III. Ocular signs Ð that is, objective evidence of ocular involvement deÞned as a positive result for at least one of the following two tests:

1.SchirmerÕs I test, performed without anesthesia (<5 mm in 5 min)

2.Rose bengal score or other ocular dye score (>4 according to van BijsterveldÕs scoring system)

IV. Histopathology: In minor salivary glands (obtained through normal-appearing mucosa) focal lymphocytic sialoadenitis, evaluated by an expert histopathologist, with a focus score >1, deÞned as a number of lymphocytic foci (which are adjacent to normal-appearing mucous acini and contain more than 50 lymphocytes) per 4 mm2 of glandular tissue

V. Salivary gland involvement: objective evidence of salivary gland involvement deÞned by a positive result for at least one of the following diagnostic tests:

1.Unstimulated whole salivary ßow (<1.5 mL in 15 min)

2.Parotid sialography showing the presence of diffuse sialectasias (punctate, cavitary or destructive pattern), without evidence of obstruction in the major ducts

3.Salivary scintigraphy showing delayed uptake, reduced concentration and/or delayed excretion of tracer

VI. Autoantibodies: presence in the serum of the following autoantibodies: 1. Antibodies to Ro(SSA) or La(SSB) antigens, or both

Revised rules for classiÞcation

For primary SS

In patients without any potentially associated disease, primary SS may be deÞned as follows:

(a)The presence of any 4 of the 6 items is indicative of primary SS, as long as either item IV (Histopathology) or VI (Serology) is positive

(b)The presence of any 3 of the 4 objective criteria items (i.e., items III, IV, V, VI)

(c)The classiÞcation tree procedure represents a valid alternative method for classiÞcation, although it should be more properly used in clinical-epidemiological survey

For secondary SS

In patients with a potentially associated disease (for instance, another well-deÞned connective tissue disease), the presence of item I or item II plus any 2 from among items III, IV, and V may be considered as indicative of secondary SS

Exclusion criteria:

Past head and neck radiation treatment Hepatitis C infection

Acquired immunodeÞciency disease (AIDS) Pre-existing lymphoma

Sarcoidosis

Graft versus host disease

Use of anticholinergic drugs (since a time shorter than fourfold the half-life of the drug)

Bengal is not available in many countries, other ocular dye scores (i.e., ßuorescein and lissamine green stains) were suggested.

The joint European and American experts deÞned a positive minor salivary gland biopsy as one or more foci of lymphocytes per 4 mm2 glandular tissue, specifying that the foci had to be adjacent to normal-appearing mucous acini. Finally, a consensus on the list of exclusion criteria was also reached. In comparison to the exclusion criteria adopted by the European preliminary criteria, the category ÒanticholinergicÓ drugs was introduced, replacing Òantidepressant, antihypertensive, parasympatholytic, parasympatholytic drugs and neuroleptic agentsÓ. In addition, the term sialoadenosis was deleted and the deÞnition Òpast head and neck radiation treatmentÓ was added. Finally, hepatitis C virus (HCV) infection was added as an exclusion criterion [24].

For sSS it was established that, in patients with a potentially associated disease, the presence of item I or II plus any two from among items III, IV and V had to be considered as indicative of the disorder [5].

The AmericanÐEuropean Revised ClassiÞcation Criteria preserved many aspects of the European Preliminary Criteria but appear to be more stringent [5, 25, 26]. In one study of the Copenhagen, San Diego, European, and AECG criteria sets conducted on 222 consecutive patients with possible SS, 90 patients (41%) fulÞlled at least one classiÞcation criteria set. The highest number of patients fulÞlled the European criteria (36%), followed by the Copenhagen criteria (28%), the AECG criteria (26%) and the San Diego criteria (9%) sets. The AECG criteria resulted therefore to be highly speciÞc and quite restrictive [25].

The AECG criteria are the most widely accepted tool presently available for the classiÞcation of patients with pSS and sSS. The AECG criteria have been applied in a number of epidemiologic studies aimed at determining the prevalence of SS (Table 29.4 summarizes the data) [27Ð39]. According to these studies, the prevalence of pSS assessed by using the AECG criteria varies from 0.09% to 0.6% [28Ð30, 33Ð35], while the incidence of the diseases was estimated to be 5.3 per 105 (4.5Ð6.1) [33]. Prevalence of pSS estimated according to the Copenhagen Criteria and to the Preliminary European criteria varied from 0.2 to 2.7 and 0.35 to 3.59, respectively [28Ð32, 36Ð39]. When studies that have assessed the prevalence of pSS prevalence by different criteria sets are analyzed, the prevalence of pSS estimated by European Preliminary Criteria resulted estimates 1.67Ð2.5 higher than estimates made by the AECG [29].

In our series, 381 out of all the 506 (75%) patients screened for SS, fulÞlled both the AECG and the 1993-EU criteria, whereas the other 125/506 (25%) fulÞlled only the latter. Among the patients who only fulÞlled European Preliminary Criteria, those with positive RF but negative Ro/SSA and/or La/SSB were clinically indistinguishable from AECG patients, over prolonged follow-up term, for both glandular and extraglandular disease manifestations.

These data suggest that excessive stringency may be the major drawback to the AECG when these criteria are employed in epidemiologic studies. The mandatory alternative items Òpositive minor salivary gland biopsyÓ or Òpositivity of Ro/SSA