Predicting adverse outcomes in primary Sjo ¨ gren’s syndrome:
identification of prognostic factors
P. Brito-Zero ´ n, M. Ramos-Casals, A. Bove, J. Sentis1and J. Fonty
Objective. To identify features present at diagnosis that were prospectively associated with adverse outcomes in a large cohort of patients
with primary Sjo ¨gren’s syndrome (SS).
Methods. Two hundred and sixty-six patients diagnosed with primary SS in our department between 1984 and 2002 were consecutively
included and followed up. Outcomes measured were vasculitis, B-cell lymphoma and death. Cox regression analysis was used to evaluate
the effect of variables at diagnosis on outcomes.
Results. Twenty-five (9%) patients developed vasculitis. Multivariate analysis identified parotid scintigraphy grades III or IV (HR 3.55,
P¼0.05) and C4 levels <0.11g/l (HR 8.26, P<0.001) as variables predicting the development of vasculitis. Nine (3%) patients developed
B-cell lymphoma. Multivariate analysis identified C3 levels <0.82g/l (HR 7.54, P¼0.016) as a predictive factor of lymphoma development.
Twenty-five (9%) patients died during follow-up. Systemic involvement (HR 4.51, P¼0.022), vasculitis (HR 4.58, P¼0.042), C4 levels
<0.11g/l (HR 5.47, P¼0.027) and cryoglobulins (HR 4.58, P¼0.013) were independently associated with death. The presence of at least
two of the above-mentioned predictive factors (parotid scintigraphy, vasculitis, hypocomplementaemia and cryoglobulinaemia) was
associated with a lower survival in comparison with patients with no factor (log rank and Breslow tests <0.001).
Conclusion. The main prognostic factors for an adverse outcome identified in our cohort of patients with primary SS were vasculitis, severe
involvement in parotid scintigraphy, hypocomplementaemia and/or cryoglobulins at diagnosis. Patients with at least two of these factors need
a closer follow-up.
KEY WORDS: Sjo ¨gren syndrome, Mortality, Vasculitis, Hypocomplementaemia, Cryoglobulins.
Sjo ¨ gren syndrome (SS) is a systemic autoimmune disease that
presents with sicca symptomatology of the main mucosa
surfaces . The histological hallmark is a focal lymphocytic
infiltration of the exocrine glands, determined by a biopsy of the
minor labial salivary glands . The spectrum of the disease
extends from sicca syndrome to systemic involvement [3, 4].
Few studies have prospectively analysed the outcome of patients
with primary SS, a disease characterized by a chronic, insidious
evolution [5–9]. However, some patients with primary SS
may present a complicated evolution of the disease due to
incidence of lymphoma, which are closely related to a higher
risk of death [10, 11]. The identification of markers prospectively
associated with a poor prognosis [6, 12] could play a significant
role in identifying those patients requiring a closer follow-up.
The aim of this study was to identify those features present at
diagnosis that were prospectively associated with adverse out-
comes (development of vasculitis, lymphoma or death) in a large
cohort of Spanish patients with primary SS.
Patients and methods
Study cohort and observation time
The study cohort included all patients diagnosed with primary SS
by our Department of Autoimmunes Diseases between 1984 and
2002 according to the 1993 European criteria . Since a new
set of classification criteria appeared in 2002, a retrospective
re-evaluation was performed to exclude patients who presented
both negative salivary gland biopsy and autoantibodies and who
did not fulfil the recently proposed classification criteria .
In 80 patients with negative anti-Ro/La autoantibodies, a salivary
gland biopsy was not carried out. In these patients, it was not
possible to retrospectively evaluate the fulfilment or not of the
2002 classification criteria.
All patients were consecutively included when the fulfilment of
criteria was confirmed by our department and thereafter followed
up prospectively with regular visits at 6–12 month intervals.
Clinical and laboratory data were collected and computerized
according to the standard protocol of our department [15, 16].
The individual observation time for every patient was from the
time of fulfilment of the 1993 classification criteria until the last
hospital visit, transfer out or death. The design of this study
conformed to the ethical standards currently applied in Spain.
Due to the anonymous nature of the study, informed patient
consent was not required.
Systemic involvement was defined as the presence of at least
one of the following features [15, 16]: non-erosive arthritis,
Raynaud’s phenomenon, lung involvement, nephropathy, vascu-
litis, peripheral neuropathy or CNS (central nervous system)
involvement. Vasculitic involvement was defined as previously
described when histological and/or arteriographic confirmation of
vascular damage was available . When a cutaneous biopsy was
topographically difficult to obtain, was clinically contraindicated
or was not possible due to lacking consent, a diagnosis of vascu-
litis was made when cutaneous lesions, evaluated by a consultant
dermatologist, were considered characteristic of vasculitis and
other processes were excluded . B-cell lymphomas were
classified according to the 2001 WHO classification for tumours
of haematopoietic and lymphoid tissues . Immunological tests
were determined as previously described [15, 16].
The outcomes measured were development of vasculitic involve-
ment, B-cell lymphoma and death. Possible predictive variables at
Department of Autoimmune Diseases and1Statistic Unit, Department of Public
Health, Department of Medicine, School of Medicine, University of Barcelona,
Institut d’Investigacions Biome `diques August Pi i Sunyer (IDIBAPS), Hospital
Clı ´nic, Barcelona, Spain.
Revised version accepted 7 December 2006.
yIn memoriam (1953–2006).
Correspondence to: M. Ramos-Casals, Servei de Malalties Autoimmunes,
Hospital Clı ´nic, C/Villarroel, 170, 08036-Barcelona, Spain. Phone: 34-93-2275774,
Fax: 34-93-2275774. E-mail: email@example.com
Advance Access publication 14 June 2007
? The Author 2007. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: firstname.lastname@example.org
by guest on June 5, 2013
the time of SS diagnosis were previous or current parotid
enlargement, previous or current vasculitis, parotid scintigraphy
grades III or IV , positive salivary gland biopsy ,
RF>25IU/l, positive cryoglobulins (cryocrit>1%), C3 levels
<0.82g/l and C4 levels <0.11g/l.
Univariate Cox regression analysis was used to evaluate the
crude effect of clinical and immunological variables at diagnosis
on adverse outcomes. Multivariate Cox regression analysis using
a backward stepwise method allowed adjustment for age, gender
and the variables that were statistically significant (P<0.05) in the
univariate analysis. The hazard ratios (HRs) and their 95%
confidence intervals (CIs) obtained in the adjusted regression
analysis were calculated. Kaplan–Meier survival curves were
compared using the log-rank and Breslow tests. The statistical
Chicago, IL, USA).
positive anti-Ro/SS-A and/oranti-La/SS-B,
The study cohort included 247 (93%) women and 19 (7%) men
with a mean age at the onset of sicca symptoms of 54.7yrs, a mean
age at the time of diagnosis of 56.9yrs and a total follow-up of
2303 patient-yrs (mean 8.66yrs).
Development of vasculitic involvement
Twenty-five (9%) patients developed vasculitis after a mean
time of follow-up of 58 months (range 12–174). Vasculitis
involved the peripheral nerves in 17 patients, the skin in 13,
kidneys in three, small bowel in two and pancreas in one.
A biopsy specimen was obtained in 20 (80%) patients. The
main histological diagnosis was small-vessel vasculitis in 16
(leucocytoclastic in 14 and lymphocytic in two) and medium-
sized vessel vasculitis (necrotizing vasculitis) in the remaining
Univariate Cox regression analysis identified the following
variables at diagnosis associated with the development of
vasculitis: parotid scintigraphy grades III or IV (HR 4.03,
P¼0.025), systemic involvement (HR 3.01, P¼0.007), vasculitis
(HR 5.27, P<0.001), anti-Ro/La antibodies (HR 2.21, P¼0.049),
RF?25UI/L (HR 2.37, P¼0.034), C3 levels <0.82g/l (HR 4.05,
P¼0.012), C4 levels <0.11g/l (HR 7.53, P<0.001) and cryoglob-
ulins (HR 5.44, P<0.001). Multivariate Cox regression analysis
identified as independent variables parotid scintigraphy grades III
or IV (HR 3.55, P¼0.05) and C4 levels <0.11g/l (HR 8.26,
P<0.001) (Table 1).
Development of B-cell lymphoma
After SS diagnosis, nine (3%) patients developed B-cell lymphoma
(eight women and one man, with a mean age at diagnosis of
haematological neoplasia of 59.66yrs).
Univariate Cox regression analysis identified parotid enlarge-
ment (HR 6.37, P¼0.006) and C3 levels <0.82g/l (HR 7.54,
P¼0.016) at diagnosis as variables associated with the develop-
ment of B-cell lymphoma. Multivariate Cox regression analysis
identified C3 levels <0.82g/l (HR 7.54, P¼0.016) as an
independent variable (Table 1).
Twenty-five (9%) patients died during follow-up. Twenty were
women and five men, with a mean age at death of 72yrs. The main
causes of death were infections in eight, cardiopulmonary diseases
in six, vasculitis in three, non-haematological neoplasia in three,
lymphoma in two and other causes in three. The standardized
mortality ratio (SMR) for the total cohort of patients (adjusted
for age and gender with the general Spanish population) was 1.22.
Univariate Cox regression analysis identified the following
variables at diagnosis associated with death: systemic involvement
(HR 3.62, P¼0.002), vasculitis (HR 5.23, P<0.001), C3 levels
<0.82g/l (HR 3.47, P¼0.049), C4 levels <0.11g/l (HR 5.52,
P<0.001) and cryoglobulins (HR 5.09, P¼0.001). Multivariate
Cox regression analysis identified as independent variables
systemic involvement (HR 4.51, P¼0.022), vasculitis (HR 4.58,
P¼0.042), C4 levels <0.11g/l (HR 5.47, P¼0.027) and cryoglob-
ulins (HR 4.58, P¼0.013) (Table 1).
Factors at diagnosis that were statistically associated with an
adverse outcome in the multivariate Cox regression analysis were
used to postulate a prognostic classification. These were vasculitis,
(low C3 and/or low C4 levels) and cryoglobulinaemia (Table 1).
Patients were classified according to the number of these factors
present at diagnosis into group A (no factor), group B (1 factor)
and group C (2 or more factors). Survival rates for each group
were 96%, 93% and 68%, respectively (log rank and Breslow tests
<0.001). Kaplan–Meier plots are shown in Fig. 1. The SMR for
each group was 0.75, 1.04 and 6.71, respectively.
The natural history of primary SS has been little studied. Some
studies performed in small series of patients agree that, although
SS is not a benign disease, it is characterized by a steady evolution
TABLE 1. Cox regression analysis of all variables for adverse outcomes (development of vasculitic involvement, B-cell lymphoma and death). Figures are unadjusted
(univariate analysis) and adjusted hazard ratios (multivariate analysis including age, gender and the significant variables in the univariate analysis)
Development of vasculitis Development of B-cell lymphomaDeath
Variables at diagnosisUnadjusted HR
Parotid scint. III/IV
SGB grades 3 or 4
C3 levels <0.82g/l
C4 levels <0.11g/l
– 6.37 (1.71–23.73)2
Scint. scintigraphy; SGB, salivary gland biopsy; ANA, antinuclear antibodies; RF, rheumatoid factor; HR, hazard ratio.
1360 P. Brito-Zero ´ n et al.
by guest on June 5, 2013
of the predominant symptoms (sicca features and general
manifestations) [7, 8, 19, 20]. Two recent studies have prospec-
tively analysed the outcome of primary SS in large series of
Greek and Swedish patients [5, 6]. Our study analysed the
factors present at diagnosis that were prospectively associated
with an adverse outcome (development of vasculitis, B-cell
lymphoma or death) in a large cohort of Spanish patients
with primary SS.
Nine per cent of our patients developed vasculitis during the
follow-up. In primary SS, vasculitis can range from a benign,
restricted process to a life-threatening systemic vasculitis .
Ioannidis et al.  found that the presence of purpura at diagnosis
was prospectively associated with a higher risk of mortality, and
was thus the first to suggest the prognostic role of vasculitis in the
outcome of patients with primary SS. In this study, univariate Cox
regression analysis identified severe scintigraphic involvement and
positive immunological markers at diagnosis as predictive factors
of vasculitis. Parotid scintigraphy was an independent prognostic
factor in the multivariate analysis, suggesting that it is not only of
diagnostic value, but could also be useful in identifying patients
with a high risk of developing extraglandular manifestations.
We have prospectively confirmed retrospective studies [21–23]
showing that patients with positive anti-Ro/La antibodies, RF,
hypocomplementaemia and cryoglobulins at diagnosis have a
higher risk of developing vasculitis during the follow-up.
In addition, multivariate analysis confirmed low C4 levels as an
independent prognostic factor for vasculitic development. Our
results suggest that patients presenting at diagnosis with these
immunological markers should be closely followed up due to their
higher risk of developing vasculitis.
In this cohort, 3% of patients with primary SS developed a
B-cell lymphoma, a similar prevalence to other reports [7, 8, 19,
24–26]. Ioannidis et al.  described parotid enlargement, palpable
purpura, anti-Ro/La antibodies and low C4 levels as predictors
of lymphoproliferation, while the predictive factors found by
Theander et al.  were vasculitis, low C3, low C4 and CD4þ
T lymphopenia. We identified parotid enlargement, low C4, low
C3 and vasculitis as predictive factors for development of B-cell
lymphoma, although for low C4 and vasculitis the statistical
significance was weaker (P-value of 0.05–0.10). These statistical
differences between studies may be due to various factors.
The low number of patients developing lymphoma in our study
probably explains why some variables showed only a trend to
significance. In addition, all three studies included a wide range of
lymphoproliferative processes. In spite of these methodological
differences,all three studies
(parotid enlargement, vasculitis and hypocomplementaemia) as
the main predictors of lymphoma development in patients with
Previous studies have suggested that some immunological
markers at diagnosis may have a key prognostic value in the
survival of patients with primary SS. Ioannidis et al.  were the
first to suggest the prognostic role of low C4 levels, and this was
confirmed, together with low C3 levels, by Theander et al. .
We also found a prognostic role for low C4 levels at diagnosis,
but, in addition, identified cryoglobulins as a new prognostic
marker in primary SS. Vasculitis and cryoglobulins were
independently associated with mortality in the multivariate
analysis, suggesting that the association between hypocomple-
mentaemia and death found in previous studies [5, 6, 28] may
be dueto the complement
cryoglobulinaemic vasculitis that some patients with primary SS
All our patients were included before the introduction of
the 2002 AECC (American European Classification Criteria),
when salivary gland biopsy and anti-Ro/La autoantibodies were
not considered as mandatory criteria. Thus, our cohort includes
a subset of patients with negative Ro/La antibodies in whom
salivary gland biopsy was not carried out at diagnosis. These
patients presented a sicca syndrome unrelated to other processes,
together with altered ocular tests and severe involvement in the
parotid scintigraphy and with positive immunological markers
including ANA and RF, and showed no features suggestive of
other systemic autoimmune diseases during the follow-up. Given
that, as explained, the 2002 criteria cannot be applied retro-
spectively in this group of patients, and given that all patients were
included homogeneously and had clinical, diagnostic and immu-
nological characteristics only compatible with primary SS, it
seems reasonable to us to maintain the homogeneity of the study,
rather than exclude one group of patients due to circumstances
caused by changes in the fulfilment criteria overtime.
Ioannidis et al.  were the first to propose a prognostic
classification of patients with primary SS according to the
presence of low C4 levels and/or palpable purpura (type I and
type II patients). This study has confirmed these data and
identified two additional risk factors (severe parotid involvement
demonstrated by scintigraphy and cryoglobulinaemia). A survival
analysis performed in our cohort according to the presence of
these four prognostic factors found that patients with at least
two adverse factors had a significantly lower survival rate in
comparison with patients with no factor. This suggests that
patients who present severe salivary gland involvement, vasculitis,
hypocomplementaemia and/or cryoglobulinaemia at diagnosis
should be considered as candidates for aggressive therapy.
Biological agents, especially those against B-cells (rituximab,
epratuzumab, belimumab), may be especially indicated according
to recent evidence [29–31]. Future clinical trials should prospec-
tively analyse the possible effect of early administration of these
agents on the survival of patients with primary SS with a high risk
of developing adverse outcomes.
In summary, the main prognostic factors for an adverse
outcome identified in our cohort of patients with primary SS
were severe involvement in parotid scintigraphy, vasculitis,
hypocomplementaemia and/or cryoglobulins at diagnosis. These
features identifya specificsubset
with primary SS in whom a closer follow-up, and probably an
earlier and more robust therapeutic management, should be
identifiedthe same factors
The authors have declared no conflicts of interest.
FIG. 1. Prognostic classification: patients were classified according to the number
of prognostic factors (vasculitis, parotid scintigraphy grades III/IV, hypocomple-
mentaemia and cryoglobulinaemia) present at diagnosis into group A (no factor),
group B (1 factor) and group C (2 or more factors). Survival rates for each group
were 96%, 93% and 68%, respectively (log rank and Breslow tests <0.001).
Morbidity and mortality in primary Sjo ¨ gren’s syndrome 1361
by guest on June 5, 2013
1 Fox RI. Sjo ¨gren’s syndrome. Lancet 2005;366:321–31.
2 Kassan SS, Moutsopoulos HM. Clinical manifestations and early diagnosis of
Sjo ¨gren syndrome. Arch Intern Med 2004;164:1275–84.
3 Ramos-Casals M, Tzioufas AG, Font J. Primary Sjo ¨gren’s syndrome: new clinical
and therapeutic concepts. Ann Rheum Dis 2005;64:347–54.
4 Ramos-Casals M, Font J. Primary Sjo ¨gren’s syndrome: current and emergent
aetiopathogenic concepts. Rheumatology 2005;44:1354–67.
5 Theander E, Manthorpe R, Jacobsson LT. Mortality and causes of death in primary
Sjo ¨gren’s syndrome: a prospective cohort study. Arthritis Rheum 2004;50:1262–9.
6 Ioannidis JP, Vassiliou VA, Moutsopoulos HM. Long-term risk of mortality and
lymphoproliferative disease and predictive classification of primary Sjo ¨gren’s
syndrome. Arthritis Rheum 2002;46:741–7.
7 Pertovaara M, Pukkala E, Laippala P, Miettinen A, Pasternack A. A longitudinal
cohort study of Finnish patients with primary Sjo ¨gren’s syndrome: clinical,
immunological and epidemiological aspects. Ann Rheum Dis 2001;60:467–72.
8 Kruize AA, Hene RJ, van der Heide A, Bodeutsch C, de Wilde PC, van Bijsterveld OP
et al. Long-term follow-up of patients with Sjo ¨gren’s syndrome. Arthritis Rheum
9 Martens PB, Pillemer SR, Jacobsson LT, O’Fallon WM, Matteson EL. Survivorship
in a population based cohort of patients with Sjo ¨gren’s syndrome, 1976–1992.
J Rheumatol 1999;26:1296–300.
10 Ramos-Casals M, Anaya JM, Garcia-Carrasco M et al. Cutaneous vasculitis in
primary Sjo ¨gren syndrome: classification and clinical significance of 52 patients.
11 Voulgarelis M, Dafni UG, Isenberg DA, Mousopoulos HM. Malignant lymphoma in
primary Sjo ¨gren’s syndrome. A multicenter, retrospective, clinical study by the
European Concerted Action on Sjo ¨gren’s syndrome. Arthritis Rheum 1999;42:
12 Skopouli FN, Dafni U, Ioannidis JP, Moutsopoulos HM. Clinical evolution, and
morbidity and mortality of Primary Sjo ¨gren’s syndrome. Semin Arthritis Rheum
13 Vitali C, Bombardieri S, Moutsopoulos HM et al. The European Study Group on
Diagnostic Criteria for Sjo ¨gren’s syndrome. Preliminary criteria for the classification
of Sjo ¨gren’s syndrome: results of a prospective concerted action supported by the
European Community. Arthritis Rheum 1993;36:340–7.
14 Vitali C, Bombardieri S, Jonsson R et al. Classification criteria for Sjo ¨gren’s
syndrome: a revised version of the European criteria proposed by the American-
European Consensus Group. Ann Rheum Dis 2002;61:554–8.
15 Ramos-Casals M, Font J, Garcia-Carrasco M et al. Primary Sjo ¨gren syndrome:
hematologic patterns of disease expression. Medicine 2002;81:281–92.
16 Garcia-Carrasco M, Ramos-Casals M, Rosas J et al. Primary Sjo ¨gren syndrome:
clinical and immunologic disease patterns in a cohort of 400 patients. Medicine
17 Jaffe ES, Harris NL, Stein H, Vardiman JW. World Health Organization Classification
of Tumors. Pathology and Genetics of Tumors of Hematopoietic and Lymphoid
Tissues. Lyon: IARC Press, 2001.
18 Schall GL, Anderson LG, Wolf RO et al. Xerostomia in Sjo ¨gren’s syndrome.
Evaluation by sequential salivary scintigraphy. JAMA 1971;216:2109–16.
19 Davidson BKS, Kelly CA, Griffiths ID. Primary Sjo ¨gren’s syndrome in the North East
of England: a long-term follow-up study. Rheumatology 1999;38:245–53.
20 Bowman SJ. Collaborative research into outcome measures in Sjo ¨gren’s syndrome.
Update on disease assessment. Scand J Rheumatol Suppl 2002;116:23–7.
21 Alexander EL, Arnett FC, Provost TT, Stevens MB. Sjo ¨gren’s syndrome: association
of anti-Ro (SS-A) antibodies with vasculitis, hematologic abnormalities and serologic
hyperreactivity. Ann Intern Med 1983;98:155–59.
22 Tsokos M, Lazarou SA, Moutsopoulos HM. Vasculitis in primary Sjo ¨gren’s syndrome:
Histologic classification and clinica presentation. Am J Clin Pathol 1987;88:26–31.
23 Ramos-Casals M, Cervera R, Yagu ¨e J et al. Cryoglobulinemia in primary SS:
prevalence and clinical characteristics in a series of 115 patients. Semin Arthritis
24 Tzioufas AG, Boumba DS, Skopouli FN, Moutsopoulos HM. Mixed monoclonal
cryoglobulinemia and monoclonal rheumatoid factor cross-reactive idiotypes as
predictive factors for the development of lymphoma in primary Sjo ¨gren’s syndrome.
Arthritis Rheum 1996;39:767–72.
25 Lazarus MN, Robinson D, Mak V, Moller H, Isenberg DA. Incidence of cancer in a
cohort of patients with primary Sjo ¨gren’s syndrome. Rheumatology 2006;45:1012–5.
26 Valesini G, Priori R, Bavoillot D et al. Differential risk of non-Hodgkin’s lymphoma in
Italian patients with primary Sjo ¨gren’s syndrome. J Rheumatol 1997;24:2376–80.
27 Theander E, Henriksson G, Ljungberg O, Mandl T, Manthorpe R, Jacobsson LT.
Lymphoma and other malignancies in primary Sjo ¨gren’s syndrome: a cohort study on
cancer incidence and lymphoma predictors. Ann Rheum Dis 2006;65:796–803.
28 Ramos-Casals M, Brito-Zeron P, Yague J et al. Hypocomplementaemia as an
immunological marker of morbidity and mortality in patients with primary Sjo ¨gren’s
syndrome. Rheumatology (Oxford) 2005;44:89–94.
29 Pijpe J, van Imhoff GW, Spijkervet FK, Roodenburg JL, Wolbink GJ, Mansour K et al.
Rituximab treatment in patients with primary Sjo ¨gren’s syndrome: an open-label
phase II study. Arthritis Rheum 2005;52:2740–50.
30 Steinfeld SD, Tant L, Burmester GR et al. Epratuzumab (humanised anti-CD22
antibody) in primary Sjo ¨gren’s syndrome: an open-label phase I/II study. Arthritis Res
31 Pers JO, Daridon C, Devauchelle V et al. BAFF overexpression is associated with
diseases. AnnNY AcadSci
1362 P. Brito-Zero ´ n et al.
by guest on June 5, 2013