Acta Pcediatrica, 2005; 94: 1566-1570

) Taylor & Francis

Taylor & Francis Croup

New insights into juvenile parotitis

KAIJA-LEENA KOLHO\ RIITTA SAARINEN^ ANNUKKA PAJU^ JACOB STENMAN^ U L F - H A K A N S T E N M A N ^ & ANNE P I T K A R A N T A ^
^Hospital for Children and Adolescents, ^Department of Otorhinolaryngology, and ^Department of Clinical Chemistry, Helsinki University Central Hospital, Helsinki, Finland, and '^Department of Surgery, Jorvi Hospital, Espoo, Finland

Abstract Aim: We enquired about the possibility of a familial trend in juvenile parotitis and evaluated the role of SPINKl mutations in juvenile parotitis. Methods: The clinical records of all children admitted to the Helsinki University Hospital during 1995 to May 2003 because of swelling in the parotid gland were reviewed. A questionnaire on possible recurrences and on familial cases was mailed. As disturbances in trypsin inhibition might be involved in the pathogenesis, we assessed the SPINKl gene encoding for Kazal-type trypsin inhibitor in voluntary patients. The study group comprised 133 children (boys 82 girls 51) with juvenile parotitis. The median age at presentation of first symptoms was 6.0 y (range 1-19 y). Results: Recurrent symptoms in the parotid gland were common (57%), and 29% of the children (38/133) had suffered from four or more episodes. A young age at the first episode of symptoms increased the likelihood of recurrences (p<0.0001). Familial cases of parotid swelling were common (22%; response rate 67%). A total of 47 patients (35%) agreed to testing for SPINKl status. Four children had a major mutation (N34S or P55S), corresponding to an 8.5% (4/47) prevalence, but this was not different from the controls (5%). Conclusion: It is likely that inherited factors are involved in the manifestation of juvenile parotitis in a subset of patients. It is tempting to speculate that disturbed proteolytic balance may play a role in the development of symptoms. Key Words: Kazal-type trypsin inhibitor, parotid gland, recurrent, familial

Introduction

medication. The use of antibiotics is controversial [8]. Prevention of recurrences is difficult, and an expectant In children, swelling in the parotid region is most likely policy is indicated [9]. Numerous pathogenic theories caused by inflammation or infection of the parotid have been suggested, such as systemic immunological gland, although occasionally lesions such as congenital diseases [10-12] and congenital parotid malformations cysts or neoplasm (benign and malignant) may be [1,13], but the aetiology for recurrences remains found [1]. Parotitis may be of bacterial origin, the most unknown [9]. To our knowledge, the molecular basis of common pathogens being Staphylococcus aureus and juvenile parotitis has never been elucidated. Kazal type 1 trypsin inhibitor, SPINKl, is a serine anaerobic bacteria [2-4], but in most cases the causative agent is unknown. Recurrent parotitis (juvenile re- protease inhibitor also known as pancreatic secretory current parotitis or chronic parotitis) has been defined trypsin inhibitor (PSTI) or tumour-associated trypsin as recurrent parotid inflammation, generally associated inhibitor (TATI) that has functional importance in the with non-obstructive sialectasia of the parotid gland pancreas by inhibiting proteolytic trypsin activity and [5,6]. The disease presents as a recurrent painful tissue destruction [14]. Recently, it was recognized swelling during mastication and/or swallowing. The that children and adolescents with chronic pancreatitis episodes of unilateral or bilateral parotid swelling may have a high incidence of mutations in the SPINKl gene occur several times per year. The usual age of onset is [15]. In subsequent studies, the association of SPINKl 3-6 y but may range from 3 mo to 16 y. For unknown mutations with idiopathic and familial chronic panreasons, symptoms vanish during puberty and most creatitis has been confirmed throughout the world patients are symptom free by the age of 22 y [7]. [16-18]. SPINKl is also expressed in the subTreatment of acute episodes consists of adequate pain mandibular gland [19] and occurs in saliva (Kolho,
Correspondence: Kaija-Leena Kolho, Hospital for Children and Adolescents, University of Helsinki, Helsinki, Finland. Tel: +358 9 471 73697. Fax: +358 9 471 75299. E-mail: kaija-leena.kolho@hus.fi
(Received 15 April 2005; accepted 11 May 2005)

ISSN 0803-5253 print/ISSN 1651-2227 online 2005 Taylor & Francis DOI: 10.1080/08035250505100399

Parotitis in children

1567

unpublished results). Because of tliis and because the salivary glands, including the parotid gland, resemble exocrine pancreatic tissue in histological structure and in enzyme production [20], it is possible that SPINKl mutations may be associated in the development of parotitis.

Table I. Clinical description of 133 children with parotitis. Parotitis (one episode) Follow-up
RcciiTTPnt

parotitis n = 76 Follow-up (mean, y) (SE)

Sex 7.0

<3y

>3y

n = 20 1.6''''= (0.16)
15 5 1 2" 11" 6" 2 13 1 0 4 12 2 10

= 37 5.22''''= (0.2)
22 15 3a 19" 8 0 18 2" 4= 13" 22 0 22"

Subjects and methods Medical charts from all children and adolescents admitted because of swelling in the parotid gland from 1 January 1995 until the 30 May 2003 at the Department of Otorhinolaryngology, Helsinki University Central Hospital, were analysed retrospectively. Helsinki University Central Hospital covers a region of 1.4 million inhabitants; all severe acute cases and recurrent cases of parotid swelling are referred to this hospital. A total of 142 children and adolescents with swelling in the parotid gland were traced for review. Two siblings had swelling of the parotid gland at the age of 1 to 2 mo and were excluded from further analysis as the family was of non-Finnish origin and communication was inadequate. Seven children had parotitis-related symptoms, but they were not classified as having juvenile parotitis (solid nodule, n = 6; osteomyelitis in clavicula presenting with swelling from clavicula up to the right parotid gland, n=\) and were excluded from further analyses. Thus, the study group comprised 133 children and adolescents with parotitis (median age at admission 8.0 y, range 1-19 y, 82 boys, 51 girls). A minority of these children (10%) had an associated disorder (asthma, w = 5; migraine, n = 2; mild retardation, n=\], Duchenne muscular dystrophy, w=l; relapsed acute lymphoblastic leukaemia, n=\; dacryostenosis, n\\, cranioplastia, n=\; insulin-dependent diabetes mellitus, n= 1). A questionnaire about the occurrence of parotitisrelated symptoms after the last visit to the outpatient clinic at our hospital and about a possible family history of (recurrent) parotitis or pancreatitis was sent in September-October 2003 to the parotitis patients/ families. Data on recurrences were gathered on chart review and on self-reported data on visits in other Outpatient Clinics and on symptoms lasting for at least 1 d. Questionnaires filled in by families of children remitted for adenoidectomy or tympanostomy during summer 2003 served as controls (w==48). The clinical description of the study group is shown in Table I. Although somewhat arbitrary, the children were classified into two groups based on their clinical symptoms: those with only one episode of parotitis (K = 57) and those with a history of recurrent parotitis ( = 76) (Table I). The mean follow-up after the first episode of parotitis was 5.7 y (SE 0.27, range 0.7-15 y). Those with recurrent symptoms had followup data from a significantly longer period (Table I).

(0.37)
46 30 20 31 19 6 6 40 20 1 9 55 17 38

Male Female
Age

0-3 y >3-6 y >6-15y >15-20 y

Ultrasonography

Normal Unilateral parotitis Bilateral parotitis Lymphadenopathy Not done

Questionnaire returned Yes

Positive family history Negative family history

Blood test for Kazal-type trypsin inhibitor-related mutations (SPINKl) 3D Yes 33 11 No. of positive findings 3 0 1

p<0.05, ''p<0.005 considered statistically significant (Fisher's exact test, '^Mann-Whitney U-test).

In addition, these children with parotitis were asked to give a blood sample for determination of mutations in the SPINKl gene encoding the Kazal-type trypsin inhibitor. The mutations N34S and P55S in the SPINKl gene were analysed as recently described [21]. Among Finnish blood donors, 2.6% are positive for the N34S mutation and 1.3% for the P55S mutation [21], but for the purposes of the study we collected control samples from 39 children (26 M, 13 F) admitted to our clinic for elective surgery. The study protocol was approved by the Ethics Committee of Helsinki University Central Hospital. Each patient/ parent gave their informed written consent.
Statistical analyses

The Mann-Whitney U-test and Spearman's rank correlation test were used for non-parametric tests. Fisher's exact test was used for other comparisons. Differences were considered significant at a p-value <0.05. Results The first episode of parotitis had occurred at a median age of 6.0 y (range 1-19 years; Figure 1). A young age

1568
16 14 12

K.-L Kolho et al. but this was not verified. Two children were wearing teeth braces. Five children underwent the excision of the plexus tympani (the Jacobson nerve) after several disabling swelling episodes. The questionnaire was answered by 67% of the families (89/133; no difference in clinical presentation in those completing the questionnaire and those not responding). Twenty-two per cent of these families (19/87) reported that a close family member (mother, . n = B>; grandmother, w = 4; aunt, n = 3; brother, n = 2; first cousin, n = 2; father, n = 2) had experienced similar symptoms in the parotid gland. The corresponding figure for families with children remitted for adenoidectomy or tympanostomy was 2% (1/48, p<0.002). There was no significant difference in the manifestation of pancreatic disease such as pancreatitis or pancreatic carcinoma between the families of the study group and the controls. Forty-seven children with parotitis (35%) were willing to give a blood sample for the analysis of mutations in the SPINKl gene. Analysis of N34S and P55S mutations was performed in 26 boys and 21 girls (median age at first episode 6.7 y, range 1.5-17). The majority of the children (33/47) had recurrent symptoms. A heterozygous N34S mutation was detected in three children and a heterozygous P55S mutation in one child. The family of a 6-y-old boy having the N34S mutation reported two first cousins (boys) with similar symptoms related to the parotid gland, and the family of a 4-y-old boy having the N34S mutation reported that his grandmother had suffered from repeated parotid swelling during childhood. Among the 39 controls, two children had a heterozygous N34S mutation. The other child was a 3-y-old girl with several episodes of otitis media and no history of symptoms related to the parotid gland in her family. Unfortunately, the family history of the other child (a 10-y-old boy admitted for tonsillectomy) was unknown. Discussion There is a general view that, as vaccination has eradicated mumps, the number of children presenting with a single episode of parotitis has decreased. Episodes of parotitis are encountered constantly in children, but the causative agent(s) are largely unknown. Some children have recurrent symptoms, but the mechanisms and risk factors for recurrences are poorly understood. To further increase understanding of this disease, we reviewed the clinical data and traced follow-up information for all of the children and adolescents who presented with parotid swelling during 1995-2003 at our hospital. In our series, the recurrence of parotid swelling was common, as half of these children had experienced repeated symptoms and as many as every third child had suffered from four or more episodes of parotid swelling. The length of follow-up, however.

81 6 4 2 0 1 3 5 7 9 11 13 15 17 19 Age at first episode of parotitis (years)

|o Number of patients I

Figure 1. Age at first presentation of parotitis.

at first episode increased the likelihood of recurrences {p< 0.0001). Recurrent symptoms in the parotid gland were found in 57% (76/133) of the children, and half of these children (38/76) had suffered from four or more episodes of parotid swelling. One child reported more than 40 episodes. Among the children older than 15 y of age, three (14%) reported recurrent symptoms. Season had no significant effect on the number of parotitis cases (tested by ANOVA). A majority of the children had unilateral symptoms (Table I), and the left parotid gland was slightly more often affected (68%). Purulent excretion was reported for 15% (20/133, aged 5-18 y) of the children but was seldom cultured for bacteria. A few had associated symptoms in the submandibular gland (2/133). Ultrasonography confirmed the clinical suspicion of parotitis in 97/110 (88%) of the children undergoing this examination. In ultrasonography, bilateral changes were found in 17% of those presenting with unilateral clinical symptoms. Three children had prominent sialectasia, but the review was not consistent concerning minor changes. The exact duration of symptoms was difficult to assess in this retrospective analysis, but the majority (81%) had symptoms for less than 5 d (median duration 3.7 d). Six children were admitted because their symptoms had lasted for more than 2 wk. The mean level of C-reactive protein was 34 mg/1 (range 2-121 mg/1, w = 59) and the mean blood leukocyte count 10.2x10^/1 (range 3.5-21x10^/1, n = 53). These laboratory parameters were measured in 6/20 of the children with purulent excretion, and there was no significant difference when compared to the parotitis patients with non-purulent findings. Oral antibiotics were prescribed for most of the children (104/133), and 16 children (12%) were treated intravenously because of a suspicion of severe infection or cellulitis. None of the children had sialolithiasis in the parotid gland. In a 6-y-old girl the swelling presented within a week after MMR vaccination. There was a suspicion of connective tissue disorder in the case of an 8-y-old girl.

Parotitis in children was shorter in those children with a single episode. Thus, it seems likely that the prevalence of repeated symptoms would be even higher if the follow-up were continued. Further, it is probable that episodes of parotitis pass unnoticed, as supported by the frequent finding of bilateral chronic parotitis even though presenting symptoms were unilateral. After a nationwide mumps (measles, mumps and rubella) vaccination programme started in Finland in 1982, mumps has become extremely rare in Finland [22]. As parotitis cases resembling mumps continue to occur, it is likely that other viruses cause parotitis as well. It has been shown that Epstein-Barr virus infections for example affect the pancreas and the parotid glands simultaneously [23]. According to our results, it is obvious that physicians do not recognize the possibility of viral aetiology in parotitis since most children received antibiotics. The majority of children showed no elevation in C-reactive protein or blood leukocyte count, and purulent excretion was rare, which support a non-bacterial aetiology. Both the causative agents of parotitis in childhood and the pathophysiological mechanisms leading to overt symptoms are largely unknown. It is assumed that an increase in the proteolytic activity within the parotid gland results in the development of interstitial oedema. A proteolytic enzyme inhibitor, aprotinin, has indeed been used with good results in treatment of the acute symptoms of parotitis [24]. In theory, it is possible that either an increased activity or decreased inhibition of proteolytic enzymes plays a key role in the development of symptoms. This theory is supported by the finding of increased proteolytic enzyme activity in the saliva of children with recurrent parotitis [25]. The Kazal-type trypsin inhibitor is a proteolytic enzyme inhibitor that protects the pancreas from tissue destruction. Recently, it has been found that a mutation in the SPINKl gene coding for Kazal-type inhibitor occurred in 23% of children with chronic recurrent pancreatitis [15], which is, however, a rare disease in childhood. As the salivary glands, including the parotid gland, resemble exocrine pancreatic tissue in their histological structure and in enzyme production [20], it is an intriguing suggestion that this mutation is enriched in patients with recurrent parotitis mimicking pancreatitis with increased proteolytic enzyme activation. One-third of the children with parotitis agreed to testing for SPINKl mutations, and three of the four children with a heterozygous mutation had recurrent symptoms. The frequency of these mutations (8.5%), however, was not significantly different from that in the study controls (5%) or the frequency among Finnish blood donors (3.9%) reported earlier [21]. Our finding of several familiar cases in a questionnaire survey, however, supports the possibility that genetic factors are implicated in the disease manifestation in a subset of patients. This is in agreement with a recent

1569

case report describing a family with four affected members [26]. It is a well-known observation that recurrent episodes of parotitis disappear when puberty is reached [1]. The reason for the relief of symptoms, however, is poorly understood. Interestingly, it has been shown in animal studies that, in puberty, the gene expressions of tissue kallikreins in the salivary glands are androgen regulated. Tissue kallikreins are serine proteases that release bioactive peptides contributing to the development of interstitial oedema in chronic parotitis [24]. Certain groups of these enzymes are expressed in immature but not in mature salivary glands, and other groups are not detectable until puberty [27]. This switch in enzyme expression in puberty suggests a regulatory role for sex hormones [28]. Thus, it is possible that pubertal maturation induces a shift in the cascade of proteolytic enzyme activity, and thereby the occurrence of idiopathic parotid swelling is diminished in adolescents. In conclusion, an early age at the first episode of parotitis is associated with an increased risk of recurrences. The majority of children with parotitis suffer several symptomatic episodes. The possibility of familial symptoms should be included in the patient history, as these aspects may reveal the nature of parotitis. Although no specific intervention is to date available for recurring symptoms, it is tempting to speculate that disturbed proteolytic balance may play a role in the development of parotid swelling. Acknowledgements
The work was supported by the Finnish Cancer Foundation, Sigrid Juselius Foundation, Finnish Academy of Sciences, Helsinki University Central Hospital, University of Helsinki, and the Stockmann Foundation.

References
[1] Myer C, Cotton RT. Salivary gland disease in children. A review. Clin Pediatr (Phila) 1986;25:314-22. [2] Brook AH. Recurrent parotitis in childhood. Br Dentol J 1969;127:271-5. [3] Nusem-Horowitz S, Wolf M, Coret A, Kronenberg J. Acute suppurative parotitis and parotid abscess in children. Int J Peditar Otorhinolaryngol 1995;32;123-7. [4] Brook I. Acute bacterial suppurative parotitis: microbiology and management. J Craniofac Surg 2003;14:37-40. [5] Galiii D, Marmary Y. Juvenile recurrent parotitis: clinicoradiologic follow-up study and beneficial effect of sialography. Oral Surg Oral Med Oral Pathol 1986;61:550-6. [6] Geterud A, Lindvall AM, Nylen O. Follow-up study of recurrent parotitis in children. Ann Otol Rhinol Laryngol 1988;74:341-2. [7] Ericson S, Zetterlund B, Ohman J. Recurrent parotitis and sialectasias in childhood. Clinical, radiological, immunologic, bacteriologic and histologic study. Ann Otol Rhinol Laryngol 1991;100:527-35. [8] Isaacs D. Instructive case. Recurrent parotitis. J Paediatr Child Health 2002;38:92-4.

1570

K.-L. Kolho et al.

[20] Terada T, Kitamura Y, Ashida K, Matsunaga Y, Kato M, Harada K, et al. Expression of pancreatic digestive enzymes in normal and pathologic epithelial cells of the human gastrointestinal system. Virchows Arch 1997;431: 195-203. [21] Lempinen M, Paju A, Kemppainen E, Smura T, Kylanpaa M-L, Nevanlinna H, et al. Mutations N34S and P55S of the SPINKI gene in patients with chronic pancreatitis or pancreatic cancer and in healthy subjects: A report from Finland. Scand J Gastroenterol 2005;40:2325-30. [22] Peltola H, Karanko V, Kurki T, Hukkanen V, Virtanen M, Penttinen K, et al. Rapid effect on endemic measles,-mumps, and rubella nationwide program. Lancet 1986;2:137-9. [23] Mor R, Pitlik S, Dux S, Rosenfeld JB. Parotitis and pancreatitis complicating infectious mononucleosis. Isr J Med Sci 1982;18:709-10. [24] Maier H, Adler D, Lenarz T, Muller-Estrel W. New concepts in the treatment of chronic recurrent parotitis. Arch Otorhinolaryngol 1985;242:321-8. [25] Ericson S, Sjoback \. Salivary factors in children with recurrent parotitis Part 2: Protein, albumin, amylase, IgA, lactoferrin lysozyme and kallikrein concentrations. Swed Dent J 1996;20:199-207. [26] Reid E, Douglas F, Crow Y, Hollman A, Gibson J. Autosomal dominant juvenile recurrent parotitis. J Med Genet 1998;35:417-9. [27] Kurabuchi S, Hosoi K, Gresik EW. Developmental and androgenic regulation of the immunocytochemical distribution of mKl, a true tissue kallikrein, in the granular convoluted tubule of the mouse submandibular gland. J Histochem Cytochem2002;50:135-45. [28] Clements JA, Matheson BA, Funder JW. Tissue-specific developmental expression of the kallikrein gene family in the rat. JBiol Chem 1990;265:1077-81.

[9] Chirte W , Premchandra DJ. Recurrent parotitis. Arch Dis Childhood 1997;77:359-63. [10] Konno A, Ito E. A study on the pathogenesis of recurrent parotitis in childhood. Ann Otol Rhinol Laryngol 1979;88 Suppl:l-20. [11] Friis B, Karup-Pedersen F, Schiodt M. Immunological studies in two children with recurrent parotitis. Acta Paediatr Scand 1983;72:265-8. [12] Marsman WA, Sukhai RN. Recurrent parotitis and isolated IgG3 subclass deficiency. Eur J Pediatr 1999;158:684. [13] Bailey H. Congenital parotid sialectasia. J Int College Surg 1945;8:109-14. [14] Pubols MH, Bartelt DC, Greene LJ. Trypsin inhibitor from human pancreas and pancreatic juice. J Bid Chem 1974;249:2235-42. [15] Witt H, Luck W, Hennies HC, Classen M, Kage A, Lass U, et al. Mutations in gene encoding the serine protease inhibitor, Kazal type 1 are associated with chronic pancreatitis. Nat Gen 2000;25:213-6. [16] Pfiitzer RH, Barmada MM, Bmnskill AP, Finch R, Hart PS, Neoptolemos J, et al. SPINKl/PSTI polymorphisms act as disease modifiers in familial and idiopathic chronic pancreatitis. Gastroenterology 2000;199:615-23. [17] Chen JM, Mercier B, Audrezet MP, Raguenes O, Quere I, Ferec C. Mutations of the pancreatic secretory trypsin inhibitor (PSTI) gene in idiopathic chronic pancreatitis. Gastroenterology 2001; 120:1061-2. [18] Drenth JP, te Morsche R, Jansen JB. Mutations in serine protease inhibitor Kazal type 1 are strongly associated with chronic pancreatitis. Gut 2002;50:687-92. [19] Hochstrasser K, Wachter B, Reisinger PW, Greim M, Albrecht GJ, Gebhard W. Amino acid sequences of mammalian kazaltype proteinase inhibitors from salivary glands. Com Biochem PhysiolB 1993;106:103-8.