Mumps outbreak in the Federation of Bosnia and Herzegovina with large cohorts of susceptibles and genetically diverse strains of genotype G , Bosnia and Herzegovina , December 2010 to September 2012

M Hukic (mirsadahukic@yahoo.com)1, A Hajdarpasic2, J Ravlija3, Z Ler3, R Baljic4, A Dedeic Ljubovic1, A Moro1, I Salimović-Besic1, A Sausy5, C P Muller5, J M Hübschen5 1. Institute of Clinical Microbiology, Clinical Centre University of Sarajevo, Sarajevo, Bosnia and Herzegovina 2. Centre for Genetics, Faculty of Medicine, University of Sarajevo, Laboratory for Molecular Medicine, Sarajevo, Bosnia and Herzegovina 3. Institute of Public Health of Federation Bosnia and Herzegovina, Sarajevo, Bosnia and Herzegovina 4. Clinic for Infectious Diseases, Clinical Centre University of Sarajevo, Sarajevo, Bosnia and Herzegovina 5. Institute of Immunology, World Health Organization European Regional Reference Laboratory for Measles and Rubella, Centre de Recherche Public de la Santé /Laboratoire National de Santé, Luxembourg, Luxembourg


Introduction
Mumps is an infection with acute onset of unilateral or bilateral self-limited swelling of the parotid or other salivary glands, which lasts at least two days.
The incubation period ranges from 12 to 25 days after exposure to the virus.Transmission from one person to another mostly occurs before and within five days of parotitis onset [1].The disease is caused by mumps virus (MuV) and can be associated with complications such as aseptic meningitis, encephalitis, hearing loss, orchitis, oophoritis, mastitis or pancreatitis [1].There is no specific treatment and vaccination is the only effective measure to prevent the disease.The presence of specific immunoglobulin M (IgM) in the serum confirms recent mumps infection.Reverse transcription-polymerase chain reaction (RT-PCR) is becoming increasingly popular for laboratory investigation of clinically suspected cases and provides also genotype information of circulating strains [1].Currently, 12 different genotypes of mumps virus are recognised and genotype G seems to be the most prevalent genotype in Europe [2].
The combined measles, mumps and rubella (MMR) vaccine is in use in Bosnia and Herzegovina (B&H) since 1980 [3,4].From 1980 to 1992 one dose of vaccine was given at 12 months of age and if this opportunity was missed, before the age of 14 years.The Republic of Srpska (RS) and some cantons of the second main entity of B&H, the Federation of Bosnia and Herzegovina (FB&H) (Bosnian Podrinje, Central Bosnia, Herzegovina-Neretva, Sarajevo, Tuzla, Una-Sana, Zenica-Doboj) continued to use the one-dose schedule until 2001.The remaining cantons (Canton 10,Posavina,West Herzegovina), two of which border Croatia, adopted in 1992 a two-dose schedule with the first dose given at 12 months and the second dose at seven years and no later than 14 years of age.In such cantons, individuals born between 1981 and 1992 were given an opportunity for a second dose of vaccine.Since 2001, the two-dose MMR vaccination schedule is implemented throughout B&H [5], whereby cantons, which had a single dose schedule until 2001, offer two doses to individuals born from 2001 onwards.
The Public Health Institute of FB&H manages the national immunisation programme in FB&H and each canton is responsible for its implementation at the level of local health centres.Mumps is one of ten infectious diseases against which vaccination is compulsory and free of charge in FB&H.The immunisation status is checked by looking at the vaccination records when children enter kindergarten, when they start primary school and again during periodic revisions in preparation of the yearly immunisation plan and/or supplementary immunisation activities.Incompletely immunised children are vaccinated in catch-up campaigns.
The war between 1992 and 1995 disrupted MMR vaccination across the entire territory of B&H and deficiencies in the programme persisted up to several years after the war [3].For instance, periodic shortages of vaccine supply and interruptions in the cold chain were reported.In addition, many refugees did not have any medical records.Before the introduction of vaccination against mumps, outbreaks occurred every three to four years and later every seven to eight years [4].Between December 2010 and September 2012 a higher disease incidence with peaks in April 2011 and January 2012 was observed.This report describes this epidemic and investigates its causes.

Methods
Any patient with acute onset of unilateral or bilateral tender, self-limited swelling of the parotid or other salivary gland, lasting two or more days and without other apparent cause was considered a clinically suspected mumps case [6].Suspected mumps cases are normally reported by medical practitioners to the Institute for Public Health of FB&H, which investigates outbreaks and reports to the Ministry of Health.Epidemiological data of 7,895 clinically suspected mumps cases reported between December 2010 and September 2012 were collected by the Institute for Public Health.The vaccination status of patients less than 18 years-old was checked in their medical records kept at the local health centre.For older patients the immunisation status was either checked in their medical records or self-reported.

Serological study
Serum was collected from 221 individuals between four and 64 years of age (mean: 21 years) who were clinically suspected mumps cases (n=71) or asymptomatic contacts to mumps cases (n=150) from the cantons of Sarajevo, Central Bosnia and Zenica-Doboj.All sera were tested with the Siemens Enzygnost Anti-Parotitis Virus kits for mumps-specific IgG and IgM antibodies.

Hospitalisation and complications
Between April 2011 and September 2012, 105 clinically suspected mumps cases from Central Bosnia and Sarajevo cantons were hospitalised at the Clinical Centre of the University of Sarajevo.Throat swabs were collected from all 105 patients and their medical records were checked.

Molecular and phylogenetic analysis of the mumps virus outbreak strains
RNA was extracted from the throat swabs of the 105 patients hospitalised at the Clinical Centre of the University of Sarajevo using the QIAamp Viral RNA Mini kit (Qiagen, Germany) according to the manufacturer's instructions.
For reverse transcription, 5 µl of RNA and random hexamers were used in a total volume of 20 µl.PCRs were performed to amplify a genetic region comprising the small hydrophobic (SH) gene using previously described primers [7].The genetic sequence was either obtained by one PCR (first PCR) or by two consecutive PCRs, whereby the first PCR was followed by a nested PCR.Starting material included 1 µl of cDNA for the first PCR, or 1 µl of the first PCR-product, for the nested PCR.The amplification steps consisted of an initial incubation for 5 minutes at 94°C, followed by 39 cycles of 94°C for 30 seconds, 56 or 52°C (first PCR) and 58 or 52°C (nested PCR) for 1 minute and 72°C for 1 minute, followed by a final extension at 72°C for 10 minutes.PCR products were analysed in a 1.5% agarose gel and only samples negative after the first PCR were further amplified by nested PCR.PCR-positive samples were sequenced using nested primers and the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, USA).
Sequences were analysed by SeqScape Software v2.5 (Applied Biosystems, USA), BioEdit version 7.0.9.0 [8] and molecular evolutionary genetics analysis (MEGA)4 software [9].Neighbour-joining phylogenetic trees based on the Kimura 2-parameter model were constructed using 316 nucleotide (nt) sequences comprising the complete SH gene.The recommended set of reference sequences [2] and some identical or similar sequences obtained by basic local alignment search tool (BLAST) were included in the phylogenetic analysis.Nt sequences of the SH gene were translated into the corresponding amino acid sequences with BioEdit.Statistical tests for significance were done using SPSS 15.0 (IBM, USA).

Outbreak description
In the period from December 2010 to September 2012, a mumps outbreak including 7,895 cases was observed in FB&H.The outbreak consisted of two distinct epidemic waves, with one epidemic peak observed in April 2011 (n=1,240 cases) and another in January 2012 (n=509) (Figure 1).
Overall, the majority of cases reported (82%: 6,481/7,895) were located in three cantons: Central Bosnia (n=2,434 cases), Zenica-Doboj (n=2,215) and Sarajevo (n=1,832).These cantons had a one dose vaccination schedule until 2001.In contrast, Canton 10, and the cantons of Posavina and West Herzegovina, all, which had a two dose schedule since 1992, included only a total of 59 reported cases.
During the first wave of the outbreak, between December 2010 and the end of October 2011, 5,677 mumps cases were reported, mainly in the cantons of Central Bosnia, Zenica-Doboj, Sarajevo and Herzegovina-Neretva (n=5,534 cases in total for the four cantons) [3].The second epidemic wave from November 2011 to September 2012, accounted for an additional 2,218 cases.During the second wave, cases continued to occur in the four previously most-affected cantons, but several hundred cases were also reported from Una-Sana (n=355) and Tuzla (n=450) cantons.Una-Sana and Tuzla cantons, which reported only sporadic cases during the first wave in 2011, observed the highest disease incidence at the beginning of 2012, during the second epidemic wave.
Considering the whole outbreak, the age of the cases ranged from less than one year to 64 years with a median age of 19 years.The age group comprising 15 to 19 year-olds was most affected during the first epidemic wave in 2011, while the highest case numbers were observed in 20 to 29 year-olds during the second wave (Figure 1).For the entire outbreak, nearly 82% (6,455/7,895) of the cases were older than 14 years (Figures 1 and 2).More males (63%: 5,005/7,895) than females were affected overall and in each age group, except for the one to four year-olds (Figure 2).Among the 20 to 29 year-old cases there were even more than twice as many males than females (Figure 2).Mumps cases reported during the outbreak were mostly unvaccinated (39%: 3,050/7,895) or had an unknown vaccination status (31%: 2,426/7,895).The other cases had been vaccinated with either one dose (15%: 1,217/7,895) or with two doses of mumps-containing vaccine (15%: 1,202/7,895).

Serology results
Among 71 patients with clinical symptoms consistent with mumps, 57 (80%) were IgM positive and IgG negative, two (3%) were positive for both IgM and IgG antibodies and 12 (17%) were only IgG positive.Among the 150 contacts without clinical signs or symptoms, five (3%) were IgM positive and IgG negative, 100 (67%) were only IgG positive and 45 (30%) were negative for both IgM and IgG antibodies.A positive serological test for mumps-specific IgM antibodies, with or without the presence of IgG, confirms recent infection.The presence of mumps-specific IgG antibodies indicates previous contact with mumps virus either by vaccination or natural infection.

Discussion
Between December 2010 and September 2012, 7,895 mumps cases were registered in FB&H and about 7,700 additional cases in RS, leading to approximately 15,600 cases across B&H.This was by far the largest mumps outbreak in the country since the introduction of the vaccine in 1980 [3] and it was also one of the largest outbreaks reported from Europe since the period between 2008 and 2009 [10].
The outbreak affected mainly regions of B&H with a single dose schedule until 2001, which is in line with the down to 60% long-term population-based effectiveness of a single dose of mumps vaccine [11].
During the 1992 to 1995 war and several years after, immunisation activities were irregular [3] with large cohorts of susceptibles, especially in regions where only a single vaccination opportunity was provided.This also explains why most patients involved in the recent mumps outbreak (72%: 5,710/7,895) belonged to the age groups 15 to 19 and 20 to 29 years (born 1981-1997) and why the majority of the patients (69%: 5,476/7,895) reported to be unvaccinated or had an unknown vaccination status.The breakdown of community immunity as a result of young adults moving on to new school or work environments may have compounded the incidence in this age group.A high mumps incidence among adolescents and young adults who were unvaccinated or received only a single dose has been reported from several other European countries [10,[12][13][14].
Also waning of protective levels of antibodies [10,15] even after two doses may have played a role.Since mumps outbreaks seem to occur even in highly vaccinated communities, a third dose of mumps-containing vaccine as a booster later in life is currently under discussion [14][15][16][17].Whether the robustness and persistence of the immune response to the various vaccine strains used in B&H during the past years (Institute for Immunology, Zagreb, Croatia (L-Zagreb strain, 1980-1999); GlaxoSmithKline Beecham, Belgium (RIT 4385 strain,1999-2012); Aventis Pasteur, France (Urabe strain, 2007 only) [4]) differed as was observed previously [11,18] deserves further attention.In response to the outbreak, all children less than 14 years of age who had received less than two doses of MMR vaccine were vaccinated, mumps cases were isolated, some schools with large numbers of cases were closed and the citizens were alerted via the media.However, the fact that each of the ten cantons of FB&H has its own health system, in addition to the anti-vaccination propaganda during the past few years [19,20] probably contributed to the overall rather weak and slow response to the outbreak.
Currently no mumps seroprevalence data are available for FB&H.The investigation of sera from a small number (n=150) of asymptomatic mumps patient contacts during the recent outbreak showed that about one third did not have specific IgG antibodies and were considered to be susceptible.A thorough seroprevalence study to identify main groups of susceptibles for targeted supplementary immunisation activities is warranted.Among the 71 clinically suspected mumps patients, 12 (17%) had IgG and no IgM antibodies, indicating that at least some of these patients may have been reinfected with mumps virus or infected after vaccination.In such cases virus detection should be attempted [21,22], but no samples besides serum were available from these patients.
The typical male complications may explain why more male patients in the 15 to 29 year age bracket were recorded.In fact, orchitis with or without meningitis was the most common complication in hospitalised patients.In contrast to a recently published study [21], we did not find any significant difference in disease severity between vaccinated and unvaccinated hospitalised patients.This may be due to low numbers  Seven sequence variants were different from any of the sequences on GenBank.Since only 57 sequences from only two cantons were obtained while approximately 15,600 mumps cases were reported from throughout B&H, it is likely that the real sequence diversity is underestimated.Nevertheless, this diversity is already considerably higher than during a one-year outbreak in the US with 3,500 cases where 221 sequences differed by no more than a single nt [21].Since at least cluster 1, 3 and 4 sequences were present already by the beginning of 2011 and genetically diverse strains were found in five different locations, the most probable scenario is that different variants were present in B&H already before the outbreak and/or correspond to multiple independent transmission chains.Some variation seems to have been generated during the epidemic since within clusters 2, 3 and 4 the strain collected at the latest time point was always the most diverse.
The differing strain from cluster 1 and all sequences from clusters 3 and 4 showed changes in the predicted amino acid sequence of the SH protein.Of note, the cluster 3 strains had a leucine to methionine substitution in position 20 of the predicted SH protein, which could potentially result in an N-terminal truncation.As the SH protein has been reported to inhibit tumour necrosis factor alpha-mediated apoptosis [23][24][25], it is conceivable that a truncated SH protein may have lost the ability to inhibit programmed cell death leading potentially to more severe clinical manifestations.In fact, six of the seven patients with the predicted leucine to methionine substitution in position 20 were hospitalised due to disease complications (4 had orchitis, 1 had orchitis and meningitis and 1 had meningitis).Further investigations are clearly warranted to elucidate the significance of the methionine substitution as well as the other predicted amino acid mutations.
It has been suggested that the gradual antigenic evolution of mumps viruses may undermine vaccine effectiveness [26].A recent study showed indeed that the genotype A strain Jeryl-Lynn induced less neutralising antibodies against the common genotype G virus than against the vaccine strain [27].The 13 patients who were infected with the genetically diverse variants in our study were between 18 and 43 years-old (mean: 25.2 years) and most of them reported that they had received two doses of vaccine (54%: 7/13) or that the vaccination history was unknown (31%: 4/13).Thus it is conceivable that waning of antibodies, vaccine quality issues, intensive exposure [21] or potentially an even lower neutralisation capacity against the present genotype G variants played a role in the outbreak.
In conclusion, we identified failures to vaccinate during and after the war and single dose immunisation policies in some cantons as important causes for the large and persistent mumps outbreak.Those may have been compounded by differential responses to vaccines and vaccine strains, waning of antibodies and potentially by the genetically diverse variants of genotype G. Molecular analysis of the mumps strains involved, revealed a level of diversity in the virus suggestive of several transmission chains, possibly as a result of long-term endemic circulation of mumps viruses in FB&H.Our study identified large cohorts of susceptibles and emphasises the need for supplementary immunisation activities in particular among adolescents and young adults who have received less than two vaccine doses.

Figure 1 4
Figure 1Number of mumps cases, by age group, in an outbreak presenting two epidemic waves in the Federation ofBosnia and  Herzegovina, Bosnia and Herzegovina, December 2010-September 2012 (n=7,895)

Figure 2
Figure 2Number of mumps cases by age and sex in an outbreak in the Federation ofBosnia and Herzegovina, Bosnia and  Herzegovina, December 2010-September 2012 (n=7,895)

Figure 3
Figure 3Complications observed among mumps patients hospitalised at the Clinical Centre of the University of Sarajevo according to age, mumps outbreak in the Federation ofBosnia and Herzegovina, Bosnia and Herzegovina, December 2010-September  2012 (n=105)

Table
Characteristics of mumps patients hospitalised at the Clinical Centre of the University of Sarajevo during a mumps outbreak in the Federation ofBosnia and Herzegovina, Bosnia and Herzegovina, December 2010-September 2012 (n=105)