Rheumatol Int (2012) 32:3985–3992

DOI 10.1007/s00296-011-2343-7

ORIGINAL ARTICLE

Four cases of ankylosing spondylitis in medieval skeletal series from Croatia

Mario Klaus · Mario Novak · Mislav Bavka

Received: 9 September 2011 / Accepted: 20 December 2011 / Published online: 31 December 2011© Springer-Verlag 2011

Abstract Osteological changes consistent with ankylos-ing spondylitis were observed in three males and onefemale skeleton recovered from four medieval sites—Velim, Koprivno, Buje, and Rijeka—all situated on Croa-tia’s eastern Adriatic coast and its immediate hinterland.The skeletons present changes in the spine, ribs, sacrum,and innominates that are typical of ankylosing spondylitisthat is a progressive, inXammatory disease of connectivetissue calciWcation. The disease most commonly aVects thesacroiliac joints, the joints of the spine, and the costoverte-bral joints. In the Wnal stages of the disease, the vertebralbodies remodel and together with the associated syndesmo-phytes form a continuous, smooth bone surface that issometimes referred to as “bamboo spine.” The prevalenceof this disorder in the analyzed Croatian samples is 4/303 or1.3% and thus corresponds with frequencies recorded inmodern European populations. DiVerential diagnosis rulesout the possibility of DISH, rheumatoid arthritis, and melo-rheostosis. These are the Wrst cases of ankylosing spondyli-tis identiWed in Croatian archaeological series.

Keywords Paleopathology · Ankylosing spondylitis · Medieval period · Croatia

Introduction

Ankylosing spondylitis (AS) is a progressive, non-infec-tious, inXammatory disease of connective tissue calciWca-tion. The joints most commonly aVected are the sacroiliacjoints, the joints of the spine, and the costovertebral joints[1–3]. In the appendicular skeleton the most commonlyaVected joints are the hips and shoulders [4]. Its prevalencevaries by population—from complete absence in Australianaborigines and Black Africans to 4.2% in adult male HaisaIndians. At present, Caucasian men appear to be the mostpredilected for the disease [1]. The disease is more com-mon in men with a 5:1 [5], and even 9:1 [1] male-to-femaleratio cited. The spinal lesion characteristic for AS is thesyndesmophyte, a bony outgrowth at the margin of the jointthat is vertically oriented—in contrast to osteoarthriticosteophytes that are horizontally oriented [6]. The diseaseusually begins in the sacroiliac joints and lumbar spine andprogressively involves the thoracic vertebrae and their cos-tovertebral joints. In later stages, the cervical spine mayalso be aVected [2]. In the Wnal stages of AS, the vertebralbodies remodel and lose their normal shape, so thattogether with the associated syndesmophytes, they form acontinuous, smooth bone surface that is sometimes referredto as “bamboo spine.” The disease is known to be highlyheritable, that is, 90% of the risk of developing the diseaseis determined genetically, with one-third of this contribu-tion owing to the histocompatibility complex antigen HLA-B27 [7].

Although the disease is rare, cases of AS have beenreported in skeletal series from diVerent parts of the worldand diVerent temporal contexts: ancient Egypt [8], AlaskanEskimo [2], Plains Indians from the USA [9], medievalHungary and Switzerland [10], and eighteenth centurySpain [11].

M. Klaus · M. Novak (&)Department of Archaeology, Croatian Academy of Sciences and Arts, Ante KovabiTa 5, 10 000 Zagreb, Croatiae-mail: mnovak@hazu.hr

M. BavkaDepartment of Diagnostic and Interventional Radiology, Dubrava University Hospital, Zagreb, Croatia

123

3986 Rheumatol Int (2012) 32:3985–3992

This paper describes four probable cases of AS fromfour Croatian medieval sites—Velim, Koprivno, Buje, andRijeka—all located on the eastern Adriatic coast or in itsimmediate hinterland (Fig. 1). The skeletal material recov-ered from these sites is part of the osteological collection ofthe Croatian Academy of Sciences and Arts in Zagreb.These four skeletal series comprise of 303 adult skeletons(individuals older than 15 years of age), and the casesdescribed in this paper represent the only examples of ASrecorded in these samples.

Materials and methods

The Wrst case was recovered from grave number 85 inVelim (Fig. 1), a site that was inhabited from the seventh tothe ninth century AD [12]. The second example was recov-ered from grave number 7 in the Koprivno site (thirteenth–fourteenth century) [13]. The third individual suVeringfrom AS was recovered from the medieval cemetery ofBuje, dated between the eleventh and fourteenth century(VinnjiT, personal communication), while the fourth skele-ton was recovered from grave number 39 in the city ofRijeka. This grave is dated to the Wfteenth century (VinnjiT,personal communication).

All skeletons are well preserved, with only several ribsand some of the small bones from the hands and feet miss-ing. Sex and age at death were estimated using standardanthropological methods. Sex was determined based onpelvic and cranial morphology [14, 15], and discriminantfunctions for the femur [16] and tibia [17] speciWcally

developed for medieval Croatian populations. Age at deathwas determined based on the degree of obliteration of thecranial sutures [18], alterations in the pubic symphysis [19],changes on the sternal ends of the ribs [20, 21], and theappearance of degenerative changes on the articular sur-faces of long bones and vertebrae [22].

The CT analysis was performed at the Department ofDiagnostic and Interventional Radiology, University Hos-pital Dubrava in Zagreb, Croatia. Axial slices wereobtained using 16 £ 0.75 collimation and 0.7-mm recon-struction increment using MDCT unit (Sensation 16; Sie-mens AG Medical Solutions, Erlangen, Germany). Thescanning parameters used were 300 mA and 120 kV. Multi-planar reformatting, maximum intensity projection recon-struction, and volume rendering technique were done on aLeonardo workstation (Siemens AG Medical Solutions,Erlangen, Germany), Aquarius workstation (Terarecon Inc,San Mateo, USA) and OsiriX Imaging Software (Pixmeo,Geneva, Switzerland).

Paleopathological analysis and radiographic imaging

The Velim skeleton belongs to an adult man aged between40 and 50 years. Most of the thoracic and lumbar vertebraeare fused (Figs. 2, 3). T4 through T11 are fused, togetherwith all ribs from both the left and right sides. T12 is free ofthe other vertebrae, but is fused with its corresponding leftrib. L1 through L5 form another fused block. All of the ver-tebrae have slightly narrowed but clearly preserved inter-vertebral spaces. The vertebral bodies have remained intact,although there is considerable development of marginalsyndesmophytes. The sacroiliac joints are fused across thejoint surfaces. Additional pathological changes recorded inthis individual include evidence of slight degenerativeosteoarthritis (osteophytes) on both shoulders and bothknees. There are also three healed antemortem fractures ofthe ribs: on the 7th and 8th right ribs, and on the 8th left rib.

The skeleton from Koprivno belongs to a 45–55-year-oldman. The following vertebrae are fused: C3 through C7, aswell as T3 through T10. In addition, T5 through T10 arefused with the corresponding right ribs, while T4 through T7are fused with the corresponding left ribs (Fig. 4). T11 andT12 are also fused together. T11 has both ribs fused, whileT12 is fused with the left one. L1 through L5 are fused, as isL5 with the sacrum. All of the vertebrae have preserved butnarrowed intervertebral spaces and developed marginal syn-desmophytes. There has been remodeling of all of the lumbarand some of the thoracic vertebral bodies that produced thecharacteristic square proWle typical of ankylosing spondylitis.Interestingly, in this case, the sacrum is not fused with theinnominate bones. This individual exhibits no other patho-logical changes on the skeleton.

Fig. 1 Map of Croatia showing the geographical locations of theanalyzed sites

123

Rheumatol Int (2012) 32:3985–3992 3987

The skeleton from Buje belongs to a 45–55-year-oldman. All cervical vertebrae as well as T1 are missingpostmortem. All vertebrae from T2 through L2 are fusedinto one block that was broken into three blocks duringexcavation (Fig. 5). T6, T7, and T8 are fused with the cor-responding right ribs; T9 and T10 are fused with both leftand right ribs, while T11 and T12 are fused with the corre-sponding left ribs. All of the described vertebrae have pre-served but narrowed intervertebral spaces and developedmarginal syndesmophytes. Unfortunately, both innomi-nates and sacrum are missing postmortem. All of the pres-ent thoracic vertebral bodies (T2 through T12) have beenremodeled producing the characteristic square proWle typ-ical of AS. This individual also presents evidence ofdegenerative osteoarthritis (osteophytes) on both hips andboth knees.

The only female skeleton exhibiting AS in these serieswas recovered in Rijeka. It belongs to a woman over55 years of age. The whole vertebral column consists ofvertebrae fused into three blocks: C1 through T2, T3through T7, and T8 through L5 (Fig. 6). Additionally, T5,T6, and T7 are fused with corresponding right ribs. All of

the vertebrae have narrowed but preserved intervertebralspaces. The vertebral bodies display a considerable devel-opment of marginal syndesmophytes. The sacrum is miss-ing postmortem, but the sacroiliac joint of the rightinnominate displays bony protrusions with recent break-ages, suggesting that it was fused with the sacrum. Thisindividual also presents evidence of degenerative osteoar-thritis (osteophytes and porosity) on the right temporoman-dibular joint, the right hip, and both knees.

DiVerential diagnosis

Morphologically, the changes observed in individuals fromthe four medieval Croatian archaeological sites are charac-teristic of AS. In all cases, the characteristics of the spinalcolumn, ribs, and the lack of peripheral joint involvementmake the diagnosis of AS highly probable. Additionally,the ankylosis of the pelvis and sacrum present in the skele-tons from Velim and Rijeka, and the “squaring” of the ver-tebral bodies in the skeletons from Koprivno and BujeconWrm the diagnosis of this disease.

Fig. 2 Anterior view of the spinal column and pelvic girdle of the man recovered from grave number 85 in Velim (a). Antero-lateral (left side)view of fused thoracic vertebrae (T3–T7) and ribs (b). Anterior view of the fused sacroiliac joints of the same individual (c)

123

3988 Rheumatol Int (2012) 32:3985–3992

DiVerential diagnosis includes diVuse idiopathic skeletalhyperostosis (DISH), rheumatoid arthritis, and melorheos-tosis.

Aufderheide and Rodríguez-Martín [1] and Roberts andManchester [3] state that AS is most frequently confusedwith DISH. DISH results in excessive bone formation at jointmargins and entheses. Most frequently the spine is aVected,although other joints can also be involved. In the spine theabnormal bone formation is created under the anterior longi-tudinal ligament, but in the thoracic vertebrae that formationis limited to the right side, probably because the descendingaorta overlies the left side of the thoracic vertebral bodies [2].Typically, DISH does not involve the diarthrodial vertebralor the costovertebral joints that are aVected in all of the skele-tons from Croatia. Compared with AS, diVerentiating fea-

tures in DISH include the following: signiWcantly lowerinvolvement of sacroiliac joints, a smaller number of verte-brae involved, non-involvement of the apophysial joints, andhorizontally oriented syndesmophytes and enthesopathies[23]. The etiology of DISH remains unknown, but studieshave shown a strong connection with obesity and insulin-independent diabetes mellitus. In bioarchaeological studies,high frequencies of DISH have been associated with seden-tary lifestyles, for instance in clergymen from monasteries[24]. While this lifestyle can theoretically be applied to theindividual from Rijeka, as this site was an urban medievalcommunity, it is not realistic in individuals that inhabitedVelim, Koprivno, or Buje as these were small, rural, settle-ments whose subsistence strategies depended on agricultureand transhumance goat and sheep herding. This hard lifestyle

Fig. 3 CT scan (anterior view) of the spinal column and pelvic girdle of the individual from grave number 85 in Velim (a). CT scan (anterior view)of the spinal column with fused vertebrae and ribs (b). Radiographic image (anterior view) of fused lumbar vertebrae and sacroiliac joints (c)

123

Rheumatol Int (2012) 32:3985–3992 3989

is reXected in the high frequencies of degenerative osteoar-thritis, vertebral osteoarthritis, and Schmorl`s nodes recordedin these sites [25, 26].

The etiology of rheumatoid arthritis is unknown, butstudies have shown a genetic predisposition for the disease[27]. Its main characteristic is the destruction of joint tissue.The underlying subchondral bone may be destroyed, andcysts may form in the spongy bone below the subchondralbone surface and adjacent to the joint margins. In contrastto AS, the disease is more common in women. Rheumatoidarthritis usually involves multiple joints and is frequentlysymmetrical. The joints of the hands are most frequently,and usually the Wrst, to be aVected. Other commonlyinvolved joints are the knee, the temporomandibular joint,the carpal joints, the shoulder, and the elbow [2, 3]. Thejoints of the axial skeleton are much less commonlyinvolved. If the spine is aVected, the changes are usuallymost pronounced in the cervical region, while the thoracicand lumbar vertebrae, as well as the sacroiliac joints areusually spared. Ankylosis is much less common in rheuma-toid arthritis than in the spondyloarthropathies such as AS.Compared to AS, the main diVerentiating features in rheu-matoid arthritis include a signiWcantly higher involvementof the small joints of the hand—none of which are aVectedin the analyzed skeletons and the destruction of joint tissue(compared to new bone formation that is noted in all fourskeletons from Croatia and is typical for AS).

Melorheostosis is a rare bony dysplasia that aVects theskeleton and the adjacent soft tissue leading to deformitiesand bony ankylosis—most frequently in the lower andupper limbs, and relatively rarely in the axial skeleton[28–30]. The typical radiographic features are a thickeningof the bone cortex resembling Xowing candle wax, sub-periosteal bone, and streaky endosteal bone formation in

the diaphyseal and epiphyseal areas with a sclerotomal pat-tern mainly involving the appendicular skeleton [31]. Untilrecently the etiology of the disease was unknown, but todayit has been established that the disorder is due to a loss-of-function mutation in the LEMD 3 gene (also known asMAN 1), which encodes an inner nuclear membrane pro-tein [32]. It can aVect children and adults and has an equalsex distribution [28]. As mentioned earlier, melorheostosisrarely involves the vertebral column and ribs—the areas ofthe skeleton most aVected by changes in the analyzed fourskeletons from Croatia. As none of the limbs in these skele-tons are aVected, and there are no traces of dripping waxchanges characteristic of the disease, it seems highlyunlikely that any of the studied individual suVered frommelorheostosis.

On the assumption that the skeletal changes described inthis report are the result of AS, the frequency of this diseasein the analyzed Croatian skeletal series is 4/303 or 1.3%(subadult skeletons were not included in this analysis as ASdoes not develop in subadults). It is important to point out,however, that the prevalence of AS at death does not neces-sarily correspond to its prevalence in the general populationas it is possible that individuals who may have developedthe disease died as young adults. Therefore, our data areprovisional and need to be taken with caution until largerskeletal databases for medieval Croatian populations areestablished, or new analytical methods are developed. Therecorded prevalence is, however, broadly comparable withfrequencies reported for modern Turks (0.25%) [33], Ital-ians (0.4%) [34], Germans (0.55%) [35], and Norwegians(1.1–1.4%) [36]. Interestingly, these frequencies suggest adistinct north–south distinction, possibly due to diVerentHLA-B27 prevalence in northern and southern countries[37] that the medieval Croatian series do not adhere to. The

Fig. 4 Anterior view of the spinal column of the man recovered from grave number 7 in Koprivno (a). CT scan (antero-lateral view, right side)of fused vertebrae and ribs (b). CT scan (antero-lateral view, left side) of fused vertebrae and ribs (c)

123

3990 Rheumatol Int (2012) 32:3985–3992

presented data does, however, suggest that the preva-lence of AS has not changed signiWcantly through time.Obviously, additional studies conducted on large well-deW-ned skeletal samples are needed in order to verify thishypothesis.

Conclusion

A disease such as ankylosing spondylitis inevitablyaVects the quality of life of the person suVering from it.Low back pain, weight loss, fever, limited chest expan-

sion, and in the Wnal stages of the disease immobility aresome of the symptoms associated with the disease. Allindividuals described in this report lived into the Wfthdecade of their lives and were, without a doubt, at thetime of their death, and possibly for a relatively longperiod before it—completely disabled. The fact that theysurvived suYciently long for the disease to express itselfin such a developed state, while being depended on thehelp and care of other members of their community, sug-gests that some medieval populations from Croatia tookcare of the disabled and chronically ill persons livingwith them.

Fig. 5 CT scan (anterior view) of the spinal column of the individual from Buje (a). Radiographic image (lateral view, left side) of fused vertebraeand ribs (b)

123

Rheumatol Int (2012) 32:3985–3992 3991

Acknowledgments The authors would like to thank the Editor-in-Chief of Rheumatology International, Ernst-Martin Lemmel, and theanonymous reviewers for constructive suggestions and comments thatsubstantially improved the manuscript. This study was Wnancially sup-ported by a grant from the Ministry of Science, Education and Sports ofthe Republic of Croatia (Grant number 101–197–0677–0670).

ConXict of interest The authors declare that they have no conXict ofinterest.

References

1. Aufderheide AC, Rodríguez-Martín C (1998) The Cambridgeencyclopedia of human paleopathology. Cambridge UniversityPress, Cambridge

2. Ortner DJ (2003) IdentiWcation of pathological conditions inhuman skeletal remains, 2nd edn. Academic Press, San Diego

3. Roberts C, Manchester K (2005) The archaeology of disease, 3rdedn. Cornell University Press, Ithaca

4. Van der Linden S, Van der Heijde D (2001) Spondyloarthropa-thies. Ankylosing spondylitis. In: Ruddy S, Harris E Jr, Sledge C(eds) Kelley’s textbook of rheumatology, 6th edn. Saunders, Phil-adelphia, pp 1039–1053

5. Resnick D, Niwayama G (1988) Diagnosis of bone and joint dis-orders, 2nd edn. Saunders, Philadelphia

6. Khan M (1998) Spondyloarthropathies. Ankylosing spondylitis:clinical features. In: Klippel J, Dieppe P (eds) Rheumatology, 2ndedn. Mosby, London, pp 1–10

7. Brown MA (2008) Breakthroughs in genetic studies of ankylosingspondylitis. Rheumatology 47:132–137

8. Feldtkeller E, Lemmel EM, Russell AS (2002) Ankylosing spon-dylitis in the pharaohs of ancient Egypt. Rheumatol Int 23:1–5

9. Bass WM, Gregg JB, Provost PE (1974) Ankylosing spondylitis(Marie Strumpel disease) in historic and prehistoric NorthernPlains Indians. Plains Anthropol 19:303–305

10. PálW G, Panuel M, Gyetvai A, Molnár E, Bende L, Dutour O(1996) Advanced-stage ankylosing spondylitis in a subject in the8th century. J Radiol 77:283–285

11. Gómez Bellard F, Sánchez Sánchez JA (1989) Spondylarthriteankylosante: un cas complet. J Paleopathol (Monographic Publi-cations) 1:117–118

12. JuriT R (2004) Velim–Velintak. Hrv arheol god 1:201–20313. Gjuranin H (2005) Zantitna arheolonka istraqivanja u selu

Koprivno sjeveroistobno od Klisa. Starohrv prosvj 32:163–19314. Phenice TW (1969) A newly developed visual method of sexing

the os pubis. Am J Phys Anthropol 30:297–30115. Krogman WM, Ãocan MY (1986) The human skeleton in forensic

medicine. Thomas, SpringWeld16. Klaus M (1997) Discriminant function sexing of fragmentary and

complete femora from medieval sites in continental Croatia.Opusc Archaeol 21:167–175

17. Klaus M, TomibiT L (2005) Discriminant function sexing of frag-mentary and complete tibiae from medieval Croatian sites. Foren-sic Sci Int 147:147–152

18. Meindl RS, Lovejoy CO (1985) Ectocranial suture closure: arevised method for the determination of skeletal age at death basedon the lateral-anterior sutures. Am J Phys Anthropol 68:57–66

19. Brooks S, Suchey JM (1990) Skeletal age determination based onthe os pubis: a comparison of the Acsádi-Nemeskéri and Suchey-Brooks methods. Hum Evol 5:227–238

20. Ãocan MY, Loth SR, Wright RK (1984) Age estimation from therib by phase analysis: white males. J Forensic Sci 29:1094–1104

21. Ãocan MY, Loth SR, Wright RK (1985) Age estimation from therib by phase analysis: white females. J Forensic Sci 30:853–863

22. PfeiVer S (1991) Estimation of age at death. In: PfeiVer S,Williamson SR (eds) An investigation of a military cemetery fromthe war of 1812. Dundurn Press, Toronto, pp 167–175

23. Zimmerman M, Kelley M (1982) Atlas of human paleopathology.Praeger, New York

24. Verlaan JJ, Oner FC, Maat GJ (2007) DiVuse idiopathic skeletalhyperostosis in ancient clergymen. Eur Spine J 16:1129–1135

25. Klaus M (2006) Bioarheologija: DemograWja, zdravlje, traumei prehrana starohrvatskih populacija. Kkolska knjiga, Zagreb

26. Novak M, Klaus M (2011) Vertebral pathologies in two earlymodern period (16th–19th century) populations from Croatia. AmJ Phys Anthropol 145:270–281

27. Fox D (2001) Etiology and pathogenesis of rheumatoid arthritis.In: Koopman W (ed) Arthritis and allied conditions. LippincottWilliams and Wilkins, Philadelphia, pp 1085–1102

28. Campbell CJ, Papademetriou T, BonWglio M (1968) Melorheosto-sis. J Bone Jt Surg (American Volume) 50A:1281–1304

29. Kalbermatten NT, Vock P, Rüfenacht D, Anderson SE (2001)Progressive melorheostosis in the peripheral and axial skeletonwith associated vascular malformations: imaging Wndings overthree decades. Skelet Radiol 30:48–52

30. Ethunandan M, Khosla N, Tilley E, Webb A (2004) Melorheosto-sis involving the craniofacial skeleton. J Craniofac Surg 15:1062–1065

Fig. 6 Lateral view (left side) of fused lumbar vertebrae of the womanrecovered from grave number 39 in Rijeka

123

3992 Rheumatol Int (2012) 32:3985–3992

31. Singh R, Singh Z, Bala R, Rana P, Sangwan SS (2010) An unusualcase of sciatic neuropraxia due to melorheostosis. Jt Bone Spine77:614–615

32. Hellemans J, Preobrazhenska O, Willaert A, Debeer P, VerdonkPCM, Costa T, Janssens K, Menten B, Van Roy N, Vermeulen SJT,Savarirayan R, Van Hul W, Vanhoenacker F, Huylebroeck D, DePaepe A, Naeyaert JM, Vandesompele J, Speleman F, VerschuerenK, Coucke PJ, Mortier GR (2004) Loss-of-function mutations inLEMD3 result in osteopoikilosis, Buschke-OllendorV syndrome andmelorheostosis. Nat Genet 11:1213–1218

33. Karkucak M, Cakirbay H, Capkin E, Topbas M, Guler M, TosunM, Baltaci D (2011) The prevalence of ankylosing spondylitis inthe eastern Black Sea region of Turkey. Eur J Genet Med 8:40–45

34. De Angelis R, SalaY F, Grassi W (2007) Prevalence of spond-yloarthropathies in an Italian population sample: a regional com-munity-based study. Scand J Rheumatol 36:14–21

35. Braun J, Listing J, Sieper J (2005) Overestimation of the preva-lence of ankylosing spondylitis in the Berlin study: comment onthe article by Braun et al. Arthritis Rheum 52:4049–4050

36. Gran JT, Husby G, Hordvik M (1985) Prevalence of ankylosingspondylitis in males and females in a young middle-aged popula-tion of Tromso, northern Norway. Annal Rheum Dis 44:359–367

37. Khan MA (1995) HLA-B27 and its subtypes in world populations.Curr Opin Rheumatol 7:263–269

123