Skeletal Manifestation TBP in Human Morden Remains
Skeletal Manifestation TBP in Human Morden Remains
ORIGINAL COMMUNICATION
Paleopathologists study the presence of diseases in the past and as such have
a vast knowledge of skeletal changes associated with different conditions.
Tuberculosis is one of the most studied diseases and still remains a major
health problem today. Its manifestations in past populations have been
extensively described, but less is known about its bony involvement in the
post-antibiotic era. The aim of this study was to assess the frequency and man-
ifestations of skeletal lesions in the post-antibiotic era in a South African sam-
ple and compare it to that found before the introduction of antibiotics.
Skeletons of 205 individuals from modern skeletal collections and who are
known to have died from TB were assessed. It was found that 39.2% of all
individuals dying in the post-antibiotic era showed skeletal changes that could
be associated with TB, while another 27.5% showed nonspecific changes. The
highest incidences were found in individuals who died after 1985, when co-
infection with HIV and drug resistance became common. While, as expected,
vertebral and rib changes were the most common, the number of individuals
who showed changes to the skull, and specifically intracranially, was surprising.
These could most probably be associated with TB meningitis, although this spe-
cific cause of death was noted in only a few individuals. It seems that individu-
als may be living longer as a result of long-term antibiotic use, leaving more
time for lesions to develop. Clin. Anat. 00:000–000, 2016. VC 2016 Wiley Periodicals, Inc.
C
V 2016 Wiley Periodicals, Inc.
2 Steyn and Buskes
single most important risk factor for contracting TB. are now living longer with their TB as a result of
HIV is also frequently the cause of progression of dor- receiving antibiotic treatment, thus allowing more
mant disease into full-blown TB and causes rapid pro- time for skeletal lesions to develop. Similar to what
gression of the disease. HIV-related TB is the most was found in another study by Holloway et al. (2011),
common in sub-Saharan countries, making up 79% of it was also found that rib lesions are becoming more
such cases known world-wide (Naidoo et al., 2010). common, while spinal lesions seem to be decreasing.
TB is now also the most common cause of HIV-related The Holloway et al. (2011) study, however, was a
deaths, and occurs in at least one-third of HIV- meta-analysis of cases reported in the literature and
positive people in sub-Saharan Africa (Kanabus, did not include any remains from post-antibiotic con-
2011; Pillay et al., 2013). texts. It should be kept in mind that there are many
Knowledge of the past incidences and signs of TB factors that may play a role in the expression of this
as well its spread can help us to better understand disease. As is the case with many other diseases and
this condition. Some recent studies seem to point to micro-organisms, Mycobacterium tuberculosis under-
Africa as the original source of this disease (Daniel, goes continuous evolution (Barnes, 2005) and the co-
2006; Wirth et al., 2008; Bos et al., 2014), but in evolution of the disease and it host has most probably
South Africa there is currently no clear skeletal evi- also changed the expression of TB in the skeleton.
dence of its existence before contact with western TB in skeletal remains is most commonly character-
societies, and its emergence and spread seems to be ized by involvement of the vertebral bodies, with
mostly associated with the mining history of the coun- destructive lesions most often occurring in the ver-
try. Of course, the absence of evidence does not nec- tebral bodies of the thoracic and lumbar regions
essarily indicate evidence of absence (Roberts and (Resnick and Niwayama, 1988). The neural arches are
Buikstra, 2003). A possible but unconfirmed case of mainly unaffected (although they may play a role in
precontact TB in South Africa was found in the North- the healing of TB lesions; Holloway et al., 2013), and
West province and dates back to the late 1700s (Pis- the subsequent collapse of the vertebral bodies often
torius et al., 1998). Of the five archaeological cases leads to kyphosis of the spine commonly referred to
identified with possible TB in the Campbell and Acker- as Pott’s disease (Ortner, 2003; Roberts and Buikstra,
mann (2010) study, three were from post-contact 2003). Virtually any part of the skeleton can be af-
contexts, one was the male mentioned above from fected, but the most commonly affected regions
the North-West Province, while the other one was only include the ribs, sacro-iliac, femoral and knee joints,
indicated to be of rural context. or any other joint. However, as is commonly the prob-
TB and its manifestations in the human skeleton lem in paleopathology, lesions are often nonspecific,
have been studied extensively in past populations, but and it is difficult to find lesions that are truly patho-
much less is known about its expression in the skele- gnomonic of TB, or for that matter, any infectious dis-
ton in the post-antibiotic era and if, in fact, the intro- ease in skeletal remains (Wilbur et al., 2009). In the
duction of antibiotics changed its expression in any study by Steyn et al. (2013), it was found that nearly
way. A notable exception here is the study by Hollo- a quarter of individuals showed nonspecific lesions
way et al. (2013), that showed that bony tuberculous which could not directly be associated with TB, sug-
lesions can heal (especially following antibiotic treat- gesting that widespread skeletal involvement may
ment), and that this may complicate the diagnosis of occur but only later develop in lesions that could spe-
TB in skeletons as, usually, TB is not associated with cifically be associated with TB, or that coexistence
new bone formation but rather bony destruction. with other conditions is common.
Recently, Steyn et al. (2013) reported on trends in The Steyn et al. (2013) study was done on a rela-
bony involvement of skeletons from patients who are tively small sample (n5147) and only included individ-
known to have died from TB. In this study, skeletons uals from the Gauteng (northern) region of South
(n 5 147) from the northern region of South Africa Africa. The purpose of this article is twofold: first to
were studied and comparisons were drawn with report on the findings from an expanded sample
regard to the skeletal manifestations of TB between which also includes specimens from the southern part
individuals who died before the introduction of antibi- (Western Cape) of the country. These skeletons come
otics (estimated to be around 1950), and those who from a different context—although it is a somewhat
died thereafter. A group of individuals dying after sweeping assumption, the skeletons from the north-
1985, broadly judged as the time when co-infection ern region that end up in collections can most prob-
with HIV became common and drug resistance devel- ably often be associated with miners from the
oped, was also included. Gauteng region (thus slightly higher SES); the skele-
In the study by Steyn et al. (2013), it was found tons from the Cape are most probably of people of
that about one third of all skeletons showed signs that very low SES and there are no extensive mining activ-
could be associated with TB—this is much higher than ities in this region. Unfortunately, little is known about
the 2–4% that is generally reported in the literature the background of the individuals in these collections,
(Vigorita, 2008; Holloway et al., 2011), but it should and assumptions on life style can only be made in a
of course be kept in mind that these are people who broad context. It is not sure if this, in itself will make
have died from TB and do not represent the patient a difference, but the Western Cape remains may to
cohort at large. Skeletal involvement increased from some extent reflect the skeletal expression of people
the pre-antibiotic period (around 21.1% of individu- of extremely low SES, possibly plagued by malnutri-
als), to 38.2% and 41.0% in the pre- and post-1985 tion and often with high levels of alcohol consumption
skeletons, respectively. It was proposed that people (Schneider et al., 2007). The second aim of the article
Skeletal Manifestations of TB in Modern Human Remains 3
TABLE 1. Western Cape and Gauteng Combined Sample: Frequency Distribution of Individuals Showing
No Signs of TB (No Skeletal Involvement), Nonspecific Signs of Infectious Disease and Signs of TB
is to reflect on some of the skeletal changes observed antibiotic treatment, those who died between 1950
in the skeletons from the post-antibiotic era and and 1985 assumed to have been treated with antibiot-
assess whether these are generally the same as those ics, and those who died after 1985 when co-infection
observed in pre-antibiotic specimens. with HIV and drug-resistant TB emerged. The date
1950 is used as post-antibiotic as Streptomycin was
introduced in the late 1940s and 1985 is used (some-
MATERIALS AND METHODS what arbitrarily) because the first cases of drug-
resistant TB and co-infection with HIV in South Africa
The study included skeletons from collections at were reported around that time (Steyn et al., 2013).
the University of Pretoria (The Pretoria Bone Collec- Unfortunately, all of the Western Cape skeletons dated
tion; L’Abbe et al., 2005), University of the Witwaters-
from the antibiotic period. Results from the various
rand (Raymond A Dart Collection; Dayal et al., 2009), time periods were then compared by means of a Chi-
University of Stellenbosch (Kirsten Collection), and squared analysis. Assessments were also made with
the University of Cape Town. These collections are regard to where lesions occurred, and if the same pat-
cadaver based, and mostly include individuals who are terns of skeletal involvement occurred in the different
unclaimed and are, therefore, donated to the univer- groups.
sities by various hospitals. They are all known individ-
uals, with known age, sex, and cause of death.
Skeletons of individuals reported to have died from RESULTS
TB were analyzed. The sample from Gauteng included
147 individuals (134 males and 13 females) (Steyn As is often the case with paleopathological investi-
et al., 2013), whereas the Western Cape sample gations, it is very difficult to assign a lesion to a spe-
included 58 individuals and comprised 43 males and cific disease. In this study all cases were donated
15 females. All available individuals were used regard- from hospitals and had firm diagnoses of TB, but
less of their age, sex, and ancestry but were not nevertheless all lesions could not directly be ascribed
included if the remains were very incomplete or frag- to TB as one should keep in mind that these individu-
mentary. As these collections are of modern individu- als may have been malnourished (contributing to sub-
als and cadaver-based, preservation was in most periosteal bone deposition) or may also have suffered
cases excellent. from other diseases including treponemal disease and
The skeletons were systematically analyzed and HIV (with nothing known about the skeletal signs of
signs that could be associated with infectious disease HIV, other than assuming that HIV in itself will prob-
were documented. No bone samples were taken and ably not lead to specific bony changes, but rather that
no destructive analyses were performed. Skeletal ele- any changes may be due to the opportunistic infec-
ments were scored as present or absent. Similar to tions associated with this condition). Unfortunately all
the methodology followed by Steyn et al. (2013), all Western Cape individuals in the collection (n 5 58)
observed lesions were judged as to whether they died after 1950, but were equally distributed between
could be due to TB. As pointed out before, it is very pre- and post-1985 specimens.
difficult to diagnose any specific disease and thus it In the combined sample, with all cases from both
was also difficult to determine whether a lesion is due regions, it can be seen that individuals with skeletal
to TB or not and many lesions could only be classified signs of TB increased from 21.1% in the pre-antibiotic
as nonspecific. These mostly comprised periostitis, period to 28.6% (individuals dying 1950–1985) to
often observed on the long bones. New bone forma- 46.7% (individuals dying after 1985) (Table 1). The
tion/plaques on the visceral surfaces of ribs or rib overall increase seen in skeletal lesions associated with
expansion (Pfeiffer, 1991), lytic lesions of the verte- TB is mostly statistically significant (pre-1950 and
brae and clear joint destruction or other lytic lesions 1950–1985: v2 value 5 0.831, P 5 0.362; pre-1950 and
were classified as lesions being due to TB. TB mostly post-1985: v2 value 5 9.170, P 5 0.0025; 1950–1985
causes destructive lesions, with limited new bone for- and post-1985: v2 value 5 5.091, P 5 0.0241). In addi-
mation (Ortner, 2003), although attempts at healing tion to these, about one quarter of all other individuals
especially in the post-antibiotic era can complicate the had other signs of skeletal involvement, mostly com-
assessment (Holloway et al., 2013). prising nonspecific periostitis on long bones.
The sample was divided into three groups based on This increase in bony involvement is especially
their date of death—those who died before 1950 obvious in the Western Cape group (Table 2), where
(1925–1949) and who probably did not receive any only 17.2% of individuals dying from 1950 to 1985
4 Steyn and Buskes
TABLE 2. Frequency Distribution of Individuals Who Showed no Skeletal Signs of Disease, Those With
Nonspecific Signs of Disease and Those With Skeletal Signs of TB in the Specific Eras in the Western
Cape Sample Only
showed signs of skeletal TB, increasing to 58.6% in cranial vault, of which some may possibly have been
those dying after 1985 (v2 value 5 10.068, P 5 0.0015 the result of other diseases (e.g., lesions resembling a
between Affected vs. Unaffected individuals; v2 5 gummatous lesion possibly resulting from treponemal
13.646; P 5 0.0002 between TB specific lesions vs. disease). A case with extensive periostitis on the out-
unaffected individuals.) In this group, about one third side of the skull on one side and a destructive lytic
of other individuals showed non-specific signs of infec- lesion on the other side is shown in Figure 1. One
tious disease. Overall, taking all individuals with TB in case of intra-cranial involvement appeared to have
the post-1950 era into account, no statistically signifi- originated from a possible otitis media as the destruc-
cant differences were found between the Western tive lesion is situated in the petrous bone (Fig. 2).
Cape and Gauteng groups for TB-specific individuals Quite frequently, though, and observed in 9 cases,
(v2 value 5 0.875; P 5 0.3496). were extensive destruction of the cranial base (Figs. 3
In the Steyn et al. (2013) study, it was proposed and 4; see also Fig. 4 in Steyn et al. 2013). In only
that there may be an increase in rib lesions (occurring one of these cases was the posterior cranial fossa
in 23.8% of all individuals; in 20% of post antibiotic involved, whereas in the others the middle and ante-
individuals but becoming more frequent if only individ- rior cranial fossae were affected. The severe and
uals with skeletal changes are taken into account) and destructive nature of these lesions is clear from these
a decrease in vertebral lesions (occurring in 8.8% of two images.
all individuals; in 4.2% of post-antibiotic individuals) Surprisingly, a high frequency (n 5 6; 11.5% out of
with time. This trend was not so clear in the Cape skel- the 52 with mandibles) of mandibular condyle des-
etons where overall 27.6% of individuals had rib truction was observed in one or both sides. In some
involvement and 13.8% had spinal involvement. of these cases, the individual was edentulous which
These individuals were all post-antibiotic and this may have contributed to the destruction of the man-
observation needs to be followed up with larger dibular condyle, and it also seems unlikely that TB of
samples. the temporomandibular joint (TMJ) would occur on
An unexpected observation was the frequent both sides. In the case shown in Figure 5 (a 33-year-
involvement of the skull, with overall 10 individuals old male with teeth), however, it seems likely that the
(7.0%; 142 individuals with skulls) from Gauteng and changes are due to TB. This individual has complete
6 individuals (10.0%; 50 individuals with skulls) from destruction of the right mandibular condyle, with sur-
the Western Cape with some sort of cranial lesion (for rounding periostitis. Few sources reported on the
this calculation it should be kept in mind that some involvement of the TMJ and TB in paleopathology
skulls were absent). Of these lesions, six were on the while in contrast, some cases have been reported in
the clinical literature. It is rare that the TMJ is affected
Fig. 1. Destructive, lytic lesion in the cranial vault of Fig. 2. Intracranial lesion possibly originating from
individual A175 (50-year-old male). [Color figure can be otitis media (40-year-old male; case number AN 881).
viewed in the online issue, which is available at wileyonline [Color figure can be viewed in the online issue, which is
library.com.] available at wileyonlinelibrary.com.]
Skeletal Manifestations of TB in Modern Human Remains 5
Fig. 4. Destruction of the middle cranial fossa in a Fig. 6. Destruction in a hip joint (38 year old male; AN
46-year-old male (Case 55-08; date of death 2008). 881; date of death 1988). [Color figure can be viewed in
[Color figure can be viewed in the online issue, which is the online issue, which is available at wileyonlinelibrary.
available at wileyonlinelibrary.com.] com.]
6 Steyn and Buskes
may form in the choroid plexus or in the walls of the ven- Aufderheide AC, Rodriguez-Martin C. 1998. The Cambridge Encyclo-
tricle, or caseating material may be released from adja- pedia of Human Paleopathology. Cambridge: Cambridge Univer-
sity Press.
cent structures such as the vertebrae or the middle ear
Barnes E. 2005. Diseases and Human Evolution. Albuquerque: Uni-
which then involves the CNS (Donald et al., 2005). versity of New Mexico Press.
Young children, mostly those younger than 4 years Bos KI, Harkins KM, Herbig A, Coscolla M, Weber N, Comas I,
old (De Souza et al., 2014), and patients co-infected Forrest SA, Bryant JM, Harris SR, Schuenemann VJ, Campbell TJ,
with HIV patients are at high risk of developing TBM. Majander K, Wilbur AK, Guichon RA, Wolfe Steadman DL, Cook
The high incidence of cranial involvement in our study DC, Niemann S, Behr MA, Zumarraga M, Bastida R, Huson D,
supports the fact that adults are also at high risk, Nieselt K, Young D, Parkhill J, Buikstra JE, Gagneux S, Stone AC,
probably associated with HIV co-infection. Other risk Krause J. 2014. Pre-Columbian mycobacterial genomes reveal
factors include malnutrition, recent measles in chil- seals as a source of New World human tuberculosis. Nature 514:
dren, alcoholism, malignancies, and the use of immu- 494–497.
Campbell TJ, Ackermann RR. 2010. Evaluating the emergence of
nosuppressive drugs in adults (Rock et al., 2008). The
tuberculosis in South Africa. Am J Phys Anthropol 141(Suppl):76.
commonest cause of adult meningitis in Soweto, Daniel TM. 2006. The history of tuberculosis. Respir Med 100:1862–
South Africa, was found to be TBM and it was related 1870.
to poor compliance or drug-resistant TB (Karstaedt
Dayal MR, Kegley ADT, Strkalj G, Bidmos MA, Kuykendall KL.2009.
et al., 1998). The highest incidence of TBM in children The history and composition of the Raymond A. Dart collection of
exists in the Western Cape in South Africa (van den human skeletons at the University of the Witwatersrand, Johan-
Bos et al., 2004). Although the literature does not nesburg, South Africa. Am J Phys Anthropol 140:324–335.
mention the exact pathogenesis of how TB causes De Souza CH, Yamane A, Pandini JC, Ceratta LB, Ferraz F, da Luz GD,
Simo ~ es PW. 2014. Incidence of tuberculous meningitis in the State
lesions in the skull, the mechanism might be similar
of Santa Catarina, Brazil. Rev Soc Bras Med Trop 47:483–489.
to that of soft tissue pathogenesis, given there is
Donald P, Schaaf H, Schoeman J. 2005. Tuberculous meningitis
enough time to allow for bony manifestations to form. and miliary tuberculosis: The Rich focus revisited. J Infect 50:
The respective cases of TB infection of the tempo- 193–195.
romandibular joint and through otitis media reported Helbling CA, Lieger O, Smolka W, Iizuka T, Kuttenberger J. 2010. Pri-
on here are unusual. Although it has been described mary tuberculosis of the TMJ: Presentation of a case and litera-
in the clinical literature, it is to our knowledge the first ture review. Int J Oral Maxillofac Surg 39:834–838.
such cases described on skeletal remains alone. Holloway KL, Henneberg RJ, De Barros Lopes M, Henneberg M.
In summary, this article demonstrated that skeletal 2011. Evolution of human tuberculosis: A systematic review and
involvement in individuals dying from TB is still very meta-analysis of paleopathological evidence. Homo 62:402–458.
Holloway KL, Link K, Ru € hli F, Henneberg M. 2013. Skeletal lesions in
much prevalent, although the frequency of lesions
human tuberculosis may sometimes heal: An aid to palaeopatho-
should be interpreted considering the fact that these logical diagnoses. PLoS ONE 8:e62798.
are people who have died from the disease, and cannot Kanabus A. 2011. TB Statistics for South Africa | National & provin-
necessarily be directly extrapolated to living people cial figures with treatment outcomes. URL: http://www.tbfacts.
with the disease. However, it has some interesting clini- org/tb-statistics-south-africa.html [accessed September 2014].
cal implications both as far as patient suffering and pos- Karjodkar F, Siddhartha V, Maideo A, Sontakke S. 2012. Osteomyeli-
sible reactivation of TB from bony lesions are tis affecting mandible in tuberculosis patients. J Clin Exp Dent 4:
considered. It seems as if the prolonged treatment with 72–76.
antibiotics may in fact contribute to an increase in skel- Karstaedt AS, Valtchanova S, Barriere R, Crewe-Brown HH. 1998.
Tuberculous meningitis in South African urban adults. QJM 91:
etal lesions, and it may be that patients are living lon-
743–747.
ger with their disease thus providing more opportunity L’Abbe EN, Loots M, Meiring JH. 2005. The Pretoria Bone Collection:
for skeletal lesions to develop. Indeed, it may be true A modern South African sample. Homo 56:197–205.
that we make them live longer but they suffer more. Naidoo K, Naidoo K, Padayatchi N, Abdool Karim Q. 2011. HIV-
associated tuberculosis. Clin Dev Immunol 2011:585919.
Ortner DJ. 2003. Identification of Pathological Conditions in Human
ACKNOWLEDGMENTS Skeletal Remains. 2nd Ed. San Diego: Academic Press. p 227–
263.
The authors would like to thank the curators of the Patel M, Scott N, Newlands C. 2012. Case of tuberculosis of the tem-
Raymond A. Dart, Pretoria Bone, Kirsten and University poromandibular joint. Br J Oral Maxillofac Surg 50:e1–e3.
of Cape Town collections for access and permission to Pawlowski A, Jansson M, Sko € ld M, Rottenberg ME, Ka € llenius G.
study the remains. The authors are also indebted 2012. Tuberculosis and HIV Co-Infection. PLoS Pathog 8:
toward Deona Botha and Yvette Scholtz for their invalu- e1002464.
Pfeiffer S. 1991. Rib lesions and New World tuberculosis. Int J
able contributions. The research of M. Steyn is sup-
Osteoarch 1:191–198.
ported by the National Research Foundation of South Pillay Y, Mametja D, Mbengashe T, Makakole-Nene S, Banda MA,
Africa (NRF). Any opinions, findings, and conclusions Gqwaru N, Mvusi L, Baron P, Mhlongo-Sigwebela N, Bhardwaj S,
expressed in this article are those of the authors and, Reid A, Ntilivamunda A, Sannie B. 2013. Joint review of HIV, TB
therefore, the NRF does not accept any liability in and PMTCT programmes in South Africa. October 2013, Main
regard thereto. Report. Pretoria: Department of Health.
Pistorius JCC, Steyn M, Nienaber WC. 1998. Two burials from Mak-
gope - A Batswana settlement in the Bankeveld. S Afr J Ethnol
REFERENCES 21:115–124.
Prasad KC, Sreedharan S, Chakravarthy Y, Prasad SC. 2007. Tuber-
Angel JL. 1981. History and development of palaeopathology. Am J culosis in the head and neck: Experience in India. J Laryngol Otol
Phys Anthropol 56:509–515. 121:979–985.
8 Steyn and Buskes
Resnick D, Niwayama G. 1988. Diagnosis of Bone and Joint Disor- Sheikh S, Pallagatti S, Gupta D, Mittal A. 2012. Tuberculous osteo-
ders. Philadelphia: Saunders. myelitis of mandibular condyle: A diagnostic dilemma. Dento-
Rippey JJ. 1994. General Pathology. Johannesburg: Witwatersrand maxillofacial Radiol 41:169–174.
Steyn M, Scholtz Y, Botha D, Pretorius S. 2013. The changing face of
University Press. p 175–183.
tuberculosis: Trends in tuberculosis-associated skeletal changes.
Roberts CA, Buikstra JE. 2003. The Bioarchaeology of Tuberculosis: Tuberculosis 93:467–474.
A Global View on a Reemerging Disease. Gainesville, FL: Univer- Van den Bos F, Terken M, Ypma L, Kimpen JLL, Nel ED, Schaaf HS,
sity Press of Florida. et al. 2004. Tuberculous meningitis and miliary tuberculosis in
Rock RB, Olin M, Baker CA, Molitor TW, Peterson PK.2008. Central young children. Trop Med Int Health. 9:309–313.
nervous system tuberculosis: pathogenesis and clinical aspects. Vigorita VJ. 2008. Orthopaedic Pathology. Baltimore, MD: Lippincott
Clin Microbiol Rev 21:243–261. Williams & Wilkins.
Wilbur AK, Bouwman AS, Stone AC, Roberts CA, Pfister LA, Buikstra
Schneider M, Norman R, Parry C, Bradshaw D, Plu € ddemann A, and
JE, Brown TA. 2009. Deficiencies and challenges in the study of
South African risk assessment collaborating group. 2007. Esti-
ancient tuberculosis DNA. J Archeol Sci 36:1990–1997.
mating the burden if disease attributable to alcohol in South guec C, Wo
€ lbeling F, Kubica T, Kremer
Wirth T, Hildebrand F, Allix-Be
Africa in 2000. S Afr Med J 97:664–672. € sch-Gerdes S, Locht C, Brisse S, Meyer A,
K, van Soolingen D, Ru
Shankaragouda B, Barjatya H, Sahu U, Savadkar A. 2013. A case of Supply P, Niemann S. 2008. Origin, spread and demography of
tuberculous meningitis presenting with cognitive defects. Int J the Mycobacterium tuberculosis complex. PLoS Pathog 4:
Nutr Pharmacol Neurol Dis 3:388–391. e1000160.