Archives of Oral Biology 44 (1999) 789±793

www.elsevier.com/locate/archoralbio

Isolation of bacteria from cervical lymph nodes in patients

with oral cancer
Haruo Sakamoto a,*, Hiroyuki Naito a, Yoshihide Ohta a, Rin Tanakna b,
Nobuko Maeda b, Jiro Sasaki a, Carl Erik Nord c
a
Department of Oral Surgery, Tokai University School of Medicine, Boseidai, Isehara, Kanagawa 259-1193, Japan
b
Department of Bacteriology and 1st Department of Anatomy, Tsurumi University School of Dental Medicine, 2-1-3 Tsurumi,
Yokohama, Kanagawa 230, Japan
c
Department of Immunology, Microbiology, Pathology and Infectious Diseases, Karolinska Institute, Huddinge University Hospital,
S-141 86 Huddinge, Sweden

Accepted 18 May 1999

Abstract

Thirty patients with oral mucosal cancer were studied in relation to oral mucosal damage and bacterial
translocation to the regional lymph nodes in the neck. All 30 patients (gingiva 11, tongue 13, cheek mucosa four,
oral ¯oor two) underwent extensive, clean-contaminated, head-and-neck surgery (including neck dissection) with free
¯ap reconstruction. A total of 153 lymph nodes was harvested for microbial and histological examination. Viable
bacteria were isolated from 70 lymph nodes (45.8%) from 25 patients (83.3%). Bacterial cells in the nodes were
detected by scanning electron microscopy. Bacterial translocation was found more often in metastatic nodes (75.0%)
than in uninvolved nodes (40.3%) ( p = 0.015; w2 test). Gingival carcinoma yielded 56.4% of bacterial growth in the
regional lymph nodes compared to tongue (40.3%), oral ¯oor (40.0%) and cheek mucosa (37.5%). As the gingival
carcinoma group includes more T4 cases (11/11; 100%) than the other three groups (7/19; 36.8%), bacterial
translocation in uninvolved nodes could be caused by the size and invasion of the primary oral tumor. Oral
streptococci (Streptococcus intermedius, Strep. constellatus, Strep. oralis, Strep. mitis, Strep. sanguis, Strep. salivarius )
were the most common isolates. Aerobic enteric bacteria (Enterococcus, Escherichia, Klebsiella etc.) were also found
in the lymph nodes. Among the anaerobic bacteria, Peptostreptococcus spp. were isolated from 12 patients.
Damaged oral mucosa in patients with oral cancer might allow the new bacterial colonization on the surface and
subsequently drain the bacteria into the regional lymph nodes as well as the general circulation. # 1999 Elsevier
Science Ltd. All rights reserved.

Keywords: Bacteria; Lymph node; Oral cancer; Translocation; Oral streptococcus

1. Introduction

The passage of viable cells from the alimentary tract

Abbreviations: PBS, phosphate bu€er solution.
* Corresponding author. Tel.: +81-463-93-1121; fax: +81- to the regional lymph nodes, the liver, the spleen and
463-91-5902. other organs is called `bacterial translocation' by Berg
E-mail address: sakamoto@is.icc.u-tokai.ac.jp (H. and Garlington (1979). Translocation is considered as
Sakamoto) a possible initial step in the pathogenesis of many bac-

0003-9969/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 0 3 - 9 9 6 9 ( 9 9 ) 0 0 0 7 9 - 5
790 H. Sakamoto et al. / Archives of Oral Biology 44 (1999) 789±793

Table 1 noma diagnosed at an earlier biopsy and were treated

Primary tumour localization and T staginga in the Department of Oral Surgery, Tokai University
School of Medicine. Written informed consent was
T2 T3 T4 cases obtained from all the patients before surgery. This
study was approved by ethical committee of Tokai
Tongue 7 3 3 13
Gingiva 0 0 11 11 University School of Medicine.
Cheek mucosa 1 0 3 4
Oral ¯oor 1 0 1 2
Total 9 3 18 30 2.2. Procedure
a
Note: staging criteria for primary oral squamous-cell carci- All 30 patients (gingiva 11, tongue 13, cheek mucosa
noma are as follows. T1, tumour (in its greatest four, oral ¯oor two) underwent extensive head-and-
dimension)Y2 cm; T2, 2 cmY tumourY4 cm; T3, neck surgery (including neck dissection) with free ¯ap
4 cmYtumour; T4, tumour invades adjacent structures,
reconstruction (Table 1). All operations were per-
including mandible, maxilla, skin, extrinsic tongue muscle of
formed by the same surgical team. Ampicillin (1 g) was
the tongue root, and soft tissue of the neck (American Joint
Committee on Cancer, 1992). given from the start of the surgery as prophylaxis. The
freshly resected surgical specimen was placed in a lami-
nar air-¯ow box to avoid microbial contamination.
The cervical tissue was dissected and the lymph nodes
terial infections (Berg and Garlington, 1983). When
were harvested. The nodes were washed three times in
the natural mucosal barrier is damaged, microorgan-
PBS and then divided into two halves, one half for his-
isms could colonize on its surface and subsequently
topathological examination and the other for micro-
penetrate at the site of the breaks (Mims et al., 1995).
biological cultivation.
After reaching submucosal tissues, microorganisms
enter lymphatic capillaries and could be transported to
the nearest lymph nodes. There are recent descriptions
2.3. Histopathological examination
of bacterial translocation in a variety of animal models
(Berg and Garlington, 1979; Deitch et al., 1987; Penn
The presence of cancer metastases in the lymph
et al., 1985; Wells et al., 1988). Compromised host
node was determined by the examination of sections
defence mechanisms, ecological changes of the gut
stained with haematoxylin±eosin. Bacterial coloniza-
micro¯ora and disruption of the mechanical barrier of
tion in the lymph node was studied by scanning elec-
the gut wall due to any reasons are factors found to
tron microscopy.
favour bacterial translocation in experimental animals.
However, the process of bacterial translocation in
humans has not been fully studied.
2.4. Microbiological cultivation
In this study, 30 patients with oral mucosal cancer
were studied in relation to oral mucosal damage and
The lymph nodes were weighed and cut into small
bacterial translocation to the regional lymph nodes in
pieces, homogenized and suspended in PBS. Ten-fold
the neck. Cervical lymph nodes are strategically placed
dilutions were prepared to 104 in PBS. A volume
between lymphatic capillaries and thoracic duct in
(100 ml) of each dilution was incubated on selective
order to recognize and ®lter o€ particles. If viable bac-
and non-selective media. For aerobic culture, sheep
teria could survive in those nodes, escaping from the
blood agar and chocolate agar (Kyokuto
immune defence system, they would become a possible
Pharmaceutical Co., Tokyo, Japan) were used. For an-
pathological source of postoperative infection. Here,
aerobic culture, four di€erent media (1) brucella HK
153 lymph nodes were harvested during radical neck
agar with haemolysed rabbit blood and de®brillated
dissection and studied microbiologically and pathologi-
sheep blood, (2) paramomycin±vancomycin brucella
cally.
HK agar with haemolysed rabbit blood, (3) pheny-
lethyl alcohol brucella HK agar with haemolysed rab-
bit blood, and (4) bacteroides bile esculin agar
2. Materials and methods (Kyokuto Pharmaceutical) were prepared. The aerobic
agar plates were incubated for 24 h at 378C and the
2.1. Patients anaerobic plates for at least 48 h at 378C in anaerobic
jars (Gas Pak; BBL, USA). The bacteria were ident-
Thirty patients with oral mucosal cancer (19 men i®ed according to the American Society for
and 11 women; mean age 63.3 years) were studied. Microbiology (Balows et al., 1991). A Rap ANA sys-
They all had histologically veri®ed squamous-cell carci- tem was also applied for identi®cation of anaerobes.
 H. Sakamoto et al. / Archives of Oral Biology 44 (1999) 789±793 791

Table 2
Tumour localization and number of lymph nodes with bacterial growth in 30 patients with oral cancer

All nodes Involved nodesa Uninvolved nodesb

Tumour No. of patients Total Bacterial growth Total Bacterial growthc Total Bacterial growth

Tongue 13 72 29 14 9 58 20
(40.3%) (64.3%) (34.5%)
Gingiva 11 55 31 5 5 50 26
(56.4%) (100%) (52.0%)
Cheek mucosa 4 16 6 1 1 15 5
(37.5%) (100%) (33.3%)
Oral ¯oor 2 10 4 4 3 6 1
(40.0%) (75.0%) (16.7%)
Total 30 153 70 24 18 129 52
(45.8%) (75.0%) (40.3%)

a
Note: involved nodes; lymph nodes with metastatic carcinoma.
b
Uninvolved nodes; lymph nodes without carcinoma.
c
Bacterial growth was more often found in involved nodes than in uninvolved nodes ( p = 0.015: w2 test, signi®cant di€erence).

3. Results harvested for microbial cultivation. Viable bacteria

A total of 153 lymph nodes from 30 patients were
were isolated in 70 nodes (45.8%) from 25 patients
(83.3%) (Table 2). Bacterial cells in the nodes were
detected by scanning electron microscopy (Fig. 1).
Bacterial translocation was found more often in meta-
static lymph nodes (75.0%) than in uninvolved nodes
(40.3%) ( p = 0.015; w2 test) (Table 1). Gingival carci-
noma yielded 56.4% of bacterial growth in the re-
gional lymph nodes compared to tongue (40.3%), oral
¯oor (40.0%) and cheek mucosa (37.5%). Isolated
microorganisms are summarized in Table 3. Oral strep-
tococci (Streptcoccus intermedius, Strep. constellatus,
Strep. oralis, Strep. mitis, Strep. sanguis, Strep. salivar-
ius ) were the most common isolates. Aerobic enteric
bacteria (Enterococcus, Escherichia, Klebsiella etc.)
were also found in the nodes. Among the anaerobic
bacteria, Peptostreptococcus spp. were isolated from 12
patients.

4. Discussion

The results demonstrated that the bacterial metasta-

Fig. 1. Bacterial colonization with coccoid cells found in a
cervical lymph node in a patient with gingival carcinoma
(T4N2M0). (Scanning electron micrograph; original magni®-
cation (a) 2000, (b) 8000).
sis to the cervical lymph nodes occurred more fre-
quently in metastatic nodes than in uninvolved nodes.
These ®ndings are similar to those in patients with col-
orectal and gynaecological malignancies (Koha et al.,
1991; Vincent et al., 1988; Wells et al., 1990). Vincent
et al. report the recovery of intestinal bacteria from
the mesenteric lymph nodes of 65% patients with col-
orectal cancer and from 15% of those who had surgery
for non-colorectal digestive diseases. More recently,
Koha et al. (1991) found that the overall incidence of
bacterial translocation to the mesenteric lymph nodes
was 23%, which was more prominent in metastatic
792 H. Sakamoto et al. / Archives of Oral Biology 44 (1999) 789±793

Table 3
Microbiological ®ndings in cervical lymph nodes in 30 patients with oral cancer

Bacteria Nos of patients and incidence with bacterial growth (%) Nos of CFUa in a lymph node mean (range)

Aerobes
Oral Streptococcus
Strep. intermedius 15 (50.0) 1.22  102 (76±4.40  102)
Strep. oralis 14 (46.7) 2.63  102 (16±2.68  103)
Strep. constellatus 11 (36.6) 1.09  102 (15±2.35  102)
Strep. mitis 9 (30.0) 73 (16±1.71  102)
Strep. sanguis 7 (23.3) 6.15  102 (15±3.12  103)
Strep. salivarius 6 (20.0) 1.12  102 (15±3.92  102)
Strep. anginosus 2 (6.7) 9.36  102
Enterococcus faecalis 4 (13.3) 1.76  102 (20±1.52  103)
Escherichia coli 1 (3.3) 15
Staphylococcus aureus 1 (3.3) 15
Serratia marscens 2 (6.7) 4.96  102
Streptcoccus equinus 1 (3.3) 7.70  102
Staphylococcus epidermidis 10 (33.3) 5.20  102 (12±1.52  103)
Staphylococcus acidominus 1 (3.3) 16
Pseudomonas aeruginosa 1 (3.3) 56
Eikenella corrodens 2 (6.7) 4.6  102
Klebsiella pneumoniae 2 (6.7) 15
Lactococcus cremoris 1 (3.3) 76
Neisseria sp. 15 (50.0) 1.13  102 (3.2  10±1.87  102)
Corynebacterium sp. 1 (3.3) 59
Capnocytophaga sp. 1 (3.3) 3.20  102
Anaerobes
Peptostreptococcus sp. 12 (40.0) 3.10  102 (18±1.56  103)
Gemella morbillorum 7 (23.3) 3.87  102 (56±1.54  103)
Prevotella melaninogenica 3 (10.0) 56 (31±80)
Fusobacterium nucleatum 3 (10.0) 1.03  102 (14±2.80  102)
Veillonella sp. 1 (3.3) 32
Bacteroides fragilis 1 (3.3) 1.54  102
Bi®dobacterium sp. 1 (3.3) 1.54  102

a
CFU, colony-forming units.

lymph nodes (32%) than uninvolved nodes (19%). In
patients with gynaecological malignancies, Wells et al.
(1990) recovered viable bacteria in 34% of the pelvic
lymph nodes harvested. Those ®ndings suggest that the
viable bacteria could survive in the drainage lymph
nodes of the digestive tract (covered with adsorptive
and secretory cells) as well as genital mucosa (lined by
squamous cells) escaping from the host defence sys-
tems. The possible mechanisms of bacterial transloca-
tion are uncertain, although phagocytosis and
intracellular endocytosis have been postulated in the
process of translocation through the intestinal wall
(Wells et al., 1988). Synergistic translocating mechan-
isms in the mesenteric lymph nodes for metastatic and
bacterial cells are also assumed to occur (Koha et al.,
1992). In the present study, translocation was more
often found in patients with gingival carcinoma than
tongue, oral ¯oor or cheek mucosa carcinoma. The
incidence of bacterial translocation in the four groups
of oral carcinoma was similar in involved nodes but
di€erent in uninvolved nodes. As the gingival carci-
noma group included more T4 cases (11/11; 100%)
than the other three groups (7/19; 36.8%), bacterial
translocation in uninvolved nodes could be caused by
the size and invasion of the primary oral tumour.
The most common isolate in the lymph node was
oral streptococci, which are members of normal oral
micro¯ora and found in odontogenic infections. We
could detect the coccoid cells (suggesting
Streptococcus, Staphylococcus ) in the nodes by scan-
ning electron microscopy. The association of an oral
mucosal defect, such as ulceration, with streptococcal
bacteraemia or septicaemia in patients under certain
pathological circumstances is reported. Heimdahl et al.
(1989) described a-streptococcal bacteraemia in
patients with bone marrow transplants. Septicaemia
with streptococci was found in 41% of microbiological
proven septicaemias associated with graft-versus-host
disease, prophylaxis with methotrexate and a sub-
sequent increased frequency of oral ulceration. Van
 H. Sakamoto et al. / Archives of Oral Biology 44 (1999) 789±793
der Lelie et al. (1991) reported 27 streptococcal septi-
caemias in a total of 76 patients with acute myeloblas-
tic leukaemia. In 20 patients was a fast recovery, but
seven patients died with adult respiratory distress syn-
drome. Van der Lelie suggested that bacteria invade
through the oral and gastrointestinal tract, which was
damaged after selective decontamination. The species
identi®cation of streptococci was Strep. mitis, Strep.
viridans, Strep. sanguis and Strep. milleri. Cervical
lymph nodes are in contact with lymphatic ¯ow before
returning to the thoracic duct. The recovery of viable
bacteria (mainly oral streptococcus) from the lymph
nodes in the present study supports the hypothesis of
the association between oral mucosal damage and bac-
terial invasion into the general circulation.
SjoÈstedt et al. (1988) found that the anaerobic milieu
in gastrointestinal tumour tissues may predispose to
growth of anaerobic intestinal bacteria, especially clos-
tridia. Peptostreptococcus, Prevotella melaninogenica,
Fusobacterium and anaerobes that belong to the oral
micro¯ora were often isolated from the lymph nodes,
which might suggest that the anaerobic milieu found in
the oral mucosal defect could promote the transloca-
tion of anaerobes into the regional nodes.
Aerobic enteric Gram-negative rods, such as
Escherichia coli, Klebsiella, or enteric Gram-positive
cocci (Enterococcus ) were found in the lymph nodes.
These enteric bacteria might be introduced from foods,
hands or other sources of contamination. In normal
circumstances, attachment and colonization of these
unexpected members are inhibited by the normal
micro¯ora of the mouth, salivary secretion or other
host defence mechanisms. However, damaged oral
mucosa in oral cancer might allow new bacterial colo-
nization of the surface and subsequently drain the bac-
teria into the regional nodes, which could become a
possible source of postoperative infection.
Further investigation would be needed to clarify the
association between bacterial colonization in the
damaged oral mucosal surface with bacterial transloca-
tion to the lymph nodes, as well as with postoperative
infection. These studies are in progress in our labora-
tory.
References
Balows, A., Hansler, W.J., Herrmann, K.L., Isenberg, H.D.,
793
Shadomy, H.J. (Eds.), 1991. Manual of Clinical
Microbiology, 5th ed. American Society for Microbiology,
Washington, DC.
Berg, R.D., Garlington, A.W., 1979. Translocation of certain
indigenous bacteria from the gastrointestinal tract to the
mesenteric lymph nodes and other organs in a gnotobiotic
mouse model. Infection and Immunity. 23, 403±411.
Berg, R.D., Garlington, A.W., 1983. Translocation of certain
indigenous bacteria from the intestinal tract. In: Hentges,
D.J. (Ed.), Human Intestinal Micro¯ora in Health and
Disease. Academic Press, London, pp. 333±349.
Deitch, E.A., Berg, R., Specian, R., 1987. Endotoxin pro-
motes the translocation of bacteria from the gut. Archives
of Surgery 122, 185±190.
Heimdahl, A., Mattson, T., Dahllof, G., Lonnquist, B.,
Ringden, O., 1989. The oral cavity as a port of entry for
early infections in patients treated with bone marrow
transplantation. Oral Surgery, Oral Medicine, Oral
Pathology 68, 711±716.
Koha, M., Brismar, B., Wikstrom, B., Nord, C.E., 1991.
Bacterial translocation in patients undergoing elective col-
orectal surgery. Microbial Ecology in Health and Disease
4, 363±367.
Koha, M., Brismal, B., Wikstrom, B., Everth, S., Nord, C.E.,
1992. Bacterial colonization and translocation in colorectal
carcinoma. Medical Microbiology Letters 1, 168±176.
Mims, C.A., Dimmock, N.J., Nash, A., Stephan, J., 1995.
Mims, C.A. (Ed.). Mim's pathogenesis of infectious dis-
ease, 4th ed. Academic Press, London, pp. 109±111 .
Penn, R.L., Maca, R.D., Berg, R.D., 1985. Increased translo-
cation of bacteria from the gastrointestinal tracts of tumor
bearing mice. Infection and Immunity 47, 793±798.
SjoÈstedt, S., Kager, L., Heimdahl, A., Nord, C.E., 1988.
Microbial colonization of tumors in relation to the upper
gastrointestinal tract in patients with gastric carcinoma.
Annals of Surgery 207, 341±346.
Van Der Lelie, H., Van Dem, Borne A.E., Van Oers, R.H.,
Thomas, B.L., Goudsmit, R., 1991. Incidence and clinical
epidermology of streptococcal septicemia during treatment
of acute myeloid leukemia. Scandinavian Journal of
Infectious Diseases 23, 163±168.
Vincent, P., Colombel, J.F., Lescut, D., Fournier, L., Savage,
C., Cortot, A., Quandalle, P., Vankemmel, M., Leclerc,
H., 1988. Bacterial translocation in patients with colorectal
cancer. Journal of Infectious Diseases 158, 1395±1396.
Wells, C.L., Maddaus, M.A., Simmons, R.L., 1988. Proposed
mechanisms for the translocation of intestinal bacteria.
Review of Infectious Diseases 10, 958±979.
Wells, C.L., Jechorek, R.P., Twinggs, L.B., Brooker, D.C.,
1990. Recovery of viable bacteria from cervical lymph-
nodes in patients with gynecological malignancies. Journal
of Infectious Diseases 162, 1216±1218.