Postextraction osteomyelitis in a bone marrow transplant recipient

Oral Diagnosis Editor PETER G. FOTOS, DDS, MS, PhD Department of Oral Pathology, Radiology, and Medicine University qf Iowa College of Dentistry, DSB Iowa City, Iowa 52242

Postextraction osteomyelitis in a bone marrow transplant recipient Andrei Barasch, DMD, FAAHD,’ Kristine M. Mosier, DMD.h Joseph .4. D’Ambrosio, DDS, MS,” Martin S. Giniger, DMD,” und Douglas E. Peterson, DMD, PhD,a Farmington, Conn. I’hIVtRSlTY

01: CONNECTICUT

HEALTH

Joao Ascensao,

MD,’

CENTER

This report describes a case of mandibular osteomyelitis after a dental extraction in a patient who subsequently underwent bone marrow transplantation (BMT) for lymphoblastic lymphoma. Surgical guidelines consistent with National Cancer Institute recommendations were followed for the extraction, which was performed before initiation of the myelosuppressive conditioning regimen. However, moderate tenderness developed at the extraction site beginning 10 days after rnarrow infusion. On day 26 the patient became febrile and blood culture-positive for Staphy/ococcus epidermidis. Radiographs exposed on day 28 demonstrated changes consistent with low-grade osteomyelitis, including diffuse loss of lamina dura and an irregular osseous rarefaction extending 1 cm posterior to the extraction site. Although the indwelling Hickman catheter was the presumed source for bacteremia, clinical and radiographic data led to consideration of mandibular osteomyelitis as an alternative cause. Characteristics of this infection in BMT recipients are reviewed. Recommendations for dental extractions and prophylactic antibiotic regimens for catheterized BMT recipients are also discussed. Although mandibular osteomyelitic lesions are not common in profoundly immunosuppressed BMT recipients, prompt recognition and treatment are essential when the disease occurs. IORW

Sr RG 0~41.

Mm

ORZI

PATHOI.

1993;75:391-6)

steomyelitis can involve both the maxilla and 0mandible.‘.’ Most of these infections develop as a consequence of trauma, radiation therapy, and/or untreated odontogenic infection, or in the presenceof “Division of Oral Medicine. Department of Oral Diagnosis, School of Dental Medicine. bDivision of Oral & Maxillofdcial Radiology, Department of Oral Diagnosis, School of Dental Medicine cFormerly. Division of Hematology/Oncology. Department of Medicine. School of Medicine: presently at Department of Medicine, Universit) of Nevada, Veterans Administration Medical Center, Reno. Copyright 1993 by Mosby-Year Book, Inc. 0030.4220/92/$1

.oo +

IO

7/ 17/39390

predisposing systemic conditions such as diabetes mellitus or sickle cell anemia.2 The clinical course of diseasecan be acute, subacute, or chronic. Although the most common symptom is localized or diffuse pain, suppuration is also common and fistulous tracts can be occasionally identified. Edema and erythema are infrequent clinical features unless periosteitis is present. Such infections tend to be localized when occurring in the mandible and more diffuse in the maxilla.’ Radiographic changes consisting of a diffuse, lytic appearance of bone are typically identifiable within 1 to 2 weeks after clinical diagnosis of the lesion.‘. 3 Osseouscomplications associated with dental ex391

ristine. adrlamycln. :lnd prednisone. and IntrLlthezai mesh otrexate and cranial irradiation. followed h\ tn:lintenana treatment consisting of c~clopho~phamide, vinzriytine, and prcdnisonc ever! h weeks and thioguaninc and cytarabinc every week for 36 \veekh. .This regimen LLXS admlnistercd t’o~ 23 month? and resulted in complete clinical remission. In September IWl llic patient noticed enlarged inguinai lymph nodes that. at subsequent hiopsq. rc~aled recurren! Iymphohlastic I~mphom~r. Noccid~nceofdi~e;ico ?\:Is loun~! in boric ni:irrob\ ,iq7irate\. I he patient \$.I\ yi?‘eii ;in inlet mittent course pi it’osfamide. mitox;Intronc. and ctoposlde. which. after 9 months, qain produced clinic;ll rcmiG>n The patient v,;I~ then e\.amined for HMI ,A11 candidate5 for BMTat the I.inivrr\lt! 01.i rinnect~cu 1ie:ill.h Center receive ,I coniprehenaive ival ~xtaniinatl~r!i hei‘ore

Fig. 1. Pcriapical radiograph of mandibular right tnol;~~ region. This lilm LV;IS exposed ?I days before extraction 01‘ tooth. Note cariuus lesion and periapical radiolucenc? ;I\so&ted with lir\t molar (ar,o~sl

tractions in patients about to receive myelosupprebsive chemotherapy have been described.‘, ’ In a stud! of I 19 extractions in 28 patients with acute nonlymphocytic leukemia, alveolar osteitis subsequent]) developed in only one patient.J It is notable that in none of the 15 patients with a granulocyte count of 2000 cells/mm’ or greater at time of surgery did postcxtraction infections develop. Similarly. other research documented only one localized infection in 26 patients who underwent 132 extractions. Granulocyte count5 in 14 of these patients were less than 1000 cells/mm~ either at time of extraction or within IO days after surgery.’ The only reported infection was initially assessed 5 days postoperatively, and the extraction site healed without complications after the USCof broadspectrum antibiotics (mezlocillin-tobramycin-clindamycin) and topical saline solution rinses. und increases in white blood cell counts. This case report describes ;I man with diagnosed lymphoblastic lymphoma who underwent extraction of the mandibular right first molar before bone marrow transplantation (BMT). Clinically documented osteomyelitis subsequently developed at the extraction

site during

myelosuppression.

Issues

of diagnosis.

management, and prevention of this complication comparable patients are discussed. CASE

in

REPORT

A 39.year-old man was referred to the University 01 Connecticut Health Center for autologous BMT. lie hud stage IlIB l~mphoblastic lymphoma. diagnosed in Ma! 1988: his medical history was otherwise unremarkable. The patient was initial11 treated with cyclophosphamide. vine-

hospital

.~ctmlssion.

On

such

ey:ltllln;ltioll

the

paCcrli

\\‘I’; ohaervcti (1) 1)~ normotrophic am! in no ;ipp;iren distres\. I le reported Inconsistent oral hygiene practices anti occasional mild smgival bleeding M hile hrushlnp his teeth ~rrd denlcti ,1n1 hi\lor\ i)i’ Ic’qion\ c:on\istcnl \\ ith hcrpe\ \iniple\ I ilu\ 611 ,r;~hthou\ ulcers. Result\ o! ,ili ,~\tr:lor:tl ex,lniination Marc iinreiliarkahic 1 hc ~ntrac~rai c:\;.tmlnation re! ealed hil;ltera! huccal 1c.u hoedema but no \tthcr oral soft tissue lesion> 1 moderate number of dental ratorztions Uere prcscnt. ;i\ \icll as p;lrtial edentulihm. multiple c‘;lriou\ lesion\. ;Ind cxtenzivc delital plaque accumulation. 4 complete intraoral radiographr~ scrieh I-cvc;~lcd .I deep cvrinus lesion on the distal-curon,i! aspect of the m,~ndlhular right lir\t molar ( Fig. / ). L.w,? (11 lamina dura ;11 the meslai root apex and the apic;ll third ot the distal root /)I’ I hi\ tooth \\a5 noted. ~\tiditlc)n:til! an ,iic,i 01 apparent h\~perobto& nia5uring .ipprcainiatcl~ \ mm X f .5 cm v.2’ iticnlilicd inferior 10 thts radiolucenq \ .-’ to 3 mm ,gencr:!Ii/zd pattern 01 :Il\eolar hone io\j L\:IS ;IIx! not cd

Hecauae

01 c\tcnsive

pcri,ipical

patliobib

and ioaii~ed

v-

i-crc periodontal ti~seasc. the mundihular right lirst mol;,r \I;Is cxtractcd or \ug. 27. 1991. with techniques con&cnt L+ith recolnnlcndaticin~ published b\ ~hc N;r!ionai C‘anar Iri\tituie.” T‘hc p Ori Scp! 1 Iclcj I i0 cI,I~\ bei‘c,re the hone ni:irro\h 11i1b \lonI. the ~;IIIL’W \\:I\ admitted to the HMT unrt for autoogous peripheral \tem cell transplantation. l-he extraction \~te ~*a\ ;IWZSX~ 11.)hc healing normall> at thix tirnc. 4 conditioning regimen 01‘ carboplalin. cyclophosphllrnide, anti stopuside W;IS iqitlatcd. The patient \&as given ;I prophylactic antimicrobial c.)ral cart regimen that included rinses 01 0. I?‘:( chlorhc\idine gluconate twice daily. 1’; hydropcn pcroxidc Ihrec t~nies daily, and n)statin-\-;in~oril!cin--gerlt:lmicin oral \uspen\lon four times dail!,. On September II (3 day5 before ini u>ion of boric niarro4,). the patient bccanic

OR 21 SL R(;FRY

OR \I

MEDICIYI: OR41.biT11010GY

Bara.rch

et al.

393

Volume 7:;. Number 3

Fig. 3. Periapical film 72 days postoperatively (day 59 after BMT), after antibiotic treatment. Radiograph reveals changes consistent with sclerosing osteitis and normal healing.

Fig. 2. A, Periapical radiograph exposed 41 days postoperatively (day 28 after BMT), illustrating diffuse loss of lamina dura at distal and inferior aspects of extraction socket iarrowx). B, Lateral-oblique film of right body of mandible exposed same day, demonstrating irregular rarefaction posterior to extraction socket. Observations noted in A and B are consistent with low-grade osteomyelitis.

febrile ( 102” F) and was given intravenous vancomycin and ceftazidime, the regimen of which was to be continued Lntil 24 days after BMT. The date of transplantation (day 0) was September 9. On September 12 (day 3) painful oral mucositis developed, which required a morphine sulfate drip for relief. On September 19 (day 10) the patient began to have tenderness to palpation of the extraction site, wil.hout any spontaneous discomfort, which persisted throughout his hospital stay. On October 3 (day 24) all antibiotics were discontinued after clinical engraftment and deferves-

cence. On October 5 (day 26), however, a fever of 103.8” F developed, and blood cultures were obtained that subsequently grew Staphylococcus epidermidis. The patient’s white blood cell count was 13,600 cells/mm3 with an ANC of 4700 cells/mm3. The fever was considered to represent a non-exit-site Hickman catheter complication, because no erythema or other dermal inflammatory signs were present at the catheter site. The patient was again given intravenous vancomycin and ceftazidime. On October 7 (day 28) mandibular radiographs revealed changes consistent with low-grade osteomyelitis (Fig. 2). A periapical view of the extraction site demonstrated a radiolucency within the socket, with diffuse loss of lamina dura at the distal and inferior aspects of the site (Fig. 2, A). A I X 1 cm irregular rarefaction posterior to the extraction site was observed in the lateral-oblique film (Fig. 2, B). A posteroanterior film of the mandible revealed a 6 mm area of mild sclerosis, inferoposterior to the rarefaction. Sequestration and periosteal bone formation were not observed in these radiographs. The patient was discharged from the hospital on October I I (day 32) with 2 gm/day intravenous vancomycin for 2 weeks, to be followed by a course of oral clindamycin. He returned to the outpatient oral medicine clinic October 24 (day 45), at which time he reported no symptoms associated with the extraction site. There were no palpable right submandibular lymph nodes, and no fever was present. Blood culture specimens taken at this time were reported as negative for bacterial growth. The next oral medicine visit was November 7 (day 59), at which time his white blood cell count was 6200 cells/mm3 with an ANC of 1600 cells/mm’. The patient had no oral symptoms, and the mandibular extraction site appeared clinically unremarkable. A periapical film exposed at this time revealed evidence of coalescing areas of sclerosing osteitis and a more normal trahecular ap-

394

Burasch et al.

pearance as compared with the radiographs exposed on October 7. These observations were interpreted to be consistent with normal healing of the postextraction dental socket (Fig.

3).

DISCUSSION This report describes a BMT recipient who underwent a dental extraction before initiation of conditioning. The oral surgical techniques were based on nationally developed guidelines for extractions in comparable patients, including use of broad-spectrum intravenous antibiotic prophylaxis. The patient’s ANC fell to less than 500 cells/mm3 I2 days after extraction. He received broad-spectrum systemic antimicrobial therapy from 3 days before BMT to da! 24, and daily oral rinses of 0.12”Y chlorhexidine gluconate twice daily, 3% hydrogen peroxide three times daily,andvancomycin-nystatin-gentamicinfourtimes daily. Despite these interventions, clinically documented mandibular osteomyelitis developed. Oral infections can often produce systemic complications in immunocompromised cancer patients.-“’ Historically. approximately 25% of systemic infections in myelosuppressed patients with acute leukemia undergoing induction of therapy have been linked to exacerbation of preexisting chronic periodontal disease.” This prevalence has fortunately decreased during the last decade in the setting of more effective infection prevention. Because of the high risk for odontogenic infection during my&suppression, it ap pears that oral treatment of patients with hematologic malignancies should include removal of severely periodontally and/or periapically involved teeth before initiation of cancer therapy.” However, virtually II0 data address the benetit of dental extractions in these patients compared with more conservative treatment or no treatment at all. because ethical issues often preclude such studies in human beings. Therefore thy decision to extract is usually based on data from noncontrolled clinical trials.‘. ‘. ’ Such studies report the occurrence of very few mild postoperative complications in chemotherapy patients after preinduction dental extractions when specific guidelines are followed. These guidelines include prophylactic antibiotics for patients with a granulocyte count of less than 2000 cells/mm’ (
ciated with an acute dental or periodontal infeclion that might develop during neutropenia if the teeth in question are not removed previously. However, the rate of occurrence of such exacerbations, the relationship between severity of preexisting periodontal or pcriapical disease and likelihood of acute flares, and the results of more conservative management of the acute complications are not well defined in large studies. Besides considerations of risks and benefits relevant to dental extractions before BMT. decisions need to be made regarding appropriate prophylactic antibi,otics in BMT recipients who have indwelling intraluminal catheters and who require oral invasive procedures. Although some regimen of antibiotic prophylaxis to reduce risk of catheter colonization is general]! .Iccepted. selections of specific antimicrobial agent\ remain contro\.ersial. Approximately two thirds of intraluminal catheter infections are caused h> .Sruph~lococcu.s species, with Staphylococcus aureu.v and S. e@derfllidis predominating.‘“-” These epidemiologic data can be used to guide selection of appropriate prophylactic agents. Also to be considered when selecting antibiotics is the qualitative change occurring in the oral flora of the myelosuppressed cancer patients. In this group there is generally a shift toward a more gram-negative microbiota. which ma! include organisms that are not commonly detected in the oral cavity.‘. ‘. ” Hartmann et al.‘3described a prophylactic regimen consisting of daily infusion of cephalosporin for the duration of antineoplastic treatment, with excellent results in preventing catheter-related infections while patients were treated with interleukin-2 for malignant disease. Although this strategy may result in adequate prevention ol’ c.atheter-related infections for a short period of time i c.g.. 15 days), a longer course such as may be needed for a BMT recipienl may pose the risk of enhancing ,gmwth of resistant organism. Other authors” have :,uggested use of the American Heart .,2ssociation cndocarditis prophylaxis regimen, currently consisling of amoxicillin or erythromycin. These two agenti; provide good coverage for oral flora coloni& ~1ncrrmal host. They do not. however. :tJcquately cober the typically altered oral flora of the BMT rccipienl including most Staplz~,/ol,oc,l,u.~ sptzties. We believe that any antibiotic prophylaxis adminiatered in association with invasive dental procedures in the patient who has an indwelling intraluminal catheter and l.\,ho will undergo BMT should cover anaerobic gram-negative species as well as SraphJjlo-

ORAL

Volume

SLWERY

ORAL

75, Number

MEDICINE

ORAL

species, including methicillin-resistant strains.16, I7 Accordingly, the regimen administered to the patient described in this report consisted of 1 gm vancomycin and 120 mg gentamicin infused slowly for 1 hour with no repeated dose administered. This approach appears to cover effectively a broad spectrum of gram-negative and gram-positive bacteria, including most of the presumed colonizing microflora. Possible exceptions to this coverage are Pseudomonas species, which do not commonly cause acute oral infections in the myelosuppressed cancer patient. The S. epidermidis bacteremia in this reported case was initially attributed to a Hickman catheter infection. As stated earlier, S. epidermidis is a common pathogen involved in such infections. However, the patient had no other catheter-related signs or symptoms consistent with this cause and no inflammati’on was present at the site of the catheter despite a white blood cell count of 13,600 cells/mm3 and an ANC of 4700 cells/mm3. The patient’s only symptom at tirne of bacteremia was persistent, moderate tenderness of the mandibular dental extraction site on palpation. The intensity of this symptom was likely affected by concurrent treatment with intravenous morphine sulfate for palliation of pain caused by oral mucositis. These clinical data, together with interpretation of the mandibular radiographs, led to inclusion of mandibular osteomyelitis in the differential diagnosis of this patient’s source of bacteremia. Culture of S. epidermidis from the patient’s blood was also consistent with this diagnosis because this organism is a common pathogen in osteomyelitis.?, I83l9 A definitive diagnosis could presumably have been achieved by obtaining a bacterial culture of the extraction site. However, the gingiva appeared reepithelialized and culturing would have involved surgical intervention. Such invasive procedures were not warranted in this patient, who was profoundly myelosuppressed at the time. Thus the patient was given antibiotic therapy and followed up clinically and radiographically for infection resolution. Radiographs of the extraction site exposed after completion of antibiotic therapy showed normal healing with no indications of infection. The development of clinically documented osteomyelitis in this patient can be attributed to either primary microbial colonization of the extraction site or to colonization caused by bacteremia resulting from other sources (anachoresis). The clinical course of the mandibular lesion, including time of onset of symptoms and radiographic changes, implicate the former possibility as more likely. I8 Treatment of osteomyelicoccus

Barasch et al.

PATHOLOGY

395

3

tis in otherwise healthy adults should include prolonged use of parenterally administered antibiotics, followed by an extended course of an oral agent with good bone penetration. 18,l9 This case report suggests that these guidelines are also appropriate for BMT recipients. Ideally, culture and sensitivity data should be obtained to achieve optimal therapeutic effect. Surgical intervention is not usually required but may become necessary if an abscess or a bony sequestrum develops. Although not a common complication, timely diagnosis of and appropriate treatment for mandibular osteomyelitis are important in BMT recipients who have undergone dental extractions within several weeks before initiation of myelosuppressive conditioning regimens. We thank manuscript.

Ms. Dawn

Wright

for preparation

of the

REFERENCES 1. Balm AJM, Tiwary RM. Osteomyelitis in the head and neck. J Laryngol Otol 1985;99:1059-65. 2. Calhoun KH, Shapiro RD, Stiernherg CM, Calhoun JH, Mader JT. Osteomyelitis of the mandible. Arch Otolaryngol Head Neck Surg 1988;114:1157-62. 3. Shafer WG, Hine MK, Levy MM, eds. Textbook of oral pathology. 4th ed. Philadelphia: WB Saunders, 1986:498-504. 4. Overholser CD, Peterson DE, Bergman SA, Williams LT. Dental extractions in patients with acute nonlymphocytic leukemia. J Oral Maxillofac Surg 1982:40:296-X. 5. Williford SK, Salishury PL III, Peacock JE, et al. The safety of dental extractions in patients with hematologlc malignancies. J Clin Oncol 1989;7:798-802. 6. Peterson DE. Pretreatment strategies for infection prevention in chemotherapy patients. NC1 Monogr 1990;9:6 I-7 I. 7. Greenberg MS, Cohen SG, McKitrick JC, Cassileth PA. The oral flora as a source of septicemia in patients with acute leukemia. ORAL SURG ORAL MED ORAL PATHOL 1982;53:32-6. 8. Peterson DE, Schubert MM. Oral toxicity. In: Perry MC, ed. The chemotherapy source hook. Baltimore: Williams & Wilkins, 1992:508-30. 9. Wingard JR. Infectious and noninfectious systemic consequences. NC1 Monogr 1990;9:21-5. IO. Carpenter CCJ. Infectious disease. In: Wyngaarden JB, Smith LH, eds. Cecil textbook of medicine; vol 2. 18th ed. Philadelphia: WB Saunders, 1988:1523. Il. Dugdale DC, Ramsey PG. Staphylococcus aureus hacteremia in patients with Hickman catheters. Am J Med I990;89: 13741. 12. Martin0 P, Micozzi A, Venditti M, et al. Catheter-related right-sided endocarditis in hone marrow transplant recipients. Rev Infect Dis 1990;12:250-7. 13. Peterson DE, Minah GE, Overholser CD, et al. Microbiology of acute periodontal infection in myelosuppressed cancer patients. J Clin Oncol 1987;9:1456-68. 14. Hartmann LC, Urha WJ, Steis RG, et al. Use of prophylactic antibiotics for prevention of intravascular catheter-related infections in interleukin-2-treated patients. .I Natl Cancer Inst 1989;81:1190-3. 15. Little JW, Falace DA, eds. Dental management of the med-

cd\. WH

C cc11 tc’itbook III med~c~nc: Saunders. I YXX: 1672-J

\iidrci Bar,lsch, I)hlD. 1. \AtII) Di\ ision of Oral Mcdicinc Ikpartment of Oral t>iagno\ts School of lkntCtl Medicine i.‘niverGt> of (.onnecticut ttcslth I armingtm. C I (16030-l 60’

Bound volumes (onI! ) I’or the

of OR\I St ~(~1 KL. OKZI MI III< IZI. \\I) OKAI 1993 issues from the Publisher. at 3 cost of $54.50

vol

1

I Sth e
C’txtcr

PZI IIC)I (I(~! arc available to subscribcv for domertic. $70.32 for Canadian. and

$66.50 for international, for Vol. 75 (.l,tnuary-.lunei and Vol. 76 (.July-December). Shipping charges are included. Each bound volume contains a cuhject and author index and all advertising is removed. Copies arc bhipped within 60 days after publication of the last issue in the volume. The binding is durable buckram with theJournal name, volume number. and )ear stamped in gold on the spine. Pu~mem JHUS~ U-wmpar~~~ a// orrlrr.~. (‘ontact Moshq, Subscription Services. I I X30 Westline Industrial Drive, St. Louis. MO 03 146.33 18. 1!S.,\: phone (800)325--I 177. e\t 435 I or (3 I3)453-415 I. Subscriptions must be in force to qualify. Bound volumes are not arailable in place of a regular journal subscription.