Candida Prosthetic Joint Infection. A Review of Treatment Methods.

Fungal microorganisms are still a rare cause of bone and joint infections. We report a new case of knee prosthetic joint infection due to Candida albicans in a patient with a previous two-stage right knee arthroplasty for septic arthritis due to S. epidermidis occurred several months ago. Moreover, the treatment in 76 cases of Candida prosthetic joint infection has been discussed. Forty patients were female and mean age at diagnosis was 65.7 (± SD 18) yrs. No risk factors for candidal infection were found in 25 patients. Infection site was the knee in 38 patients and hip in 36; pain was present in 44 patients and swelling in 24. The most frequent species was C. albicans, followed by C. parapsilosis. Eleven patients were only treated with antifungal drugs being the outcome favourable in all of them. Two-stage exchange arthroplasty was performed in 30 patients, and resection arthroplasty in other 30; in three patients one-stage exchange arthroplasty was done. A favourable outcome was found in 58 patients after antifungal plus surgical treatment, in 11 after antifungal treatment alone and in one after surgery alone. The type of treatment is still not clearly defined and an algorithm for treatment in fungal PJI should be established, but various types of surgical procedures may be applied.

Introduction

Prosthetic joint infection (PJI) involves the joint prosthesis and contiguous tissue and is one of the main reasons for total arthroplasty failure 1, 2. A wide range of microorganisms may cause PJI, most often Gram-positive bacteria, especially staphylococcal species, and Gram-negative pathogens 3. However, other microorganisms can also be responsible for PJI, including fungi, particularly Candida species. PJI due to Candida is rare and represents a therapeutic challenge because no specific guidelines have been already established and published case reports vary widely in therapeutic approach 4. Currently, the gold standard for treatment consists in a two-stage revision surgery 4, 5, although it is doubtful whether multiple procedures are able to provide any improvements and it is also unknown if other techniques such as one-stage exchange may be successful.

Here, we describe one patient seen at the Orthopaedics and Trauma Department of the Hospital Universitario Virgen de las Nieves (Granada, Spain) with PJI due to C. albicans which is being treated with antifungal drugs alone. Moreover, we have reviewed the medical literature searching case reports with Candida PJI discussing about the treatment methods applied.

Case Report

A 66-year-old man had a right knee arthroplasty due to osteoarthritis suffered for several years. In January 2015, the patient underwent a two-stage right knee arthroplasty for septic arthritis due to S. epidermidis, as well as treatment with vancomycin + gentamycin. He was immunocompromised due to a splenectomy performed several years ago. In July 2016, the patient was attended at the Emergency Department of our Hospital due to pain, inflammation and joint leak for two weeks. The physical examination revealed inflammatory signs and swelling on the knee. The complete blood count, chemical profile and urinalysis were normal, except for a C-reactive protein (CRP) of 100 mg/L. A joint fluid (JF) was drawn by puncture from the affected knee and sent to the microbiology laboratory. The sample was inoculated after centrifugation in aerobic and anaerobic blood agar (BD Columbia Agar 5% Sheepblood®, Becton Dickinson), chocolate agar (BD Choco Agar, Becton Dickinson) and thioglycollate broth (BDTM Fluid Thioglycollate Medium), all incubated at 37º C. Previously, 1 mL of the JF was inoculated into an aerobic blood culture bottle (BACTEC, 9240 BD, Becton Dickinson, Franklin Lakes, NJ, USA), being positive after 24 hours of incubation. Pathogen growth was observed on aerobic blood agar and chocolate agar. Identification of C. albicans and susceptibility to this strain were then tested using the Vitek system (BioMérieux, Mercy L'Etoile, France) as well as mass spectrometry (Bruker Biotyper, Billerica, MA, USA). The isolate was susceptible to anidulafungin (0.015 µg/ml), micafungin (<0.008 µg/ml) caspofungin (0.06 µg/ml), voriconazole (<0.008 µg/ml), itraconazole (0,03 µg/ml) fluconazole (0,125 µg/ml), and amphotericin B (1 µg/ml). Breakpoints from CLSI were used for the majority of antifungal drugs, but only from EUCAST for amphotericin B 6, 7. No blood cultures were taken at this stage. The patient rejected a prosthesis exchange, so a surgical procedure was then done with local debridement of the lesion, and five intraoperative periprosthetic tissue samples were taken from different locations, following recommendations of Kamme and Lindberg 8. In the laboratory, each sample (1 cm2) was placed in 3 ml of sterile saline solution and vortexed for 30 seconds. Then, the sample was inoculated in aerobic and anaerobic blood agar (BD Columbia Agar 5% Sheepblood®, Becton Dickinson), chocolate agar (BD Choco Agar, Becton Dickinson) and thioglycolate broth (BDTM Fluid Thioglycollate Medium, Becton Dickinson), all incubated at 37º C., and chromogenic candida agar (CandiSelect TM, Bio-Rad, Redmond, WA, USA) incubated at 30º C. After 18 hours of incubation, microorganism growth was again observed and further identified as C. albicans.

Treatment with caspofungin (50 mg/day) was administered for 14 days, and the patient was then discharged under oral suppressive treatment with fluconazole (200 mg/12 h.) for 6 months. At 2 months of follow-up, the patient remained clinically stable, and laboratory findings were normal. At the moment, the patient has rejected prosthesis reimplantation and is currently waiting for a 6 months of antifungal treatment.

Literature Review

We describe one patient recently seen at the Orthopaedics and Trauma Department of the Hospital Universitario Virgen de las Nieves (Granada, Spain) with PJI due to C. albicans.

Using the key words “fungal prosthetic joint infection” and “candida prosthetic joint infection” we searched MEDLINE (National Library of Medicine, Bethesda, MD), Web of Science, CINAHL, and Cochrane systematic review databases for case reports of this condition. We also checked the references cited in the papers for additional case reports published before 1966.

We traced 75 cases caused by Candida species and described in sufficient detail. These cases, along with our patient, are the basis of the present report. Among others, data on treatment, outcome and follow-up were recorded. A patient was considered to have Candida infection and was then included when a positive preoperative aspiration culture and/or a positive intraoperative culture were obtained. We did not include cases with Candida infection accompanied by another pathogen or cases with insufficient details for comparisons (clinical and laboratory data).

The basis for this review was recently published 9, but we have added three new cases of Candida PJI. Table 1 summarizes the treatment, outcome and follow-up of all cases here reviewed.

Table 1

Treatment, outcome and follow-up of 76 patients with Candida species PJI.

Reference/author	TreatmentAntifungal treatment Surgical treatment		Outcome	Follow-up (months)
51/Zhu Y	Amphotericin BVoriconazole	NR	Cure	3
52/Reddy KJ	FluconazoleAmphotericin B ICS	TEA	Cure	24
46/Artiaco S	FluconazoleMiconazole	Drainage of abscess	Cure	12
37/Lidder S	Amphotericin B	TEA	Cure	24
34/Ueng SWN	FluconazoleAmphotericin B ICS	TEA	Cure	NR
34/Ueng SWN	Fluconazole	TEA	Cure	NR
34/Ueng SWN	Fluconazole	TEA	Cure	NR
34/Ueng SWN	FluconazoleAmphotericin B ICS	TEA	Cure	NR
34/Ueng SWN	FluconazoleAmphotericin B ICS	TEA	Cure	NR
34/Ueng SWN	Fluconazole	RA	NR	NR
34/Ueng SWN	Fluconazole	RA	NR	NR
34/Ueng SWN	Fluconazole	RA	Dead	NR
42/Anagnostakos K	Caspofungin	TEA	Cure	28
42/Anagnostakos K	Fluconazole	TEA	Cure	22
42/Anagnostakos K	Fluconazole	TEA	Cure	70
42/Anagnostakos K	Fluconazole	TEA	Cure	15
42/Anagnostakos K	Fluconazole	TEA	Cure	36
42/Anagnostakos K	Fluconazole	TEA	Cure	47
47/Bartalesi F	VoriconazoleCaspofungin + amphotericin BAmphotericin B	TEA	Cure	48
35/Wu MH	FluconazoleAmphotericin B ICS	RA	Cure	12
11/ Kelesidis T	Fluconazole	NR	Cure	12
12/Graw B	Fluconazole	TEA	CureDead (UD)	240
13/Bland CM	Liposomal amphotericin B + micafungin + fluconazole	RA	NR	NR
43/Dumaine V	Caspofungin + flucytosineFluconazole + flucytosine	RAArthrodesis	Cure	15
53/Lejko-Zupanc T	Liposomal amphotericin B + fluconazoleCaspofungin	RA	Cure	36
54/Fabry K	Voriconazole (3 days)Oral voriconazole (7 months)	NR	CureDead (UD)	24
27 (/2004) Gaston G	VoriconazoleAmphotericin BAmphotericin B ICS	RA	Amputation	6
48/Lazzarini L	Amphotericin B	RA	Cure	48
15/Wyman J	FluconazoleAmphotericin B	TEA	Cure	36
16/Phelan DM	Amphotericin BKetoconazoleFluconazole	TEA	Cure	73
16/Phelan DM	NR	TEA	Cure	51
16/Phelan DM	Amphotericin BFluconazole	TEA	Cure	70
16/Phelan DM	FluconazoleFluconazole ICS	TEA	Cure	17
55/Açkgöz CZ	Fluconazole	RAArthrodesis	Cure	7.5
38/Bruce ASW	FluconazoleFluconazole IB	TEA	Cure	84
38/Bruce ASW	FluconazoleFluconazole IB	TEA	CureDead (UD)	48
56/Marra F	FluconazoleAmphotericin B ICS	RA (twice)	E. coli infectionNR final outcome	NR
44/Merrer J	Fluconazole	NR	CureDead (UD)	11
36/Yang SH	Fluconazole	TEA	Cure	48
39/Ramamohan N	Amphotericin B + 5-flucytosine	TEA	Cure	24
57/Badrul B	Fluconazole	TEA	MRSA infectionCure	60
49/Wada M	Fluconazole	NR	Cure	36
17/Brooks DH	Amphotericin BFluconazole	NR	Cure	24
40/Selmon GPF	Amphotericin BItraconazole + fluconazole	OEA	Cure	48
18/Simonian PT	Ketoconazole	NR	Cure	72
50/Fukasawa N	Fluconazole	NR	P. aeruginosainfectionCure	24
19/Cushing RD	Fluconazole	NR	Cure	12
58/Nayeri F	5-flucytosine + amphotericin B5-flucytosine + itraconazole	RA	Cure	22
20/Hennesy MJ	Amphotericin B5-flucytosine	TEA	Cure	24
21/Cardinal E	Amphotericin B	RA	CureDead (UD)	NR
21/Cardinal E	Amphotericin B	RA	CureDead (UD)	12
21/Cardinal E	Fluconazole	RA	Cure	6
22/White A	FluconazoleAmnphotericin BItraconazole	RA	Cure	24
23/Tunkel AR	Amphotericin BKetoconazoleFluconazole	RA	Amputation	NR
41/Paul J	Amphotericin B + 5-fluorocytosineKetoconazole	RAArthrodesis	Cure	24
24/Evans RP	Amphotericin B	TEA	Cure	24
24/Evans RP	Amphotericin B	TEA	S. aureus infectionCure	60
25/Darouiche RO	Amphotericin B	RA	Cure	8
25/Darouiche RO	Amphotericin B	RA	Cure	1.5
25/Darouiche RO	Amphotericin BKetoconazole	RAArthrodesis	Cure	36
25/Darouiche RO	Amphotericin B	RA	Cure	5
26/Lambertus M	Amphotericin B	RA	S. epidermidis infectionCure	24
26 /Lambertus M	Amphotericin BKetoconazole	RAArthrodesis	Cure	14
27/Levine M	Amphotericin B	RAArthrodesis	Cure	24
28/Iskander MK	Amphotericin BKetoconazole	RAArthrodesis	Cure	NR
29/Koch AE	Amphotericin B5-flucytosineKetoconazole	RAArthrodesis	Cure	21
30/Lim EVA	Amphotericin B	RA	Cure	28
31/Younkin S	5-fluocytosine + amphotericin B	TEA	Cure	24
32/Lichtman EA	Amphotericin BKetoconazole	RA	Cure	3
33/Goodman JS	Amphotericin B	RA	Cure	NR
33/Goodman JS	Amphotericin B	TEARAArthrodesis	Cure	12
10/MacGregor RR	Amphotericin B + 5-flucytosine	RA	Cure	12
45/Jenny JY	Caspofungine + 5-flucytosineVoriconazole + 5-flucytosine	OEA	Cure	24
45/Jenny JY	Voriconazole + 5-flucytosineFluconazole + 5-flucytosine	OEA	Cure	36
9/Cobo F	CaspofunginFluconazole	No surgical treatment	Cure	6
PR/Cobo F	Caspofungin 14 daysFluconazole (expected 6 months)	No surgical treatment	No relapse	3

UD: unrelated disease; NR: not reported; PR: present report; TEA: two-stage exchange arthroplasty; RA: Resection arthroplasty; OEA: one-stage exchange arthroplasty

ICS: impregnated cement spacer; IB: impregnated beads

General Characteristics

There were 40 (52.6%) women, while the sex was not reported in two patients. The mean age of patients was 65.7 (± SD 18) yrs (range 35-93 yrs). Thirty-five cases were from the USA 10-33, 10 from Taiwan 34-36, six from the United Kingdom 37-41, six from Germany 42, four from France 43-45, three from Italy 46-48, two each from Japan 49, 50 and Spain [9, and present report], and one each from China 51, India 52, Slovenia 53, Belgium 54, Turkey 55, Canada 56, Malaysia 57 and Sweden 58. No risk factors for candidal infection were found in 25 patients (32.8%). The joint involved was the knee in 38 (50%), the hip in 36 (47.3%) and the shoulder in two cases. Pain was reported by 44 (57.8%) patients, and the second most frequent symptom was swelling in 24 (31.5%) patients. Symptoms were not reported for 18 (23.6%) patients.

The most frequently isolated Candida species was C. albicans, found in 36 (47.3%), followed by C. parapsilosis in 17 (22.3%), C. glabrata in 12 (15.7%) and C. tropicalis in 8 (10.5%), with infection by both C. albicans and C. glabrata in one patient 40. Candida species were diagnosed by culture of joint fluid (JF) aspirate in 33 cases (43.4%), culture of intraoperative sample (IoS) in 18 (23.6%) and culture of both JF and IoS samples in 19 (25%). Blood cultures were taken only in eight (10.5%) patients, and were positive for Candida in four of these (50%).

Antifungal treatment

Seventy-five (98.6%) patients underwent antifungal treatment, with a single drug in 46 cases (60.5%), with two drugs in 16 cases (21%) and more than two in 13 (17.1%). Twenty-nine of the patients with monotherapy (63%) were treated with fluconazole and 14 (30.4%) with amphotericin B. One patient treated with amphotericin B alone suffered recurrence of the infection 33, while another patient treated with fluconazole alone died as a consequence of the infection 34.

Application of antifungal spacer cement was applied in ten (13.1%) patients (seven with amphotericin B and three with fluconazole).

Surgical treatment

Surgery was performed in 65 (85.5%) patients, 30 of whom (39.4%) underwent two-stage exchange arthroplasty; resection of arthroplasty without reimplantation was undertaken in other 30 (39.4%). One-stage exchange arthroplasty was undergone in 3 (3.9%) patients, and other procedures in two patients. Surgical treatment was not reported in 11 (14.4%) patients.

Outcome

The final outcome was not reported in four patients. A favourable outcome was found in 58 (76.3%) patients after antifungal plus surgical treatment, in 11 (14.4%) after antifungal treatment alone and in one after surgery alone. A patient experienced recurrence of infection with fluconazole therapy, but his outcome was positive with miconazole plus drainage of fluid abscess 46.

Regarding to the type of surgery, all patients who underwent a two-stage exchange arthroplasty cured, although two of them suffered a bacterial infection 24, 57. From the patients treated with resection arthroplasty, one dead after fluconazole treatment 34, two suffered amputation 14, 23 and two were found to have a bacterial infection 26, 56. All three patients treated with one-stage exchange arthroplasty cured 40, 45. Our two patients are currently well with antifungal treatment only, and no relapse of disease has been observed in the short follow-up.

Discussion

PJI caused by Candida species is still a rare disease. However, the incidence is expected to rise because of the increasing number of patients implanted worldwide with joint arthroplasties 59, 60.

Risk factors for candidal infection, including immunosuppression, systemic disease and/or long-term antibiotics use, may play an essential role in the development of invasive candidal infections although other factors could be involved in triggering the infection, specially the presence of biofilm on bioprosthetic surfaces. Biofilm formation is considered the most prevalent growth form of microorganisms 61 and plays a key role in the development of clinical infections 62. The majority of C. albicans infections are associated with biofilm formation on the host or on the surfaces of medical devices or prostheses 63. Other factors, such as the adherence of C. albicans and their hydrolytic enzyme secretion may also have a strong influence on the development of PJI 64, and their modification may serve as possible targets for antifungal drugs against these infections.

Pain and swelling are the main symptoms of PJI due to Candida species, although the onset of symptoms can be insidious and development of the disease can be slow. Because symptoms are mild and there is frequently no diagnostic suspicion of PJI caused by Candida, the diagnosis can often be delayed. Another important problem is to elucidate whether the presence of Candida species in samples can be considered as a contaminant or not, because there is still no standard definition of PJI.

The treatment of choice for PJI caused by Candida species has not yet been established. The use of antifungal agents locally (mixed with cement) or systemically administered is a challenging issue. Locally, amphotericin B appears to be the ideal drug, but some studies have reported several problems 65, 66, while there is no report on the use of novel antifungal drugs. However, this option has not been usually used, because from 76 cases reviewed, antifungal spacers were applied in only 10 (the majority of them with amphotericin B). In all cases the outcome was positive, but curiously in two patients with resection arthroplasty plus amphotericin B impregnated cement spacer application, an amputation 14 and a bacterial infection 56 was observed.

For systemic administration, lipid formulations of amphotericin B and fluconazole are the drugs of choice for this type of infection, and echinocandins may be an option 67.

Various authors have analysed the activity of some antifungal drugs against Candida biofilms. Two reports described resistance to fluconazole in these structures 68, 69, while another study found that it interfered with the development of C. albicans biofilms 70. On the other hand, lipid formulations of amphotericin B have shown activity against C. albicans biofilms 70. Anidulafungin was more active than amphotericin B against C. albicans biofilms of 24-h maturation, but amphotericin B was more active than anidulafungin against C. albicans biofilms of 48-h maturation 71.

In this review, various types of antifungal drugs have been used for treatment, and the majority of them with a positive result. In table 1, it can see that eleven patients (14.4%) were treated only with antifungal drugs [9, 11, 17-19, 44, 49, 50, 51, 54, and present report]; all these patients obtained a positive outcome, although the follow-up range from 3 to 72 months. According to these results, a correct and long treatment with antifungal drugs may be a good option, but due to the heterogeneity of the studies further research is required on this important issue, although it is highly recommended that these patients should be treated with drugs selected after antifungal susceptibility tests. In addition, a longer follow-up of these patients should be performed.

A two-stage arthroplasty exchange is currently considered the best approach in terms of eradication of the infection and preservation of the joint function in PJI caused by Candida species 4, 5. Furthermore, when infection is chronic, this type of surgery is generally also recommended 72. However, the success rate of this technique is controversial. One study reported a success rate of 93% for short-term infection control with 6 months of oral antifungal drugs after reimplantation 73, while another found that two-thirds of patients with PJ resection for fungal infection underwent reimplantation and that the infection was abolished in less than half of them 4.

On the other hand, resection arthroplasty was performed in around half of patients here reviewed (n=30). From these patients, in five of them the outcome was not initially favourable, and in three of them the final outcome was not reported. These results indicate that this procedure should not be the initial approach, also due to the important joint functional loss.

Regarding the third surgical method, some years ago Selmon et al reported a case with a positive outcome after a one-stage exchange arthroplasty 40. Moreover, some authors have recently reported favourable outcome after one-stage exchange in selected cases of fungal PJI 45, 74. In these cases, the responsible microorganism was identified post-operatively, with delayed specific antifungal treatment. In spite of this fact, the scientific evidence about these cases suggests that this technique may be appropriate in terms of eradication of infection, although the number of cases is still scarce. Further research should be done in order to establish the possible indications of one-stage exchange arthroplasty in cases of Candida PJI.

In summary, PJIs caused by Candida species are rare but fastidious infections that require a high index of suspicion because of their mild symptoms and insidious evolution. The diagnosis must be confirmed microbiologically and antifungal susceptibility testing of Candida strains is also highly recommended. The treatment is still not clearly defined and, although the association of long-term antifungal use with two-stage exchange arthroplasty is currently the gold standard to eradicate the infection, the analysis of the data of this review suggests the possibility of using one-stage exchange arthroplasty or antifungal treatment alone in order to obtain a favourable outcome for these patients. The next challenge for the scientific community is to establish the adequate algorithm for treatment in fungal PJI.

Informed consent

The patient described in this case report gave her informed consent for the inclusion in this publication.