A comparative study of intraoperative frozen section and alpha defensin lateral flow test in the diagnosis of periprosthetic joint infection

Figures & data

Figure 1. Sensitivity (

) and specificity (

) of different tests in diagnosing PJI when using the MSIS criteria.

Table 1. Distribution of microorganisms

Isolated pathogens	Conventional culture (n = 29)
Staphylococcus aureus	5
Coag. negative Staphylococci	4
Escherichia coli	3
Cutibacterium acnes	1
Streptococcus agalactiae	1
Enterococcus faecalis	1
Finegoldia magna	1
Actinomyces neuii	1
Parvimonas micra	1
Moraxella osloensis	1
Candida albicans	1
Negative culture	9

Table 2. Performance of the frozen section and alpha defensin using MSIS criteria for diagnosis of PJI

Performance	Frozen section	Alpha defensin
Sensitivity (%)	86.2 (68.7–95.0)	69.0 (50.6–82.8)
Specificity (%)	93.1 (84.3–97.3)	94.4 (86.0–98.2)
Accuracy (%)	91.1 (85.5–96.6)	87.1 (80.6–93.7)
Positive predictive value (%)	83.3 (70.0–96.7)	83.3 (68.4–98.2)
Negative predictive value (%)	94.4 (89.0–99.7)	88.3 (81.1–95.5)
Positive likelihood ratio	12.4 (5.3–29.3)	12.4 (4.6–33.2)
Negative likelihood ratio	0.15 (0.06–0.37)	0.33 (0.19–0.57)
Area under the curve	0.90 (0.83–0.97)	0.82 (0.73–0.91)

The 95% confidence interval is presented in parentheses

Table 3. False negative and false positive alpha defensin test or frozen section based on Musculoskeletal Infection Society (MSIS) criteria

Patient	Sex	Age	Joint	MSIS	Alpha defensin	Frozen section	Permanent section^a	Sinus tract	CRP (mg/L)	WBC (cells/µL)	PMN (%)	Culture^b	Antibiotics
False negatives
1	M	59	Knee	PJI	Negative	Positive	Type 3	No	11.2	1,456	67	Staph. epidermidis	No
2	F	78	Hip	PJI	Negative	Negative	Type 3	No	6.6	645	54	Enterococcus faecalis	No
3	M	81	Hip	PJI	Negative	Negative	Type 1	No	23.0	ND	ND	MRSA	No
4	M	44	Hip	PJI	Negative	Positive	Type 2	No	65.1	91,343	92	No growth	No
5	M	22	Hip	PJI	Negative	Positive	Type 3	No	45.0	44,853	91	No growth	No
6	F	86	Hip	PJI	Positive	Negative	Type 1	No	17.5	12,987	86	Cutibacterium acnes	No
7	F	82	Hip	PJI	Negative	Positive	Type 2	No	40.2	26,640	97	Staph. sacchrolyticus	No
8	F	71	Knee	PJI	Negative	Positive	Type 3	No	46.6	14,638	82	No growth	No
9	F	83	Knee	PJI	Negative	Positive	Type 2	No	160.0	5,068	87	No growth	Yes
10	F	61	Hip	PJI	Negative	Negative	Type 1	No	8.9	ND	ND	Moraxella osloensis	No
False positives
11	F	71	Hip	AF	Positive	Positive	Type 3	No	4.0	ND	ND	No growth	No
12	F	60	Hip	AF	Positive	Positive	Type 2	No	2.9	< 1.0	–	No growth	No
13	F	69	Knee	AF	Positive	Negative	Type 1	No	5.2	2,212	88	No growth	No
14	F	71	Hip	AF	Negative	Positive	Type 2	No	2.6	16,420	71	No growth	No
15	M	74	Hip	AF	Negative	Positive	Type 2	No	1.6	< 1.0	–	No growth	No
16	F	69	Hip	AF	Positive	Negative	Type 1	No	11.8	< 1.0	–	No growth	No
17	M	45	Knee	AF	Negative	Positive	Type 3	No	3.4	< 1.0	–	No growth	Yes

a According to Krenn and Morawietz classification (Morawietz et al.2009).

b 2 or more periprosthetic cultures grew phenotypically identical organisms. None of these patients showed bacterial growth in only 1 sample (synovial fluid, tissue, sonication fluid).

PJI = periprosthetic joint infection; AF = aseptic failure; CRP = serum C-reactive protein; WBC = synovial fluid leukocyte count, PMN = polymorphonuclear neutrophils.

Figure 2. Receiver operating characteristic curves for diagnostic accuracy of periprosthetic joint infection based on the alpha defensin lateral flow test and the frozen section when using the MSIS criteria. There is a statistically significant difference between the two ROC curves (p = 0.006).

Table 4. Preoperative and postoperative diagnosis of PJI according to the MSIS criteria and intraoperative diagnosis by using alpha defensin lateral flow test and frozen section. Values are frequency (percentage)

	Patients (n)	Preop.PJI	Postop.PJI	Intraoperative positive
				alpha defensin test	frozen section
Minor criteria
0 positive	61	0 (0)	4 (7)	3 (75)	3 (75)
1 positive	19	0 (0)	4 (21)	1 (25)	2 (50)
2 positive	13	0 (0)	13 (100)	9 (69)	12 (92)
3 positive	7	7 (100)	7 (100)	6 (86)	7 (100)
Major criterion	1	1 (100)	1 (100)	1 (100)	1 (100)

Minor criterion (serum CRP, synovial fluid leukocyte count, or percentage of polymorphonuclear neutrophils, synovial fluid culture). Major criterion (sinus tract). PJI = periprosthetic joint infection.

Morawietz L, Tiddens O, Mueller M, Tohtz S, Gansukh T, Schroeder J H, Perka C, Krenn V. Twenty-three neutrophil granulocytes in 10 high-power fields is the best histopathological threshold to differentiate between aseptic and septic endoprosthesis loosening. Histopathology 2009; 54(7): 847–53.

 PubMed Web of Science ®Google Scholar