Introduction

The use of alloplastic material in implantation surgery has become a standard procedure. Whether it involves mesh implantation in hernia surgery, heart valve replacement in cardiology or total joint implantation in orthopedic surgery, these procedures belong to the most frequently performed operations. Through the increasing use of alloplastic materials implant-associated complications have naturally risen. One of the most severe complications is infection of the implant. Diagnosing implant infection includes clinical examination, laboratory parameters, microbiological samples, radiological imaging and is followed by antibiotic and surgical treatment often leading to explantation of the implant [1]. While acute infections can be diagnosed relatively easy, diagnosing chronic low-grade infection is challenging [2]. In periprosthetic joint implantation surgery low-grade infections are known to cause severe complications such as loosening of the prosthesis, pain, pseudoarthrosis and immobilization leading to secondary complications [3]. This extends the hospital stay and increases financial costs for the health system and the patient’s mortality [4,5]. In unclear cases and to secure diagnosis a biopsy of periprosthetic tissue is necessary, following histopathological examination[1]. Histopathological examination thereby relies on the number of neutrophils per high-power-field [6]. Discriminating low-grade from high-grade infection is possible using the CD-15Focus-Score [7]. However, applying this method to other implants and tissue presents notable challenges. While periprosthetic tissue is cuttable, alloplastic material embedded in the tissue impedes the sample preparation resulting in poor sample quality and suboptimal examination conditions. Especially in unclear cases poor sample quality is devastating since treatment of septic and aseptic implant failure differs.

Matrix-assisted laser desorption/ionization time-of-flight (MALDI TOF) is a technique originally used in forensic science for identification of unknown substances and drugs [8,9]. In the following years MALDI TOF was used in microbiology to identify subspecies of bacteria as well as their antibiotic resistances [10,11]. In the progress MALDI TOF imaging mass spectrometry (IMS) emerged as a tool to unite histopathological examination with mass spectrometry, combining the molecular assessment with morphology of the tissue and showing their spatial distribution within the tissue. MALDI TOF IMS has shown promising results in tumor typing and classification, offering the possibility to use specific m/z values as biomarkers to discriminate tissue types [12]. Gravius et al successfully detected specific neutrophil peptides using MALDI TOF IMS in periprosthetic tissue [13].In the present study we used infected explanted meshes as an example for implants to investigate the potential of MALDI TOF IMS in diagnosis of low-grade implant infection.

Materials and Methods

Sample Preparation

15 formalin-fixed paraffin-embedded (FFPE) specimens with explanted meshes were cut as serial sections. For each case serial sections were cut for MALDI and mounted onto an ITO-slide, following two sections for CD-15 immunohistochemistry and histological examination, which were stained with hematoxylin and eosin. All 15 cases were histopathologically diagnosed as infected tissues and neutrophil-rich areas were annotated by a pathologist on the H&E slide (B.K.). Diagnosis of infection was performed by counting the number of neutrophils per high power field using the synovitis score established by Krenn et al. [14] Infected tissues were histologically distinguished in low-grade and high-grade infected tissues. For sample preparation for MALDI IMS we performed deparaffinization, antigen retrieval and in situ digestion with trypsin as shown in Figure 1.

Figure 1: Workflow of MALDI TOF IMS: the specimen is cut as serial sections. The hematoxylin and eosin (H&E) samples and CD-15-immunohistochemistry samples are used for histopathological examination (HE) and annotation of areas with a high number of neutrophils. Parallel the serial section is used for MALDI TOF IMS. After sample preparation including deparaffinization, antigen retrieval, trypsin and Matrix deposition MALDI TOF IMS can be performed. Predictive m/z values (here shown for 788.405 and 944.533) working as biomarkers were applied to detect neutrophil activity in tissue. Afterwards we compared areas with high intensity for applied biomarkers to histologically annotated areas

MALDI Analysis

Following the instructor manual, CHCA matrix solution (7mg/ ml in 50/50 acetonitrile/0,5% TFA) was applied onto the digested sections by the same ImagePrep devise. Afterwards MALDI imaging analysis was performed using rapifleX^® (Bruker Daltonik, Bremen, Germany). Mass spectra within a range of mass to charge ratio (m/z values) from 500-4000 were acquired. After data processing, according to the manual including baseline subtraction, realignment, and normalization, we searched for specific m/z values (Table 1) as biomarkers to detect neutrophils with MALDI TOF and compared regions with high intensity for the applied m/z values to the histologically annotated areas.

Table 1

Results

Using meshes as an example for explanted alloplastic material we were able to demonstrate factors that are impeding the diagnosis of infection by standard histological and immunohistological procedures. Even though tissues with the embedded meshes are still cuttable, the sample preparation is challenging, sometimes resulting in poor sample qualtiy (Figures 2(a,b) and 3(a-c)). Due to squeezing artifacts and poor cutting quality, PMNs cannot be adequately identified using light microscopy. The visualization of PMNs is also difficult using CD15 immunohistochemical staining due to the strong background staining. A reliable determination of the density of PMNs for the diagnosis of a low-grade infection is therefore not possible. As a proof of principle, we demonstrate in the present study that detection of neutrophil peptides in tissue with explanted meshes is possible with MALDI TOF IMS independent from the section or tissue quality. Specific m/z values as biomarkers to detect neutrophil activity indicated high intensity in infected meshes. The histologically annotated regions matched with areas with high intensity for predictive proteins listed in Table 1, as well as with the spatial distribution of the neutrophils in the CD15 immunohistochemistry (Figure 4(a-c)). Comparing the intensity of applied m/z values from neutrophil-rich to neutrophil-poor areas significant differences were observed. Using SCiLS Lab software, we were able to automatically find a co-localized m/z value as presented in Figure 4b within the tissue. The co-localized m/z value has been identified as the Calgranulin B (S100A9). In addition to the already annotated regions we detected further regions with high numbers of neutrophils using MALDI TOF IMS.

Figure 2a: hematoxylin-eosin section presenting the spatial distribution of neutrophil granulocytes in the investigated area of interest.

Figure 2b: hematoxylin-eosin section of a different sample, showing the varying quality of the sample

Figure 3a: CD-15-Immunohistochemistry section with tissue folding.

 Figure 3b and Figure 3 c: CD-15 immunohistochemistry section in different samples, showcasing the varying quality of the staining

Figure 4a: Distribution of the applied m/z values, working as biomarkers for neutrophil activity and infection (m/z 660.2 ± 0.59, 788.468 ± 0.59, 944.5 ± 0.59 and 1349.7± 0.59).