Clinical Archives of Bone and Joint Diseases Metabolic Diseases and Crystal Induced Arthropathies Technic of Non-Staining Histologic Sections - A Comparative Study of Stan dard Stains and Histochemical Reactions

Arthropathy induced by monosodium salt of uric acid [C 5 H 4 N 4 O 3 ] (MSU) (gout), by calcium pyrophosphate dihy- drate [Ca 2 P 2 O 7 .2H 2 O] (CPPD) crystals (chondrocalcinosis, pseudogout, pyrophosphate arthropathy) and arthropathy induced by hydroxyapatite [Ca 5 (PO 4 ) 3 (OH)] (HA) crystals (apatite rheumatism, hydroxyapatite arthritis, calcifying te-nosynovitis, Milwaukee syndrome, calcific tendinitis, calcific periarthritis) are regarded as distinct clinical entities. The solubility of MSU, CPPD and HA crystals in conventional fixatives (aqueous formaldehyde solution), in alcohol, ace -tone, and xylene or in solutions of dyes vary. The crystals in of 105) tissue samples of 37 (78.72% of 47) patients; CPPD in 15 (60.00% of 25) tissue samples of 10 (62.50% of 16) patients. HA crystals were detected exclu sively with this method: in all tissue samples (in 19 of 19; 100.0%) of all patients (in 4 of 4; 100.0%), none with the traditional stains and reactions. The non-staining technique is a simple and very effective method to demonstrate crystal deposits in tissue samples. Handbooks of histologic methods and histochemistry do not mention this simple technique. In case of clinically or histologically suspected meta bolic or crystal-induced diseases it is advisable to analyze unstained tissue sections as well, supplemented with the traditional hematoxylin-eosin stained ones. This approach may also be useful in other crystal deposition induced diseases or identification of foreign bodies and refractile arte -facts.


Introduction
Metabolic diseases and crystal induced arthropathies are characterized by deposits of crystal and/ or non-crystalline (amorphous) calcium phosphates in synovial membranes (synovium) and periarticular soft tissues. Bone and cartilage may also be involved, and crystals may be present in the synovial fluid as well. Foreign bodies and refractile artefacts may vary the histopathological findings. Different crystals (with or with-   submicroscopic [1] 50-500 nm [2] clusters of crystals:  [3] Liquid lipid crystals- Figure 5 Maltese cross, positive variable spherules 0.5-30 μm [3] Lithium heparin positive weak polymorphous 2-5 μm in alcohol, acetone, and xylene or in solutions of dyes vary. The crystals in tissues may dissolve during fixation in aqueous formaldehyde solution, embedding in paraffin or during staining [15,16]. Only those minerals or crystals can be detected microscopically by staining or histochemical reactions which remain in tissue sections after fixation, paraffin embedding and staining [17][18][19][20][21]. The probability of identification of crystals is much higher in unstained sections viewed under polarized light than in traditionally stained ones [22][23][24][25]26]. The aim of this study was to compare the "non-staining" technique according to Bély and Apáthy (2013) [22] with worldwide accepted stains and histochemical methods such as hematoxylin-eosin (H-E) staining, Gömöri's methenamine silver method, Schultz staining, Alizarin red S staining, and von Kossa's reaction [17][18][19][20][21] in tissue samples of patients with clinically diagnosed gout, chondrocalcinosis and apatite rheumatism in order to demonstrate the effectivity of crystal detection with these standard methods in comparison with non-staining technique [22].

Patients and Methods
One hundred and five (105) tissue samples of 47 patients with clinically diagnosed gout, 25 tissue samples of 16 patients with clinically diagnosed chondrocalcinosis, and 19 tissue samples of 4 patients with clinically diagnosed apatite rheumatism were studied. Demographics are summarized in Table 2.
The biopsy material (tissue samples, surgical specimens) was varied: articular and periarticular soft tissues of different joints (elbow, knee, carpal synovium, subcutaneous nodules etc.); analysis of clinical relationships, localization, symptoms, radiological signs etc., were not considered. The tissue blocks were fixed in an 8% aqueous solution of formaldehyde [CH 2 (OH) 2 ] at pH 7.6 for > 24 hours at room temperature (20 °C) and embedded in paraffin. Tissue samples of gout fixed in alcohol (in clinically recognized cases) were excluded from this study, in order to compare the results under the same circumstances.
Using Bély and Apáthy's "non-staining" technique, the fixation of tissue blocks in 8% aqueous solution of formaldehyde at pH 7.6 for > 24 hours at room temperature (20 °C) and embedding in paraffin were the same as with the standard stainings and reactions. Unstained tissue sections were deparaffinized, mounted and cover slipped with Canada balsam.
The standard and unstained sections were examined with a professional polarizing light microscope (Olympus BX51). In selected cases the nature of crystals was confirmed by a JEM 100CX electron microscope and electron diffraction as well (Figure 1a, Figure 1b, Figure  1c and Figure 1d).

Results
The average age of patients with gout was significantly lower than the average age of patients with chondrocalcinosis (p < 0.050). There was no significant difference between the average age of patients with chondrocalcinosis and with apatite rheumatism (p < 0.129).  and hydroxyapatite arthritis (in case of standard stains and reaction in comparison with Bély and Apáthy's "non-staining" technique) is summarized in Table 3.

Detection of monosodium salt of uric acid [C 5 H 4 N 4 O 3 ] (MSU) crystals
Comparing the classical staining methods and histochemical reactions, there was a significant difference between their effectivity (sensitivity). The Gömöri or Schultz stains were more effective in detection of MSU  Figure 1: a) Urate crystals, surface electron micrograph, x1600; b) Urate crystals, transmission electron micrograph, x6000; c) CPPD crystals, surface electron micrograph, x10000; d) Hydroxy apatite crystals, surface electron micrograph, x50000; e) Cholesterol crystals, surface electron micrograph, x1300, Original magnifications correspond to the 600 × 900 mm negative. The printed size may be different; therefore, it is necessary to indicate the original magnifications corresponding to a fixed size (in case of electron micrographs this is the 6 × 9 cm analogue negative).

Detection of calcium pyrophosphate dihydrate [Ca 2 P 2 O 7 .2H 2 O] (CPPD) crystals
CPPD crystals were not detected in combination with MSU crystals in our patient cohort's. The HE staining was more effective in detection of CPPD crystals than Alizarin red S staining or von Kossa reaction, and Bély and Apáthy's non-staining technique detected many more CPPD crystals than HE staining (Table 5). Using these methods there was a difference in the number of detected crystals, but regarding the effectivity (sensitiv-than with H-E, and the non-staining technique of Bély and Apáthy's was much more sensitive than all of these ( Table 4). The differences between these methods, regarding the effectivity and sensitivity, were significant, except for the number of patients. Though more patients were positive for urate crystals with the none staining technique than with Schultz's stain, but this difference was statistically not significant. The statistical correlations ("p" values of significance) are summarized in Table 4, comparing tissue samples and involved patients with gout with different stains and techniques.  [19,21] NA NA 4 (25.0) 7 of 24 * (29.17) 0 (0.0) 0 (0.0) von Kossa [17,21] NA NA 2 (12.5) 2 of 24 * (8.33) 0 (0.0) 0 (0.0) Bély and Apáthy [22,25] 37 (78.72) 83 (79.05) 10 (62.5) 15 (60.00) 4 (100.0) 19 (100.0) Table 3: Only the presence of crystals was registered in Pts and Ts (yes or no); the amount of crystal deposits was not evaluated.

Remark to
Pts-Patients; Ts-Tissue samples.
All tissue samples were fixed in an 8% aqueous solution of formaldehyde at pH 7.6 for > 24 hours at room temperature (20 °C) and embedded in paraffin.
NA-Not Analyzed.
Gouty tophi were not analyzed with Alizarin red S staining or von Kossa's reaction. Tissue samples with chondrocalcinosis or apatite rheumatism were not evaluated with Gömöri's methenamine silver method, because CPPD and HA crystals do not stain with this method [18,20]. Schultz's stain is specific for MSU (monosodium salt of uric acid), uric acid and cholesterol, therefore all tissue samples with gout, chondrocalcinosis and apatite rheumatism were analyzed for cholesterol with Schultz's stain [17].
*Some tissue sections were lost during histological processing or deposits were not present in deeper sections of the tissue blocks.  clusters were not detected in combination with MSU crystals.

Detection of cholesterol [C 27 H 46 O] crystals
Cholesterol [C 27 H 46 O] crystals were present only in tissue sections stained according to Schultz, and in unstained sections according to Bély and Apáthy; cholesterol crystals were not detected with H-E, Gömöri, and Alizarin Red S staining or with von Kossa's reaction. In tissue sections stained according to Schultz, cholesterol was detected in case of gout in 10 (21.27%) of 47 patients (with or without with MSU) and in case of chondrocalcinosis or apatite rheumatism, in 14 (77.77%) of 18 patients without CPPD or HA crystals (CPPD or HA crystals ity) of these methods these differences were not significant statistically in most cases ( Table 5). The statistical correlations ("p" values of significance) are summarized in Table 5, comparing tissue samples and involved patients with chondrocalcinosis with different stains and techniques.

Detection of hydroxyapatite [Ca 5 (PO 4 ) 3 (OH)] (HA) crystals
Clusters of HA crystals and aggregates were detected only in unstained sections according to Bély and Apáthy in combination with more or less CPPD crystals together (significance was not calculated; there were no comparable values). Clusters of HA crystals and aggregates of a b c d e f in tissue sections stained according to Schultz.

Discussion
The histological diagnosis of metabolic disorders is based on the presence crystals in tissue sections with or without non-crystalline (amorphous), calcium phosphate and/or carbonate containing mineral deposits. There is a difference in the shape, size, intensity of birefringence, and optical breaking direction of MSU, CPPD, HA and cholesterol crystals. The solubility of these crystals in conventional fixatives (aqueous formaldehyde solution), in alcohol, acetone, and xylene or in solutions of dyes is also different. The crystals in tissues may dissolve during fixation in aqueous formaldehyde solution (formalin), embedding in paraffin or during staining.
In clinically known or suspected cases of gout the surgical tissue specimens should be fixed in absolute ethyl alcohol, because urate crystals are soluble in 8% form-do not stain according to Schultz). In unstained tissue section according to Bély and Apáthy, cholesterol was present with variable prevalence in the entire patient group, but exact analysis was not possible in patients with the clinical diagnosis of gout or chondrocalcinosis, because of large amounts of MSU or CPPD crystals. Under polarized light CPPD crystals show positive birefringence (parallel to the long axis of the crystals analogous to the birefringence of collagen fibers, see: Figure 3c and Figure 3g).
(i) Alizarin red S, viewed with the light microscope, same as (a) x100; (j) same as (c) x200 Non-crystalline calcium containing mineral deposits are staining with calcium specific Alizarin red S. Alizarin red S does not stain the CPPD crystals, and the masses of amorphous calcium phosphate and carbonate may mask the crystals, with no detectable birefringence; (k) von Kossa's reaction, viewed with the light microscope, same as (a) x100; (l) same as (c) x200 Non-crystalline phosphate or carbonate containing mineral deposits show a positive reaction according to von Kossa. The CPPD crystals are negative with von Kossa's reaction, and the masses of amorphous calcium phosphate and carbonate may mask the crystals, with no detectable birefringence.
(m) Intact CPPD crystals and fragments, unstained section, viewed under polarized light, same field as (e-f) x600 The intact CPPD crystals have a rhomboid shape, they range in size is from 5 to 40μm and show a strong birefringence; (n) Unstained section, Red I compensator, viewed under polarized light, same field as (g-h and m) x600 Axis parallel direction of birefringence of CPPD crystals is positive. Our results indicate that the very simple "not-staining" technique is a most effective method to demonstrate crystal deposits in tissue samples [15,16,22,25]. Theoretically the largest amounts of crystals may be best preserved in unstained frozen sections. Indeed, large amounts of cholesterol or fatty acid crystals may be visualized in frozen sections under polarized light [35]. The frozen sections are not suggested for diagnosis of metabolic disorders in everyday practice, because large amounts of cholesterol crystals may conceal other crystals.
A disadvantage of unstained sections is that parallel (serial) tissue sections have to be stained traditionally, since detailed histology cannot be studied adequately in unstained sections with the light microscope or under polarized light. Another disadvantage (or advantage) is that in unstained sections other crystals can be found which differ in shape, size, arrangement or quality of birefringence from the well-known crystals, and their aldehyde solution [17,18,20,21]. To quote McManus and Mowry "since urates are slightly soluble in water, alcohol fixation is preferable" [21], but we found that in gout most urate crystals dissolve during the hematoxylin-eosin staining procedure [15,16,[32][33][34]. In present study the alcohol fixed cases of gout were excluded, in order to compare the traditional staining and reaction with Bély and Apáthy's non-staining technique under the same circumstances.
CCPD crystals are less soluble than urate and are likely to be detected in traditionally processed tissue sections. The small and soluble HA or the highly soluble cholesterol crystals are not detected in traditionally fixed, embedded and stained tissue samples. In case of cholesterol deposition, the absence of crystals and characteristic empty spaces accompanied by a typical inflammatory reaction of macrophages and multinucleated giant cells are a reminder of the dissolved crystal deposits. The HA (and CPPD) crystals are accompanied by amorphous calcium phosphate, or calcium carbonate deposits of blue-violet colour. The absence of an inflammatory reaction is characteristic [27].

(c) H-E viewed under polarized light, same as (a) x20; (d) same as (c) x40
In traditionally fixed tissue specimens the HA crystals (crystal clusters and aggregates of clusters) dissolved and are not demonstrable (the sporadic CPPD crystals or fragments are also not visible).
(e) Unstained section viewed under polarized light, same field as (a) x100; (f) same as (e) x200 The individual HA crystals are small, 50-500 nm, rod-shaped and are arranged typically in 1-5 μm spheroid microaggregates, which are not in visible (detectable) range with polarizing microscopy [2]. The Figures m-p demonstrate crystal clusters and aggregates of clusters of 6.5 and 20 mm size, which may appear under plain light microscopy, but according to Forster, et al. [28] without birefringence. Using a professional polarizing microscope with high brightness, the clusters show of a week birefringence. Under polarized light the direction of birefringence is positive according to the long axis of HA crystals, like that of collagen fibers.
The HA crystal clusters (microaggregates) and aggregates of clusters are associated sporadically with much larger and partially fragmented CPPD crystals. The CPPD crystals are less soluble in comparison with HA crystals or crystal aggregates. The birefringence of CPPD crystals is stronger than that of HA crystals or crystal aggregates. The masses of amorphous calcium phosphate and carbonate may mask the crystals, with no detectable birefringence, even in in synovial fluid "identifying individual calcium hydroxyapatite crystals can be elusive" [28].
(g) Unstained section, Rot I compensator, viewed under polarized light, same field as (a and e) x100, (h) same as (g) x200.
Under polarized light HA and CPPD crystals show positive birefringence parallel to the long axis of the crystals, but the intensity of birefringence of HA is much weaker than that of CPPD crystals.
(m) Unstained section viewed under polarized light, same field as (e) x200, (n) same as (m) x600 The 50-500 nm small, rod-shaped individual HA crystals are arranged typically in 1-5 μm spheroid clusters (microaggregates) and larger aggregates of clusters [28], sporadically associated with fragmented CPPD crystals. The CPPD crystals are larger, have a rhomboid shape, and show a strong birefringence in comparison with HA. The probability of identifying crystals is much higher in unstained sections viewed under polarized light than in haematoxylin-eosin stained ones. Textbooks of histologic methods and histochemistry do not mention this simple technique.
In case of clinically or histologically suspected metabolic or crystal induced diseases the analysis of tissue samples is suggested with unstained tissue sections as well, supplemented with the traditional hematoxylin-eosin staining. This approach may also be useful in other crystal deposition induced diseases or identification of foreign bodies and refractile artefacts.

Disclosure/Conflict of Interest
There is no conflict of interest.

Disclosure Statement
This work did not receive financial support from any source.
Authors contribution is equal.
The manuscript has been changed according to the required order of guidelines. identification would require further specific (electron microscopic, electron diffraction, etc.) studies.
Major textbooks of histochemistry discuss many techniques and staining methods to demonstrate preserved crystals and mineral deposits in tissue, but none mention the simplest method, namely viewing of unstained tissue sections with polarized light [17][18][19][20][21].
In his book Mohr demonstrated crystals in unstained tissue sections (independent of us) but does not mention the advantage of this method in comparison with traditional stainings [9]. According to our best knowledge a detailed analysis or comparative study of our non-staining technique and its comparison with traditional stainings and reactions has not been available in the literature.
In case of suspected metabolic or crystal induced disorders, we suggest analyzing the tissue samples with unstained tissue sections as well, supplemented with traditional stainings and reactions. Crystals remain detectable in unstained sections viewed under polarized light in the great majority of cases which appear negative with H-E staining [22][23][24][32][33][34].

Conclusions
Bély and Apáthy's non-staining technique is a simple and sensitive method and may help in the microscopic demonstration and analysis of crystalline deposits.