Aims In bone marrow (BM) biopsies, tartrate-resistant acid phosphatase (TRAP) staining represents the gold standard for the characterisation of osteoclasts. TRAP is one of the few enzymes that is histochemically detectable on formalin-fixed paraffin-embedded tissue. This study investigated whether TRAP is also able to visualise BM osteoclasts in autopsy tissue. It was hypothesised that, due to a progressive loss of enzymatic activity in osteoclasts post-mortem, TRAP staining could allow the time of death of a patient to be determined.
Methods TRAP-stained BM slides of 96 cases including 51 pathology and 23 forensic autopsies and 22 biopsies were histologically evaluated and their staining intensity (SI) semi-quantitatively graded. In the autopsy cases, the results were correlated with the post-mortem interval (PMI, time span in days between death and autopsy).
Results TRAP staining intensities (TRAP-SIs) did not differ between men and women and showed a steady decrease with age. TRAP-SIs were significantly stronger in biopsies than in autopsy cases. Among the autopsies, TRAP-SIs were highly variable and not dependent on PMI, except for three forensic cases with PMI ≥7 days which showed a complete loss of TRAP stainability. On the whole, the TRAP-SIs of pathology and forensic cases did not differ significantly.
Conclusions This study clearly shows that BM osteoclasts stay TRAP-positive for 7 days post-mortem, although with markedly reduced TRAP-SIs compared with biopsies. Since TRAP-SIs were not correlated with the duration of PMI, TRAP staining of BM osteoclasts cannot serve as a tool to determine the time of death of a patient.
Statistics from Altmetric.com
In the microscopic examination of bone marrow (BM) biopsies, tartrate-resistant acid phosphatase (TRAP; EC 184.108.40.206 5b) staining is used as a marker of osteoclasts in bone.1 TRAP is one of the few enzymes that is histochemically detectable on formalin-fixed paraffin-embedded tissue. We have noticed a discrepancy in TRAP staining intensity (TRAP-SI) between BM biopsies in which osteoclasts usually show a strong reaction, and samples of autopsy cases whose osteoclasts have often lost most of their positivity. To our knowledge, detailed TRAP staining studies of post-mortem BM samples have not been previously reported in the literature.
The aim of the present study was to investigate whether or not, and to what extent, TRAP can visualise BM osteoclasts in autopsy cases. We correlated TRAP-SIs of autopsy cases with the post-mortem interval (PMI; ie, the time span in days between death and autopsy) and compared the results with the TRAP-SIs of BM biopsies. We hypothesised that, owing to a suspected degradation of TRAP post-mortem, it could be possible to determine retrospectively the time of death in accordance with decreasing TRAP-SIs. In order to widen the database, we included autopsy cases from the Institute of Legal Medicine, Freiburg University Medical Center in our study. As is generally known, forensic and pathological autopsies differ considerably, especially with regard to the patients’ histories and causes of death.
The study included a total of 74 autopsies which were performed between February 2010 and December 2011 at the Freiburg University Medical Center; 51 of the autopsies were carried out at the Institute of Pathology and 23 at the Institute of Legal Medicine. The case collection consisted exclusively of adults (aged 20–97) and included 40 men and 34 women. There were no criteria for case selection except for the PMI. From experience, PMI is about 1–2 days in pathology cases but may be considerably longer in forensic cases. In order to diversify the PMIs in the study group, we preferentially included cases with ultrashort (<1 day) and very long PMIs (>3 days). Thus, the spectrum of PMIs ranged from 0 to 5 days in pathology cases and from 2 to 42 days in forensic cases. The histories and diagnoses of the patients had no influence on the case selection. In 23 pathology autopsies the patient had died from a malignancy. Among these, five cases with haematological tumours contained infiltrates of acute myeloid leukaemias (cases 23 and 43) or malignant lymphomas (cases 32, 46 and 50) in their BM specimens.
This group consisted of 22 patients (aged 7–90) of both sexes who were BM donors or were biopsied because of an iron deficiency anaemia. Four of the patients had a history of tumour (malignant lymphomas or prostatic cancer) but were without tumour in the BM sample.
In each case a thin rod-like sample containing haematopoietic marrow spaces was taken from the iliac spine. The specimens measured up to 1 cm in length and 0.5 cm in diameter. All samples were fixed with Schaefer's fixative,2 decalcified in EDTA, buffered to pH 7.0 and then sectioned after paraffin embedding. After deparaffinisation, TRAP staining was carried out as described elsewhere.3 In brief, 100 ml acetate buffer solution pH 5.6 was mixed with 0.5 ml hexazotised pararosaniline and with 16 mg naphthol AS-BI phosphate disodium salt (with a substrate Sigma N-8505) dissolved in 2 ml DMSO/Triton. In order to stain TRAP specifically as an isoenzyme of acid phosphatase (EC 220.127.116.11 5b), 150 mg tartaric acid as inhibitor was added and dissolved in 1.8 ml 1 mmol/l NaOH with a resulting pH of 5.2–5.4. The paraffin slides were taken into water through a series of xylenes and graded alcohols, incubated in the staining solution for 2.5 h at 37°C and then rinsed with tap water. Nuclei were counterstained with haemalaun. TRAP positivity in osteoclasts became manifest as a diffusely distributed intense red staining of the cytoplasm. In each case, naphthol AS-D chloroacetate esterase (NACE) staining was also available. NACE is known to show no fading of staining intensity (SI) in BM specimens regardless of decalcification, tissue fixation and paraffin embedding.4
The TRAP-SIs were investigated semi-quantitatively by a 4-grade scoring system (table 1). In order to calculate a mean value of SI, at least 3–10 osteoclasts were evaluated per case. The staining procedures were generally carried out for a couple of cases simultaneously to provide internal positive controls.
Statistical analysis was performed using the Wilcoxon test5 to determine the levels of significance and the Kendall's tau (τ) coefficient to determine the correlation coefficient.6 All tests were two-tailed and a p value ≤0.05 was considered significant.
The TRAP-SIs varied considerably among the autopsy cases and ranged from 0 to 2.7 in pathological cases (median 0.5, 95% CI 0 to 0.7) and from 0 to 2.8 in forensic cases (median 0.5, 95% CI 0 to 1.0) (figure 1). In biopsies the TRAP-SIs ranged from 1 to 3 (median 3.0, 95% CI 2.5 to 3.0). Table 2 shows the individual data and staining results of each case. Typical examples of TRAP staining in the BM are shown in figure 2. The results of a quantitative analysis of TRAP staining procedures are shown in detail in table 3. NACE staining was strong in all BM samples except for forensic case 69 which showed advanced decay at PMI 42.
The TRAP-SIs in the different groups were analysed in correlation with sex, age and origin.
Median TRAP-SIs of all cases (autopsies and biopsies) did not differ significantly between the two sexes (p=0.386). The median TRAP-SI score was 0.75 (rank sum=2210.5) in men and 0.8 (rank sum=2445.5) in women.
There was a gradual overall decrease in TRAP-SIs with age which reached a significant level (p=0.041, Kendall's tau coefficient τ=−0.128). In younger adults (<50 years of age) the TRAP-SI score was considerably higher (median 1.3) than in the older age group (50–69 years, median 0.8) or in elderly subjects (≥70 years, median 0.7) (p=0.020, τ=−0.172).
Biopsy samples had significantly higher TRAP-SIs than pathology autopsy cases (median 3.0 vs 0.5; p<0.001; rank sum 1351.5 vs 1349.5) or forensic cases (median 0.5; p<0.001; rank sum 746.0 vs 289.0). Among the autopsy cases, TRAP-SIs of forensic and pathology cases did not differ significantly (p=0.847; rank sum 879.0 vs 1896.0).
PMI in autopsy cases
The mean TRAP-SIs of autopsy cases classified according to PMI in general were remarkably constant (range 0.6–0.76) regardless of the duration of PMI (figure 3). The only exceptions were three forensic cases with PMIs ≥7 in which advanced decay had deteriorated the BM morphology, resulting in decreased TRAP-SIs (). The parameters TRAP-SI and continuous PMI showed no significant association (p=0.607, τ=−0.047). Also, a comparison of grouped PMIs with TRAP-SIs showed no statistical association (PMI six categories: p=0.746; PMI three categories: p=0.980).
In the morphological examination of BM biopsies the detection of TRAP may help to characterise multinucleated osteoclasts and their mononuclear precursors that can be important, especially in cases of increased osteolytic activity.1 ,7–9 Irrespective of their functional activity, the osteoclasts show a strong diffuse cytoplasmic TRAP staining. Osteoclasts can therefore be detected even if only part of the cytoplasm is present in the cut section. Moreover, chondroclasts, Gaucher cells and hairy cells are also usually TRAP-positive whereas, in ordinary phagocytic reticulum cells, the acid phosphatase is completely inhibited by tartrate.6 ,10 In our study, none of the specimens contained cartilage and none of the patients had Gaucher's disease or hairy cell leukaemia.
In contrast to parenchymal organs such as liver or kidneys which, on autopsy, mostly show progressive autolysis with a deterioration of anatomical structures, the BM post-mortem often remains fully diagnostic.
As we defined the PMI as the sole criterion for the selection of the cases, we disposed of a wide range of different PMIs (0–42 days) to monitor the course of TRAP-SIs following death. As shown in table 2 and figure 1, the TRAP-SIs varied considerably within all groups regardless of the age and sex of the patients. Four forensic autopsy cases with ultralong PMIs ≥7 days showed advanced tissue decay with a complete loss of TRAP positivity, except for forensic case 70. As this 70-year-old woman had died of hypothermia in early winter, her TRAP-SI value of 1.2 with PMI=42 days suggested that constant cooling of a corpse may conserve TRAP positivity for a long time. TRAP-SIs of biopsy patients were significantly (p=0.000058) stronger than those of autopsy cases with only a small overlap of values. The median TRAP-SI values of forensic and pathology autopsy cases were the same.
TRAP-SIs of autopsy cases did not correlate with the duration of PMI. We observed cases with low TRAP-SIs on the day of death (ie, PMI=0 days, case 1, TRAP score 0) as well as cases with high TRAP-SIs on day 4 post-mortem (such as case 23, TRAP score 2.7). Thus, the original hypothesis that there might be a progressive decrease in TRAP-SIs due to a degradation of TRAP post-mortem was clearly disproved in our study. On the basis of our findings, it was not possible to delimit the time of death by determining the SI of TRAP. Since TRAP-SIs were already variable in BM biopsies (scores ranging from 1 to 3), we speculated that individual, so far unknown, factors might influence TRAP positivity in vivo. The variability of TRAP-SI values post-mortem may simply reflect the conditions before death.
Earlier studies on the delimitation of the time since death by immunohistochemical staining of inner organs detecting, for example, insulin,11 thyroglobulin,12 glucagon13 and calcitonin14 in forensic corpses stated that it was possible to at least define a maximum time since death in the case of a negative reaction. In these studies, the authors described a time-dependent loss of stainability within the different immunoreactions which was also influenced by various external factors such as cause of death or surrounding conditions of the individual case. To our knowledge, this kind of study has not yet been performed in the BM.
In conclusion, our results show that TRAP staining is an appropriate tool to detect BM osteoclasts also in autopsy samples. However, in cases with PMI ≥7 days, TRAP staining was usually negative. TRAP-SIs in the BM of autopsy cases were significantly reduced compared with biopsies. The considerable variability in TRAP-SI scores prevents a reliable assessment of time since death from a forensic point of view.
Osteological examinations of autopsy specimens are feasible because tartrate-resistant acid phosphatase (TRAP) remains histochemically detectable in bone marrow (BM) osteoclasts for about 7 days post-mortem.
Within these 7 days post-mortem, the TRAP staining intensity (SI) of BM osteoclasts is not closely correlated with the duration of the time since death.
Owing to the individual variability of TRAP positivity in BM osteoclasts present in vivo, TRAP staining alone cannot serve as a tool to determine the time since death span.
Contributors All five authors contributed substantially to the conception and design, analysis and interpretation of data, drafting the article, revising it critically for important intellectual content and final approval of the version to be published.
Ethical approval Ethics approval was obtained from the Ethics Commission of the Freiburg University Medical Center.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.