If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Nine studies (n = 725 adult patients) were included in the meta-analysis.
•
The serum procalcitonin (PCT) assay is a highly accurate test for distinguishing the cause of meningitis.
•
PCT is a stronger diagnostic indicator than C-reactive protein for bacterial meningitis.
Summary
Objective
Clinically, it is often difficult to differentiate between bacterial and viral aetiologies in adults with suspected meningitis. Several studies have demonstrated the potential use of serum procalcitonin (PCT) in making this differentiation. The aim was to pool these studies into a meta-analysis to determine the diagnostic accuracy of PCT.
Methods
Major electronic databases were searched for articles studying the use of serum PCT in the differentiation of bacterial and viral meningitis in adult patients. No date or language restrictions were applied. Data analysis was performed using Meta-DiSc 1.4 and MIX 2.0.
Results
Nine studies (n = 725 patients) were included in the meta-analysis. Serum PCT was found to be a highly accurate test for diagnosing meningitis. The pooled sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, and diagnostic odds ratio (DOR) for PCT were 0.90 (95% confidence interval (CI) 0.84–0.94), 0.98 (95% CI 0.97–0.99), 27.3 (95% CI 8.2–91.1), 0.13 (95% CI 0.07–0.26), and 287.0 (95% CI 58.5–1409.0), respectively. PCT was found to be far superior to C-reactive protein, which had a pooled DOR of only 22.1 (95% CI 12.7–38.3).
Conclusions
Serum PCT is a highly accurate diagnostic test that can be used by physicians for rapid differentiation between bacterial and viral causes of meningitis in adults.
Due to the high mortality rate and potential neurological sequelae in survivors, there is an urgent need for rapid diagnosis with near 100% sensitivity.
Serum procalcitonin level and other biological markers to distinguish between bacterial and aseptic meningitis in children: a European multicenter case cohort study.
Cerebrospinal fluid (CSF) analysis is the current gold standard for the diagnosis of BM, along with biomarkers such as C-reactive protein (CRP) and white blood cell count (WBC). However, none of these tests achieve 100% sensitivity and confer a high enough specificity to distinguish between bacterial and viral meningitis (VM).
Serum procalcitonin level and other biological markers to distinguish between bacterial and aseptic meningitis in children: a European multicenter case cohort study.
CRP has traditionally been used as the biomarker for inflammation. However, CRP may show a delayed increase during the course of bacterial infection, resulting in false-negative tests in the early stages of the disease.
Procalcitonin (PCT) is now considered to be the best candidate to replace CRP due to its high diagnostic accuracy in various infectious pathologies, including sepsis, acute infectious endocarditis, and pancreatitis.
Fibronectin, interleukin 6 (IL-6), and tumour necrosis factor alpha (TNF-α) have also been proposed as potential biomarkers, but have not thus far been accepted widely for clinical use.
It has been hypothesized that this increase is due to the overexpression of the CALC-1 gene and increased release of PCT from various tissues in response to bacterial endotoxins and inflammatory cytokines such as TNF-α, IL-6, and IL-1β.
Unlike CRP, PCT has not been reported to be elevated in viral infections, thus conferring it the important ability to distinguish easily between bacterial and viral aetiologies.
This delay in diagnosis, combined with the traditional 72-h wait for the results of Gram stains, often results in patients receiving empiric antibiotic therapy.
The reported results of these studies are varied and a consensus has yet to be reached on the diagnostic value of PCT in meningitis.
The aim of this study was to pool and analyze the results of all the reported studies to determine the true diagnostic accuracy of PCT in adult patients with suspected meningitis.
2. Methods
2.1 Search strategy
In accordance with the PRISMA guidelines, a systematic literature search of the major electronic databases including PubMed, Scopus, EMBASE, Science Direct, Web of Science, and the Cochrane Library was performed to identify studies eligible for the meta-analysis. The search was specifically tailored to each of the electronic databases. Search terms included procalcitonin, PCT, S-PCT, ProCT, meningitis, and meningism. No date or language restrictions were set. Case reports, letters to the editor, and conference posters and abstracts were searched, but not included in the meta-analysis.
2.2 Selection of studies
Studies were eligible for inclusion in the meta-analysis if they (1) investigated the diagnostic accuracy of PCT in adult patients with suspected meningitis to distinguish between bacterial and viral aetiologies, (2) measured serum PCT on admission, and (3) reported data necessary to construct 2 × 2 tables (true-positives, false-positives, false-negatives, true-negatives). Studies were excluded if they (1) reported incomplete data or did not provide data necessary to construct 2 × 2 tables, (2) did not compare BM vs. VM, (3) had a poor methodological quality as determined using the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) tool, (4) studied exclusively paediatric populations, or (5) only measured the levels of PCT in CSF. Three authors (J.V., B.M.H., and P.K.R.) independently assessed each full-text article for inclusion in the analysis. Disagreements were resolved by a consensus among the authors. When necessary, articles were translated from their original text into English for further review by medical professionals who are native English speakers and fluent in the language of the original text.
2.3 Data extraction
Two authors (J.V. and J.R.) independently extracted data from the selected studies. Extracted data included sample size, mean age, PCT and CRP cut-offs, sensitivity, specificity, testing method, time of measurement, serum levels of PCT and CRP, and definitions of BM. Two by two tables were then constructed to calculate values of true-positives, false-positives, false-negatives, and true-negatives, to be pooled into the meta-analysis. In the case of studies that reported multiple cut-offs with sensitivities and specificities, the cut-off with the highest Youden's J statistic was used in the meta-analysis.
This tool assesses risk of bias and applicability through a series of signalling questions related to methodological quality of the study. Risk of bias is assessed in four domains – patient selection, index test, reference standard, and flow and timing. Applicability is assessed by the first three of the aforementioned domains. Each domain was ranked as high risk, unclear risk, or low risk individually by two authors (J.V. and J.R.).
2.5 Statistical analysis
The statistical analysis was performed using Meta-DiSc 1.4 and MIX 2.0 by one of the authors (B.M.H.). Using a random-effects model, the pooled sensitivities, specificities, positive likelihood ratios (LR+), negative likelihood ratios (LR−), and diagnostic odds ratios (DOR) were calculated. Summary receiver operating characteristic (SROC) curves were generated and the area under the curve (AUC) and Q* index (the point on the SROC curve where sensitivity and specificity are equal) were then calculated appropriately. Heterogeneity was assessed using the Higgins I2 test, with values of 25%, 50%, and 75% indicating low, moderate, and high degrees of heterogeneity, respectively. The threshold effect was measured using Spearman's correlation coefficient.
In order to further explore heterogeneity, meta-regression was performed using a single covariate with logDOR as the dependent variable. To assess publication bias, an asymmetrical funnel plot was constructed. Due to the current lack of accurate and reliable tests for publication bias in diagnostic studies, no other assessments were performed.
An overview of the study identification process is given in Figure 1. Through database searching, 2379 articles were initially identified. After screening and removal of duplicates, 45 articles were assessed by full text for eligibility in the meta-analysis. Of the full text articles, nine were deemed eligible for the meta-analysis and 36 were excluded from the meta-analysis. Two studies by Viallon et al.
– were included in the meta-analysis. Of the studies included, seven were in English, one in Spanish, and one in Korean. The papers were also from a wide geographical spectrum, including Egypt, Slovenia, Denmark, Iran, Korea, Spain, France, and Germany. Only studies on adult patients were included. The mean age in the studies ranged from 39 to 55 years.
Table 1Characteristics of studies included in the meta-analysis
All studies measured serum PCT levels upon admission in patients with suspected meningitis to distinguish between a bacterial or viral aetiology. The cut-off for PCT varied amongst the studies and ranged between 0.25 ng/ml to 2.13 ng/ml. PCT assay methods also varied between studies and included the LUMItest (BRAHMS Diagnostika, Berlin, Germany),
The reported sensitivities for diagnosing BM in the studies that measured PCT ranged from 68.8% to 100% and specificities ranged from 83.3% to 100%. Seven of the nine studies also measured CRP as a biomarker, with cut-off values ranging from 0.8 mg/dl to 9 mg/dl. The reported sensitivities of CRP ranged from 67.5% to 90% and specificities ranged from 57% to 92%.
Overall, the methodological quality among the studies included was good, and the risk of bias was low. The study quality assessment using the QUADAS-2 tool is summarized in Figure 2.
Figure 2Summary of the QUADAS-2 assessment of the studies included.
The results indicate that serum PCT is a highly accurate test for the detection of BM. The pooled sensitivity for PCT was 0.90 (95% CI 0.84–0.94), while the pooled specificity for PCT was 0.98 (95% CI 0.97–0.99) (Figure 3). The pooled LR+ was 27.3 (95% CI 8.2–91.1) and the pooled LR− was 0.13 (95% CI 0.07–0.26) (Figure 4A , B). PCT was found to be a very powerful diagnostic test, with a pooled DOR equal to 287.0 (95% CI 58.5–1409.0) (Figure 4C). A SROC curve was also constructed for PCT and the AUC was 0.97 (standard error (SE) = 0.03) and the Q* index was 0.91 (SE = 0.04) (Figure 4D).
Figure 3Pooled sensitivity (A) and specificity (B) for serum procalcitonin for the diagnosis of bacterial meningitis in adults.
Figure 4Pooled positive likelihood ratio (LR+) (A), negative likelihood ratio (LR−) (B), diagnostic odds ratio (DOR) (C), and summary receiver operating characteristic (SROC) (D) for serum procalcitonin for the diagnosis of bacterial meningitis in adults.
The pooled analysis showed that PCT is the more accurate and effective diagnostic test. The pooled sensitivity and specificity for CRP were 0.82 (95% CI 0.75–0.88) and 0.81 (95% CI 0.77–0.84), respectively (Figure 5). The difference in specificities indicates that PCT is far stronger at ruling in the diagnosis of BM as compared to CRP. The AUC for CRP was 0.89 (SE = 0.02) and the Q* index was 0.83 (SE = 0.02). The pooled LR+ of CRP (4.2, 95% CI 2.8–6.0) was lower than that of PCT, but the pooled LR− of CRP (0.23, 95% CI 0.15–0.35) was similar to that of PCT. However, the DOR indicates the superiority of PCT, with the DOR of CRP equal to only 22.1 (95% CI 12.7–38.3).
Figure 5Pooled sensitivity (A) and specificity (B) for C-reactive protein for the diagnosis of bacterial meningitis in adults.
Significant heterogeneity was detected for the pooled DOR of serum PCT (I2 = 66.2%). However, heterogeneity was also calculated for the pooled DOR of CRP and was found to be substantially lower than the heterogeneity associated with pooled PCT (I2 = 3.6%). To explore the source of heterogeneity, the threshold effect was assessed. The calculated Spearmen correlation coefficient for PCT was r = −0.151 (p = 0.699), indicating that the threshold effect was not the source of heterogeneity in the PCT analysis.
It was suspected that the cause of heterogeneity may have been variation in the type of serum PCT assay used in the studies. To further explore this hypothesis, a meta-regression was done to determine if the type of PCT testing assay was responsible for the heterogeneity. The calculated relative DOR (RDOR) was 3.42 (95% CI 1.49–7.81, p = 0.01), indicating that the PCT testing assay was a significant source of heterogeneity between studies. As such, a subgroup analysis was performed for each testing method used by at least two different studies in the analysis. Three studies (n = 426 patients)
used the Kryptor test (BRAHMS Diagnostika), which had the highest pooled DOR of all PCT assays (DOR 1125.2, 95% CI 77.4–16362.0), but also the most significant heterogeneity of the testing assays (I2 = 77.1%). Two studies (n = 75 patients)
The pooled sensitivity (0.87, 95% CI 0.76–0.94), specificity (0.94, 95% CI 0.86–0.984), and DOR (173.4, 95% CI 30.3–992.9) were all lower than in the general analysis. A second subgroup analysis was performed on two studies that had a lower PCT cut-off value ranging from 0.25 ng/ml to 0.28 ng/ml (n = 275 patients).
The pooled sensitivity (0.96, 95% CI 0.85–0.99) and specificity (0.99, 95% CI 0.98–1.0), as well as the DOR (945.3, 95% CI 10.2–87874.0), were higher than in the three studies with the higher common cut-off of 0.5 ng/ml.
3.5 Publication bias
Assessment for potential publication bias was performed using a funnel plot, which revealed asymmetry among the studies included (Figure 6).
Figure 6Funnel plot presenting study heterogeneity and potential for publication bias.
In recent years, however, it has become increasingly appreciated as a reliable marker of bacterial infection due to extensive non-thyroidal production in response to proinflammatory mediators such as TNF-α and IL-6.
Through a proposed cytokine-induced ubiquitous extrathyroidal upregulation of CALC-1 gene expression, bacterial infections induce a pervasive PCT release from hepatocytes, adipocytes, and various other parenchymal cells.
Due to the serious, time-sensitive nature and high mortality of BM, diagnostic tests with a high sensitivity and fast turnaround time are needed. This analysis demonstrates that serum PCT is a highly accurate diagnostic test for distinguishing between bacterial and viral aetiologies in patients with suspected meningitis. PCT was found to be a more specific than sensitive marker of BM. The pooled PCT specificity was 0.98 and LR+ was 27.3, indicating that PCT is a strong marker for ruling in BM in adult patients. This is compared to CRP, which had a specificity of 0.81 and LR+ of only 4.2. The likelihood ratio is a diagnostic test calculated from the sensitivity and specificity, and is an indicator of the value of performing a diagnostic test. It expresses the degree to which a diagnostic procedure modifies the probability of a disease. The range of LR is from 0 to infinity, therefore any values above or below 1, increase or decrease the likelihood of a disease, respectively.
Pooled sensitivities between PCT and CRP were more similar, however the pooled sensitivity of PCT was still greater than that of CRP (0.90 vs. 0.82), albeit not significantly. AUC, an overall assessment of the quality of a diagnostic test, was 0.97 for PCT, as compared to 0.89 for CRP.
The true power of serum PCT is reflected in the DOR, which is a single indicator of the accuracy of a diagnostic test, combining data on both specificity and sensitivity.
The DOR for PCT was 287.0, while the DOR for CRP was only 22.1. The strong DOR for serum PCT reflects its clinical utility in patients with suspected meningitis.
In suspected meningitis, the combined clinical history, physical examination findings, basic laboratory tests (such as WBC count), and lumbar puncture results of a patient form the gold standard for diagnosis. However, clinical symptoms and physical signs are often not present, and lack the ability to differentiate the aetiology of meningitis. Nakao et al. reported the sensitivities of fever, headache, vomiting, and rash for predicting pleocytosis in the CSF to be only 0.20, 0.91, 0.04, and 0.02, respectively.
reported a pooled sensitivity of Kernig's sign in adults to be merely 0.11, while the reported pooled sensitivities of Brudzinski's sign and nuchal rigidity were 0.09 and 0.31, respectively. Due to the low sensitivities of physical signs, it is important to evaluate the entire clinical picture when formulating a differential diagnosis and determining further appropriate diagnostic tests.
In order to confirm a clinical suspicion of meningitis and determine its aetiology, CSF analysis is considered the gold standard. However, it was found that the pooled serum PCT diagnostic accuracy indices in the present study were superior to those reported for CSF markers by Jereb et al.
. Their reported serum PCT sensitivity of 0.9 was found to be superior to all measured CSF parameters, including CSF leukocyte count (0.35), CSF polymorphonuclear leukocytes (0.45), CSF:blood glucose ratio (0.85), and CSF protein concentration (0.70). Their reported specificity of serum PCT of 1.0 was, however, comparable to the other CSF markers, which ranged from 0.96 to 1.0. This demonstrates that serum PCT allows for more accurate ruling out of BM as compared to traditional CSF markers.
As such, we recommend the use of a serum PCT assay to supplement the traditional gold standard combination of clinical history, physical examination, basic laboratory tests, and CSF analysis, in order to increase the overall diagnostic accuracy in differentiating the aetiology of suspected meningitis. Furthermore, owing to its high sensitivity and specificity, PCT may be helpful in situations where the conventional tests are unable to reach a conclusive diagnosis, such as in cases with non-conclusive CSF findings.
The level of PCT in the CSF has also been studied as a potential diagnostic biomarker of BM, albeit to a lesser degree than serum PCT. Three studies have assessed the diagnostic accuracy of PCT in CSF using the same cut-off value of 0.5 ng/ml.
The reported sensitivities ranged from 0.55 to 1.0, and the reported specificities in the studies ranged from 0.93 to 1.0. The mechanism by which the PCT concentration rises in the CSF during BM has been debated.
showed that increased CSF PCT levels correlate with increased disease severity, duration, and mortality. Further studies are needed to confirm the efficacy of CSF PCT as a diagnostic marker of BM and its potential role as a prognostic marker of mortality.
In order to be widely implemented in the clinical setting, the economic implications of serum PCT and CRP testing must be considered. CRP is tested using a wide variety of inexpensive assays costing between USD $1 and $10.
Comparison of procalcitonin with C-reactive protein and serum amyloid for the early diagnosis of bacterial sepsis in critically ill neonates and children.
However, according to a study by Nabulsi et al., routine CRP testing inflated hospital bills by USD $26 715.9, while failing to impact the decision-making process in a majority of cases.
Comparison of procalcitonin with C-reactive protein and serum amyloid for the early diagnosis of bacterial sepsis in critically ill neonates and children.
This can be advantageous for the implementation of appropriate treatment, as well as for avoiding unnecessary empirical antibiotic therapy. For example, according to a review by Agarwal and Schwartz,
patients being treated for sepsis in the intensive care unit under PCT-guided interventions experienced a 28% to 31% reduction in the duration of antibiotic therapy.
Due to the superior accuracy and power of PCT as compared to that of CRP, its clinical implementation can be a cost-effective tool in differentiating between BM and VM. The use of PCT to supplement the traditional gold standard would allow for the accurate and rapid differentiation of the aetiology of meningitis, early in the course of the disease, which could potentially limit the need for costly interventions such as blood cultures and empiric antibiotic therapy. The rapid confirmation of a viral aetiology would also reduce the costs of unnecessary hospitalization and the treatment of resulting potential nosocomial infections.
Furthermore, serum PCT levels, unlike those of CRP, have not been shown to be affected by non-steroidal anti-inflammatory drugs, steroids, or inflammatory comorbidities such as inflammatory bowel disease, systemic lupus erythematosus, or gout.
Diagnostic value of procalcitonin for differentiation between bacterial infection and non-infectious inflammation in febrile patients with acute adult-onset Still's disease.
Additionally, PCT may potentially be useful in discriminating between bacterial and non-infectious aetiologies of meningitis. Neuro-Behçet's disease can mimic BM, presenting with acute meningeal symptoms and showing polymorphonucleic pleocytosis in the CSF.
the authors found that all seven patients had elevated CRP levels, but none had an elevated PCT level. These show yet additional benefits of using serum PCT as a biomarker, over CRP, in potential bacterial infections.
A few limitations on the clinical use of PCT should also be mentioned. The use of PCT to differentiate between BM and other causes of febrile encephalopathy due a bacterial aetiology, such as a brain abscess, is most likely limited, as PCT is often elevated in the case of most bacterial infections.
Procalcitonin for diagnosis of bacterial pneumonia in critically ill patients during 2009 H1N1 influenza pandemic: a prospective cohort study, systematic review and individual patient data meta-analysis.
and other bacterial infections, the diagnostic use of PCT in patients presenting with acute meningitis in the presence of other bacterial infections is also potentially limited when compared to CSF analysis. Thus, physicians should take into account the entire clinical picture when formulating their diagnosis.
Additionally, PCT levels would be expected to decrease in patients receiving antibiotics prior to testing, thus giving misleading results.
Procalcitonin algorithms for antibiotic therapy decisions: a systematic review of randomized controlled trials and recommendations for clinical algorithms.
found that PCT levels decreased significantly 3 days after the initiation of antibiotic treatment in children with BM. Furthermore, the authors investigated how serum PCT levels correlated to disease severity.
Additionally, in patients who did not experience clinical improvement after the initiation of antibiotic therapy, PCT levels did not decrease as significantly as compared to those who experienced clinical improvement.
Further studies should examine the use of PCT to assess prognosis, mortality, and the efficacy of antibiotic therapy in patients with BM, as well as how antibiotic therapy affects the use of PCT as a diagnostic marker.
This meta-analysis was limited by the number of studies available, the potential for publication bias, and the moderate heterogeneity between the studies included. In this analysis, the funnel plot did reveal asymmetry, however it is suspected that this was due to the limited number of studies included in the meta-analysis. Traditional publication bias assessments, such as Begg's test and Egger's test, are designed to assess bias in interventional studies, not diagnostic studies, and as such have low power and can give seriously misleading results.
As such, the potential for publication bias in the studies included cannot be ruled out.
Heterogeneity in the present analysis appears to be due primarily to the type of PCT assay used in the study. The newest of the testing methods used in the studies, the Kryptor test (BRAHMS Diagnostika), had the highest pooled DOR of all the testing assays used in the analysis, but also had a moderate level of heterogeneity between the studies using this assay. However, with a rapid turnaround time of only 20 min and the high DOR, it represents an excellent assay that can be used clinically to assess PCT levels.
Despite the limitations of the meta-analysis, serum PCT is an excellent tool that can be used to enable quick and accurate differentiation between BM and VM in adults, and supplement the clinical history, physical examination, basic laboratory tests, and CSF analysis for a more accurate diagnosis.
Serum PCT cut-off values in the meta-analysis ranged from 0.25 ng/ml to 2.13 ng/ml. Interestingly, in subgroup analyses, a reduction in cut-off value from 0.5 ng/ml to a cut-off range of 0.25 ng/ml to 0.28 ng/ml, demonstrated an increase in sensitivity, specificity, and DOR from 0.87, 0.94, and 173.4 to 0.96, 0.99, and 945.3, respectively. It is suspected that this is due to the use of the Kryptor test (BRAHMS Diagnostika) PCT assay by both of the studies with the lower cut-off values, while no studies in the 0.5 ng/ml cut-off subgroup used this PCT assay. The increase in both sensitivity and specificity at this lower cut-off range as compared to the general analysis, increases the clinical utility of this test by increasing the ruling-out power of the PCT assay needed to accurately exclude BM. As such, the use of a low PCT cut-off of around 0.25 ng/ml with the Kryptor test (BRAHMS Diagnostika) PCT assay is recommended, which gives both the high sensitivity and high specificity needed clinically to accurately diagnose the cause of meningitis. However, further studies are required to determine the optimal cut-off values for the other PCT testing assays. Additionally, future studies should assess the diagnostic role of PCT specifically in suspected cases of meningitis with non-conclusive CSF findings, in cases of non-infectious meningitis, and in cases of meningitis caused by opportunistic pathogens such as Mycobacterium tuberculosis.
5. Conclusions
Serum PCT is a powerful diagnostic test for the assessment of suspected meningitis, allowing rapid differentiation between bacterial and viral aetiologies, and earlier initiation of appropriate and necessary therapies. The evidence suggests that while serum PCT offers a similar specificity to the traditionally used CSF markers of meningitis, it confers a higher sensitivity. Thus PCT, when used as a supplement to clinical signs and CSF analysis, allows for a more accurate overall diagnosis in patients with suspected meningitis. Lastly, this analysis indicates that PCT is a far superior diagnostic biomarker than the traditionally used marker of inflammation, CRP. The use of PCT serum assays in adult patients with suspected meningitis is therefore recommended.
Acknowledgements
This study and its publication was financed by the Jagiellonian University Faculty of Medicine KNOW funds. Krzysztof A. Tomaszewski was supported by the Foundation for Polish Science (FNP).
Conflict of interest: The authors declare that they have no conflict of interest.
Appendix A. Supplementary data
The following are the supplementary data to this article:
Serum procalcitonin level and other biological markers to distinguish between bacterial and aseptic meningitis in children: a European multicenter case cohort study.
Comparison of procalcitonin with C-reactive protein and serum amyloid for the early diagnosis of bacterial sepsis in critically ill neonates and children.
Diagnostic value of procalcitonin for differentiation between bacterial infection and non-infectious inflammation in febrile patients with acute adult-onset Still's disease.
Procalcitonin for diagnosis of bacterial pneumonia in critically ill patients during 2009 H1N1 influenza pandemic: a prospective cohort study, systematic review and individual patient data meta-analysis.
Procalcitonin algorithms for antibiotic therapy decisions: a systematic review of randomized controlled trials and recommendations for clinical algorithms.
00178-2&id=gr1.jpg){kind=link}
00178-2&id=gr2.jpg){kind=link}
00178-2&id=gr3.jpg){kind=link}
00178-2&id=gr4.jpg){kind=link}
00178-2&id=gr5.jpg){kind=link}
00178-2&id=gr6.jpg){kind=link}