Usefulness of Procalcitonin in Diagnosing Diabetic Foot Osteomyelitis: A Pilot Study

Infected diabetic foot soft tissue (cellulitis) and bone (osteomyelitis) are the leading causes of hospital admissions for people with diabetes mellitus and precede 90% of lower limb amputations.¹ Diabetic foot osteomyelitis (DFO) with or without cellulitis causes high morbidity and significant mortality of 50% at 5 years after major amputation.¹ These substantial issues become more prevalent with aging and are coupled with substantial economic burden. Each year in Australia, there are 5600 minor amputations (toe, midfoot) and 1500 major amputations (above/below knee) due to diabetes.² Admission alone costs the hospital system $16 000 for a minor amputation and $34 000 for a major amputation,² which increases stepwise with advancing age. Part of the reason for the high cost of care, morbidity, and mortality is due to the lack of a simple, effective test to differentiate the 2 common foot infections—cellulitis and osteomyelitis.³ Current diagnostic tests for the 2 conditions are either too expensive (magentic resonance imaging [MRI], bone scans), involve an invasive procedure (bone biopsy), limited availability (MRI/bone scan), or are of low diagnostic yield (x-ray).⁴ 

Serum procalcitonin (PCT), a propeptide of the hormone calcitonin released by non-neuroendocrine parenchymal cells, was reported to be elevated in both cellulitis^5,6 and osteomyelitis,^7,8 with evidence showing lower values for cellulitis compared with osteomyelitis. Despite the apparent differences in serum levels of this biomarker in diabetic foot infections, to the authors’ knowledge, no studies have been conducted to determine cutoff values of serum PCT to distinguish diabetic foot cellulitis from osteomyelitis. The aim of the study was to determine the clinical usefulness of serum PCT in diagnosing/differentiating diabetic foot cellulitis from osteomyelitis. 

Study design and participants

This prospective cohort study was carried out at the Townsville University Hospital and Kirwan Community Health Center, both in Kirwan, Australia, comparing biomarkers of patients with diabetic foot osteomyelitis (group 1) and soft tissue infection of the foot (group 2). The study received Institutional Review Board/Independent Ethics Committee/Research Ethics Board approval HREC/17/QTHS/113 & 65. This study was conducted according to the principles of Good Clinical Practice, the declaration of Helsinki, and national laws and regulations regarding clinical studies.

Patients with type 2 diabetes mellitus, aged 18 years or older, who had a moderate or severe infected ulcer based on the Infectious Diseases Society of America classification were included in the study after informed consent was obtained.⁹ Patients who had planned surgical intervention, clinically significant lower-extremity ischemia (as defined by an ankle/brachial index of less than 0.65), or ulcer of non-diabetic pathophysiology were excluded from the study. Additional exclusion criteria included other infectious diseases, immunosuppressive therapy, organ and/or hematological malignancies, and end-stage renal disease requiring dialysis. 

The bones involved included 3 areas of the foot, namely hindfoot (tarsal bones of calcaneus and talus), midfoot (other tarsals), and forefoot (metatarsals and phalanges). The diagnosis of osteomyelitis was confirmed by positive histopathological examination and culture of bone at the clear surgical margin or as indicated from a percutaneous bone biopsy. In patients who had clinically proven osteomyelitis, antibiotics were administered after obtaining diagnostic blood and tissue samples. The DFO group was defined by intraoperative bone sample showing histologic findings of osteomyelitis, presence of probe-able bone underlying ulcer or by imaging (radiograph, MRI). Clinical history of patients was collected, including age, sex, ethnicity, alcohol and smoking status, medications, comorbidities such as retinopathy, coronary artery disease, and hypertension, and previous amputation.

Laboratory procedures

Morning fasting venous blood samples were taken for PCT and other inflammatory markers. After 30 minutes of collection, the sample was centrifuged for 12 minutes at 3000 rpm; 0.1 mL of venous plasma was stored at –80°C for analysis as a batch to minimize variance. Serum levels of PCT, interleukin 6 (IL-6),  tumor necrosis factor-α, adiponectin, CD253 (TRAIL; Affymetrix eBioscience), osteoprotegerin (OPG; Thermo Fisher Scientific), Dickkopf-related protein 1 (DKK1), osteocalcin, osteopontin (OPN), and sclerostin (Map Human Bone Magnetic Panel; MILLIPLEX) were measured using the ELISA kits as per manufacturer’s instructions. 

Statistical analysis

All data were expressed as mean ± SEM. Statistical analysis was performed using SPSS version 25 (IBM Corp). Continuous data were checked for normality followed by a parametric test, Student’s t test, or non-parametric Man-Whitney U test to compare the groups. Categorical variables were tested using χ² test or Fisher's exact test. Receiver operating characteristic curves (ROC) were generated, and the best cutoff point with highest prognostic value presented based on Youden’s index was used to determine the diagnostic levels. Systemic immune-inflammation index (SII), neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte (MLR) ratio, and platelet-to-lymphocyte ratio (PLR) were calculated.¹⁰ The clinical utility index calculator was utilized to compute positive and negative predictive values.¹¹ A P value of .05 was considered statistically significant.

A total of 37 participants were enrolled in the study, with 19 participants in the osteomyelitis group and 18 participants in the control group. The baseline characteristics were similar in both groups, as detailed in Table 1. The mean age for the osteomyelitis group was 62.0 ± 12 years, and the control group was 65.0 ± 8.0 years (P = .5). The majority of participants in both groups were patients with type 2 diabetes; both groups had a similar percentage of participants taking oral hypoglycemic agents and insulin as well as similar percentages of non-traumatic lower-limb amputations (59% vs 74%, respectively) (P = .7). Fifty-seven ulcers were studied, comprising 28 and 29 ulcers in the DFO group and control group, respectively. The majority of patients in both groups (81%) had forefoot ulcers, including 78% in the DFO group and 86% in the control group as shown in Table 2. Further analysis of the ulcer location (ie, forefoot, midfoot, and hindfoot) revealed similar distribution in both groups. No statistically significant differences for alcohol consumption and smoking were observed between the groups. Comorbidities, such as diabetic retinopathy, dyslipidemia, coronary artery disease, and peripheral vascular diseases, were similar in both groups. There were no differences between the counts for neutrophils, platelets, monocytes, or lymphocytes between the groups.

Differences between serum levels of PCT and IL-6 were found to be statistically significant. The median serum levels of PCT in the DFO group were 108.5 pg/mL (range, 65.0–124.0 pg/mL) and 57.0 pg/mL (range, 37.2–77.0 pg/mL) (P = .02), and IL-6 levels in DFO group were 22.0 pg/mL (range, 3.0–86.5 pg/mL) and 3.5 pg/mL (range, 1.0–9.5 pg/mL) (P = .02) (Figure). 

The ROC analysis showed that out of all the candidates tested, PCT had a sensitivity of 79% and specificity of 70% at a positive diagnostic test cutoff of 64.0 pg/mL (P = .02). Interleukin 6 had sensitivity of 75% and specificity of 64% at a cutoff of 4.2 pg/mL (P = .02). The clinical utility index of both was found to be fair. Analysis reference intervals on ratios of NLR, PLR, MLR, LNR, and SII revealed a fair test candidate for MLR with a sensitivity of 74% and specificity of 67% at a positive diagnostic test cut-off of 0.3 (P = .02) (Table 3). 

Diabetic foot ulcers are of major concern among patients with long-term diabetes mellitus. Most diabetic foot infections are diagnosed based on clinical findings and may need invasive tests to confirm DFO.¹² Therefore, a suitable diagnostic marker with a high clinical utility, high sensitivity (true positive correctly identified by a test), and sensitivity (true negative correctly identified by a test) would be useful to confirm the diagnosis, rationalize requirement for invasive testing, and prevent delays in modified treatment regimens due to delay in diagnosis. The current study demonstrated significantly high levels of PCT and IL-6 in the DFO group compared with the control group. Procalcitonin has been proposed as an ideal candidate for diagnosing diabetic foot infection; the serum level of PCT has been shown to be elevated in diabetic foot infections, as reviewed by Velissaris et al.¹³ Similar to the present study, Uzun et al¹⁴ reported a significant difference in levels of PCT in patients with DFO, with a sensitivity and specificity of 77% and 100%, respectively, at the cutoff value of PCT greater than or equal to 0.08 ng/ml and 0.06 ng/ml. The authors also reported comparable sensitivity and specificity as in the present study. Similarly, a significant difference in levels of PCT was found in patients with infected foot ulcer compared with non-infected foot ulcer with high specificity (100%) and a sensitivity of 23.3%,¹⁵ although it is unclear from the study if the infected foot ulcers were confirmed to have DFO or cellulitis without bone infection. On the other hand, Asirvatham et al¹⁶ found PCT to have a high sensitivity and specificity of 80% and 75%, respectively, in diagnosing DFO at a serum level of 60.0 pg/mL, very similar to the findings in the present study. 

In contrast to the above findings, Korkmaz et al¹⁷ did not report elevated levels of PCT in patients with DFO. Interestingly, the present study also showed increased levels of IL-6 levels in patients with DFO, as previously reported by others.¹⁸ This study showed MLR to be a candidate for predicting osteomyelitis at a cutoff point of 0.3, with a sensitivity and specificity of 74% and 67%, respectively, and a fair test based on clinical utility index. This study did not observe PLR to have good utility, unlike the study reported by Demirdal et al.¹⁹

The present study suggests PCT and IL-6 could be practical diagnostic tests for differentiating DFO from cellulitis based on clinical utility index. The cost of the PCT test works out to be approximately $20 per test compared with performing a bone scan or an MRI at a cost of about $270 in Australia.³ A preliminary test for PCT would be beneficial and could be conducted before an MRI or bone scan to confirm the findings. This is particularly relevant for patients in resource-deprived and/or economically developing countries or those from rural and remote areas in economically developed countries who would not have to travel to a tertiary hospital in urban areas to undergo an MRI or a bone scan to confirm the DFO diagnosis. Additionally, preliminary testing may also help save costs to the health care provider.

It is important to note that the present study is, to the authors’ knowledge, the first study in patients with DFO compared with non-DFO with cellulitis. Previous studies demonstrated differences in serum levels of PCT between DFO and diabetic non-infected foot controls.^16,17 The present results faced potential limitations, including small sample size, though they fitted with the outlined objectives of a pilot study. Secondly, a bone biopsy was not conducted on all patients involved in the study. Although bone biopsy is usually considered a gold standard for the diagnosis of osteomyelitis, such a procedure may not be routinely performed in clinical practice due to ethical issues of creating a new wound by excising adjacent healthy tissue and worsening the healing process, particularly in patients with a low index of suspicion for DFO. Instead, the authors used the recommendation of the International Working Group on the Diabetic Foot in defining DFO in this study.¹² Also, a detailed study would be needed to understand the impact of ischemia and dialysis on the biomarkers as indicator for osteomyelitis, as this cohort has a higher risk for amputation. Despite these, the present results are consistent with others’ findings of the diagnostic usefulness of PCT in DFO.^16,17

The sensitivity of PCT in differentiating DFO from cellulitis was modest, which may be due to the small sample size in the study. Also, a larger sample size with specific location of osteomyelitis would enable a better analysis of diagnostic parameters in different regions of the foot; for example, forefoot vs midfoot or hindfoot infections with locations having a correlation with risk of amputations could be analyzed. 

In conclusion, this research reports a significantly high level of PCT in patients with DFO with good sensitivity and specificity in differentiating it from diabetic foot cellulitis. A larger study should be conducted to confirm the findings. In addition, a detailed economic analysis would be helpful to determine the use of PCT as a diagnostic test in patients with infected diabetic foot ulcers. 

Authors: Venkat N. Vangaveti, PhD¹; Oliver G. Heyes, MD^1,2; Shaurya Jhamb, MD^1,2; Nagaraja Haleagrahara, PhD³; and Usman H. Malabu, MD, FACP, FRACP^1,2

Affiliations: ¹Translational Research in Endocrinology and Diabetes, College of Medicine and Dentistry, James Cook University, Douglas, Queensland, Australia; ²Department of Endocrinology and Diabetes, Townsville University Hospital, Douglas, Queensland, Australia; ³College of Public Health, Medical and Veterinary Sciences, James Cook University, Douglas, Queensland, Australia

Correspondence: Usman H. Malabu, Queensland Health, Townsville University Hospital, 100 Angus Smith Drive, Douglas, Queensland, 4814, Australia; usman.malabu@jcu.edu.au 

Disclosure: The authors disclose no financial or other conflicts of interest.