Discussion:
The effect of DM on NOE is not clearly understood. One may intuitively speculate that glycemic control prior and during admission is of paramount importance in the treatment of NOE. However,is this in fact true? Is an uncontrolled patient more likely to be infected with a certain pathogen or is he at a greater risk for a worse outcome? Perhaps the main reason for not having a validated answer is the relative rarity of the disease and the small case series published from different centers, during the last few decades. Our group previously reported on 83 ears in 81 patients with NOE3. We found that duration of complaints prior to admission and the presence of aural discharge were associated with longer hospitalization length. Additionally, advanced age was associated with conservative treatment failure and the need for surgery. In this research, we have slightly enlarged the study group, and focused on the effects of DM and glycemic control on NOE progression, using the same end-point parameters- length of admission and the need for surgery.To the best of our knowledge this study represents the largest and most detailed research evaluating the relationship between DM and NOE.
The diagnosis of NOE is still not in complete agreement and there are more than 20classification systems4reported in the literature. In our institute we diagnosed NOEbased on the Friedman and Cohen criteria9, which uses obligatory and occasionalclinical parameters. Although not all patients exhibited all the required parameters at admission, repeated physical examination during hospitalization revealed all obligatory criteria’s, confirming the diagnosis of NOE.
Similar tothe findings in previous studies, our patients with NOE were mostly elderly and 93.2% of themwere known diabetics. Also in agreement with the literature, otalgia and aural discharge were the most common complaints at admission, and edema of the external ear canal and granulation tissue were the most common clinical findings at admission.
PA is considered the most common pathogen in NOE. Despite that, there have been numerous publications indicating a shift in the incidence of the offending pathogen and an increase in the rate of fungal NOE and sterile NOE3,4. In the current study PA-NOE was the most common pathogen (45.0%), followed by sterile culture (20.8%) and fungal-NOE (15.3%). When comparing these results with previously reported data,we have observed a continuing decline in PA-NOE and an increased in fungal-NOE3.
Our understanding in regards to the effects of DM on NOE progression is not clear, however several key elements in the pathophysiology and treatment of NOE share resemblance to diabetic foot11, which is caused due to DM related peripheral vascular pathology. Initially, both NOE and diabetic foot occur as a result of direct spread of soft tissue inflammation into the bone, causing local osteomyelitis. Second, in both entities PA is a leading pathogen in the inflammatory process. Last, treatment protocol is similar and comprised of long term antibiotic treatment and surgery in selected cases. These similarities support previous assumptions that microangiopathies are probably the main pathology leading to NOE among DM patients7,11. If this is the case, it is reasonable to assume that DM duration and severity effects NOE progression.
There are limited studied on the relationship between DM and NOE, using different parameters to assess both DM severity, and NOE progression. Joshua et al6found that NOE patients that exhibited all obligatory parameters by the Friedman and Cohen criteria, had a higher incidence of DM, higher use of oral antidiabetic medications, and higher incidence of DM related complications in comparison to NOE patients that did not show all obligatory parameters. Duration of diabetes prior to admission, HbA1c levels and microalbuminuria however, were not evaluated. Stern-Shavitet al12reported on patients from the same center and found that DM correlated and predicted disease specific mortality, however DM duration and severity were not analyzed. Leeet al.13reported that DM duration was associated with uncontrolled NOE, but HbA1c was not associated with NOE progression. Similarly, Lohet al.7reported that DM severity, defined by HbA1c >7.0% was not associated with the disease outcome. Our study focused on DM severity by analyzing three parameters – DM duration, HbA1c levels and microalbuminuria, which is commonly used to assess renal dysfunction as a result of long lasting vascular pathology.
In contrast to previous studies our results found that HbA1c levels were associated with a longer hospitalization duration among NOE patients (p=.027). DM duration and microalbuminuriadid not correlate with need for surgery and duration of hospitalization. This might indicate that diabetic control at the time of disease onset plays a more substantial role in the insemination of bacteria into the surrounding bone, in comparison to overall DM duration, thus leading to a severe disease requiring longer in hospital treatment.
Apart from DM severity, the importance of glycemic control during hospitalization was also reported by several publications. Carfrae et al.14 reported that strict glycemic control is an essential principle in NOE treatment. Similarly, Hollies5 reported that tight glycemic control is one of the key management strategies for the treatment of NOE. In contrast, Chen et al.15reported that average glycemic levels among NOE patients were not associated with mortality in NOE patients. Also, Lee et al.13 reported that mean glucose levels were not statistically significant between controlled and uncontrolled NOE. In the presented study, mean glycemic levels during hospitalization were associated with DM duration (p=.005) and HbA1C levels (p-value 0.001), but were not associated with hospitalization duration or need for surgery. This may show that strict glycemic control during hospitalization is not as important as previously mentioned in the treatment and outcome of NOE patients.
Interestingly, higher glycemic levels during hospitalization were associated with PA-NOE (p=.045). This might be explained by the fact that DM is known to cause a dysfunction in the phagocytotic activity of polymonuclear cells and macrophages leading to increased sensitivity of PA infection16, 17.The possible relevance of this finding might be important in sterile NOE. In such patients, poor glycemic control during hospitalization can support the continuing use of anti-PA antibiotics