DISCUSSION
The current study involving a large number of patients showed that a CG cut-off value of <2.0 mg/dl had a sensitivity of 83.8% and a specificity of 74.9% in order to detect AD in Turkish adult patients. Our results also showed that CG is useful to detect low IgG levels in parallel to low gamma globulin values in protein electrophoresis. Furthermore, its usefulness was confirmed in identifying almost 80% of adult PAD patients tested.
Since CG is easily accessible, simple, and inexpensive, it is suitable for usage as a screening test12,13. Furthermore, it is frequently used as part of routine laboratory testing in patients. It has a high potential to facilitate early diagnosis of PAD or SAD because of its widespread use, high sensitivity and specificity. The use of CG as a screening test in AD was first suggested by Jolles S. et al2. In their study, CG was obtained by subtracting albumin, which is measured by both BCG and bromocresol purple (BCP) methods, from total protein. They defined different cut-off values for these two methods2. Similar to the study of Pecoraro A. et al. 8, the Italian version of this study, we measured albumin only with the BCG method in our study.
In the study of Jolles S. et al., a CG value of <1.8 obtained by using the BCG method, predicted a sensitivity of 66% and a specificity of 78% for IgG levels lower than 500 mg/dL2. In the current study, we accepted the IgG value of 600 mg/dL as a limit for hypogammaglobulinemia14. Therefore, the aim of our CG cut-off value is to predict patients with a IgG value lower than 600 mg/dL. In addition, we also calculated cut-off values for different IgG levels. The most appropriate CG cut-off value for predicting patients with IgG <500 mg/dl was <1.9 which showed a sensitivity of 82.7% and a specificity of 79.3%. Our study revealed better sensitivity and specificity but with consistent results when compared to the original study by Jolles S. et al2.
In the study of Pecoraro, A. et al., a CG value of <1.9 predicted a sensitivity of 70.0% and a specificity of 75.0% for IgG values lower than 600 mg/dL8. In our study, the most appropriate CG cut-off value, which we selected as <2.0, had a sensitivity of 83.8% and a specificity of 74.9% for predicting patients with this IgG value. Since CG value of 1.9 with a sensitivity of 75.4% and a specificity of 84.4% was also an appropriate cut-off value according to Youden’s index, we could have selected it similar to Pecoraro et al. However, for screening tests, since it is recommended to select a sensitivity higher than 80.0%11, we accepted the cut-off value of <2.0 as the most appropriate.
CG contains other globulins such as alpha 1, alpha 2, beta 1 and beta 2 as well as gamma globulins which are mostly formed by immunoglobulins15,16. Considering the presence of proteins in alpha and beta bands which can act as acute phase reactants, the CG value may remain high despite the presence of low gamma globulin value in cases with chronic disease associated with inflammation and or infection15-17. In addition, the gamma globulin level might be normal while the CG value is low due to the conditions that can decrease the production or increase the catabolism of the alpha and beta bands15. In order to observe this difference, we also measured the gamma globulin values ​​of the patients participating in our study, unlike previous studies performed on adults. Gamma globulin cut-off value of <0.7 which best predicted IgG levels lower than 600 mg/dL, had a higher sensitivity and specificity than the CG cut-off value of <2.0 as expected. In the study of Piza C. et al. conducted on pediatric and adolescent patients, in parallel with our study, it was stated that gamma globulin had a higher sensitivity and specificity than CG and it would be appropriate to use it as a screening test like CG9. However, since CG is obtained in clinics much more frequently and cheaper than gamma globulin, it has the potential to screen a larger number of patients.
In our study, we were able to analyze the distribution of the clinical characteristics of the patients according to the cut-off value of CG. We observed that patients with a CG cut-off value of <2.0 had a significantly higher rate of recurrent parenteral antibiotic use than the others (p=0.025). To date, the CG has been defined as a screening test by the previous studies to predict the possibility of AD. In the current study, we have extended this to demonstrate the clinical implications, importance and utility CG.
The most frequent source clinics of the patients with low IgG values ​ were nephrology and hematology. Excluding diseases such as hypertension and diabetes, which were similar in the whole patient population and not directly related to antibody deficiency, the patients were mostly diagnosed as having chronic renal disease, nephrotic syndrome, multiple myeloma and renal transplantation. In the study of Jolles S. et al., the hematology clinic was the biggest source for the patients with IgG values ​​lower than 400 mg/dL2. However, nephrology was the biggest source for this patient group in our study. We attribute this result to the fact that Istanbul Faculty of Medicine has a large renal transplantation center because patients can receive intensive immunosuppression especially in the early stages of renal transplantation. Not surprisingly, apart from antihypertensives and statins, the most frequent medications were corticosteroids, chemotherapeutics and other immunosuppressive drugs.
The retrospective analyses of the data of 47 patients with primary AD (IgG<600) followed in our clinic showed that 78.7% had a CG value of <2.0 at the time of diagnosis. Pecoraro A. et al. found this rate to be 97.3% (37/38) in a similar analysis performed in their study8. This difference may be due to the clinical conditions of the patients such as the presence of infection and acute phase elevation at the time of admission. Furthermore, in all the patients whose protein electrophoresis were measured at the time of the diagnosis, gamma globulin level was less than 0.7 supporting this possibility.
Besides infections, the wide range of heterogeneous clinical noninfectious presentations of CVID may cause delays in diagnosis2,18. These delays may result in ongoing recurrent infections and irreversible organ damage2. However, early diagnosis which can be achieved by using the CG method might prevent chronic and irreversible complications associated with AD in this patient group19,20. On the other hand, hematological diseases including B cell-associated lymphoproliferative disease, chronic lymphocytic leukemia, multiple myeloma; diseases associated with impaired lymphoid circulation or increased immunoglobulin catabolism; treatments including mofetil mycophenolate, cyclophosphamide, corticosteroids, antiepileptics, rituximab and chemotherapeutics can all lead to SAD. CG represents an unbiased screening approach and could prevent complications that may occur due to delayed diagnosis, not only in PAD but also in SAD7,8.
A potential limitation of our study is that the homogeneity of the patient population could have been affected due to the fact that Istanbul Faculty of Medicine is a tertiary hospital which is considered as a reference center for many diseases. Most of the people whose biochemical analyses were performed in our hospital were ill and there were almost not any healthy people who came just for a check-up. Therefore, numbers of less acutely unwell patients such as those seen most often in primary care may be less well represented in our study.
In conclusion, the current study indicated that a CG level of less than 2.0 mg/dl should be a warning sign for undetected AD in our population. Therefore, given the various manifestations of antibody deficiency and the potential presentations to both primary care and different secondary care services, low CG can alert clinicians for possible AD to prevent diagnostic delay and optimize therapy. CG screening represents a cheap and simple means of detecting primary and secondary antibody deficiency across a range of healthcare settings and we have established cut-off values for both CG and electrophoretic gamma globulin in out Turkish population.