Original Paper Received: December 16, 2013 Accepted after revision: May 5, 2014 Published online: June 4, 2014
Int Arch Allergy Immunol 2014;164:83–88 DOI: 10.1159/000363383
Immunoglobulin G Values before Treatment Are Correlated With the Responsiveness to Initial Intravenous Immunoglobulin Therapy for Kawasaki Disease Kosuke Yanagimoto a Yuichi Nomura a Kiminori Masuda b, c Masako Hirabayashi c Yasuko Morita b, c Michiko Yoshishige b, c Kentaro Ueno a Taisuke Eguchi a Yoshifumi Kawano a
a
Department of Pediatrics, Kagoshima University Graduate School of Medical and Dental Sciences, b Department of Pediatrics, Kagoshima City Hospital and c Department of Pediatrics, Kagoshima Medical Association Hospital, Kagoshima, Japan
Key Words Kawasaki disease · Immunoglobulin G · Coronary artery abnormalities · Intravenous immunoglobulin therapy
Abstract Background: Low levels of serum immunoglobulin G (IgG) before intravenous immunoglobulin (IVIG) therapy for Kawasaki disease (KD) have been reported as one of the risk factors for coronary artery abnormalities (CAAs). This risk factor needs to be re-evaluated because the dosage of IVIG has changed from 0.2–0.4 g/kg/day for 5 days to a single high dose of 2 g/kg. Methods: We reviewed the clinical records of KD patients admitted to our hospital from January 2001 to August 2011. Patients who were given a single high dose of IVIG within 7 days of illness, and who had blood collected for serum immunoglobulin values before treatment, were selected. The serum immunoglobulin levels and coronary artery diameters measured by echocardiogram were transformed to z-scores. Results: The subjects were 197 KD patients, including 22 IVIG nonresponders and 16 patients with CAAs. Of these, 150 (76%) had a z-score for IgG (IgGz) of ≤0. There were no differences in IgGz values between patients with CAAs and those without CAAs. However, nonresponders
© 2014 S. Karger AG, Basel 1018–2438/14/1642–0083$39.50/0 E-Mail [email protected] www.karger.com/iaa
had higher IgGz values than responders (median, 25th percentile and 75th percentile: −0.26, −0.83 and 0.34 vs. −0.79, −1.40 and −0.03; p = 0.020). Logistic regression analysis showed that the IgGz value was an independent risk factor for resistance to IVIG (OR 1.36, 95% CI 1.002–1.849; p = 0.048). Conclusions: Low IgGz values were not a risk factor for CAAs in this study. However, KD patients with relatively high IgGz values before treatment may have an increased risk of resistance to initial IVIG therapy. © 2014 S. Karger AG, Basel
Introduction
Kawasaki disease (KD) is an acute febrile illness that is attributed to systemic vasculitis and it is common in infants and young children [1, 2]. Coronary artery abnormalities (CAAs) are a serious complication in patients with KD, with the potential to cause ischemic heart disease [3]. The current standard treatment for KD is a combination of a single high dose of intravenous immunoglobulin (IVIG) and aspirin, and the incidence of CAAs has decreased to 3.2% from an incidence of 20–25% when patients were treated with aspirin alone [4, 5]. However, Correspondence to: Dr. Kosuke Yanagimoto Department of Pediatrics, Kagoshima University Graduate School of Medical and Dental Sciences, Sakuragaoka 8-35-1 Kagoshima City, Kagoshima Prefecture 890-8520 (Japan) E-Mail ah6k-yngm @ asahi-net.or.jp
16.5% of KD patients do not respond to initial IVIG therapy [4] and these patients have an increased risk of developing CAAs [6]. It is therefore important to investigate the management of IVIG nonresponders. Predictive scores, including the Kobayashi [7], Egami [8] and Sano [9] scores, are widely used in Japan to identify IVIG nonresponders. Although the sensitivity and specificity are high, these predictive scores do not identify all nonresponders. The identification of patients at risk of resistance to IVIG therapy needs to be improved. Better predictive scores would help in the identification and treatment of severely affected patients, potentially reducing the incidence of CAAs. Sawaji et al. [10] reported that low levels of serum immunoglobulin G (IgG) before treatment is one of the risk factors for the development of CAAs. However, the patients in their study were given a small daily dose of IVIG (0.1–0.2 g/ kg/day). Recently, the standard dosage of IVIG has been changed from 0.2–0.4 g/kg/day for 5 days to a single dose of 2 g/kg administered over 1–2 days, and the incidence of CAAs has been decreasing. It is therefore important to investigate whether low serum IgG levels before treatment are still a predictor of the development of CAAs in KD patients who are given a single high dose of IVIG. Furthermore, because there is a correlation between CAAs and nonresponsiveness to IVIG therapy, low serum IgG levels before treatment might be a prognostic factor not only for the risk of CAAs but also for resistance to IVIG therapy. One of the most important pathophysiological features in KD is the increase in capillary permeability caused by systemic vasculitis [11]. Serum proteins such as albumin and IgG leak into extravascular spaces in inflammatory conditions [12], and serum IgG levels in KD patients are likely to be negatively correlated with an increase in capillary permeability. We previously reported that levels of platelet vascular endothelial cell growth factor (VEGF), which increases capillary permeability [13], were positively correlated with the severity of KD [14]. It is thus also important to investigate the relationship between the levels of VEGF and IgG. The aims of this study were to determine whether serum IgG levels in KD patients before high-dose IVIG therapy are a prognostic factor for the development of CAAs or for resistance to IVIG therapy.
who were referred, admitted to the hospital and eventually diagnosed with KD were enrolled. Those who had blood collected for serum immunoglobulin (IgG, IgA and IgM) values before treatment and who were given IVIG (2 g/kg over 1–2 days) within 7 days were selected for the study. Patients given prednisolone in addition to initial IVIG therapy were excluded. The study was approved by the Institutional Review Board of the Kagoshima University Graduate School (20-66, 21-83). Informed consent was obtained from the parents of all study participants before treatment. Data Collection To investigate the relationship between the levels of each serum immunoglobulin and the severity of KD, we assigned patients to groups based on the response or nonresponse to initial IVIG therapy, and on the presence or absence of CAAs. We defined the first day of illness as the day on which fever developed, and made the diagnosis according to the guidelines of KD in Japan [15]. Patients who had persistent fever (>37.5 ° C) for longer than 24 h after the completion of IVIG therapy were defined as nonresponders. All KD patients were administered aspirin at an initial dosage of 30 mg/kg/day (a low dose recommended by the guidelines of the JCS Working Group [15]), reduced to 5 mg/kg/day when levels of C-reactive protein returned to the normal range. CAAs were evaluated using 2-dimensional echocardiography. This was performed before IVIG treatment, every few days during hospitalization, and once a week after hospital discharge until 1 month after the onset of the illness. All coronary artery diameters were transformed to z-scores based on the Japanese normal values of coronary artery dimensions [16]. The z-score was calculated using the following equation: coronary z-score = (observed diameter − mean normal diameter)/(standard deviation of the normal diameter). The presence of CAAs was defined as a coronary zscore of >3.0 at 1 month after the onset of the illness. The levels of each serum immunoglobulin were transformed to z-scores based on Japanese normal values [17]. The z-score was calculated using the following equation: immunoglobulin zscore = (observed value – mean normal value)/(standard deviation of the normal value). We analyzed the differences in the level of each serum immunoglobulin between KD patients with CAAs and without CAAs, and also between IVIG responders and nonresponders. We investigated the correlations between the z-score for each serum immunoglobulin and the duration of fever after the completion of IVIG therapy, and between the z-score for each serum immunoglobulin and the day of illness. We also investigated the correlations between the serum platelet VEGF levels and the day of illness, and between the serum platelet VEGF levels and the z-score for IgG (IgGz). Serum VEGF levels were analyzed by Special Reference Laboratories Inc. (Kagoshima, Japan) using an enzyme-linked immunosorbent assay. Serum samples were collected before the initial IVIG therapy and were stored at –40 ° C until the time of assay. The level of VEGF in a single platelet was calculated using the following equation: serum VEGF (pg/ml) × serum volume (ml)/total platelet count [18].
Materials and Methods Study Participants We retrospectively reviewed the medical records of pediatric patients admitted to Kagoshima Medical Association Hospital in Kagoshima, Japan, from January 2001 to August 2011. Patients
84
Int Arch Allergy Immunol 2014;164:83–88 DOI: 10.1159/000363383
Statistical Analysis All analyses were carried out using SPSS 21.0 statistical software (SPSS, Chicago, Ill., USA). The correlations between serum immunoglobulin levels and the day of illness and between serum immunoglobulin levels and the duration of fever after completion of IVIG
Yanagimoto et al.
Table 1. The comparison between KD patients and healthy controls in serum IgG level
1 month 2 months 3–6 months 7–9 months 10–12 months 1–3 years 3–6 years 6 years
Patients, IgG level, mg/dl n KD patients healthy controls [17] 1 7 19 19 12 80 49 10
465 465±49.0 432±114 510±135 543±148 736±172 850±210 951±131
572±148 390±133 470±180 748±297 792±377 810±96.2 940±184 1,071±165
2 IgGz
Age
4
0
–2 –4 0
1
2
3
4
5
6
7
8
Age (years)
therapy were analyzed using the Spearman rank correlation coefficient. The median levels of each serum immunoglobulin for patients with CAAs and patients without CAAs and for responders and nonresponders were compared using the Mann-Whitney U test. Logistic regression analysis was carried out to assess the risk factors for resistance to IVIG therapy. Receiver operating characteristic (ROC) analysis was used to assess the accuracy of potential prognostic markers for responsiveness to initial IVIG therapy.
Fig. 1. The relationship between IgGz and age. Of 197 KD patients, 150 (76%) had a low IgGz of ≤0.
6
Results
Three hundred and three KD patients were admitted to the hospital during the study period. Of these, 49 were excluded for reasons relating to treatment: 34 were treated with aspirin alone, 5 with a smaller dose of IVIG (0.4 g/ kg/day), 3 were given prednisolone in addition to IVIG and 7 were administered IVIG after the 7th day of illness. Fifty-seven patients were excluded because they did not have blood collected for serum immunoglobulin values. One hundred and ninety-seven patients fulfilled the inclusion criteria (65% of all the KD patients), including 22 nonresponders and 16 patients with CAAs. The median coronary z-score in patients with CAAs was 4.5 (3.0–10). The CAAs reported were coronary artery dilatation or small aneurysms (2.1–4.1 mm in diameter); no patients had giant aneurysms. Absolute IgG mean values dependent on age were lower in patients than in healthy controls except for the group of patients of 2 months of age (table 1). The relationship between IgGz and age is shown in figure 1. As for the absolute IgG mean values, most of the subjects (76%) had an IgGz of ≤0 (median, 25th percentile and 75 percentile: −0.72, −1.4 and 0.0). The values of the z-scores for IgA (IgAz) and IgM (IgMz) were −0.46, −0.94 and 0.15 and 0.79, 0 and 1.82, respectively. IgG as a Prognostic Marker for IVIG Nonresponders
IgGz
4 2 0 –2 –4
0
1
2
3
4
5
6
7
Day of illness
Fig. 2. The relationship between IgGz and the day of illness. The value of IgGz showed two distinct tendencies. The value of IgGz increased slightly during the first 3 days of illness and then decreased.
The relationship between serum IgGz before treatment and the day of illness is shown in figure 2. The pretreatment IgGz values tended to be higher during the first 3 days of illness and then decrease. Overall, IgGz values showed a weak positive correlation with the day of illness, but the trend was not significant (ρ = 0.10, p = 0.16). There was no correlation between the day of illness and IgAz or IgMz values (data not shown). The serum immunoglobulin z-scores for KD patients with and without CAAs are shown in figure 3a. There were no significant differences between z-scores for IgG, Int Arch Allergy Immunol 2014;164:83–88 DOI: 10.1159/000363383
85
6
4
2
z-Score
for IVIG responders and nonresponders (b). a There were no differences in immunoglobulin z-score values between KD patients with CAAs and those without CAAs. b Nonresponders had significantly higher values of IgGz than IVIG responders. * p = 0.020. There were no significant differences in IgAz or IgMz values.
*
6
4 z-Score
Fig. 3. Immunoglobulin z-score values for KD patients with or without CAAs (a), and
0 –2 –4
a
IgG
IgA
0 –2 –4
Patients without CAAs Patients with CAAs
–6
2
IVIG responder IVIG nonresponder
–6
b
IgM
IgG
IgA
IgM
50
6
40
Platelet VEGF 10–8 pg
4
IgGz
2 0 –2
30
20
–4 10
0
2
4
6
8
10
12
14
Duration of fever after completion of IVIG (days) 1
2
3
4
5
6
7
Day of illness
Fig. 4. The relationship between IgGz and the duration of fever
after completion of IVIG therapy. IgGz did not correlate with the duration of fever.
Fig. 5. The relationship between serum platelet VEGF and the day of illness. Platelet VEGF was significantly positively correlated with the day of illness. Platelet VEGF indicates the VEGF contained in a single platelet.
IgA or IgM for patients with or without CAAs. However, as shown in figure 3b, IVIG nonresponders had significantly higher values of IgGz than responders (median, 25th percentile and 75 percentile: −0.26, −0.83 and 0.34 vs. −0.79, −1.40 and −0.03; p = 0.020). There were no significant differences in the values of IgAz or IgMz between responders or nonresponders. The relationship between IgGz and the duration of fever after the completion of IVIG therapy is shown in figure 4. The duration of fever after IVIG is an indicative marker of the responsiveness to treatment. When we ex-
amined the scatter-plots after day 2 that indicated a part of the IVIG nonresponsiveness, the IgGz values seemed to be higher in patients with a longer duration of fever, but the IgGz over the whole duration did not correlate with the duration of fever (ρ = 0.10, p = 0.15). We obtained VEGF values from 138 patients, and investigated the relationship between VEGF and IgGz values. The levels of platelet VEGF had a significant positive correlation with the day of illness (ρ = 0.24, p < 0.01; fig. 5). However, the levels of platelet VEGF did not correlate significantly with the IgGz values (ρ = 0.053, p = 0.53).
86
Int Arch Allergy Immunol 2014;164:83–88 DOI: 10.1159/000363383
Yanagimoto et al.
Logistic regression analysis revealed that IgGz values were an independent risk factor for resistance to IVIG (OR 1.36, 95% CI 1.002–1.849; p = 0.048). ROC analysis showed that an IgGz of −0.60 was a cut-off value with a sensitivity of 73% and a specificity of 59%, and an IgGz of zero was a cut-off value with a sensitivity of 41% and a specificity of 78%. Following the ROC analysis, we probatively examined the sensitivity and the specificity of the combination of IgGz values and the Kobayashi score, which is one of the predictive scores for resistance to IVIG. The Kobayashi score alone applied to the subjects of our study had a sensitivity of 59% and a specificity of 77% at a cut-off point of 5. When screening nonresponders with IgGz values of ≥0 or Kobayashi scores of ≥5 points, the sensitivity and specificity was 82 and 60%, respectively.
Discussion
The incidence of CAAs and resistance to IVIG in this study was similar to that reported in the Japanese national survey [4]; therefore, our subjects were considered to be representative of KD patients in Japan. We determined that most KD patients (76%) have a low value of IgGz (≤0) in the acute phase of the disease. This finding may be attributed to an increase in capillary permeability caused by vasculitis because IgG can leak into extravascular spaces in the inflammatory state [12]. Sawaji et al. [10] reported that low IgGz values were a risk factor for CAAs in KD patients who were given a small daily dose of IVIG. However, in our patients given a single high dose of IVIG, there was no difference in the value of IgGz between those with CAAs and those without CAAs. Newburger et al. [19] also found no differences in IgGz values associated with the presence of CAAs in KD patients given a small daily dose of IVIG (0.4 g/kg/day) or a single high dose (2 g/kg). Therefore, we suggest that, at least in KD patients given a high dose of IVIG, IgGz values are not a prognostic marker for CAAs. The IgGz values of nonresponders were significantly higher than those of responders. We had assumed that the values of IgGz and IgAz might decrease progressively as a result of the persistent increase in capillary permeability with more severe vasculitis. IgGz values showed a tendency to decrease after the third day of illness. The values of IgMz (which has a greater molecular weight) but also the values of IgGz and IgAz did not correlate negatively with the day of illness. Serum levels of platelet VEGF, which increases capillary permeability, were positively correlated with the day of illness. Therefore, we could not IgG as a Prognostic Marker for IVIG Nonresponders
explain the variability of IgGz in terms of capillary permeability alone. There are other factors that may have an association with the variability of serum IgG. Inflammatory cytokines, such as IL-6 and TNF-α, increase in the acute phase of KD and the serum levels of these cytokines are higher in KD patients with CAAs and in IVIG nonresponders [20, 21]. IL-6, produced by CD 4+ T cells, stimulates B cell activation and the production of immunoglobulins [22, 23]. These additional factors may be associated with the relatively high values of IgGz recorded in nonresponders affected by more severe vasculitis, but the variability of IgGz values in the acute phase of KD is complex and cannot be explained as simply. IVIG nonresponders have a particularly high risk of developing CAA [6]. However, there were less KD patients with CAAs than without CAAs even among the IVIG nonresponders. Therefore, the association with IgGz and the responsiveness to IVIG do not necessarily correspond to the association of IgGz and the development of CAAs. However, we think that improving accuracy of a predictive score for IVIG nonresponders would be useful for treatment, along with a strong focus on more severe conditions, and could, furthermore, reduce the incidence of CAAs. Further examination is needed to be able to strongly assert IgG as a prognostic marker. Our study showed that IgGz values, which have not previously been applied to conventional predictive scores, could serve as a prognostic marker for IVIG resistance. In addition, the combination of IgGz values and other predictive scores could improve the accuracy of predicting a nonresponse to IVIG therapy.
Disclosure Statement The authors declare no financial or personal conflict of interest which could interfere with the study outcome.
References
1 Kawasaki T: Acute febrile mucocutaneous syndrome with lymphoid involvement with specific desuamation of the fingers and toes in children. Arerugi 1967;16:178–222. 2 Fujiwara H, Hamashima Y: Pathology of the heart in Kawasaki disease. Pediatrics 1978;61: 100–107. 3 Burns JC, Glode MP: Kawasaki syndrome: Lancet 2004;364:533–544. 4 Nakamura Y, Yashiro M, Uehara R, et al: Epidemiologic features of Kawasaki disease in Japan: results of the 2007–2008 nationwide survey. J Epidemiol 2010;20:302–307.
Int Arch Allergy Immunol 2014;164:83–88 DOI: 10.1159/000363383
87
5 Roweley AH, Shulman ST: Kawasaki disease; in Kliegman RM, Behrman RE, Jenson HB, Stanton BF (eds): Textbook of Pediatrics, ed 18. Philadelphia, WB Saunders, 2007, pp 1036–1042. 6 Uehara R, Belay ED, Maddox RA, et al: Analysis of potential risk factors associated with nonresponse to initial intravenous immunoglobulin treatment among Kawasaki disease patients in Japan. Pediatr Infect Dis J 2008;27: 155–160. 7 Kobayashi T, Inoue Y, Takeuchi K, et al: Prediction of intravenous immunoglobulin unresponsiveness in patients with Kawasaki disease. Circulation 2006;113:2606–2612. 8 Egami K, Muta H, Ishii M, et al: Prediction of resistance to intravenous immunoglobulin treatment in patients with Kawasaki disease. J Pediatr 2006;149:237–240. 9 Sano T, Kurotobi S, Matsuzaki K, et al: Prediction of non-responsiveness to standard highdose gamma-globulin therapy in patients with acute Kawasaki disease before starting initial treatment. Eur J Pediatr 2007;166:131– 137. 10 Sawaji Y, Haneda N, Yamaguchi S, et al: Coronary risk factors in acute Kawasaki disease: Correlation of serum immunoglobulin levels with coronary complications. Acta Paediatr Japon 1998;40:218–225.
88
11 Terai M, Honda T, Yasukawa K, et al: Prognostic impact of vascular leakage in acute Kawasaki disease. Circulation 2003; 108: 325– 330. 12 Egawa G, Nakamizo S, Natsuaki Y, et al: Intravital analysis of vascular permeability in mice using two-photon microscopy. Sci Rep 2013;3:1932. 13 Terai M, Yasukawa K, Narumoto S, et al: Vascular endothelial growth factor in acute Kawasaki disease. Am J Cardiol 1999; 83: 337– 339. 14 Ueno K, Nomura Y, Hashiguchi T, et al: Platelet vascular endothelial growth factor is a useful predictor for prognosis in Kawasaki syndrome. Br J Haematol 2010;148:285–292. 15 JCS Joint Working Group: Guidelines for diagnosis and management of cardiovascular sequelae in Kawasaki disease (JCS 2008). Circ J 2010;74:1989–2020. 16 Kurotobi S, Nagai T, Kawakami N, et al: Coronary diameter in normal infants, children and patients with Kawasaki disease. Pediatr Int 2002;44:1–4.
Int Arch Allergy Immunol 2014;164:83–88 DOI: 10.1159/000363383
17 Fujiwara H, Kawakami K: The research of dynamic state of immunoglobulin in childhood. Nihonsyonikagakkaizassi 1969;73:564–576. 18 Hashiguchi T, Arimura K, Matsumuro K, et al: Highly concentrated vascular endothelial growth factor in platelets in Crow-Fukase syndrome. Muscle Nerve 2000;23:1051–1056. 19 Newburger JW, Takahashi M, Beiser AS, et al: A single intravenous infusion of gamma globulin as compared with four infusions in the treatment of acute Kawasaki syndrome. N Engl J Med 1991;324:1633–1639. 20 Matsubara T: Interleukin 6 activates and tumor necrosis factor-alpha levels in serum of patients with Kawasaki disease. Arerugi 1991; 40:147–154. 21 Lin CY, Lin CC, Hwang B, et al: Cytokines predict coronary aneurysm formation in Kawasaki disease patients. Eur J Pediatr 1993; 152:309–312. 22 Cohen SBA, Clayton J, Londei M, et al: T cells and cytokines; in Balkwill FR (ed): Cytokines: A Practical Approach, ed 2. New York, Oxford University Press, 1995, pp 186–187. 23 Meenu W, Christopher B, Lisa P, et al: Quantitative biological assays for individual cytokines; in Balkwill FR (ed): Cytokines: A Practical Approach, ed 2. New York, Oxford University Press, 1995, p 365.
Yanagimoto et al.
Copyright: S. Karger AG, Basel 2014. Reproduced with the permission of S. Karger AG, Basel. Further reproduction or distribution (electronic or otherwise) is prohibited without permission from the copyright holder.
Int Arch Allergy Immunol 2014;164:83–88 DOI: 10.1159/000363383
Immunoglobulin G Values before Treatment Are Correlated With the Responsiveness to Initial Intravenous Immunoglobulin Therapy for Kawasaki Disease Kosuke Yanagimoto a Yuichi Nomura a Kiminori Masuda b, c Masako Hirabayashi c Yasuko Morita b, c Michiko Yoshishige b, c Kentaro Ueno a Taisuke Eguchi a Yoshifumi Kawano a
a
Department of Pediatrics, Kagoshima University Graduate School of Medical and Dental Sciences, b Department of Pediatrics, Kagoshima City Hospital and c Department of Pediatrics, Kagoshima Medical Association Hospital, Kagoshima, Japan
Key Words Kawasaki disease · Immunoglobulin G · Coronary artery abnormalities · Intravenous immunoglobulin therapy
Abstract Background: Low levels of serum immunoglobulin G (IgG) before intravenous immunoglobulin (IVIG) therapy for Kawasaki disease (KD) have been reported as one of the risk factors for coronary artery abnormalities (CAAs). This risk factor needs to be re-evaluated because the dosage of IVIG has changed from 0.2–0.4 g/kg/day for 5 days to a single high dose of 2 g/kg. Methods: We reviewed the clinical records of KD patients admitted to our hospital from January 2001 to August 2011. Patients who were given a single high dose of IVIG within 7 days of illness, and who had blood collected for serum immunoglobulin values before treatment, were selected. The serum immunoglobulin levels and coronary artery diameters measured by echocardiogram were transformed to z-scores. Results: The subjects were 197 KD patients, including 22 IVIG nonresponders and 16 patients with CAAs. Of these, 150 (76%) had a z-score for IgG (IgGz) of ≤0. There were no differences in IgGz values between patients with CAAs and those without CAAs. However, nonresponders
© 2014 S. Karger AG, Basel 1018–2438/14/1642–0083$39.50/0 E-Mail [email protected] www.karger.com/iaa
had higher IgGz values than responders (median, 25th percentile and 75th percentile: −0.26, −0.83 and 0.34 vs. −0.79, −1.40 and −0.03; p = 0.020). Logistic regression analysis showed that the IgGz value was an independent risk factor for resistance to IVIG (OR 1.36, 95% CI 1.002–1.849; p = 0.048). Conclusions: Low IgGz values were not a risk factor for CAAs in this study. However, KD patients with relatively high IgGz values before treatment may have an increased risk of resistance to initial IVIG therapy. © 2014 S. Karger AG, Basel
Introduction
Kawasaki disease (KD) is an acute febrile illness that is attributed to systemic vasculitis and it is common in infants and young children [1, 2]. Coronary artery abnormalities (CAAs) are a serious complication in patients with KD, with the potential to cause ischemic heart disease [3]. The current standard treatment for KD is a combination of a single high dose of intravenous immunoglobulin (IVIG) and aspirin, and the incidence of CAAs has decreased to 3.2% from an incidence of 20–25% when patients were treated with aspirin alone [4, 5]. However, Correspondence to: Dr. Kosuke Yanagimoto Department of Pediatrics, Kagoshima University Graduate School of Medical and Dental Sciences, Sakuragaoka 8-35-1 Kagoshima City, Kagoshima Prefecture 890-8520 (Japan) E-Mail ah6k-yngm @ asahi-net.or.jp
16.5% of KD patients do not respond to initial IVIG therapy [4] and these patients have an increased risk of developing CAAs [6]. It is therefore important to investigate the management of IVIG nonresponders. Predictive scores, including the Kobayashi [7], Egami [8] and Sano [9] scores, are widely used in Japan to identify IVIG nonresponders. Although the sensitivity and specificity are high, these predictive scores do not identify all nonresponders. The identification of patients at risk of resistance to IVIG therapy needs to be improved. Better predictive scores would help in the identification and treatment of severely affected patients, potentially reducing the incidence of CAAs. Sawaji et al. [10] reported that low levels of serum immunoglobulin G (IgG) before treatment is one of the risk factors for the development of CAAs. However, the patients in their study were given a small daily dose of IVIG (0.1–0.2 g/ kg/day). Recently, the standard dosage of IVIG has been changed from 0.2–0.4 g/kg/day for 5 days to a single dose of 2 g/kg administered over 1–2 days, and the incidence of CAAs has been decreasing. It is therefore important to investigate whether low serum IgG levels before treatment are still a predictor of the development of CAAs in KD patients who are given a single high dose of IVIG. Furthermore, because there is a correlation between CAAs and nonresponsiveness to IVIG therapy, low serum IgG levels before treatment might be a prognostic factor not only for the risk of CAAs but also for resistance to IVIG therapy. One of the most important pathophysiological features in KD is the increase in capillary permeability caused by systemic vasculitis [11]. Serum proteins such as albumin and IgG leak into extravascular spaces in inflammatory conditions [12], and serum IgG levels in KD patients are likely to be negatively correlated with an increase in capillary permeability. We previously reported that levels of platelet vascular endothelial cell growth factor (VEGF), which increases capillary permeability [13], were positively correlated with the severity of KD [14]. It is thus also important to investigate the relationship between the levels of VEGF and IgG. The aims of this study were to determine whether serum IgG levels in KD patients before high-dose IVIG therapy are a prognostic factor for the development of CAAs or for resistance to IVIG therapy.
who were referred, admitted to the hospital and eventually diagnosed with KD were enrolled. Those who had blood collected for serum immunoglobulin (IgG, IgA and IgM) values before treatment and who were given IVIG (2 g/kg over 1–2 days) within 7 days were selected for the study. Patients given prednisolone in addition to initial IVIG therapy were excluded. The study was approved by the Institutional Review Board of the Kagoshima University Graduate School (20-66, 21-83). Informed consent was obtained from the parents of all study participants before treatment. Data Collection To investigate the relationship between the levels of each serum immunoglobulin and the severity of KD, we assigned patients to groups based on the response or nonresponse to initial IVIG therapy, and on the presence or absence of CAAs. We defined the first day of illness as the day on which fever developed, and made the diagnosis according to the guidelines of KD in Japan [15]. Patients who had persistent fever (>37.5 ° C) for longer than 24 h after the completion of IVIG therapy were defined as nonresponders. All KD patients were administered aspirin at an initial dosage of 30 mg/kg/day (a low dose recommended by the guidelines of the JCS Working Group [15]), reduced to 5 mg/kg/day when levels of C-reactive protein returned to the normal range. CAAs were evaluated using 2-dimensional echocardiography. This was performed before IVIG treatment, every few days during hospitalization, and once a week after hospital discharge until 1 month after the onset of the illness. All coronary artery diameters were transformed to z-scores based on the Japanese normal values of coronary artery dimensions [16]. The z-score was calculated using the following equation: coronary z-score = (observed diameter − mean normal diameter)/(standard deviation of the normal diameter). The presence of CAAs was defined as a coronary zscore of >3.0 at 1 month after the onset of the illness. The levels of each serum immunoglobulin were transformed to z-scores based on Japanese normal values [17]. The z-score was calculated using the following equation: immunoglobulin zscore = (observed value – mean normal value)/(standard deviation of the normal value). We analyzed the differences in the level of each serum immunoglobulin between KD patients with CAAs and without CAAs, and also between IVIG responders and nonresponders. We investigated the correlations between the z-score for each serum immunoglobulin and the duration of fever after the completion of IVIG therapy, and between the z-score for each serum immunoglobulin and the day of illness. We also investigated the correlations between the serum platelet VEGF levels and the day of illness, and between the serum platelet VEGF levels and the z-score for IgG (IgGz). Serum VEGF levels were analyzed by Special Reference Laboratories Inc. (Kagoshima, Japan) using an enzyme-linked immunosorbent assay. Serum samples were collected before the initial IVIG therapy and were stored at –40 ° C until the time of assay. The level of VEGF in a single platelet was calculated using the following equation: serum VEGF (pg/ml) × serum volume (ml)/total platelet count [18].
Materials and Methods Study Participants We retrospectively reviewed the medical records of pediatric patients admitted to Kagoshima Medical Association Hospital in Kagoshima, Japan, from January 2001 to August 2011. Patients
84
Int Arch Allergy Immunol 2014;164:83–88 DOI: 10.1159/000363383
Statistical Analysis All analyses were carried out using SPSS 21.0 statistical software (SPSS, Chicago, Ill., USA). The correlations between serum immunoglobulin levels and the day of illness and between serum immunoglobulin levels and the duration of fever after completion of IVIG
Yanagimoto et al.
Table 1. The comparison between KD patients and healthy controls in serum IgG level
1 month 2 months 3–6 months 7–9 months 10–12 months 1–3 years 3–6 years 6 years
Patients, IgG level, mg/dl n KD patients healthy controls [17] 1 7 19 19 12 80 49 10
465 465±49.0 432±114 510±135 543±148 736±172 850±210 951±131
572±148 390±133 470±180 748±297 792±377 810±96.2 940±184 1,071±165
2 IgGz
Age
4
0
–2 –4 0
1
2
3
4
5
6
7
8
Age (years)
therapy were analyzed using the Spearman rank correlation coefficient. The median levels of each serum immunoglobulin for patients with CAAs and patients without CAAs and for responders and nonresponders were compared using the Mann-Whitney U test. Logistic regression analysis was carried out to assess the risk factors for resistance to IVIG therapy. Receiver operating characteristic (ROC) analysis was used to assess the accuracy of potential prognostic markers for responsiveness to initial IVIG therapy.
Fig. 1. The relationship between IgGz and age. Of 197 KD patients, 150 (76%) had a low IgGz of ≤0.
6
Results
Three hundred and three KD patients were admitted to the hospital during the study period. Of these, 49 were excluded for reasons relating to treatment: 34 were treated with aspirin alone, 5 with a smaller dose of IVIG (0.4 g/ kg/day), 3 were given prednisolone in addition to IVIG and 7 were administered IVIG after the 7th day of illness. Fifty-seven patients were excluded because they did not have blood collected for serum immunoglobulin values. One hundred and ninety-seven patients fulfilled the inclusion criteria (65% of all the KD patients), including 22 nonresponders and 16 patients with CAAs. The median coronary z-score in patients with CAAs was 4.5 (3.0–10). The CAAs reported were coronary artery dilatation or small aneurysms (2.1–4.1 mm in diameter); no patients had giant aneurysms. Absolute IgG mean values dependent on age were lower in patients than in healthy controls except for the group of patients of 2 months of age (table 1). The relationship between IgGz and age is shown in figure 1. As for the absolute IgG mean values, most of the subjects (76%) had an IgGz of ≤0 (median, 25th percentile and 75 percentile: −0.72, −1.4 and 0.0). The values of the z-scores for IgA (IgAz) and IgM (IgMz) were −0.46, −0.94 and 0.15 and 0.79, 0 and 1.82, respectively. IgG as a Prognostic Marker for IVIG Nonresponders
IgGz
4 2 0 –2 –4
0
1
2
3
4
5
6
7
Day of illness
Fig. 2. The relationship between IgGz and the day of illness. The value of IgGz showed two distinct tendencies. The value of IgGz increased slightly during the first 3 days of illness and then decreased.
The relationship between serum IgGz before treatment and the day of illness is shown in figure 2. The pretreatment IgGz values tended to be higher during the first 3 days of illness and then decrease. Overall, IgGz values showed a weak positive correlation with the day of illness, but the trend was not significant (ρ = 0.10, p = 0.16). There was no correlation between the day of illness and IgAz or IgMz values (data not shown). The serum immunoglobulin z-scores for KD patients with and without CAAs are shown in figure 3a. There were no significant differences between z-scores for IgG, Int Arch Allergy Immunol 2014;164:83–88 DOI: 10.1159/000363383
85
6
4
2
z-Score
for IVIG responders and nonresponders (b). a There were no differences in immunoglobulin z-score values between KD patients with CAAs and those without CAAs. b Nonresponders had significantly higher values of IgGz than IVIG responders. * p = 0.020. There were no significant differences in IgAz or IgMz values.
*
6
4 z-Score
Fig. 3. Immunoglobulin z-score values for KD patients with or without CAAs (a), and
0 –2 –4
a
IgG
IgA
0 –2 –4
Patients without CAAs Patients with CAAs
–6
2
IVIG responder IVIG nonresponder
–6
b
IgM
IgG
IgA
IgM
50
6
40
Platelet VEGF 10–8 pg
4
IgGz
2 0 –2
30
20
–4 10
0
2
4
6
8
10
12
14
Duration of fever after completion of IVIG (days) 1
2
3
4
5
6
7
Day of illness
Fig. 4. The relationship between IgGz and the duration of fever
after completion of IVIG therapy. IgGz did not correlate with the duration of fever.
Fig. 5. The relationship between serum platelet VEGF and the day of illness. Platelet VEGF was significantly positively correlated with the day of illness. Platelet VEGF indicates the VEGF contained in a single platelet.
IgA or IgM for patients with or without CAAs. However, as shown in figure 3b, IVIG nonresponders had significantly higher values of IgGz than responders (median, 25th percentile and 75 percentile: −0.26, −0.83 and 0.34 vs. −0.79, −1.40 and −0.03; p = 0.020). There were no significant differences in the values of IgAz or IgMz between responders or nonresponders. The relationship between IgGz and the duration of fever after the completion of IVIG therapy is shown in figure 4. The duration of fever after IVIG is an indicative marker of the responsiveness to treatment. When we ex-
amined the scatter-plots after day 2 that indicated a part of the IVIG nonresponsiveness, the IgGz values seemed to be higher in patients with a longer duration of fever, but the IgGz over the whole duration did not correlate with the duration of fever (ρ = 0.10, p = 0.15). We obtained VEGF values from 138 patients, and investigated the relationship between VEGF and IgGz values. The levels of platelet VEGF had a significant positive correlation with the day of illness (ρ = 0.24, p < 0.01; fig. 5). However, the levels of platelet VEGF did not correlate significantly with the IgGz values (ρ = 0.053, p = 0.53).
86
Int Arch Allergy Immunol 2014;164:83–88 DOI: 10.1159/000363383
Yanagimoto et al.
Logistic regression analysis revealed that IgGz values were an independent risk factor for resistance to IVIG (OR 1.36, 95% CI 1.002–1.849; p = 0.048). ROC analysis showed that an IgGz of −0.60 was a cut-off value with a sensitivity of 73% and a specificity of 59%, and an IgGz of zero was a cut-off value with a sensitivity of 41% and a specificity of 78%. Following the ROC analysis, we probatively examined the sensitivity and the specificity of the combination of IgGz values and the Kobayashi score, which is one of the predictive scores for resistance to IVIG. The Kobayashi score alone applied to the subjects of our study had a sensitivity of 59% and a specificity of 77% at a cut-off point of 5. When screening nonresponders with IgGz values of ≥0 or Kobayashi scores of ≥5 points, the sensitivity and specificity was 82 and 60%, respectively.
Discussion
The incidence of CAAs and resistance to IVIG in this study was similar to that reported in the Japanese national survey [4]; therefore, our subjects were considered to be representative of KD patients in Japan. We determined that most KD patients (76%) have a low value of IgGz (≤0) in the acute phase of the disease. This finding may be attributed to an increase in capillary permeability caused by vasculitis because IgG can leak into extravascular spaces in the inflammatory state [12]. Sawaji et al. [10] reported that low IgGz values were a risk factor for CAAs in KD patients who were given a small daily dose of IVIG. However, in our patients given a single high dose of IVIG, there was no difference in the value of IgGz between those with CAAs and those without CAAs. Newburger et al. [19] also found no differences in IgGz values associated with the presence of CAAs in KD patients given a small daily dose of IVIG (0.4 g/kg/day) or a single high dose (2 g/kg). Therefore, we suggest that, at least in KD patients given a high dose of IVIG, IgGz values are not a prognostic marker for CAAs. The IgGz values of nonresponders were significantly higher than those of responders. We had assumed that the values of IgGz and IgAz might decrease progressively as a result of the persistent increase in capillary permeability with more severe vasculitis. IgGz values showed a tendency to decrease after the third day of illness. The values of IgMz (which has a greater molecular weight) but also the values of IgGz and IgAz did not correlate negatively with the day of illness. Serum levels of platelet VEGF, which increases capillary permeability, were positively correlated with the day of illness. Therefore, we could not IgG as a Prognostic Marker for IVIG Nonresponders
explain the variability of IgGz in terms of capillary permeability alone. There are other factors that may have an association with the variability of serum IgG. Inflammatory cytokines, such as IL-6 and TNF-α, increase in the acute phase of KD and the serum levels of these cytokines are higher in KD patients with CAAs and in IVIG nonresponders [20, 21]. IL-6, produced by CD 4+ T cells, stimulates B cell activation and the production of immunoglobulins [22, 23]. These additional factors may be associated with the relatively high values of IgGz recorded in nonresponders affected by more severe vasculitis, but the variability of IgGz values in the acute phase of KD is complex and cannot be explained as simply. IVIG nonresponders have a particularly high risk of developing CAA [6]. However, there were less KD patients with CAAs than without CAAs even among the IVIG nonresponders. Therefore, the association with IgGz and the responsiveness to IVIG do not necessarily correspond to the association of IgGz and the development of CAAs. However, we think that improving accuracy of a predictive score for IVIG nonresponders would be useful for treatment, along with a strong focus on more severe conditions, and could, furthermore, reduce the incidence of CAAs. Further examination is needed to be able to strongly assert IgG as a prognostic marker. Our study showed that IgGz values, which have not previously been applied to conventional predictive scores, could serve as a prognostic marker for IVIG resistance. In addition, the combination of IgGz values and other predictive scores could improve the accuracy of predicting a nonresponse to IVIG therapy.
Disclosure Statement The authors declare no financial or personal conflict of interest which could interfere with the study outcome.
References
1 Kawasaki T: Acute febrile mucocutaneous syndrome with lymphoid involvement with specific desuamation of the fingers and toes in children. Arerugi 1967;16:178–222. 2 Fujiwara H, Hamashima Y: Pathology of the heart in Kawasaki disease. Pediatrics 1978;61: 100–107. 3 Burns JC, Glode MP: Kawasaki syndrome: Lancet 2004;364:533–544. 4 Nakamura Y, Yashiro M, Uehara R, et al: Epidemiologic features of Kawasaki disease in Japan: results of the 2007–2008 nationwide survey. J Epidemiol 2010;20:302–307.
Int Arch Allergy Immunol 2014;164:83–88 DOI: 10.1159/000363383
87
5 Roweley AH, Shulman ST: Kawasaki disease; in Kliegman RM, Behrman RE, Jenson HB, Stanton BF (eds): Textbook of Pediatrics, ed 18. Philadelphia, WB Saunders, 2007, pp 1036–1042. 6 Uehara R, Belay ED, Maddox RA, et al: Analysis of potential risk factors associated with nonresponse to initial intravenous immunoglobulin treatment among Kawasaki disease patients in Japan. Pediatr Infect Dis J 2008;27: 155–160. 7 Kobayashi T, Inoue Y, Takeuchi K, et al: Prediction of intravenous immunoglobulin unresponsiveness in patients with Kawasaki disease. Circulation 2006;113:2606–2612. 8 Egami K, Muta H, Ishii M, et al: Prediction of resistance to intravenous immunoglobulin treatment in patients with Kawasaki disease. J Pediatr 2006;149:237–240. 9 Sano T, Kurotobi S, Matsuzaki K, et al: Prediction of non-responsiveness to standard highdose gamma-globulin therapy in patients with acute Kawasaki disease before starting initial treatment. Eur J Pediatr 2007;166:131– 137. 10 Sawaji Y, Haneda N, Yamaguchi S, et al: Coronary risk factors in acute Kawasaki disease: Correlation of serum immunoglobulin levels with coronary complications. Acta Paediatr Japon 1998;40:218–225.
88
11 Terai M, Honda T, Yasukawa K, et al: Prognostic impact of vascular leakage in acute Kawasaki disease. Circulation 2003; 108: 325– 330. 12 Egawa G, Nakamizo S, Natsuaki Y, et al: Intravital analysis of vascular permeability in mice using two-photon microscopy. Sci Rep 2013;3:1932. 13 Terai M, Yasukawa K, Narumoto S, et al: Vascular endothelial growth factor in acute Kawasaki disease. Am J Cardiol 1999; 83: 337– 339. 14 Ueno K, Nomura Y, Hashiguchi T, et al: Platelet vascular endothelial growth factor is a useful predictor for prognosis in Kawasaki syndrome. Br J Haematol 2010;148:285–292. 15 JCS Joint Working Group: Guidelines for diagnosis and management of cardiovascular sequelae in Kawasaki disease (JCS 2008). Circ J 2010;74:1989–2020. 16 Kurotobi S, Nagai T, Kawakami N, et al: Coronary diameter in normal infants, children and patients with Kawasaki disease. Pediatr Int 2002;44:1–4.
Int Arch Allergy Immunol 2014;164:83–88 DOI: 10.1159/000363383
17 Fujiwara H, Kawakami K: The research of dynamic state of immunoglobulin in childhood. Nihonsyonikagakkaizassi 1969;73:564–576. 18 Hashiguchi T, Arimura K, Matsumuro K, et al: Highly concentrated vascular endothelial growth factor in platelets in Crow-Fukase syndrome. Muscle Nerve 2000;23:1051–1056. 19 Newburger JW, Takahashi M, Beiser AS, et al: A single intravenous infusion of gamma globulin as compared with four infusions in the treatment of acute Kawasaki syndrome. N Engl J Med 1991;324:1633–1639. 20 Matsubara T: Interleukin 6 activates and tumor necrosis factor-alpha levels in serum of patients with Kawasaki disease. Arerugi 1991; 40:147–154. 21 Lin CY, Lin CC, Hwang B, et al: Cytokines predict coronary aneurysm formation in Kawasaki disease patients. Eur J Pediatr 1993; 152:309–312. 22 Cohen SBA, Clayton J, Londei M, et al: T cells and cytokines; in Balkwill FR (ed): Cytokines: A Practical Approach, ed 2. New York, Oxford University Press, 1995, pp 186–187. 23 Meenu W, Christopher B, Lisa P, et al: Quantitative biological assays for individual cytokines; in Balkwill FR (ed): Cytokines: A Practical Approach, ed 2. New York, Oxford University Press, 1995, p 365.
Yanagimoto et al.
Copyright: S. Karger AG, Basel 2014. Reproduced with the permission of S. Karger AG, Basel. Further reproduction or distribution (electronic or otherwise) is prohibited without permission from the copyright holder.
