- Email: [email protected]
Tonsillitis index: An objective tool for quantifying the indications for tonsillectomy for recurrent acute tonsillitis
International Journal of Pediatric Otorhinolaryngology (2005) 69, 1515—1520
www.elsevier.com/locate/ijporl
Tonsillitis index: An objective tool for quantifying the indications for tonsillectomy for recurrent acute tonsillitis Keiji Fujihara a,*, Hironobu Goto a, Masanobu Hiraoka a, Masaki Hayashi a, Muneki Hotomi a, Shinji Tamura a, Kiyonori Kuki a, Noboru Yamanaka a, Peter J. Koltai b a
Department of Otolaryngology–—Head and Neck Surgery, Wakayama Medical University, 811-1 Kimiidera, Wakayama City 641-8509, Japan b Division of Pediatric Otolaryngology, Stanford University School of Medicine, Stanford, CA, USA Received 8 December 2004; accepted 18 April 2005
KEYWORDS Tonsillitis index; Tonsillectomy; Recurrent acute tonsillitis; Natural history of tonsillitis; The co-stimulatory signals; CD80; CD86
Summary This report is a preliminary exploration of the concept of a ‘‘Tonsillectomy Index’’ (TI) as an objective tool for quantifying the indications for tonsillectomy for recurrent acute tonsillitis (AT). The TI is derived by multiplying the number of episodes of AT by the number of years during which the episodes of AT occurred. Our objective in this study was to investigate whether there is a relationship between the natural history of AT, the immunological functions of tonsils and our proposed TI. For the natural history of AT, we medically followed 11 children with a history of AT for 5 years. When TI was equal to or greater than 8 (TI 3 8), the children suffered a significantly greater number of episodes of AT. For the immunological portion of our study, we enrolled 36 children and 46 adults undergoing tonsillectomy for either AT (study group) or tonsillar hypertrophy (control group, CG). We analyzed the costimulatory signals, CD80 and CD86 on tonsillar B-lymphocytes. The expression rates of CD80 and CD86 in the AT group with TI 3 8 were significantly decreased compared to those with TI was less than 8 (TI < 8), as well as with those in control (tonsillar hypertrophy) group. Our preliminary findings suggest that when the TI 3 8, the tonsils have deteriorated immunologically and spontaneous resolution of recurrent AT is less likely to occur, hence tonsillectomy is appropriate. TI may be a useful tool for surgical decision making. # 2005 Elsevier Ireland Ltd. All rights reserved.
* Corresponding author. Tel.: +81 73 441 0651; fax: +81 73 446 3846. E-mail address: [email protected] (K. Fujihara). 0165-5876/$ — see front matter # 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2005.04.007
1516
K. Fujihara et al.
1. Introduction Tonsillectomy has been shown to be effective in the management of recurrent acute tonsillitis (AT) [1,2], nevertheless, the indications of tonsillectomy for recurrent AT are controversial, unsettled and there is no consensus as to which children would benefit from such surgery. Tonsillectomy has gone through periods of enthusiasm, as well as uncertainty, as to its overall benefit in children. There is a natural decline in the frequency of upper respiratory infections as children age and many children will ‘outgrow’ their recurrent throat infections [3,4]. The development of antibiotic therapy has greatly influenced our ability to treat recurrent throat infections and prevent the serious sequelae that can arise from them. Some published reports have suggested that the removal of the tonsils was of little, if any, benefit in children [5]. In the United States, the American Academy of Otolaryngology–—Head and Neck Surgery suggests that children with three or more infections of the tonsils and/or adenoids per year, despite adequate medical therapy, are candidates for tonsillectomy [6]. Other national health organizations have published guidelines for tonsillectomy based on the frequency of recurrent infections [7,8]. In Japan, the consensus indication for tonsillectomy is four episodes of AT in each of preceding 2 years. While there is no world wide agreement about a specific number of bouts of AT that warrants tonsillectomy, most clinicians agree that some range of the number of infections over a period of time would be an indication for surgical intervention. Based on this insight, we explore the preliminary concept of ‘‘Ton-
sillitis Index’’ (TI), defined as the number of episodes of AT multiplied by the number of years during which the episodes of AT occurred. Our goal in suggesting this concept is an attempt to establish an objective tool for quantifying the indications for tonsillectomy in AT. The tonsil is an immunological organ and it seems reasonable to consider how recurrent infection affects its immuno-physiology when considering the indication for tonsillectomy. Tonsillar lymphocytes from recurrently infected tonsil have a significantly lower uptake of [3H] thymidine than those from tonsillar hypertrophy [9]. Patients with frequent episodes of tonsillitis have a significant increase of IgD-positive cells and IgG-positive cells in interfollicular and subepithelial compartments and a decrease of CD4-positive T-cells in the germinal centers and subepithelial areas [10]. Antigen specific activation of Tcells in host defense system is important and T cell activation requires two signaling events. The first, an Ag-specific signal delivered by the MHC/Ag complex interacting with TCR. The complete T cell activation depends on a second signal, defined as co-stimulation, by a molecule known as CD28 [11]. The ligands for CD28 are CD80 and CD86, which appear on B cells and antigen presenting cells. When the CD28—CD80 and CD86 signaling pathway are blocked, it results in inhibition of T cell proliferation, as well as reduction of a subset of T cell clones, inducing cellular Ag-specific unresponsiveness, termed anergy [12]. It is our hypothesis that these tonsillar immunological abnormalities can be correlated with the frequency of recurrent infection, the natural history of AT and our proposed TI.
Table 1 The patients enrolled in the study of spontaneous resolution of recurrent acute tonsillitis (AT) No.
Sex
TI
1 2 3 4 5 6 7 8 9 10 11
M M F F F M F F F M M
28 21 20 17.5 14 12 10 7.5 7 5 2.5 13.1 7.8
Mean S.D.
First episode
The most frequent episodes
The latest episodes
Follow up period
Age#
Age *
Frequncy$
Age
Frequncy
1 0 2 1.5 4 1 1 2.5 1 2 1
5 3 7 4.5 6 4 3 4.5 2.5 3 1
7 7 4 7 7 4 5 5 7 5 5
5 5 7 7 7 5 8 7 4 7 4
6 7 4 3 5 3 2 2 5 3 1
4 5 5 5.5 3 4 7 4.5 3 5 3
1.5 1.1
4.0 1.7
5.7 1.3
6.0 1.4
3.7 1.8
4.5 1.2
TI: Tonsillitis index; TI is derived by multiplying frequency$ at the most frequent episode by the difference between age# at the first episode and age* at the most frequent episodes.
Tonsillitis index for tonsillectomy
2. Materials and methods 2.1. Part 1 — spontaneous resolution of recurrent acute tonsillitis We retrospectively reviewed 11 children (4—8 years old, 5 males and 6 females) in a uncontrolled fashion, with a history of recurrent AT who had been managed medically for their infections for no less than 3 years at the Wakayama Medical University Hospital and referred clinics in Wakayama, Japan (Table 1). Our criteria for an episode of ATwas acute pharyngeal pain and fever equal to or greater than 38 8C. Despite being treated nonsurgically, the only patients included in the review were those who met the American Academy of Otolaryngology clinical practice guidelines for consideration for tonsillectomy [1]. At the end of the minimum of a 3-year follow up, their care givers were asked to respond to a questionnaire, which included age at the first episode, frequency of tonsillitis and age at the most frequent episodes and at the most recent year of follow up (Table 1). We then statistically analyzed by simple regression, the correlation of TI at the most frequent episodes of AT and the number of episodes at the latest year of follow up.
2.2. Part 2 — immunological assay of tonsillar function 2.2.1. Patient selection We studied the immunological function of the tonsillar specimens of 36 children (ages 3—15 years old, 20 males and 16 females) and 46 adults (ages 19—45, 26 males and 20 females) obtained during tonsillectomy for hypertrophic tonsil or recurrent AT at Wakayama Medical University Hospital or affiliated hospitals. Informed consent was obtained from patients or their parents before the enrollment in this study. The specimens from the patients with a diagnosis of tonsillar hypertrophy and without a history of AT were designated as the control group (CG). Specimens from the patients with recurrent AT were designated as tonsillitis group (TG). We then divided the TG into those that had TI equal to or more than 8 (TI 3 8) and a TI less than 8 (TI < 8). Our decision to use the TI = 8 as the dividing point for the patients in the TG was based on the initial part 1 of this study which suggested that patients who had less than 8 episodes of AT were more likely to have spontaneous resolution of their AT than those who had more than 8 episodes. 2.2.2. B cell purification Tonsillar mononuclear cells were purified by the standard Ficoll—Hypaque gradient method. The
1517
lymphocytes were immunostained with monoclonal antibodies as described below. 2.2.3. Antibodies The monoclonal antibodies used for phenotypic and functional studies were as follows; FITC conjugated anti-CD80 (PharMingen, Franklin Lakes, NJ), FITC conjugated anti-CD86 (PharMingen, Franklin Lakes, NJ) and PE conjugated anti-CD19 (PharMingen, Franklin Lakes, NJ). 2.2.4. Flow cytometric analysis Antibody binding was analyzed on a FACS caliber (Becton Dickinson, Franklin Lakes, NJ). The FACS caliber was used for producing dot plot. We analyzed the data of CD80 and CD86 in each of three groups (the CG, the TG with TI < 8 and the TG with TI 3 8) with ANOVA post-hoc test. The correlations of CD80 and CD86 were also analyzed in the three groups by simple regression assay. 2.2.5. Statistics ANOVA post-hoc test and simple regression analysis were performed using StatView Version 5.0 (SAS Institute, Cary, NC, USA). P-values less than 0.05 were designated as significant.
3. Results 3.1. Part 1 — spontaneous resolution of recurrent acute tonsillitis Among the children who were medically managed for AT, the average follow up time was 4.5 1.2 years. The first episode of AT occurred at 1.5 1.1 years of age. The most frequent bouts of AT, 5.7 1.3 episodes a year, occurred at 4.0 1.7 years of age. The ages at the last year of follow up were 6.0 1.4 years of age. There were 3.7 1.8 episodes of AT during the last year of follow up. The TI was derived by
Fig. 1 The correlation between TI and the number of sore throat with fever at the latest year.
1518
K. Fujihara et al.
multiplying the number of episodes of AT by the number of years between age at the first episode and age at the most frequent episodes. TI varied from 2.5 to 28 (Table 1). The frequency of AT was not correlated with age. TI was correlated with the frequency of AT during the last year of follow up. Regression analysis yielded a simple regression line (y = 1.52 + 0.168x, R2 = 0.503, P < 0.05) indicating that when the TI was equal to or more than 8 (TI 3 8), the children continued to suffer from AT, 3 or more times a year up to 4.5 years after the onset of infections (Fig. 1).
3.2. Immunological functions of the tonsil 3.2.1. Flow cytometric analysis Tonsillar lymphocytes were harvested from the surgical specimens in the immediate perioperartive period and flow cytometric analysis was performed in the same day. In the pediatric control group, the CD80 positive rate of tonsillar B-lymphocytes was 11.26 1.21% (mean S.D.). In the pediatric group with TI < 8, the positive rate of CD80 was significantly lower (7.38 1.53%) than that of the control group. In the pediatric group with TI 3 8, the positive rate of CD80 was significantly lower (6.1 0.68%) than that of both the control and the TI < 8 groups. In the pediatric control group, the CD86 positive rate of tonsillar B-lymphocytes was 11.45 0.77%. In the pediatric group with TI < 8, the positive rate of CD86 was significantly lower (6.06 1.50%) than that of the control group. In the pediatric group with TI 3 8, the positive rate of CD86 was significantly lower (3.67 1.22%) than that of both the control and the TI < 8 groups (Fig. 2a). In adult control group, the CD80 positive rate of tonsillar B-lymphocytes was 9.66 0.7%. In the adult group with TI < 8, the positive rate of CD80 was significantly lower (4.04 0.97%) than that of the control group. In the adult group with TI 3 8, the positive rate of CD80 was significantly lower (3.04 1.17%) than that of both the control and the TI < 8 groups. In adult control group, CD86 positive rate of tonsillar B-lymphocytes was 8.98 1.23%. In the adult group with TI < 8, the positive rate of CD86 was significantly lower (3.95 1.24%) than that of the control group. In the adult group with TI 3 8, the positive rate of CD86 was significantly lower (3.01 0.92%) than that of both the control and the TI < 8 group (Fig. 2b). 3.2.2. Simple regression assay In both the pediatric and adult control groups, the expression rates of CD80 and CD86 in tonsillar B-
Fig. 2 (a) The positive rates of CD80 and CD86 in children. TI < 8: TI was less than 8; TI 3 8: TI was equal to or greater than 8. **P < 0.01, *P < 0.05. (b) The positive rates of CD80 and CD86 in adults. **P < 0.01, *P < 0.05.
Fig. 3 (a) The correlation of CD80 and CD86 in tonsillar B-lymphocytes in the control, TI < 8 and TI 3 8 in children. **P < 0.01. (b) The correlation of CD80 and CD86 in tonsillar B-lymphocytes in the control, TI < 8 and TI 3 8 in adults. *P < 0.05.
Tonsillitis index for tonsillectomy
lymphocytes were almost identical with significant correlations between CD80 and CD86 in both children (Fig. 3a) and adults (Fig. 3b). However, in both the pediatric and adult groups with TI < 8 and with TI 3 8, the expression rates of CD80 and CD86 were decreased and showed less correlation between them.
4. Discussion The palatine tonsils are strategically located at the entrance of both the respiratory and gastrointestinal tracts. They are primarily immune organs for protection against external microorganisms and as such, a source of health. However, because of their specific cryptic architecture, they can become recurrently infected and thus become a source of disease. When the balance between immunologic protection and recurrent infection tilts persistently toward illness, tonsillectomy should be considered. Defining the fulcrum of that balance with precision has been remarkably difficult. In the tonsils, functional interactions between B and T lymphocytes are known to depend on the expression of co-stimulatory molecules (CD80 and CD86) and their counter-receptors (CD28 and CTLA4). CD86 induces naive T cells to become IL-4 producers, thereby directing the immune response more towards Th2, whereas CD80 is a more neutral differentiative signal for Th1 and Th2 [12]. The derangement of the CD80 and CD86—CD28 co-stimulatory pathway may change the clinical form of the disease by depressing cell-mediated immunity and inducing T cell anergy [13]. In this study, CD80 and CD86 were expressed at almost same rate and were significantly correlated in the healthy control tonsils. However, in the groups with AT, CD80 and CD86 were significantly decreased and did not correlate as they did in the controls. Furthermore, the expression of both CD80 and CD86 was more decreased and imbalanced in the groups with more persistent and severe disease. To the best of our knowledge this is the first data which demonstrates a clinical correlation between the indication for tonsillectomy and an immunological derangement in the function of tonsils. Studies performed at the University of Pittsburgh have sought to answer many of the questions regarding the efficacy of tonsillectomy in children afflicted with recurrent pharyngotonsillitis [1,2]. In order to be entered into the first study [1] children were required to have at least three episodes of tonsillitis in each of the prior 3 years or 5 episodes in each of 2 years or seven episodes in 1 year. In the first and second years of follow up, those children who received tonsillectomy had statistically fewer epi-
1519
sodes of throat infection when compared to the children in the nonsurgical arm. The third year of follow up resulted in a trend towards improvement in the surgical arm. The nonsurgical group experienced a decline in the frequency of throat infections in each of the follow up years. These studies are widely accepted as the methodologically best investigations regarding the efficacy of tonsillectomy for children with AT, and their criteria for surgical intervention are considered both judicious and appropriate. Applying our proposed concept of TI to the Pittsburgh criteria would yield a TI of 7 for 7 infections in any 1 year, a TI of 10 for 5 infections 2 years in a row and a TI of 9 for 3 infections 3 years in a row, rendering an average TI of 8.67. This is remarkably consistent with our own findings, derived by very different methodology. Our results suggest that when the TI 3 8, there is a significantly greater likelihood that recurrent ATwill persist and not resolve spontaneously, perhaps as a consequence of the tonsil having deteriorated immunologically. This may explain why spontaneous resolution of recurrent AT is less likely to occur and also suggests that a TI 3 8 is a reasonable indication for tonsillectomy. Nevertheless, at this point our study’s shortcomings preclude a strong conclusion in this regard. The number of subjects is too small and there is a lack of prospective validated diagnostic criteria. With the limited sample size, it is not possible to derive reliable sensitivity and specificity information or confidence limits. Furthermore, in the immunological part of our study, using specimens from patients with tonsillar hypertrophy as our controls is methodologically problematic. This is also the case with bundling pediatric and adult patients, given the possibility that there may be important clinical differences in tonsillitis between children and grownups. The concept of a TI should be an accurate and useful tool for surgical decision making. Our preliminary data points in this direction.
Acknowledgements We would like to thank Dr. T. Jinnin, Dr. T. Kimura and Dr. T. Fujihara in Wakayama for follow up data of recurrent AT at their clinics.
References [1] J.L. Paradise, C.D. Bluestone, R.Z. Bachman, D.K. Colborn, B.S. Bernard, F.H. Taylor, et al. Efficacy of tonsillectomy for recurrent throat infections in severely affected children, N. Engl. J. Med. 310 (1984) 674—683.
1520
[2] J.L. Paradise, C.D. Bluestone, D.K. Colborn, B.S. Bernard, H.E. Rockette, M. Kurs-Lasky, Tonsillectomy and adenotonsillectomy for recurrent throat infection in moderately affected children, Pediatrics 110 (1) (2002) 7—15. [3] T.J. Woodford, A. Ahmed, D.J. Willatt, et al. Spontaneous resolution of tonsillitis in children on the waiting list for tonsillectomy, Clin. Otolaryngol. 25 (2000) 428—430. [4] M.P. Prim, J.I. de Diego, M. Larrauri, C. Diaz, N. Sastre, J. Gavilan, Spontaneous resolution of recurrent tonsillitis in pediatric patients on the surgical waiting list, Int. J. Pediatr. Otorhinolaryngol. 65 (1) (2002) 35—38. [5] J.L. Paradise, Tonsillectomy and adenoidectomy, in: C.D. Bluestone, S.E. Stool, M.A. Kenna (Eds.), Pediatric Otolaryngology, third ed., WB Saunders, Philadelphia, 1996, pp. 1054—1065. [6] American Academy of Otolaryngology–—Head and Neck Surgery, 2000 Clinical Indicators Compendium, vol. 9, issue no. 6, Alexandria, VA, 2000, p. 19. [7] Finnish Medical Society Duodecim: Sore Throat and Tonsillitis, EBM Guidelines, Duodecim Medical Publications Ltd., 2001, pp. 1—4.
K. Fujihara et al.
[8] Scottish Intercollegiate Guideline Network (SIGN): Management of Sore Throat and Indications for Tonsillectomy, A National Clinical Guideline, 1999, p. 23. [9] N. Yamanaka, K. Kobayashi, T. Himi, F. Shido, A. Kataura, Spontaneous DNA synthesis in tonsillar lymphocytes and its clinical implications, Acta Otolaryngol. 96 (1—2) (1983) 181—187. [10] N. Yamanaka, H. Matsuyama, Y. Harabuchi, A. Kataura, Distribution of lymphoid cells in tonsillar compartments in relation to infection and age. A quantitative study using image analysis, Acta Otolaryngol. 112 (1) (1992) 128—137. [11] R.H. Schwartz, A cell culture model for T lymphocyte clonal anergy, Science 248 (1990) 1349—1356. [12] M.K. Jenkins, P.S. Taylor, S.D. Norton, K.B. Urdahl, CD28 delivers a co-stimulatory signal involved in antigen-specific 1L-2 production by human T cells, J. Immunol. 147 (1991) 2461—2466. [13] H. Reiser, M.J. Stadecker, Costimulatoy B7 molecules in the pathogenesis of infectious and autoimmune diseases, N. Engl. J. Med. 335 (1996) 1369—1377.
www.elsevier.com/locate/ijporl
Tonsillitis index: An objective tool for quantifying the indications for tonsillectomy for recurrent acute tonsillitis Keiji Fujihara a,*, Hironobu Goto a, Masanobu Hiraoka a, Masaki Hayashi a, Muneki Hotomi a, Shinji Tamura a, Kiyonori Kuki a, Noboru Yamanaka a, Peter J. Koltai b a
Department of Otolaryngology–—Head and Neck Surgery, Wakayama Medical University, 811-1 Kimiidera, Wakayama City 641-8509, Japan b Division of Pediatric Otolaryngology, Stanford University School of Medicine, Stanford, CA, USA Received 8 December 2004; accepted 18 April 2005
KEYWORDS Tonsillitis index; Tonsillectomy; Recurrent acute tonsillitis; Natural history of tonsillitis; The co-stimulatory signals; CD80; CD86
Summary This report is a preliminary exploration of the concept of a ‘‘Tonsillectomy Index’’ (TI) as an objective tool for quantifying the indications for tonsillectomy for recurrent acute tonsillitis (AT). The TI is derived by multiplying the number of episodes of AT by the number of years during which the episodes of AT occurred. Our objective in this study was to investigate whether there is a relationship between the natural history of AT, the immunological functions of tonsils and our proposed TI. For the natural history of AT, we medically followed 11 children with a history of AT for 5 years. When TI was equal to or greater than 8 (TI 3 8), the children suffered a significantly greater number of episodes of AT. For the immunological portion of our study, we enrolled 36 children and 46 adults undergoing tonsillectomy for either AT (study group) or tonsillar hypertrophy (control group, CG). We analyzed the costimulatory signals, CD80 and CD86 on tonsillar B-lymphocytes. The expression rates of CD80 and CD86 in the AT group with TI 3 8 were significantly decreased compared to those with TI was less than 8 (TI < 8), as well as with those in control (tonsillar hypertrophy) group. Our preliminary findings suggest that when the TI 3 8, the tonsils have deteriorated immunologically and spontaneous resolution of recurrent AT is less likely to occur, hence tonsillectomy is appropriate. TI may be a useful tool for surgical decision making. # 2005 Elsevier Ireland Ltd. All rights reserved.
* Corresponding author. Tel.: +81 73 441 0651; fax: +81 73 446 3846. E-mail address: [email protected] (K. Fujihara). 0165-5876/$ — see front matter # 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2005.04.007
1516
K. Fujihara et al.
1. Introduction Tonsillectomy has been shown to be effective in the management of recurrent acute tonsillitis (AT) [1,2], nevertheless, the indications of tonsillectomy for recurrent AT are controversial, unsettled and there is no consensus as to which children would benefit from such surgery. Tonsillectomy has gone through periods of enthusiasm, as well as uncertainty, as to its overall benefit in children. There is a natural decline in the frequency of upper respiratory infections as children age and many children will ‘outgrow’ their recurrent throat infections [3,4]. The development of antibiotic therapy has greatly influenced our ability to treat recurrent throat infections and prevent the serious sequelae that can arise from them. Some published reports have suggested that the removal of the tonsils was of little, if any, benefit in children [5]. In the United States, the American Academy of Otolaryngology–—Head and Neck Surgery suggests that children with three or more infections of the tonsils and/or adenoids per year, despite adequate medical therapy, are candidates for tonsillectomy [6]. Other national health organizations have published guidelines for tonsillectomy based on the frequency of recurrent infections [7,8]. In Japan, the consensus indication for tonsillectomy is four episodes of AT in each of preceding 2 years. While there is no world wide agreement about a specific number of bouts of AT that warrants tonsillectomy, most clinicians agree that some range of the number of infections over a period of time would be an indication for surgical intervention. Based on this insight, we explore the preliminary concept of ‘‘Ton-
sillitis Index’’ (TI), defined as the number of episodes of AT multiplied by the number of years during which the episodes of AT occurred. Our goal in suggesting this concept is an attempt to establish an objective tool for quantifying the indications for tonsillectomy in AT. The tonsil is an immunological organ and it seems reasonable to consider how recurrent infection affects its immuno-physiology when considering the indication for tonsillectomy. Tonsillar lymphocytes from recurrently infected tonsil have a significantly lower uptake of [3H] thymidine than those from tonsillar hypertrophy [9]. Patients with frequent episodes of tonsillitis have a significant increase of IgD-positive cells and IgG-positive cells in interfollicular and subepithelial compartments and a decrease of CD4-positive T-cells in the germinal centers and subepithelial areas [10]. Antigen specific activation of Tcells in host defense system is important and T cell activation requires two signaling events. The first, an Ag-specific signal delivered by the MHC/Ag complex interacting with TCR. The complete T cell activation depends on a second signal, defined as co-stimulation, by a molecule known as CD28 [11]. The ligands for CD28 are CD80 and CD86, which appear on B cells and antigen presenting cells. When the CD28—CD80 and CD86 signaling pathway are blocked, it results in inhibition of T cell proliferation, as well as reduction of a subset of T cell clones, inducing cellular Ag-specific unresponsiveness, termed anergy [12]. It is our hypothesis that these tonsillar immunological abnormalities can be correlated with the frequency of recurrent infection, the natural history of AT and our proposed TI.
Table 1 The patients enrolled in the study of spontaneous resolution of recurrent acute tonsillitis (AT) No.
Sex
TI
1 2 3 4 5 6 7 8 9 10 11
M M F F F M F F F M M
28 21 20 17.5 14 12 10 7.5 7 5 2.5 13.1 7.8
Mean S.D.
First episode
The most frequent episodes
The latest episodes
Follow up period
Age#
Age *
Frequncy$
Age
Frequncy
1 0 2 1.5 4 1 1 2.5 1 2 1
5 3 7 4.5 6 4 3 4.5 2.5 3 1
7 7 4 7 7 4 5 5 7 5 5
5 5 7 7 7 5 8 7 4 7 4
6 7 4 3 5 3 2 2 5 3 1
4 5 5 5.5 3 4 7 4.5 3 5 3
1.5 1.1
4.0 1.7
5.7 1.3
6.0 1.4
3.7 1.8
4.5 1.2
TI: Tonsillitis index; TI is derived by multiplying frequency$ at the most frequent episode by the difference between age# at the first episode and age* at the most frequent episodes.
Tonsillitis index for tonsillectomy
2. Materials and methods 2.1. Part 1 — spontaneous resolution of recurrent acute tonsillitis We retrospectively reviewed 11 children (4—8 years old, 5 males and 6 females) in a uncontrolled fashion, with a history of recurrent AT who had been managed medically for their infections for no less than 3 years at the Wakayama Medical University Hospital and referred clinics in Wakayama, Japan (Table 1). Our criteria for an episode of ATwas acute pharyngeal pain and fever equal to or greater than 38 8C. Despite being treated nonsurgically, the only patients included in the review were those who met the American Academy of Otolaryngology clinical practice guidelines for consideration for tonsillectomy [1]. At the end of the minimum of a 3-year follow up, their care givers were asked to respond to a questionnaire, which included age at the first episode, frequency of tonsillitis and age at the most frequent episodes and at the most recent year of follow up (Table 1). We then statistically analyzed by simple regression, the correlation of TI at the most frequent episodes of AT and the number of episodes at the latest year of follow up.
2.2. Part 2 — immunological assay of tonsillar function 2.2.1. Patient selection We studied the immunological function of the tonsillar specimens of 36 children (ages 3—15 years old, 20 males and 16 females) and 46 adults (ages 19—45, 26 males and 20 females) obtained during tonsillectomy for hypertrophic tonsil or recurrent AT at Wakayama Medical University Hospital or affiliated hospitals. Informed consent was obtained from patients or their parents before the enrollment in this study. The specimens from the patients with a diagnosis of tonsillar hypertrophy and without a history of AT were designated as the control group (CG). Specimens from the patients with recurrent AT were designated as tonsillitis group (TG). We then divided the TG into those that had TI equal to or more than 8 (TI 3 8) and a TI less than 8 (TI < 8). Our decision to use the TI = 8 as the dividing point for the patients in the TG was based on the initial part 1 of this study which suggested that patients who had less than 8 episodes of AT were more likely to have spontaneous resolution of their AT than those who had more than 8 episodes. 2.2.2. B cell purification Tonsillar mononuclear cells were purified by the standard Ficoll—Hypaque gradient method. The
1517
lymphocytes were immunostained with monoclonal antibodies as described below. 2.2.3. Antibodies The monoclonal antibodies used for phenotypic and functional studies were as follows; FITC conjugated anti-CD80 (PharMingen, Franklin Lakes, NJ), FITC conjugated anti-CD86 (PharMingen, Franklin Lakes, NJ) and PE conjugated anti-CD19 (PharMingen, Franklin Lakes, NJ). 2.2.4. Flow cytometric analysis Antibody binding was analyzed on a FACS caliber (Becton Dickinson, Franklin Lakes, NJ). The FACS caliber was used for producing dot plot. We analyzed the data of CD80 and CD86 in each of three groups (the CG, the TG with TI < 8 and the TG with TI 3 8) with ANOVA post-hoc test. The correlations of CD80 and CD86 were also analyzed in the three groups by simple regression assay. 2.2.5. Statistics ANOVA post-hoc test and simple regression analysis were performed using StatView Version 5.0 (SAS Institute, Cary, NC, USA). P-values less than 0.05 were designated as significant.
3. Results 3.1. Part 1 — spontaneous resolution of recurrent acute tonsillitis Among the children who were medically managed for AT, the average follow up time was 4.5 1.2 years. The first episode of AT occurred at 1.5 1.1 years of age. The most frequent bouts of AT, 5.7 1.3 episodes a year, occurred at 4.0 1.7 years of age. The ages at the last year of follow up were 6.0 1.4 years of age. There were 3.7 1.8 episodes of AT during the last year of follow up. The TI was derived by
Fig. 1 The correlation between TI and the number of sore throat with fever at the latest year.
1518
K. Fujihara et al.
multiplying the number of episodes of AT by the number of years between age at the first episode and age at the most frequent episodes. TI varied from 2.5 to 28 (Table 1). The frequency of AT was not correlated with age. TI was correlated with the frequency of AT during the last year of follow up. Regression analysis yielded a simple regression line (y = 1.52 + 0.168x, R2 = 0.503, P < 0.05) indicating that when the TI was equal to or more than 8 (TI 3 8), the children continued to suffer from AT, 3 or more times a year up to 4.5 years after the onset of infections (Fig. 1).
3.2. Immunological functions of the tonsil 3.2.1. Flow cytometric analysis Tonsillar lymphocytes were harvested from the surgical specimens in the immediate perioperartive period and flow cytometric analysis was performed in the same day. In the pediatric control group, the CD80 positive rate of tonsillar B-lymphocytes was 11.26 1.21% (mean S.D.). In the pediatric group with TI < 8, the positive rate of CD80 was significantly lower (7.38 1.53%) than that of the control group. In the pediatric group with TI 3 8, the positive rate of CD80 was significantly lower (6.1 0.68%) than that of both the control and the TI < 8 groups. In the pediatric control group, the CD86 positive rate of tonsillar B-lymphocytes was 11.45 0.77%. In the pediatric group with TI < 8, the positive rate of CD86 was significantly lower (6.06 1.50%) than that of the control group. In the pediatric group with TI 3 8, the positive rate of CD86 was significantly lower (3.67 1.22%) than that of both the control and the TI < 8 groups (Fig. 2a). In adult control group, the CD80 positive rate of tonsillar B-lymphocytes was 9.66 0.7%. In the adult group with TI < 8, the positive rate of CD80 was significantly lower (4.04 0.97%) than that of the control group. In the adult group with TI 3 8, the positive rate of CD80 was significantly lower (3.04 1.17%) than that of both the control and the TI < 8 groups. In adult control group, CD86 positive rate of tonsillar B-lymphocytes was 8.98 1.23%. In the adult group with TI < 8, the positive rate of CD86 was significantly lower (3.95 1.24%) than that of the control group. In the adult group with TI 3 8, the positive rate of CD86 was significantly lower (3.01 0.92%) than that of both the control and the TI < 8 group (Fig. 2b). 3.2.2. Simple regression assay In both the pediatric and adult control groups, the expression rates of CD80 and CD86 in tonsillar B-
Fig. 2 (a) The positive rates of CD80 and CD86 in children. TI < 8: TI was less than 8; TI 3 8: TI was equal to or greater than 8. **P < 0.01, *P < 0.05. (b) The positive rates of CD80 and CD86 in adults. **P < 0.01, *P < 0.05.
Fig. 3 (a) The correlation of CD80 and CD86 in tonsillar B-lymphocytes in the control, TI < 8 and TI 3 8 in children. **P < 0.01. (b) The correlation of CD80 and CD86 in tonsillar B-lymphocytes in the control, TI < 8 and TI 3 8 in adults. *P < 0.05.
Tonsillitis index for tonsillectomy
lymphocytes were almost identical with significant correlations between CD80 and CD86 in both children (Fig. 3a) and adults (Fig. 3b). However, in both the pediatric and adult groups with TI < 8 and with TI 3 8, the expression rates of CD80 and CD86 were decreased and showed less correlation between them.
4. Discussion The palatine tonsils are strategically located at the entrance of both the respiratory and gastrointestinal tracts. They are primarily immune organs for protection against external microorganisms and as such, a source of health. However, because of their specific cryptic architecture, they can become recurrently infected and thus become a source of disease. When the balance between immunologic protection and recurrent infection tilts persistently toward illness, tonsillectomy should be considered. Defining the fulcrum of that balance with precision has been remarkably difficult. In the tonsils, functional interactions between B and T lymphocytes are known to depend on the expression of co-stimulatory molecules (CD80 and CD86) and their counter-receptors (CD28 and CTLA4). CD86 induces naive T cells to become IL-4 producers, thereby directing the immune response more towards Th2, whereas CD80 is a more neutral differentiative signal for Th1 and Th2 [12]. The derangement of the CD80 and CD86—CD28 co-stimulatory pathway may change the clinical form of the disease by depressing cell-mediated immunity and inducing T cell anergy [13]. In this study, CD80 and CD86 were expressed at almost same rate and were significantly correlated in the healthy control tonsils. However, in the groups with AT, CD80 and CD86 were significantly decreased and did not correlate as they did in the controls. Furthermore, the expression of both CD80 and CD86 was more decreased and imbalanced in the groups with more persistent and severe disease. To the best of our knowledge this is the first data which demonstrates a clinical correlation between the indication for tonsillectomy and an immunological derangement in the function of tonsils. Studies performed at the University of Pittsburgh have sought to answer many of the questions regarding the efficacy of tonsillectomy in children afflicted with recurrent pharyngotonsillitis [1,2]. In order to be entered into the first study [1] children were required to have at least three episodes of tonsillitis in each of the prior 3 years or 5 episodes in each of 2 years or seven episodes in 1 year. In the first and second years of follow up, those children who received tonsillectomy had statistically fewer epi-
1519
sodes of throat infection when compared to the children in the nonsurgical arm. The third year of follow up resulted in a trend towards improvement in the surgical arm. The nonsurgical group experienced a decline in the frequency of throat infections in each of the follow up years. These studies are widely accepted as the methodologically best investigations regarding the efficacy of tonsillectomy for children with AT, and their criteria for surgical intervention are considered both judicious and appropriate. Applying our proposed concept of TI to the Pittsburgh criteria would yield a TI of 7 for 7 infections in any 1 year, a TI of 10 for 5 infections 2 years in a row and a TI of 9 for 3 infections 3 years in a row, rendering an average TI of 8.67. This is remarkably consistent with our own findings, derived by very different methodology. Our results suggest that when the TI 3 8, there is a significantly greater likelihood that recurrent ATwill persist and not resolve spontaneously, perhaps as a consequence of the tonsil having deteriorated immunologically. This may explain why spontaneous resolution of recurrent AT is less likely to occur and also suggests that a TI 3 8 is a reasonable indication for tonsillectomy. Nevertheless, at this point our study’s shortcomings preclude a strong conclusion in this regard. The number of subjects is too small and there is a lack of prospective validated diagnostic criteria. With the limited sample size, it is not possible to derive reliable sensitivity and specificity information or confidence limits. Furthermore, in the immunological part of our study, using specimens from patients with tonsillar hypertrophy as our controls is methodologically problematic. This is also the case with bundling pediatric and adult patients, given the possibility that there may be important clinical differences in tonsillitis between children and grownups. The concept of a TI should be an accurate and useful tool for surgical decision making. Our preliminary data points in this direction.
Acknowledgements We would like to thank Dr. T. Jinnin, Dr. T. Kimura and Dr. T. Fujihara in Wakayama for follow up data of recurrent AT at their clinics.
References [1] J.L. Paradise, C.D. Bluestone, R.Z. Bachman, D.K. Colborn, B.S. Bernard, F.H. Taylor, et al. Efficacy of tonsillectomy for recurrent throat infections in severely affected children, N. Engl. J. Med. 310 (1984) 674—683.
1520
[2] J.L. Paradise, C.D. Bluestone, D.K. Colborn, B.S. Bernard, H.E. Rockette, M. Kurs-Lasky, Tonsillectomy and adenotonsillectomy for recurrent throat infection in moderately affected children, Pediatrics 110 (1) (2002) 7—15. [3] T.J. Woodford, A. Ahmed, D.J. Willatt, et al. Spontaneous resolution of tonsillitis in children on the waiting list for tonsillectomy, Clin. Otolaryngol. 25 (2000) 428—430. [4] M.P. Prim, J.I. de Diego, M. Larrauri, C. Diaz, N. Sastre, J. Gavilan, Spontaneous resolution of recurrent tonsillitis in pediatric patients on the surgical waiting list, Int. J. Pediatr. Otorhinolaryngol. 65 (1) (2002) 35—38. [5] J.L. Paradise, Tonsillectomy and adenoidectomy, in: C.D. Bluestone, S.E. Stool, M.A. Kenna (Eds.), Pediatric Otolaryngology, third ed., WB Saunders, Philadelphia, 1996, pp. 1054—1065. [6] American Academy of Otolaryngology–—Head and Neck Surgery, 2000 Clinical Indicators Compendium, vol. 9, issue no. 6, Alexandria, VA, 2000, p. 19. [7] Finnish Medical Society Duodecim: Sore Throat and Tonsillitis, EBM Guidelines, Duodecim Medical Publications Ltd., 2001, pp. 1—4.
K. Fujihara et al.
[8] Scottish Intercollegiate Guideline Network (SIGN): Management of Sore Throat and Indications for Tonsillectomy, A National Clinical Guideline, 1999, p. 23. [9] N. Yamanaka, K. Kobayashi, T. Himi, F. Shido, A. Kataura, Spontaneous DNA synthesis in tonsillar lymphocytes and its clinical implications, Acta Otolaryngol. 96 (1—2) (1983) 181—187. [10] N. Yamanaka, H. Matsuyama, Y. Harabuchi, A. Kataura, Distribution of lymphoid cells in tonsillar compartments in relation to infection and age. A quantitative study using image analysis, Acta Otolaryngol. 112 (1) (1992) 128—137. [11] R.H. Schwartz, A cell culture model for T lymphocyte clonal anergy, Science 248 (1990) 1349—1356. [12] M.K. Jenkins, P.S. Taylor, S.D. Norton, K.B. Urdahl, CD28 delivers a co-stimulatory signal involved in antigen-specific 1L-2 production by human T cells, J. Immunol. 147 (1991) 2461—2466. [13] H. Reiser, M.J. Stadecker, Costimulatoy B7 molecules in the pathogenesis of infectious and autoimmune diseases, N. Engl. J. Med. 335 (1996) 1369—1377.