To cite this article: Radulovic S, Wilson D, Calderon M, Durham S. Systematic reviews of sublingual immunotherapy (SLIT). Allergy 2011; 66: 740–752.
Allergic rhinitis is common worldwide, with significant morbidity and impact on quality of life. In patients who don’t respond adequately to anti-allergic drugs. Subcutaneous allergen immunotherapy is effective although requires specialist administration. Sublingual immunotherapy may represent an effective and safer alternative. This Cochrane systematic review is an update of one published in 2003. We searched Cochrane ENT Group Trials Register, Central, PubMed, EMBASE, CINAHL, Web of Science, Biosis Previews, Cambridge Scientific Abstarcts, mRCT and additional sources. We included randomised, double-blind, placebo- controlled trials of sublingual immunotherapy in adults and children. Two authors selected studies and assessed them for quality. Data were put into RevMan 5.0 for a statistical analysis. We used standardised mean difference (SMD), with a random effect model to combine data. Sixty studies were included, with 49 suitable for meta-analysis. We found significant reductions in symptoms (SMD −0.49; 95%CI (−0.64 to −0.34, P < 0.00001)) and medication requirements (SMD −0.32; 95%CI (−0.43 to −0.21, P < 0.00001)) compared with placebo. None of the trials reported severe systemic reactions, anaphylaxis or use of Adrenaline. This updated review reinforces the conclusion of the original 2003 Cochrane Review that sublingual immunotherapy is effective for allergic rhinitis and appears a safe route of administration.
As evidenced by the publication of this special edition of the Journal, allergen-specific immunotherapy has been in clinical use for a century. For patients with severe allergic rhinitis who do not respond to usual therapy, specific immunotherapy, given by injection, induces immunomodulation resulting in significant reductions in symptoms and medication requirements (1) that continue beyond the cessation of treatment (2)
Although highly effective, allergen injection immunotherapy has significant drawbacks – local adverse events such as injection site itch or swelling are fairly common and, although rare, systemic reactions can be severe or even life-threatening. For this reason, the use of injection IT has been restricted to specialist centres and requires significant resources for its delivery.
In an attempt to eliminate these difficulties, alternative routes for the delivery of immunotherapy, ideally with equivalent efficacy and a better safety profile, have been sought. Over the last three decades, much attention has been focused on the sublingual route. Initial studies utilizing this route were small but suggested a favourable efficacy and safety profile leading to the design of bigger trials that continued this trend but were individually insufficiently powered to give definitive answers.
The systematic reviews
In 2002, we undertook a systematic review of the literature in sublingual immunotherapy (SLIT) up to September 2002, using the methodology of the Cochrane Collaboration. Twenty-two randomized, placebo-controlled trials of adequate quality, involving 979 patients, were identified. The findings of this review and its meta-analysis were published in the Cochrane library (3) and subsequently in Allergy (4).
Despite significant heterogeneity between the trials, most likely due to widely differing scoring systems, the meta-analysis showed a significant reduction in both symptoms [standardized mean difference (SMD) −0.42, 95% confidence interval −0.69 to −0.15; P = 0.002] and medication requirements [SMD −0.43 (−0.63, −0.23); P = 0.00003] following immunotherapy. Subgroup analyses failed to identify any difference according to the allergen administered, and there was no significant reduction in symptoms and medication scores in those studies involving only children. Adverse events reported in these trials were minor and local, and no systemic reactions were reported.
Subsequent to this review many further studies, of increasing size and quality, were published. In anticipation of producing an even more comprehensive evaluation of the efficacy and safety of SLIT and of answering some of the outstanding questions from the original review it was decided to repeat the systematic review and meta-analysis, updating it to 2009. The updated review is published in full in the Cochrane library and an abridged version is reproduced below (5).
To evaluate the efficacy of SLIT compared with placebo in the following:
- 1 reducing symptoms and/or medication requirements during naturally occurring allergic rhinitis;
- 2 altering immunological markers in blood and immunological markers and allergen sensitivity in target organs (nose, eye, skin).
To evaluate the safety of SLIT.
Criteria for considering studies for this review
Types of studies Randomized, double-blind, placebo-controlled clinical trials.
Types of participants The studies included participants of any age (children and adults). All patients had a history of allergic rhinitis, with or without allergic conjunctivitis and with or without allergic asthma. In all studies, the allergen was clearly identified. Patients’ sensitivity was proven by positive skin-prick tests and/or high specific IgE to a particular allergen. The existence of other clinically relevant sensitivities was one of the exclusion criteria in the majority of studies.
Trials dealing with asthma only were excluded from this review.
Types of interventions Included studies were those investigating the efficacy and safety of SLIT. Trials were analysed regardless of treatment dose, duration or whether the allergen was swallowed or spat out.
Types of outcome measuresPrimary outcomes. Symptom scores, however, recorded (either daily or weekly, via symptom score diaries, visual analogue scales, number of well days or overall assessment).
Medication scores referring to the use of relevant anti-allergic medications, however, recorded and scored.
Secondary outcomes. Measurement of serum IgE and IgG (total and specific), assessment of allergen sensitivity (eye, nose or skin), quality of life and adverse event reports.
Search methods for identification of studies
Systematic searches for randomized controlled trials were conducted. There were no language, publication year or publication status restrictions. The date of the last search was 14 August 2009. Full details of the search strategy can be found in the Cochrane Library version of the review (5).
Data collection and analysis
Selection of studies Two authors (SR and MC) independently screened the search results and selected studies that appeared to meet the review inclusion, and all such studies were obtained in full text for further assessment. Any disagreements about which studies to include in the review were resolved by further discussion with the other two authors (SRD and DW).
Data extraction and management Data were extracted from the included studies onto a standard form, covering study type and methodology, number and description of participants, details of type, dosage, schedule, duration of SLIT used, as well as the results, types, timing and method of outcome measures. One author (SR) extracted all data, and values were checked by MC and SRD. If published manuscripts did not report data in format suitable for meta-analysis, or if data were insufficient, further information was sought directly from the authors.
As all the review authors were previously familiar with the content of most of the studies, we did not remove the study author names before assessment and data extraction.
Assessment of risk of bias in included studies For the original 2003 review, quality assessment of the trials was performed. Concealment of allocation and blinding of study participants and investigators was assessed according to the guidelines of The Cochrane Collaboration (6). For the 2010 update, all the originally included studies were scored using the Jadad scale (7) and then reassessed for risk of bias using the Cochrane Collaboration ‘Risk of bias’ tool as guided by The Cochrane Handbook for Systematic Reviews of Intervention (8).
Data synthesis Apart from adverse events, all the outcome data analysed were continuous. The most common way of recording data was through daily diary cards, recording and scoring symptoms (nasal, eye or less frequently chest) and medication use (antihistamine tablets, nasal sprays, eye drops). These data were subsequently totalled and averaged.
A wide range of different scoring systems and scales were employed by trial authors for both, primary and secondary outcomes. This creates problems of heterogeneity but is unavoidable. Data were entered into RevMan 5 for statistical analysis (9) using the SMD (the difference in means between two treatment groups, immunotherapy and placebo, in units of pooled standard deviation).
Random-effect models were used for statistical analysis of the overall efficacy of SLIT, and the results are presented as SMDs with 95% confidence intervals (CI).
Heterogeneity between studies was analysed using the chi-squared test, with a P-value of <0.1, indicating significant heterogeneity between studies. In addition, to further assess heterogeneity, we used the I2, statistic variable that describes the percentage of total variation across trials that is because of heterogeneity rather than sampling error. Threshold values recommended in the Cochrane Handbook for Systematic Reviews of Interventions (8) were used (0% to 40%: might not be important; 30% to 60%: moderate heterogeneity; 50% to 90%: substantial heterogeneity; 75% to 100%: considerable heterogeneity).
Subgroup analysis was carried out according to the review protocol as follows:
- 1 seasonal vs perennial allergens;
- 2 children vs adults;
- 3 dosage of major allergen (<5 μg of major allergen protein vs 5–20 μg vs >20 μg;
- 4 duration of immunotherapy (<6 months vs 6–12 months vs 12 months, to cover preseasonal, perennial and prolonged treatment);
- 5 sublingual spit vs sublingual swallow
- 6 sublingual drops vs tablets.
Adverse events were analysed as discontinuous data, and only a descriptive analysis is presented.
Results of the search
Description of studies The updated searches in 2009 identified 628 papers of potential interest. A total of 498 papers were discarded after reading the abstracts (review articles or descriptive studies, papers investigating other routes of immunotherapy or not investigating allergic rhinitis). Hundred and thirty papers were therefore evaluated in detail. Following this evaluation, we discarded another 68 papers (Fig. 1).
We thus identified 62 papers as potentially appropriate for the review and meta-analysis. We identified two studies as ongoing (Ingels 2002; O’Hehir 2005).
A total of 60 studies (10–71) are now included in the 2010 update of this review, of which 49 studies are included in the updated meta-analyses (Fig. 1). Eleven studies that did not contain efficacy data eligible for meta-analysis contained useful adverse event data.
|Subgroup||Number of studies||n (active)||n (placebo)||SMD (95% CI)||P|
|All seasonal||39||2081||2003||−0.34 (−0.44; −0.25)||<0.00001|
|All perennial||10||252||253||−0.93 (−1.69; −0.17)||<0.00001|
|Grass||23||1549||1464||−0.35 (−0.45; −0.24)||<0.00001|
|Parietaria||5||74||77||−0.36 (−0.69; −0.04)||0.03|
|Ragweed||2||85||90||−0.44 (−0.74; −0.14)||0.004|
|Tree||9||197||183||−0.42 (−0.77; −0.06)||0.02|
|House dust Mite||9||232||232||−0.97 (−1.8; −0.3)||0.02|
|Adults||34||1631||1566||−0.44 (−0.56; −0.31)||0.0001|
|Children||15||702||690||−0.52 (−0.94; −0.1)||0.02|
|<6 months||17||890||882||−0.54 (−0.86; −0.21)||0.001|
|6–12 months||16||867||869||−0.31 (−0.46; −0.16)||<0.0001|
|>12 months||16||580||509||−0.63 (−0.92; −0.34)||<0.0001|
|Dose of major allergen|
|<5 μg||8||141||134||−0.32 (−0.65; 0.05)||0.09|
|5–20 μg||12||1006||966||−0.34 (−0.43; −0.24)||<0.00001|
|>20 μg||12||541||500||−0.33 (−0.49; −0.17)||<0.00001|
|Drops||35||1270||1194||−0.35 (−0.42; −0.28)||<0.00001|
|Tablets||11||945||936||−0.48 (−0.58; −0.38)||<0.00001|
Most trials were performed with grass pollen (23 studies). Other allergens used were Parietaria (five trials), ragweed (two trials), trees (nine trials: two olive, three cypress, two birch pollen, two mixed trees), house dust mite (eight trials) and cat (one trial). One of the trials investigated the efficacy of grass and birch pollen immunotherapy.
Thirty-four studies were performed in adults and 15 investigated efficacy and safety in children. Treatment lasted for <6 months in 17 studies; 6–12 months in 16 studies and longer than 12 months in 16 studies.
Of the 49 studies, 32 reported the major allergen dose in a manner suitable for meta-analysis. The rest of the trials either did not provide the sufficient data or reported the cumulative dose (weekly, monthly or a total cumulative dose over the complete treatment). Eight trials used daily doses of <5 μg, in 12 studies the dose was between 5 and 20 μg per day, and 12 papers reported a daily dose of more than 20 μg.
Nine trials reported data on skin sensitivity, but only six (skin prick test after treatment) could be included in the meta-analysis. Seven trials reported data eligible for meta-analysis of nasal reactivity. Data on conjunctival reactivity were not sufficient for meta-analysis.
A total of 24 studies are excluded from this review – these are detailed under ‘characteristics of excluded studies’ in the full review.
Risk of bias in included studies All included studies were double-blind, placebo-controlled trials of parallel-group design. Concealment of treatment allocation was considered adequate in all studies, based on statements made by the original authors.
Blinding of study subjects and investigators was almost universally maintained by the use of identical placebo preparations. It should, however, be noted that most investigators reported high levels of minor oral side-effects (tingling, itching and swelling beneath the tongue) in actively treated subjects, which could influence blinding.
Full risk of bias assessments can be found in the characteristics of included studies’ table and Figs 2 and 3 of the full review.
Effects of interventions
Symptom scores A total of 2333 active treatment (SLIT) and 2256 placebo patients were included. The combined SMD following SLIT was −0.49 (95% CI −0.64 to −0.34) favouring active treatment (P < 0.00001). There was significant heterogeneity between the studies (χ2 = 256.76, P < 0.00001, I2 = 81%) (Fig. 2)
Medication scores Thirty-eight trials reported medication score results, with a total of 1737 patients in the SLIT group and 1642 in the placebo group. The combined SMD was −0.32 (95% CI −0.43 to −0.21, P < 0.00001). Significant heterogeneity was indicated (χ2 = 73.32, P = 0.0003, I2 = 50%) (Fig. 3).
Data from subgroup analyses for symptom and medication scores are summarized in Tables 1 and 2.
|Subgroup||Number of studies||n (active)||n (placebo)||SMD (95% CI)||P|
|All seasonal||32||1557||1457||−0.3 (−0.41; −0.19)||<0.00001|
|All perennial||6||180||185||−0.43 (−0.89; 0.02)||0.06|
|Grass||17||1201||1107||−0.23 (−0.37; −0.1)||0.0008|
|Parietaria||5||74||77||−0.62 (−1; −0.24)||0.001|
|Tree||9||197||183||−0.38 (−0.62; −0.13)||0.002|
|House dust mite||5||95||94||−0.52 (−1.09; −0.03)||0.07|
|Adults||26||1168||1067||−0.4 (−0.53; −0.26)||<0.00001|
|Children||12||569||575||−0.16 (−0.32; 0)||0.06|
|<6 months||15||694||672||−0.32 (−0.43; 0.18)||<0.00001|
|6–12 months||13||705||693||−0.31 (−0.5; −0.12)||0.002|
|>12 months||10||338||277||−0.34 (−0.62; −0.04)||0.03|
|Dose of major allergen|
|<5 μg||6||110||109||−0.59 (−0.94; 0.24)||0.0008|
|5–20 μg||11||992||948||−0.21 (−0.35; −0.07)||0.003|
|>20 μg||10||360||301||−0.22 (−0.43; 0)||0.05|
|Drops||27||865||788||−0.01 (−0.05; 0.04)||0.74|
|Tablets||9||799||779||−0.33 (−0.46; −0.2)||<0.00001|
Specific serum antibodies Fourteen studies reported increases in serum-specific IgE levels in a manner suitable for the meta-analysis, with 675 participants in the SLIT and 659 in the placebo group. The combined SMD was 0.27 (95% CI −0.01 to 0.55, P = 0.05). Significant heterogeneity was indicated between studies (χ2 = 68.63, P < 0.00001, I2 = 81%).
Total serum-specific IgG was measured in three studies, with 286 participants in the SLIT and 304 in the placebo group. The combined SMD was 0.95 (95% CI 0.78 to 1.12, P < 0.00001). There was no significant heterogeneity between studies (χ2 = 1.01, P = 0.60, I2 = 0%).
Serum-specific IgG4 was measured in 13 trials, with a total number of 588 in the SLIT and 599 in the placebo group. The total SMD was 0.46 (95% CI 0.29 to 0.63, P < 0.00001). Highly significant heterogeneity was indicated (χ2 = 174.24, P < 0.00001, I2 = 93%).
Adverse events All sixty studies in the review were analysed for safety results.
Not all studies included data on adverse events. The lack of a standardized grading system for reporting of adverse events associated with SLIT required that events be presented as descriptive data only. Of the 60 studies evaluated for safety, twenty-five studies reported different local events and systemic reactions were reported by 18 studies (Table 3). Local adverse events were more common in immunotherapy-treated patients, whereas the majority of reported adverse events were mild to moderate and did not require any treatment. Fifteen studies reported adverse events leading to withdrawal from studies. In the SLIT group, 41 of 824 (5%) withdrew because of adverse events compared to 12 out of 861 (1%) of placebo-treated participants. Systemic adverse events were common in both placebo- and SLIT-treated participants and of mild-to-moderate intensity. There were no apparent differences between sublingual and placebo-treated participants (Table 3), with the exception of gastrointestinal symptoms that were more common in the sublingual group (88 of 630 participants, 14%) compared to placebo-treated participants (10 of 561, 1.7%). It is debatable whether gastrointestinal adverse events following a therapy administered by the sublingual route should be reported as systemic rather than local adverse events. There were no reports of serious adverse systemic events nor anaphylaxis and no recorded use of adrenaline in any of the studies.
|Type of reaction||Studies||Sublingual immunotherpay||Placebo|
|Patients||Total events (per patient)||Patients||Total events (per patient)|
|Labial oedema||11||604||55 (0.09)||536||7 (0.01)|
|Buccal pruritis||21||1126||1798 (1.6)||1075||492 (0.46)|
|Buccolingual oedema||8||648||143 (0.22)||606||2 (0.003)|
|Throat irritation||10||770||243 (0.3)||747||29 (0.04)|
|Oral (nonspecified)||3||68||143 (2.1)||71||24 (0.34)|
|Local nonspecified||3||119||7 (0.06)||116||3 (0.03)|
|Urticaria||8||204||7 (0.03)||199||9 (0.04)|
|Pruritis/rash||10||363||13 (0.04)||222||9 (0.04)|
|Conjunctivitis||8||262||774 (2.95)||238||786 (3.3)|
|Rhinitis||16||965||1403 (1.45)||912||1034 (1.13)|
|Rhinoconjunctivitis||6||184||60 (0.33)||176||58 (0.33)|
|Asthma/wheeze||15||488||51 (0.1)||450||42 (0.09)|
|Cough||8||337||313 (0.93)||304||211 (0.69)|
|Gastrointestinal||20||630||88 (0.14)||561||10 (0.02)|
|Headache||6||535||70 (0.2)||548||68 (0.12)|
|Systemic nonspecified||5||330||4 (0.01)||36||0|
Quality of life Quality of life was reported by three studies, but assessment in those studies differed greatly, and we considered that these data could not be included in our analysis.
This systematic review of sublingually administered allergen immunotherapy (SLIT) represents an update of a review first published in The Cochrane Library in 2003 (3). The original review included data from 22 randomized controlled trials (979 patients) and demonstrated the efficacy of this form of treatment based on meta-analysis of symptom severity scores (SMD −0.42; 95% confidence interval (CI) −0.69 to −0.15). Ongoing research in this area has been considerable, and this review has now been updated to include studies published since 2003. The number of studies included has almost trebled to 60 (with 49 being suitable for pooling in meta-analyses), and the number of patients in meta-analysis has increased over four:fold, reflecting a trend towards larger, better-designed and more powerful trials.
The overall results of the meta-analysis differ little from those seen in 2003, with the overall effect for symptom scores SMD (−0.49; 95% CI −0.64 to −0.34) being of a similar magnitude, with tighter CI reflecting the greatly increased number of study subjects. The same is true for the analysis of medication scores, with SMD −0.32 (95% CI −0.43 to −0.21). These data continue to support the clinical efficacy of SLIT for allergic rhinitis.
In contrast to the original review, the greater number of studies has allowed more meaningful analyses of some of the predetermined subgroups. In particular, there are now 15 studies looking exclusively at children, some of which are large studies in their own right (16, 69). The treatment effect within this subgroup of trials appears to be similar to that seen in adults, especially when considering symptom scores. SLIT represents a particularly attractive alternative to injection immunotherapy in this patient group and our findings are entirely consistent with those reported elsewhere (72).
The protocol for the original review reflected the then classification of aero-allergens into seasonal and perennial. The Allergic Rhinitis and its Impact on Asthma classification (73) now uses the terms intermittent and persistent, but for this review this change makes no difference. In this meta-analysis, there does appear to be a greater effect with perennial allergens (predominantly house dust mite) when compared to seasonal allergens although this is based on fewer studies. More studies of perennial rhinitis are needed to confirm or exclude this possibility.
It is not possible to differentiate between different doses on the basis of this meta-analysis. The difficulty in determining dose in terms of micrograms of major allergen and standardizing this information across a range of studies utilizing allergen extracts from different sources was acknowledged in the original review and remains problematic.
Although the difference is small, this review has shown a trend in symptom score reduction in trials that lasted for longer than 12 months, when compared with shorter treatment periods. Indeed, SLIT is now given for longer time periods (over 12 months in 32% of included studies compared with 19% in 2003). More recent studies have shown that treatment for longer than 12 months provides consistent clinical improvement in symptom and medication scores (74, 75). These data are encouraging and should be taken into consideration in future recommendations or guidelines for the use of SLIT in allergic rhinitis.
Looking at the total effect and SMD (95% CI) of SLIT for individual allergens, house dust mite appears to be more effective than treatment with other types of allergen and even more effective than treatment with grass pollen. However, the majority of these trials are small, with five (out of nine) trials involving fewer than 20 participants. Heterogeneity in this group is amongst the highest in all the meta-analyses; when comparing P-values for the overall effect, the level of significance appears to be lower than for the majority of other allergens. We therefore conclude that this finding should be interpreted with caution.
This review has shown that SLIT provokes significant changes in terms of allergen-specific IgG and IgG4 antibodies, which coincide with a clinical response in terms of symptom and medication scores. These findings are in complete concordance with the findings of the previous SLIT review (3), as well as injection immunotherapy (76). Unfortunately, changes detected in IgG and IgG4 values were not supported by changes in allergen sensitivity, and meta-analysis of data for skin and nasal reactivity after treatment showed no difference between the immunotherapy and placebo group (5).
The exact role of IgG and IgG4 antibodies is still not completely clear. Although they are likely to have a ‘protective’ role, there is still an ongoing debate as to whether this increase in (particularly) IgG4 is just a consequence of exposure to a high dose of allergen or the real immunomodulatory effect of SLIT. This review could not draw a definitive conclusion and only further mechanistic studies can enable us to answer this question.
This review explored the possible differences between different sublingual preparations (i.e. sublingual drops vs tablets). Although tablets proved to be more effective in terms of medication scores and had similar efficacy in terms of symptoms, overlapping CI and the substantial heterogeneity between studies did not allow us to draw any firm conclusions. It seems both preparations are similarly effective.
An increasing number of studies report quality of life as a primary or secondary outcome measure as is the case in clinical trials in general. Nevertheless, there were big differences in quality-of-life scoring systems such that we were unable to analyse these data by meta-analysis.
It was acknowledged in the original review that many of the studies included were small, early publications that did not conform to the CONSORT (1996) guidelines for the publication of randomized controlled trials (77, 78). The methodological quality of the studies included on this occasion has been scrutinized more closely using the new Cochrane criteria, including assessment of randomization and allocation concealment. The possible confounder of publication bias that exists in all meta-analysis (72, 79) was acknowledged in the original review and has again been addressed on this occasion through extensive consultation with those active in the field of SLIT research. We are confident that no data have been excluded purely on the basis of negative outcome.
Heterogeneity between studies was acknowledged as a significant problem in the last review and is a known problem in systematic reviews. This results largely from methodological and clinical heterogeneity (i.e. differences in scoring systems, sample sizes, type and dose of allergen, age groups, etc.) used across studies. Selection of the studies for this review was defined in our protocol, and studies that satisfied our criteria were chosen. The method used by The Cochrane Collaboration for assessing heterogeneity has however changed and can now be expressed as an I2 statistic. It remains the case that studies in this field are heterogeneous, and data are expressed in a wide variety of different ways. However, certain subgroup analyses (e.g. seasonal allergens, individual allergens, subgroup analysis for major allergen content) have shown the significant reduction in heterogeneity even though these groups were prespecified by a protocol. This means that there will always be a degree of interpretation required when amalgamating studies in meta-analysis, and further subgroup analysis could be performed to address better the problem of heterogeneity. Despite this, there is remarkable consistency in the outcomes of related systematic reviews (3, 80, 81).
Although considered as a secondary outcome, we felt analysis of adverse events to be crucial as a low incidence confers advantage on SLIT as an alternative to injection immunotherapy. Adverse events data were, by their nature, noncontinuous. Authors mainly reported data as total number of events for a number of patients, rather than number of certain events per patient. Some papers reported their most common events as a percentage of total events or as a percentage of patients who experienced particular events. These were therefore not suitable for meta-analysis, and we were able to perform only descriptive analysis. A further problem is that, unlike for the subcutaneous route, there is currently no internationally standardized methodology for reporting local adverse events associated with SLIT. With these reservations, local reactions are again shown to be common and reported much more frequently in SLIT recipients than in those receiving placebo. These are clearly unavoidable but are usually seen as an inconvenience and cause little distress and have no lasting effect; though, rarely these adverse events were distressing enough to warrant withdrawal of treatment.
Systemic reactions were again largely confined to the upper respiratory tract and associated organs (rhinitis, conjunctivitis or rhinoconjunctivitis) and were more frequent in the SLIT than in the placebo groups. Asthma or wheeze was no more likely in SLIT recipients than in placebo recipients. Gastrointestinal effects (nonspecified) were uncommon but more apparent in SLIT recipients and were largely confined to paediatric trials. None were considered serious. There were no reports in clinical trials of severe systemic reactions or anaphylaxis, and none of the systemic reactions needed the use of adrenaline. No fatalities were reported.
This review evaluates a large number of double-blind, placebo-controlled studies that, in total, give us a large number of doses. No incidences of life-threatening reactions were reported in the studies analysed. We conclude that SLIT remains a safe treatment with an extremely low incidence of significant side-effects. We have been unable to correlate adverse events with allergen dose.
There are no reports of fatalities following SLIT. Six isolated cases of severe reactions have been reported independently of clinical trials and all involved deviation from current recommended practice according to international guidelines (82–86). Numbers are too few to allow identification of risk factors for severe systemic reactions; however, it can be noted that five out of six occurred in young women, five out of six had asthma, and two out of six had previously experienced severe reactions during subcutaneous immunotherapy. These data should be viewed in the context of the number of doses of SLIT that have been prescribed and administered worldwide.
Since the original systematic review in 2003, SLIT has become established as an effective and low-risk alternative to allergen injection immunotherapy, which carries a significant morbidity and a requirement for delivery within specialist centres capable of meeting CSM recommendations. SLIT is recommended to be initiated in secondary care and the first dose taken under medical supervision, whereas maintenance treatment is recommended to be self-administered in the patient’s home.
Only two studies in the original review compared injection immunotherapy with SLIT directly (87, 88). In the current review, only one study compared injection immunotherapy vs SLIT. Although comparison of those two treatment options was not the objective of this review, the search process enabled us to identify papers comparing the efficacy of these treatment options. We found very few such papers: there appears to be insufficient data available to draw any conclusions and more definitive head-to-head trials are needed.
A Cochrane review of allergen injection immunotherapy, which included 51 trials with 2871 participants (76), showed a SMD of −0.73 (95% CI −0.97 to −0.5) for symptom scores compared with placebo and a SMD of −0.57 (95% CI −0.82 to −0.33) for medication scores. Although the SMDs are numerically different, the CI overlap with those for SLIT, indicating no apparent difference between the two therapies on this basis. However, it is not correct to perform a direct statistical comparison between these two meta-analyses. These data raise the importance of future double-blind, double-dummy trials that directly compare these two routes of immunotherapy.
With similar reservations, the latest Cochrane review of subcutaneous immunotherapy for asthma (89) has shown levels of efficacy of the same order in terms of reduction of symptom scores and need for rescue medication (Table 4).
|Authors||Age group||N participants active/placebo|
N for symptom score (medication score)
|Symptom Scores SMD random (95% CI) P value heterogeneity (I2)||Medication Scores SMD random (95% CI) P value heterogeneity (I2)|
|SLIT||Radulovic et al. (5)||Adults and children||2333/2256 (1737/1642)||−0.49 (−0.64, −0.34)|
P < 0.00001
I2 = 81%
|−0.32 (−0.43, −0.21)|
P < 0.00001
I2 = 50%
|SCIT||Calderon et al. (76)||Adults||597/466 (549/414)||−0.73 (−0.97, −0.50)|
P < 0.00001
I2 = 63%
|−0.57 (−0.82, −0.33)|
P < 0.00001
I2 = 64%
|Abramson et al. (89)||Adults and children||727/557 (485/384)||−0.59 (−0.83, 0.35)|
P < 0.00001
I2 = 73%
|−0.53 (−0.80, −0.27)|
P = 0.000065
I2 = 67%
The original systematic review by Wilson et al. (3) changed the field of SLIT for allergic rhinitis – encouraging significant investment in the further evaluation of a treatment with proven efficacy. This latest review includes the subsequent highly powered clinical trials and includes more than four times the number of patients. The data establish SLIT as a viable alternative to allergen injection immunotherapy, with a significantly lower risk profile. There is support for the use of SLIT in children and additional support for its use in allergic rhinitis because of seasonal allergens and perennial disease because of house dust mite. The sublingual route appears safer and has a better side-effect profile compared to the subcutaneous route, although a standardized grading system for side-effects is needed to permit more accurate and consistent reporting in the future. A recent study of SLIT using grass allergen tablets for seasonal pollinosis demonstrated long-term benefit for at least one year following 3 years of treatment (75), implying the induction of both clinical and immunological tolerance. Further long-term studies in adults and children are needed; also adequately powered head-to-head trials of sublingual vs subcutaneous immunotherapy. Inclusion of standardized quality-of-life measures and pharmacoeconomic evaluations as major outcomes will further define the place of SLIT in the treatment of allergic rhinoconjunctivitis.
Conflicts of interest
SR Durham has received lecture and consultancy fees and research funding via Imperial College from ALK Abello, Horsholm Denmark, a manufacturer of allergy vaccines. Moises Calderon has received lecture fees from ALK Abello, Horsholm Denmark, a manufacturer of allergy vaccines.