Details on inclusion and exclusion criteria: Only subjects with known active infectious mononucleosis were excluded, not those reporting a past infection (exclusion criterion #4). Exclusion criterion #10 was reworded to history of “one or more” tick bites (instead of “tick bites”) within the last 60 days (to clarify that just one tick bite in the history was sufficient to exclude the subject). Subjects must be withdrawn if they develop Lyme Disease /Borreliosis. Furthermore, it was detailed, that patients received the necessary therapy in case they developed Lyme Disease / Borreliosis. The statistical aspects of the study were detailed to fully comply with the confirmatory character of this phase III study. A detailed description of the adaptive interim assessment was added. The sample size was re-calculated according to the three stage group sequential design (two interim analyses and one final analysis).

Definition of the primary endpoint was clarified. A new secondary efficacy endpoint was introduced based on seroconversion as measured by (i) IgM or (ii) IgM and IgG or (iii) IgM and IgG and appearance of EM in baseline-seronegative subjects. Efficacy analysis is only conducted on IgM and / or IgG baseline-seronegative subjects. Secondary efficacy analysis 1 will be performed for all treated subjects, the modified ITT and PP sets on the primary efficacy endpoint (rate of treatment failures at Day 57) and will include the same analyses as for the primary efficacy Analysis. Secondary efficacy analysis 2 will check for the reduction in the rate of seroconversion as demonstrated by (i) IgM seroconversion only, (ii) IgM and IgG seroconversion and (iii) IgM and IgG seroconversion and appearance of EM in the All Treated Subjects, ITT, modified ITT and PP set, respectively. The testing procedure will be the same as for the primary efficacy endpoint. Note that (i), (ii) and (iii) are secondary alternatives for the definition of treatment failures. The previous version did not allow entry for subjects with a skin reaction at the tick bite scored at 2 or higher than 2 . However, after having enrolled approximately 300 subjects in 2011, it was recognized, that this criterion prevented several subjects from participation, as their skin reaction to the tick bite scored ‘2’. Missing data conventions were updated. Changes in statistical section of the study protocol.

The protocol stated, that the analysis of treatment failures based on seroconversion should be according to MIQ 12 (MiQ12 (2000). Lyme-Borreliose by B. Wilske, L. Zöller, V. Brade, H. Eiffert, U. B. Göbel, G. Stanek). However, the MIQ was not correctly reflected in the previous version and this is corrected accordingly. The previous protocol failed to count IgM based seroconversion as Treatment Failure (TF) and this is corrected. In the previous version of the protocol, the IDMC was responsible to decide on doubtful cases of treatment failures (TF) and agree if the doubtful case is to be counted as TF or not. This is only detailed in this protocol to ensure, that the current MIQ (see above) is reflected. The protocol is particularly pointing the IDMC’s attention to cases within which serological results switch from “negative” (deemed as ‘negative’ by the MIQ; Visit 1) to “borderline” (deemed as ‘questionable’ by the MIQ; Visit 4). A stricter futility boundary was introduced (futility stop in case of a 1-sided p-value ≥ 0.1587, compared to ≥ 0.5 in protocol version 3), and this boundary is considered as binding. The mITT set was changed. The previous protocol used the serostatus of the index tick to define the mITT set. The new mITT set introduced here is as the ITT set but TF defined on IgM only seroconversion or IgG only seroconversion (e.g. not in combination with an EM) are excluded. Changes to the PP set became necessary accordingly and the Borrelia status of the index tick is integrated within this analysis set. It has been detailed, at which point the second interim analysis analysis is to be conducted. Details on the asymptotic distribution of the test statistic have been added. Determination of sample size: A larger treatment effect is assumed in the sample size calculation (relative risk of 0.4, corresponding to treatment failures rates of 3.1% for placebo and 1.24% for SHB004).