Response and immunoresistance to botulinum toxin injections

Injections of botulinum toxin type A (BTX) provide effective, symptomatic relief for several disorders characterized by abnormal muscular contractions, such as dystonia, spasticity, tremors, tics, and other motor disorders. ^[1] BTX injections are considered the treatment of choice for many of the focal dystonias, particularly blepharospasm, cervical dystonia (torticollis), laryngeal dystonia (spasmodic dysphonia), and task-specific dystonias (eg, writer's cramp). Although the benefits persist in the vast majority of patients after repeated injections, ^[2] in some patients the condition becomes unresponsive to subsequent treatments. ^[3,4] One reason for the development of resistance to BTX treatment is the development of blocking antibodies (ABS). In this study, we sought to determine the causes of resistance and to identify risk factors for the formation of ABS.

Methods.

Eighty-six of 1,321 patients who received BTX injections on 6,549 visits at the Baylor College of Medicine Movement Disorder Clinic were tested for BTX ABS. Twenty-two were randomly selected for this study, which was approved by the Baylor Institutional Board for Human Research. The remainder were tested because of lack of or inadequate response to BTX injections (60 patients) or at the request of the patients (four patients), even though they had had a response to the treatment.

The response to treatment was rated on a 0 to 4 ``peak effect'' scale (0: no effect; 1: mild effect, no functional improvement; 2: moderate improvement, no change in functional disability; 3: moderate change in severity and function; and 4: marked improvement in severity and function). ``Lack of or inadequate response'' was defined as peak effect of 0 or 1. This study used the Allergan, Inc, preparation of BTX (BOTOX).

The presence of ABS was tested using a bioassay originally described by Hatheway et al ^[5,6] at the Center for Disease Control and performed by the Northview Pacific Laboratories, Berkeley, CA. Results of this qualitative bioassay are reported as either positive or negative. A positive result, indicating the presence of ABS in the patient's serum, occurs when the patient's serum neutralizes the effects of BTX and prevents the death of mice inoculated with the toxin. In a negative result, mice inoculated with BTX die, presumably because they are not protected by circulating ABS.

Results.

Of the 86 patients who were tested for BTX ABS, 20 had positive results and 66 had negative; 22 who had negative results were randomly tested. Five patients were tested more than once: in one, results turned negative 1 1/2 years after initial positive test results; in the remainder results were initially negative and subsequently became positive. These latter patients were included only in the ABS-positive group. All patients who had positive tests had no response to BTX (peak effect of 0 or 1 on at least two consecutive treatment sessions). All these patients were receiving BTX injections for cervical dystonia, except for one who had oromandibular dystonia manifested by forceful, spasmodic jaw closure. That patient's condition eventually failed to improve even with botulinum toxin type F. The other four patients treated with botulinum toxin type F obtained satisfactory but short-lasting improvement. None of the randomly tested patients had a positive BTX ABS test. This group was composed of 16 patients with cervical dystonia, five with cranial dystonia, and one with hand/arm dystonia.

Although there was a marked overlap between the two groups, the patients with positive BTX ABS results had an earlier age at onset (mean age: 31.8 plus minus 16.7 years versus 43.4 plus minus 10.5; p less than 0.05), higher mean dose per visit (249.2 plus minus 32.5 U versus 180.8 plus minus 68.7, p less than 0.0005), and higher total cumulative dose (mean dose: 1,709 plus minus 638 U versus 1,066 plus minus 938; p less than 0.01) as compared with the randomly selected patient group Table 1. There was no statistical difference between patients with negative and positive BTX ABS results for any other variable, including gender, age, duration of symptoms, total visits, number of visits, inter-injection interval, and duration of treatment.

Of 26 patients tested because of poor response to at least one treatment session and who received additional injections, 21 had good response (peak effect more than equals 2) after subsequent injection and five had minimal or no effect (peak effect less than equals 1). Six patients were later treated surgically with cervical rhizotomy: two subsequently received BTX injections and had good responses, two had only slight improvement, and two were lost to follow-up.

Discussion.

This study shows a 100% correlation between the presence of BTX ABS, as determined by the mouse bioassay, and complete failure of response to BTX injections. The patients with positive ABS results had absolutely no response and no atrophy after BTX injections although all but one had had responses in the past. The one patient with oromandibular dystonia, who did not have an improvement even with dosages as high as 400 U per treatment visit, later also did not have a response to botulinum toxin type F. Our study also establishes a link among the mean dose per treatment session, total cumulative dose, and the development of BTX antibodies Table 1. Since we had always waited at least 1 month between injections, we did not observe the effects of ``boosters'' on ABS formation. There was no difference in the frequency of treatment sessions between our patients with and without ABS. This contrasted with another study <sup>[4]</sup> that showed a direct correlation between the presence of ABS and ``booster'' injections, usually administered within 2 to 3 weeks after the initial treatment. In that study, ABS were detected in 24 of 559 patients (4.3%), and 10.5% of the 559 were estimated to have developed resistance to BTX. Since the investigators did not indicate how many patients were actually tested, the frequency of ABS could not be determined from that study. ^[4] A sub-analysis of their results on patients with BTX resistance identified three potential risk factors for the development of BTX resistance: (1) frequent injections, (2) ``booster'' injections, and (3) high doses of BTX per treatment.

The frequency of BTX ABS has been reported to be in the 0 to 10% range, based on the results of the in vivo mouse bioassay. ^[3,4,6-8] In contrast, Siatkowski et al ^[9] reported positive ABS in 57% of 42 patients treated with BTX for blepharospasm, hemifacial spasm, and cervical dystonia. This Figure wasbased on a new test using a sphere-linked immunodiagnostic assay (SLIDA). Although this assay apparently has a four-fold greater sensitivity than ELISA, the presence of ABS detected by the SLIDA technique failed to correlate with the patient's clinical response. ^[9] Since the method fails to detect neutralizing ABS and the nature of the ABS is unknown, the SLIDA method has limited value in assessing lack of clinical response to BTX treatment. ELISA has been used for the detection of BTX ABS, but clinical correlation between the presence of such ABS and a lack of response to BTX injections has not been established. ^[6,10,11] In one study, ^[12] only three of 96 patients with focal dystonia who received repeated injections of the preparation (Porton Down, UK) had ABS detected by an in vivo toxin neutralization test. The inter-injection interval was shorter in the patients with positive ABS tests, but there was no correlation with dosage or with clinical response. Again, the assay apparently did not detect blocking, clinically relevant ABS. In addition to lack of atrophy of the muscles that underwent injection, other physical examination techniques can be used to detect BTX resistance. For example, the lack of depression of transverse lines in the forehead after a unilateral injection of 10 to 15 U of BTX in the frontalis muscle, a site remote from that of the therapeutic injection, has been suggested as evidence for BTX resistance. ^[13] The validity of this observation, however, has not yet been tested. We are currently evaluating the effects of unilateral injection into the corrugator muscle. In patients who are responsive to BTX, such injection produces asymmetrical contraction, whereas in immunoresistant patients the symmetry of eyebrow contractions is preserved.

We believe that the chief reasons for the marked differences in the frequencies of ABS are: (1) heterogeneous patient population with various conditions requiring different dosages of BTX and (2) utilization of different techniques for detecting ABS. ABS can be directed against different components of the BTX preparation; while some are directed against the BTX molecule, others are directed against the associated proteins. Only those ABS that effectively block the biologic activity of the toxin are clinically relevant. Although cumbersome and expensive, the mouse neutralization bioassay is currently considered the most reliable assay for biologically relevant immunoresistance. ^[6] None of our patients with positive ABS results, measured by the mouse bioassay, had responses to subsequent BTX injections. In some of those with no response, however, the mouse bioassay failed to detect BTX ABS. This may indicate lack of sensitivity of the assay or other reasons for poor response, such as selection of wrong muscles or inadequate dose.

Patients who develop ABS to BTX type A, and thus become resistant to this type of toxin, usually experience benefits with other types of toxins, such as botulinum toxins B and F. ^[14-17] Although the magnitude of improvement with these toxins is similar to that with type A, the duration of the benefit tends to be shorter. Ten of 15 torticollis patients ^[17] who became resistant to BTX A improved with botulinum toxin F, but the benefits lasted only 1 month rather than the 3 months of benefits experienced before they became resistant to BTX type A. While the duration of benefit from botulinum toxin type B appears longer than from type F, only long-term studies will determine whether the alternative types of botulinum toxin offer any advantages other than their clinical benefits in patients who become immunoresistant to BTX type A. Alternating type A with one of the other types of botulinum toxin might reduce the long-term risk of immunoresistance.

In summary, BTX ABS formation is one explanation for the occasional lack of response to BTX injections. Treatment with alternate serotypes may offer clinical benefit to this group of patients. Absence of detectable BTX ABS may occur in patients with poor response to BTX injections because of inadequate dosage or selection of inappropriate muscles; this condition warrants reinjection at higher dosages or at alternative sites.