- Email: [email protected]
Incidence of Vocal Cord Dysfunction After Fluoroscopically Guided Steroid Injections in the Axial Skeleton
1330
ORIGINAL ARTICLE
Incidence of Vocal Cord Dysfunction After Fluoroscopically Guided Steroid Injections in the Axial Skeleton Atul L. Bhat, MD, David W. Chow, MD, Michael J. DePalma, MD, Cyndi Garvan, PhD, Larry Chou, MD, David Lenrow, MD, Curtis W. Slipman, MD ABSTRACT. Bhat AL, Chow DW, DePalma MJ, Garvan C, Chou L, Lenrow D, Slipman CW. Incidence of vocal cord dysfunction after fluoroscopically guided steroid injections in the axial skeleton. Arch Phys Med Rehabil 2005;86:1330-2. Objective: To report on the incidence of dysphonia and/or associated throat symptoms after steroid injections in the axial skeleton. Design: A prospective cohort study. Setting: Academic spine center. Participants: Patients (N⫽100) undergoing a diagnostic injection followed by a therapeutic injection in the axial skeleton. Interventions: Not applicable. Main Outcome Measure: Presence or absence of dysphonia and/or associated throat symptoms, as determined by completion of a predetermined 10-item questionnaire. This questionnaire was administered before participation in the study and at specific intervals after a diagnostic injection and, again, after a therapeutic spinal injection. Results: There was a 12% incidence of transient dysphonia and/or associated throat symptoms in this study population after a therapeutic injection. Conclusions: Throat symptoms are a potential, albeit transient side effect, after an epidural space corticosteroid injection. The mechanism through which dysphonia or other throat symptoms develop is unknown but appears to be mediated by a systemic steroid effect. Key Words: Dysphonia; Injections, epidural; Rehabilitation; Steroids. © 2005 by American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation LUOROSCOPICALLY GUIDED INJECTIONS of steroid preparations are commonly used in the axial skeleton to F address symptoms originating from the nerve roots, zygapophyseal joint, intervertebral disk, and the sacroiliac joint. These agents can cause minor and transient side effects, but rarely result in serious complications.1,2 Slipman et al3 previously reported the first case of transient dysphonia with throat
From the Orthopaedic Surgical Associates of Lowell PC, Lowell, MA (Bhat); Tufts University Medical Center, Boston, MA (Bhat); Bay Area Spine Institute, San Francisco, CA (Chow); Penn Spine Research Group, The Penn Spine Center, Department of Physical Medicine and Rehabilitation, Hospital of University of Pennsylvania, Philadelphia, PA (DePalma, Chou, Lenrow, Slipman); and Department of Statistics, Division of Biostatistics, University of Florida, Gainesville, FL (Garvan). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. Reprint requests to Curtis W. Slipman, MD, Penn Spine Center, Dept of Rehabilitation Medicine, Hospital of University of Pennsylvania, 3400 Spruce St, Ground Fl, White Bldg, Philadelphia, PA 19104, e-mail: [email protected]. 0003-9993/05/8607-9389$30.00/0 doi:10.1016/j.apmr.2004.11.042
Arch Phys Med Rehabil Vol 86, July 2005
symptoms and vocal cord abnormality as a result of a fluoroscopically guided lumbosacral epidural space steroid injection. These findings prompted a prospective study looking at all patients receiving fluoroscopically guided steroid injections in the axial skeleton. The primary objective of our study was to report on any dysphonia and/or associated throat symptoms after steroid injections into the axial skeleton. The secondary goal was to report on any possible association between these symptoms with the location of the injection and to elucidate a possible correlation between the 2. METHODS This prospective study was conducted at a multidisciplinary spine center located in an academic medical center. All patients receiving a diagnostic injection into the axial skeleton, using 2% lidocaine (Xylocaine), were asked to answer a preinjection questionnaire. This included questions about any history of voice hoarseness or huskiness; weakening of voice; excessive throat clearing or irritation; steroid use (past, present); tobacco use; asthma, bronchitis, chronic obstructive pulmonary disease, or emphysema; family deafness; gastric reflux; sleep apnea; frequent episodes of sore throat; frequent throat infections; previous laryngoscopy; previous throat or vocal cord surgery; and previous tracheostomy or intubations. Patients with a positive history were considered to have vocal cord abuse and were not eligible for inclusion in this study. As determined by the clinical algorithm, a diagnostic injection was performed under fluoroscopic guidance into either the epidural space, the zygapophyseal joint, or the sacroiliac joint. Contrast (Omnipaque) was used for all injections to confirm needle placement. In each instance the injectate flow was observed in real time, to ensure that there was no vascular flow. An independent registered nurse conducted telephone interviews at 24 hours and 72 hours after this diagnostic injection, querying patients for the presence of any weakening of voice, voice hoarseness or huskiness, sore throat, frequent throat clearing, excessive coughing, facial flushing, persistent hiccups, or nausea or lethargy. The telephone interviewer initiated the verbal assessment by using open-ended questions about any new complaints and then focused on these specific questions to maximize patient self-reporting. The first 100 consecutive patients who agreed to participate were enrolled into the study because each provided a negative response for each of these questions both before and after lidocaine injection and was thus eligible for inclusion. These patients subsequently received a therapeutic injection using a steroid preparation (Celestone) combined with 1% lidocaine and, therefore, each served as his/her own control. The same nurse contacted these subjects via telephone, before, at 24 hours, at 72 hours, and 7 days after the therapeutic injection. All patients were once again questioned about the presence of any dysphonia or associated throat symptoms. A percentile affirmative response was obtained from this cohort. Statistical analysis was performed using the SAS, version 8.a The FREQ procedurea was used to estimate the 95% exact confidence
VOCAL CORD DYSFUNCTION, Bhat
interval (CI) for the incidence of dysphonia. Our method was based on the probablilities of the binomial distribution instead of an approximation from the normal distribution. This method of binominal distribution provides a more conservative (ie, wider) CI, without making assumptions about the data that cannot be verified. Before participation in the study, all subjects provided written informed consent. RESULTS Of the 100 subjects, 48 were men and 52 were women. The average age was 48.4 years (range, 22– 83y). This cohort underwent a total of 62 lumbar selective nerve root injections (SNRIs) or transforaminal epidural injections (TEIs), 29 cervical SNRIs or TEIs, 6 lumbar zygapophyseal joint injections, and 3 sacroiliac joint injections. The volume of betamethasone used was 2mL for the lumbar epidural space, nerve root, or sacroiliac joint; 1mL for the cervical epidural space or nerve root; and 0.8mL for the lumbar zygapophyseal joint. The equivalent dosage of betamethasone acetate was 12, 6, and 4.8mg, respectively. Twelve patients reported the presence of dysphonia and/or associated throat symptoms. No patient reported these symptoms after a diagnostic injection; therefore, the incidence of dysphonia or other throat symptoms in this study sample was 12%. The lower and upper limits of the 95% CI were 6.4% and 20.0%, respectively. Of these 12 patients, 4 had undergone a cervical SNRI and 8 received a lumbar SNRI. No patient receiving a glucorticoid injection in the zygapophyseal joint or the sacroiliac joint experienced any dysphonia or throat symptoms. Positive throat symptoms were noted at 24 hours (n⫽9) and at 72 hours (n⫽3) after the therapeutic injection. A complete resolution of symptoms occurred in all but 1 patient, at the end of 1 week after the therapeutic steroid injection. The single patient who continued to experience symptoms at the end of 1 week was questioned during the next week. Two weeks after the therapeutic (betamethasone) injection, that subject reported a complete resolution of symptoms. Three of these 12 patients also underwent a videostroboscopic evaluation, during the dysphonic episode, approximately 48 hours after the therapeutic injection. In each case, there was vocal cord edema, with thick surrounding mucus. Another stroboscopic evaluation 1 week after the steroid injection showed complete resolution of these abnormal findings in all 3 patients. DISCUSSION Dysphonia reportedly affects 0% to 50% of patients using aerosolized steroid preparation to treat chronic asthma,4 and this variability may be a consequence of the differences among the study designs.4-9 The inhalation of aerosolized steroids has been hypothesized to affect vocal cord adductor function resulting in dysphonia.8 A local steroid effect has been postulated to be the culprit because the symptoms are directly related to total dose and potency.8 Since the early 1950s, epidural steroids have been used in medical practices. Subsequently, other types of spinal injections using corticosteroid, including zygapophyseal joint injections and sacroiliac joint injections, were performed. Various corticosteroid preparations have been used, including methylprednisolone acetate (Depo-Medrol), triamcinolone acetonide (Kenalog), triamcinolone diacetate (Aristocort), and betamethasone acetate and phosphate mixture (Celestone Soluspan). More recently, an algorithmic decision-making process has helped to determine which spinal structure ought to be treated with a therapeutic (corticosteroid) injection. An aspect
1331
of this methodologic stepwise analysis is the performance of a diagnostic injection at the site of suspected pathology, to confirm the source of the symptoms. When these procedures are performed, only a local anesthetic preparation elucidates the pain generator. In our study, all patients undergoing a diagnostic injection, using 2% Xylocaine, completed a preinjection checklist to exclude previous vocal cord abuse. Our use of these inclusion criteria removed subjects who had the potential of reporting dysphonia and/or throat symptoms for reasons other than the provision of corticosteroids into the axial skeleton. Although we hoped to minimize the risk that the throat symptoms were due to the use of local anesthetic or contrast, this possibility was not completely eliminated. This raises the concern of the possibility of throat symptoms secondary to an allergic reaction to contrast or lidocaine because sensitization to either or both of these agents could have transpired after the initial diagnostic injection. Yet, it is unlikely that an allergic response to either of these agents accounts for our results because no immediate symptoms were noted and type I hypersensitivity reactions are the most common adverse reactions to lidocaine10 and contrast media.11 The shortest interval between steroid injection and patient symptom manifestation was 24 hours. Hence, these reactions did not exemplify immediate or type I hypersensitivity reactions. Delayed adverse reactions, or type IV hypersensitivity, are less frequent than type I reactions and occur with an incidence of 0.7% in lidocaine exposures10 and 2.1% with exposure to contrast media.11 In both instances, mild dermatologic alterations are the most common clinical signs and symptoms, which are typically expressed within 24 hours.10,11 Although possible, it is improbable that the observed throat symptoms in our study cohort could be attributed to the administration of lidocaine or contrast during the diagnostic injections because no symptoms manifested during 72 hours after injection. The rationale for the use of epidural steroid in the spine is predicated on the benefits of blocking the inflammatory cascade and, possibly, neural blockade. Such a notion seems to be supported by the observation that pain relief outlasts the transient pharmacologic action of corticosteroids and local anesthetics. It is postulated that corticosteroids may reduce inflammation by inhibiting the synthesis or release of a number of proinflammatory substances and by causing a reversible local anesthetic effect. Indeed, systemic and oral steroids have been prescribed to treat dysphonia or laryngeal inflammation.12,13 Published cases seem to describe patients with inflammatory laryngeal conditions caused by an environmental allergen or overuse. In 1 case report, a decrement in laryngeal edema was directly observed in response to a 6-day course of oral methylprednisolone.12 The therapeutic role of corticosteroids to address a presumably inflammatory condition is illustrated by these articles. Yet, the patients included in these investigations were excluded from our study because of the history of risk factors for vocal cord abuse. Hence, our cohort represents patients without a predisposition to develop a laryngeal, edematous, inflammatory condition. Rather, our study exposes a potential side effect of epidural steroid injections, possibly caused by a systemic effect, because the symptomatic organ was distant to the deposition of steroid. Our findings do not help clarify why this phenomenom occurs. They simply show the existence of a positive correlation between epidural placement of steroid and the development of throat symptoms. Only 9% of our subjects did not undergo epidural deposition of steroid, and all 9 did not report throat symptoms after therapeutic injections. Perhaps the larger volume of injectant and the denser vasculature of the epidural Arch Phys Med Rehabil Vol 86, July 2005
1332
VOCAL CORD DYSFUNCTION, Bhat
space allow more systemic spread of steroid. The epidural space is known to contain an extensive venous lattice structure, whereas, by comparison, the facet and sacroiliac joint spaces are essentially void of vascular elements. Indeed, the systemic absorbtion of epidurally placed medication has been shown elsewhere.14 Although vascular uptake has been shown during facet joint and sacroiliac joint block,15 whether this spread can occur in the absence of documented intravascular uptake is undetermined. It is possible that extravasation of steroid through recesses in the facet joint capsule can result in epidural extension.16 However, it is unlikely that the small volume used in our study escaped into the epidural space. Yet, the distribution of epidurally placed medications is more easily explained than any correlation between this extension and dysphonia or throat abnormalities. A cogent explanation of a relationship between epidural administration of corticosteroids and signs and symptoms localized to a distant structure requires the existence of a systemic effect. Suppression of the hypopituitary axis, resulting in altered fluid and electrolyte balance or changes in autonomic activity, may be responsible for a more viscous laryngeal mucous production or edema. An extensive review of the literature failed to reveal any studies investigating the relationship between suppression of the hypothalamic-pituitary-adrenal (HPA) axis and laryngeal mucosal characteristics. However, suppression of the HPA axis by intramuscular corticosteroid doses similar to those introduced into our patients has been shown.17 Whether this depression results in abnormalities responsible for our findings is speculative. Currently, it appears that an epidural injection of corticosteroid is associated with a 12% incidence of throat symptoms that will manifest by 24 to 72 hours and resolve completely by 7 to 14 days. We are unable to comment on the systemic effect after zygapophyseal joint or the sacroiliac joint injection because the sample size was small. The measurement tool used to query patients about their postinjection symptoms is not without flaws. Patients may be inclined to overestimate symptoms when their presence is elicited by a questionnaire rather than simply observed by a trained health care professional. However, 3 patients had undergone videostroboscopic evaluation showing vocal cord edema that subsequently resolved concurrent with clinical resolution of their throat symptoms. We chose to not have each symptomatic patient undergo videostroboscopic evaluation because we anticipated that this side effect would be self-limited, thereby rendering this diagnostic evaluation clinically superfluous; using it would not impact our ultimate treatment. Although videostroboscopy was only performed in 25% of symptomatic patients, it supports the notion that a Hawthorne effect did not skew our observations. We consider this as a preliminary study only and are currently enrolling patients to increase the sample size for a larger prospective study. CONCLUSIONS Dysphonia and/or associated throat symptoms are a potential side effect of fluoroscopically guided steroid injections in the epidural space. The spine interventionalist needs to be aware and the patient needs to be educated about this potential but
Arch Phys Med Rehabil Vol 86, July 2005
transient side effect of epidural steroid injections. The mechanism of dysphonia and other throat symptoms is unknown but may be secondary to a systemic steroid effect. References 1. Huston CW, Slipman CW, Meyers JS, Yang ST, Anghel BN. Side effects and complications of fluoroscopically guided nerve root injections [abstract]. Arch Phys Med Rehabil 1996;77:937. 2. Botwin KP, Gruber RD, Bouchlas CG, Torres-Ramos FM, Freeman TL, Slaten WK. Complications of fluoroscopically guided transforaminal lumbar epidural injections. Arch Phys Med Rehabil 2000; 81:1045-50. 3. Slipman CW, Chow DW, Lenrow DA, Blaugrund JE, Chou LH. Dysphonia associated with spidural steroid injection: a case report. Arch Phys Med Rehabil 2002;83:1309-10. 4. Williamson IJ, Matusiewicz SP, Brown PH, Greening AP, Crompton GK. Frequency of voice problems and cough in patients using pressurized aerosol inhaled steroid preparations. Eur Resp J 1995;8: 590-2. 5. Adelroth E, Rosenhall L, Glennow C. High dose inhaled budesonide in the treatment of severe steroid-dependent asthmatics. A two-year study. Allergy 1985;40:58-64. 6. Cooper JE, Grant IW. Beclomethasone dipropionate aerosol in treatment of chronic asthma. Q J Med 1977;183:295-308. 7. Smith MJ, Hodson ME. High dose beclosmethasone inhaler in the treatment of asthma. Lancet 1983;i:265-9. 8. Williams AJ, Baghat MS, Stableforth DE, Cayton RM, Shenoi PM, Skinner C. Dysphonia caused by inhaled steroids: recognition of a characteristic laryngeal abnormality. Thorax 1983;38:813-21. 9. Williams MH, Kane C, Shim CS. Treatment of asthma with triamcinolone acetonide delivered by aerosol. Am Rev Respir Dis 1974;109:538-43. 10. Mackley CL, Marks JG, Anderson BE. Delayed-type hypersensitivity to lidocaine. Arch Dermatol 2003;139:343-6. 11. Yasuda R, Munechika H. Delayed adverse reactions to nonionic monometric contrast-enhanced media. Invest Radiol 1998;33:1-5. 12. Watts CR, Clark R, Early S. Acoustic measures of phonatory improvement secondary to treatment by oral corticosteroids in a professional singer: a case report. J Voice 2001;15:115-21. 13. Jackson-Menaldi CA, Dzul AI, Holland RW. Hidden respiratory allergies in voice users: treatment strategies. Logoped Phoniatr Vocol 2002;27:74-9. 14. Bonica JJ, Berges PU, Morikawa K. Circulatory effects of peridural block: I. Effects of level of anesthesia and dose of lidocaine. Anesthesiology 1970;33:619-26. 15. Sullivan WJ, Willick SE, Chira-Adisai W, et al. Incidence of intravascular uptake in lumbar spinal injection procedures. Spine 2000;25:481-6. 16. Destouet JM, Gilula LA, Murphy WA, Monsees B. Lumbar facet joint injection: indication, technique, clinical correlation, and preliminary results. Radiology 1982;145:321-5. 17. Lazarevic MB, Skosey JL, Djordjevic-Denic G, et al. Reduction of cortisol levels after single intra-articular and intramuscular steroid injection. Am J Med 1995;99:370-3. Supplier a. SAS Institute Inc, 100 SAS Campus Dr, Cary, NC 27513.
ORIGINAL ARTICLE
Incidence of Vocal Cord Dysfunction After Fluoroscopically Guided Steroid Injections in the Axial Skeleton Atul L. Bhat, MD, David W. Chow, MD, Michael J. DePalma, MD, Cyndi Garvan, PhD, Larry Chou, MD, David Lenrow, MD, Curtis W. Slipman, MD ABSTRACT. Bhat AL, Chow DW, DePalma MJ, Garvan C, Chou L, Lenrow D, Slipman CW. Incidence of vocal cord dysfunction after fluoroscopically guided steroid injections in the axial skeleton. Arch Phys Med Rehabil 2005;86:1330-2. Objective: To report on the incidence of dysphonia and/or associated throat symptoms after steroid injections in the axial skeleton. Design: A prospective cohort study. Setting: Academic spine center. Participants: Patients (N⫽100) undergoing a diagnostic injection followed by a therapeutic injection in the axial skeleton. Interventions: Not applicable. Main Outcome Measure: Presence or absence of dysphonia and/or associated throat symptoms, as determined by completion of a predetermined 10-item questionnaire. This questionnaire was administered before participation in the study and at specific intervals after a diagnostic injection and, again, after a therapeutic spinal injection. Results: There was a 12% incidence of transient dysphonia and/or associated throat symptoms in this study population after a therapeutic injection. Conclusions: Throat symptoms are a potential, albeit transient side effect, after an epidural space corticosteroid injection. The mechanism through which dysphonia or other throat symptoms develop is unknown but appears to be mediated by a systemic steroid effect. Key Words: Dysphonia; Injections, epidural; Rehabilitation; Steroids. © 2005 by American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation LUOROSCOPICALLY GUIDED INJECTIONS of steroid preparations are commonly used in the axial skeleton to F address symptoms originating from the nerve roots, zygapophyseal joint, intervertebral disk, and the sacroiliac joint. These agents can cause minor and transient side effects, but rarely result in serious complications.1,2 Slipman et al3 previously reported the first case of transient dysphonia with throat
From the Orthopaedic Surgical Associates of Lowell PC, Lowell, MA (Bhat); Tufts University Medical Center, Boston, MA (Bhat); Bay Area Spine Institute, San Francisco, CA (Chow); Penn Spine Research Group, The Penn Spine Center, Department of Physical Medicine and Rehabilitation, Hospital of University of Pennsylvania, Philadelphia, PA (DePalma, Chou, Lenrow, Slipman); and Department of Statistics, Division of Biostatistics, University of Florida, Gainesville, FL (Garvan). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. Reprint requests to Curtis W. Slipman, MD, Penn Spine Center, Dept of Rehabilitation Medicine, Hospital of University of Pennsylvania, 3400 Spruce St, Ground Fl, White Bldg, Philadelphia, PA 19104, e-mail: [email protected]. 0003-9993/05/8607-9389$30.00/0 doi:10.1016/j.apmr.2004.11.042
Arch Phys Med Rehabil Vol 86, July 2005
symptoms and vocal cord abnormality as a result of a fluoroscopically guided lumbosacral epidural space steroid injection. These findings prompted a prospective study looking at all patients receiving fluoroscopically guided steroid injections in the axial skeleton. The primary objective of our study was to report on any dysphonia and/or associated throat symptoms after steroid injections into the axial skeleton. The secondary goal was to report on any possible association between these symptoms with the location of the injection and to elucidate a possible correlation between the 2. METHODS This prospective study was conducted at a multidisciplinary spine center located in an academic medical center. All patients receiving a diagnostic injection into the axial skeleton, using 2% lidocaine (Xylocaine), were asked to answer a preinjection questionnaire. This included questions about any history of voice hoarseness or huskiness; weakening of voice; excessive throat clearing or irritation; steroid use (past, present); tobacco use; asthma, bronchitis, chronic obstructive pulmonary disease, or emphysema; family deafness; gastric reflux; sleep apnea; frequent episodes of sore throat; frequent throat infections; previous laryngoscopy; previous throat or vocal cord surgery; and previous tracheostomy or intubations. Patients with a positive history were considered to have vocal cord abuse and were not eligible for inclusion in this study. As determined by the clinical algorithm, a diagnostic injection was performed under fluoroscopic guidance into either the epidural space, the zygapophyseal joint, or the sacroiliac joint. Contrast (Omnipaque) was used for all injections to confirm needle placement. In each instance the injectate flow was observed in real time, to ensure that there was no vascular flow. An independent registered nurse conducted telephone interviews at 24 hours and 72 hours after this diagnostic injection, querying patients for the presence of any weakening of voice, voice hoarseness or huskiness, sore throat, frequent throat clearing, excessive coughing, facial flushing, persistent hiccups, or nausea or lethargy. The telephone interviewer initiated the verbal assessment by using open-ended questions about any new complaints and then focused on these specific questions to maximize patient self-reporting. The first 100 consecutive patients who agreed to participate were enrolled into the study because each provided a negative response for each of these questions both before and after lidocaine injection and was thus eligible for inclusion. These patients subsequently received a therapeutic injection using a steroid preparation (Celestone) combined with 1% lidocaine and, therefore, each served as his/her own control. The same nurse contacted these subjects via telephone, before, at 24 hours, at 72 hours, and 7 days after the therapeutic injection. All patients were once again questioned about the presence of any dysphonia or associated throat symptoms. A percentile affirmative response was obtained from this cohort. Statistical analysis was performed using the SAS, version 8.a The FREQ procedurea was used to estimate the 95% exact confidence
VOCAL CORD DYSFUNCTION, Bhat
interval (CI) for the incidence of dysphonia. Our method was based on the probablilities of the binomial distribution instead of an approximation from the normal distribution. This method of binominal distribution provides a more conservative (ie, wider) CI, without making assumptions about the data that cannot be verified. Before participation in the study, all subjects provided written informed consent. RESULTS Of the 100 subjects, 48 were men and 52 were women. The average age was 48.4 years (range, 22– 83y). This cohort underwent a total of 62 lumbar selective nerve root injections (SNRIs) or transforaminal epidural injections (TEIs), 29 cervical SNRIs or TEIs, 6 lumbar zygapophyseal joint injections, and 3 sacroiliac joint injections. The volume of betamethasone used was 2mL for the lumbar epidural space, nerve root, or sacroiliac joint; 1mL for the cervical epidural space or nerve root; and 0.8mL for the lumbar zygapophyseal joint. The equivalent dosage of betamethasone acetate was 12, 6, and 4.8mg, respectively. Twelve patients reported the presence of dysphonia and/or associated throat symptoms. No patient reported these symptoms after a diagnostic injection; therefore, the incidence of dysphonia or other throat symptoms in this study sample was 12%. The lower and upper limits of the 95% CI were 6.4% and 20.0%, respectively. Of these 12 patients, 4 had undergone a cervical SNRI and 8 received a lumbar SNRI. No patient receiving a glucorticoid injection in the zygapophyseal joint or the sacroiliac joint experienced any dysphonia or throat symptoms. Positive throat symptoms were noted at 24 hours (n⫽9) and at 72 hours (n⫽3) after the therapeutic injection. A complete resolution of symptoms occurred in all but 1 patient, at the end of 1 week after the therapeutic steroid injection. The single patient who continued to experience symptoms at the end of 1 week was questioned during the next week. Two weeks after the therapeutic (betamethasone) injection, that subject reported a complete resolution of symptoms. Three of these 12 patients also underwent a videostroboscopic evaluation, during the dysphonic episode, approximately 48 hours after the therapeutic injection. In each case, there was vocal cord edema, with thick surrounding mucus. Another stroboscopic evaluation 1 week after the steroid injection showed complete resolution of these abnormal findings in all 3 patients. DISCUSSION Dysphonia reportedly affects 0% to 50% of patients using aerosolized steroid preparation to treat chronic asthma,4 and this variability may be a consequence of the differences among the study designs.4-9 The inhalation of aerosolized steroids has been hypothesized to affect vocal cord adductor function resulting in dysphonia.8 A local steroid effect has been postulated to be the culprit because the symptoms are directly related to total dose and potency.8 Since the early 1950s, epidural steroids have been used in medical practices. Subsequently, other types of spinal injections using corticosteroid, including zygapophyseal joint injections and sacroiliac joint injections, were performed. Various corticosteroid preparations have been used, including methylprednisolone acetate (Depo-Medrol), triamcinolone acetonide (Kenalog), triamcinolone diacetate (Aristocort), and betamethasone acetate and phosphate mixture (Celestone Soluspan). More recently, an algorithmic decision-making process has helped to determine which spinal structure ought to be treated with a therapeutic (corticosteroid) injection. An aspect
1331
of this methodologic stepwise analysis is the performance of a diagnostic injection at the site of suspected pathology, to confirm the source of the symptoms. When these procedures are performed, only a local anesthetic preparation elucidates the pain generator. In our study, all patients undergoing a diagnostic injection, using 2% Xylocaine, completed a preinjection checklist to exclude previous vocal cord abuse. Our use of these inclusion criteria removed subjects who had the potential of reporting dysphonia and/or throat symptoms for reasons other than the provision of corticosteroids into the axial skeleton. Although we hoped to minimize the risk that the throat symptoms were due to the use of local anesthetic or contrast, this possibility was not completely eliminated. This raises the concern of the possibility of throat symptoms secondary to an allergic reaction to contrast or lidocaine because sensitization to either or both of these agents could have transpired after the initial diagnostic injection. Yet, it is unlikely that an allergic response to either of these agents accounts for our results because no immediate symptoms were noted and type I hypersensitivity reactions are the most common adverse reactions to lidocaine10 and contrast media.11 The shortest interval between steroid injection and patient symptom manifestation was 24 hours. Hence, these reactions did not exemplify immediate or type I hypersensitivity reactions. Delayed adverse reactions, or type IV hypersensitivity, are less frequent than type I reactions and occur with an incidence of 0.7% in lidocaine exposures10 and 2.1% with exposure to contrast media.11 In both instances, mild dermatologic alterations are the most common clinical signs and symptoms, which are typically expressed within 24 hours.10,11 Although possible, it is improbable that the observed throat symptoms in our study cohort could be attributed to the administration of lidocaine or contrast during the diagnostic injections because no symptoms manifested during 72 hours after injection. The rationale for the use of epidural steroid in the spine is predicated on the benefits of blocking the inflammatory cascade and, possibly, neural blockade. Such a notion seems to be supported by the observation that pain relief outlasts the transient pharmacologic action of corticosteroids and local anesthetics. It is postulated that corticosteroids may reduce inflammation by inhibiting the synthesis or release of a number of proinflammatory substances and by causing a reversible local anesthetic effect. Indeed, systemic and oral steroids have been prescribed to treat dysphonia or laryngeal inflammation.12,13 Published cases seem to describe patients with inflammatory laryngeal conditions caused by an environmental allergen or overuse. In 1 case report, a decrement in laryngeal edema was directly observed in response to a 6-day course of oral methylprednisolone.12 The therapeutic role of corticosteroids to address a presumably inflammatory condition is illustrated by these articles. Yet, the patients included in these investigations were excluded from our study because of the history of risk factors for vocal cord abuse. Hence, our cohort represents patients without a predisposition to develop a laryngeal, edematous, inflammatory condition. Rather, our study exposes a potential side effect of epidural steroid injections, possibly caused by a systemic effect, because the symptomatic organ was distant to the deposition of steroid. Our findings do not help clarify why this phenomenom occurs. They simply show the existence of a positive correlation between epidural placement of steroid and the development of throat symptoms. Only 9% of our subjects did not undergo epidural deposition of steroid, and all 9 did not report throat symptoms after therapeutic injections. Perhaps the larger volume of injectant and the denser vasculature of the epidural Arch Phys Med Rehabil Vol 86, July 2005
1332
VOCAL CORD DYSFUNCTION, Bhat
space allow more systemic spread of steroid. The epidural space is known to contain an extensive venous lattice structure, whereas, by comparison, the facet and sacroiliac joint spaces are essentially void of vascular elements. Indeed, the systemic absorbtion of epidurally placed medication has been shown elsewhere.14 Although vascular uptake has been shown during facet joint and sacroiliac joint block,15 whether this spread can occur in the absence of documented intravascular uptake is undetermined. It is possible that extravasation of steroid through recesses in the facet joint capsule can result in epidural extension.16 However, it is unlikely that the small volume used in our study escaped into the epidural space. Yet, the distribution of epidurally placed medications is more easily explained than any correlation between this extension and dysphonia or throat abnormalities. A cogent explanation of a relationship between epidural administration of corticosteroids and signs and symptoms localized to a distant structure requires the existence of a systemic effect. Suppression of the hypopituitary axis, resulting in altered fluid and electrolyte balance or changes in autonomic activity, may be responsible for a more viscous laryngeal mucous production or edema. An extensive review of the literature failed to reveal any studies investigating the relationship between suppression of the hypothalamic-pituitary-adrenal (HPA) axis and laryngeal mucosal characteristics. However, suppression of the HPA axis by intramuscular corticosteroid doses similar to those introduced into our patients has been shown.17 Whether this depression results in abnormalities responsible for our findings is speculative. Currently, it appears that an epidural injection of corticosteroid is associated with a 12% incidence of throat symptoms that will manifest by 24 to 72 hours and resolve completely by 7 to 14 days. We are unable to comment on the systemic effect after zygapophyseal joint or the sacroiliac joint injection because the sample size was small. The measurement tool used to query patients about their postinjection symptoms is not without flaws. Patients may be inclined to overestimate symptoms when their presence is elicited by a questionnaire rather than simply observed by a trained health care professional. However, 3 patients had undergone videostroboscopic evaluation showing vocal cord edema that subsequently resolved concurrent with clinical resolution of their throat symptoms. We chose to not have each symptomatic patient undergo videostroboscopic evaluation because we anticipated that this side effect would be self-limited, thereby rendering this diagnostic evaluation clinically superfluous; using it would not impact our ultimate treatment. Although videostroboscopy was only performed in 25% of symptomatic patients, it supports the notion that a Hawthorne effect did not skew our observations. We consider this as a preliminary study only and are currently enrolling patients to increase the sample size for a larger prospective study. CONCLUSIONS Dysphonia and/or associated throat symptoms are a potential side effect of fluoroscopically guided steroid injections in the epidural space. The spine interventionalist needs to be aware and the patient needs to be educated about this potential but
Arch Phys Med Rehabil Vol 86, July 2005
transient side effect of epidural steroid injections. The mechanism of dysphonia and other throat symptoms is unknown but may be secondary to a systemic steroid effect. References 1. Huston CW, Slipman CW, Meyers JS, Yang ST, Anghel BN. Side effects and complications of fluoroscopically guided nerve root injections [abstract]. Arch Phys Med Rehabil 1996;77:937. 2. Botwin KP, Gruber RD, Bouchlas CG, Torres-Ramos FM, Freeman TL, Slaten WK. Complications of fluoroscopically guided transforaminal lumbar epidural injections. Arch Phys Med Rehabil 2000; 81:1045-50. 3. Slipman CW, Chow DW, Lenrow DA, Blaugrund JE, Chou LH. Dysphonia associated with spidural steroid injection: a case report. Arch Phys Med Rehabil 2002;83:1309-10. 4. Williamson IJ, Matusiewicz SP, Brown PH, Greening AP, Crompton GK. Frequency of voice problems and cough in patients using pressurized aerosol inhaled steroid preparations. Eur Resp J 1995;8: 590-2. 5. Adelroth E, Rosenhall L, Glennow C. High dose inhaled budesonide in the treatment of severe steroid-dependent asthmatics. A two-year study. Allergy 1985;40:58-64. 6. Cooper JE, Grant IW. Beclomethasone dipropionate aerosol in treatment of chronic asthma. Q J Med 1977;183:295-308. 7. Smith MJ, Hodson ME. High dose beclosmethasone inhaler in the treatment of asthma. Lancet 1983;i:265-9. 8. Williams AJ, Baghat MS, Stableforth DE, Cayton RM, Shenoi PM, Skinner C. Dysphonia caused by inhaled steroids: recognition of a characteristic laryngeal abnormality. Thorax 1983;38:813-21. 9. Williams MH, Kane C, Shim CS. Treatment of asthma with triamcinolone acetonide delivered by aerosol. Am Rev Respir Dis 1974;109:538-43. 10. Mackley CL, Marks JG, Anderson BE. Delayed-type hypersensitivity to lidocaine. Arch Dermatol 2003;139:343-6. 11. Yasuda R, Munechika H. Delayed adverse reactions to nonionic monometric contrast-enhanced media. Invest Radiol 1998;33:1-5. 12. Watts CR, Clark R, Early S. Acoustic measures of phonatory improvement secondary to treatment by oral corticosteroids in a professional singer: a case report. J Voice 2001;15:115-21. 13. Jackson-Menaldi CA, Dzul AI, Holland RW. Hidden respiratory allergies in voice users: treatment strategies. Logoped Phoniatr Vocol 2002;27:74-9. 14. Bonica JJ, Berges PU, Morikawa K. Circulatory effects of peridural block: I. Effects of level of anesthesia and dose of lidocaine. Anesthesiology 1970;33:619-26. 15. Sullivan WJ, Willick SE, Chira-Adisai W, et al. Incidence of intravascular uptake in lumbar spinal injection procedures. Spine 2000;25:481-6. 16. Destouet JM, Gilula LA, Murphy WA, Monsees B. Lumbar facet joint injection: indication, technique, clinical correlation, and preliminary results. Radiology 1982;145:321-5. 17. Lazarevic MB, Skosey JL, Djordjevic-Denic G, et al. Reduction of cortisol levels after single intra-articular and intramuscular steroid injection. Am J Med 1995;99:370-3. Supplier a. SAS Institute Inc, 100 SAS Campus Dr, Cary, NC 27513.