Midline nasal and hard palate destruction in cocaine abusers and cocaine's role in rhinologic practice.

Abstract

Intranasal cocaine abuse can lead to destruction of the palate and perforation of the nasal septum. The pathophysiology of cocaine-induced midline destructive lesions is multifactorial and includes local ischemia secondary to vasoconstriction, chemical irritation from adulterants put in "cut" cocaine, and infection secondary to trauma, impaired mucociliary transport, and decreased humoral and cell-mediated immunity. Cocaine abuse should be suspected in patients with a palatal or septal perforation of unknown etiology.

Introduction

Cocaine, which is derived from the leaves of Erythroxylon coca, has been used for its stimulating and euphoric effects for centuries; reports of its use by the Incas date back to 1230 A.D. (1) Today, cocaine is among the most addictive and dangerous of the commonly abused drugs in our society. (2) Approximately 23 million people in the United States have used cocaine at least once. (3) It was estimated that in 1992, 2.9 million Americans used cocaine occasionally (i.e., less than once a month), 1.3 million people used it more than once a month, and 640,000 people used it at least once a week. (3) Street cocaine is ingested by intranasal snorting, free-base smoking, and intravenous injection.

The legal use of cocaine as an anesthetic is common in rhinology practices. (4) Cocaine is widely used because its ability to vasoconstrict is unique among local anesthetics. (5) However, the use of cocaine for otolaryngologic procedures has come under scrutiny because of its potential side effects (see "Questions arise over the use of cocaine in rhinologic practice," page 176).

In this article, we describe the cases of two patients who were seen during August 1997 at the Manhattan (N.Y.) Eye, Ear, and Throat Hospital. Both patients had experienced osteocartilaginous necrosis secondary to cocaine abuse. We also review the pathophysiology and differential diagnosis of a cocaine-induced midline destructive lesion (CIMDL), and we discuss alternatives to the use of cocaine in rhinologic practice.

Case reports

Patient 1. We evaluated a 47-year-old woman who had a 14-year history of intranasal use of street cocaine that had begun in 1983. She described her early use as recreational, using 0.5 grams "a couple of times a month" during the first 4 years. During this time, she began to experience hypernasal speech, nasal obstruction, anosmia, and a diminished sense of taste.

Her cocaine intake increased dramatically during the next 2 years to between 0.5 and 2.0 g/day, and in 1989 she began to experience intermittent epistaxis, chronic facial and periorbital pain, purulent nasal discharge, and episodes of fever associated with sore throat and swollen neck glands. She recalled that during this period, "pieces of cartilage and small bits of bone" would fall from her nose when she cleaned it with a Q-Tip. Despite the severe discomfort that this caused, she did not seek professional medical care. She later became pregnant and quit using cocaine, but by 1996 she had resumed snorting 1.0 to 1.5 g/day. During the year prior to her visit, she began to notice a defect in her hard palate along with oronasal regurgitation of both solids and liquids.

In July 1997, she finally sought medical attention for facial swelling. After she did not respond to a 2-week trial of oral amoxicillin/clavulanate, she was admitted to the hospital with a diagnosis of facial cellulitis. On admission, she disclosed the nature of her cocaine abuse. Examination revealed left facial swelling and erythema, a nasal septal perforation with extensive crusting, and an oronasal fistula (figure, A). Laboratory measurements of antinuclear antibody (ANA), antineutrophil cytoplasmic antibody (c-ANCA), C-reactive protein, and rheumatoid factor levels were within normal limits. The patient's erythrocyte sedimentation rate (ESR) was slightly elevated, her packed cell volume was 30.2 ml/100 ml (slightly anemic), and her peripheral smear showed eosinophilia.

The woman's symptoms resolved with intravenous antibiotics and frequent nasal saline irrigations. Four weeks after being discharged, she was taken to the operating room for surgical closure of her oronasal fistula, but the defect could not be closed because of extensive granulation tissue and bony necrosis surrounding the fistula (figure, B). Therefore, the palate perforation was obturated with a denture. The patient also underwent multiple follow-up biopsies, which revealed evidence of chronic granulomatosis and areas of necrosis in the tissue.

Patient 2. A 44-year-old woman who was a known cocaine abuser reported a 2-day history of left facial swelling associated with fever and chills. She was diagnosed with left facial cellulitis. On physical examination, we noted a nasal septal perforation and a large amount of nasal crusting. Laboratory tests revealed that her white blood cell count was elevated (13,100/[mm.sup.3]) and that her nasal cultures were positive for Staphylococcus aureus. She was also positive for c-ANCA (140 U/ml), and her ESR was elevated. Findings on chest x-ray were normal.

After the patient failed to improve on intravenous antibiotics (metronidazole, ampicillin/sulbactam, and gentamicin) and nasal spray irrigations with saline and mupirocin, she was taken to the operating room. Endoscopy was performed to debride the nasal cavity and to obtain several biopsy specimens. A number of these specimens contained evidence of necrotic granulation tissue without vascular involvement.

Discussion

The first case of nasal septal perforation secondary to cocaine use was reported by Owens in 1912. (6,7) In 1997, Helie and Foumier reported that 17 cases of CIMDL had been described in the literature, four of which had involved the hard palate. (8) Two other case reports have been published--one of a patient who had only a nasal septal perforation (7) and one of a patient who had a nasal septal perforation and an oronasal fistula. (9) We suspect that the incidence of these lesions is greater than what has thus far been reported because many cocaine abusers who have a CIMDL do not seek medical attention.

For the purpose of this article, a patient is considered to have a CIMDL if he or she has at least two of these three conditions: (1) a nasal septal perforation, (2) lateral nasal wall destruction of the inferior or middle turbinate in the maxillary or ethmoid sinus, or (3) hard palate involvement, either clinically or radiographically.

Pathophysiology. Several factors are involved in the pathophysiology of a CIMDL. They include local ischemia secondary to vasoconstriction, chemical irritation from adulterants put in "cut" cocaine, and infection secondary to trauma, decreased humoral and cell-mediated immunity, and impaired mucociliary transport:

* Ischemia. The potent vasoconstricting properties of cocaine can induce ischemic necrosis of cartilage and mucosa. In addition, ischemia is often aggravated when abusers also use an over-the-counter vasoconstrictor to prevent the associated rebound hyperemia and thereby prolong the "high." (8,10)

* Chemical irritation. Pure cocaine is diluted with a wide variety of substances before it is sold on the street. These adulterants include mannitol, dextrose, lactose, inositol (a vitamin B compound), lidocaine, procaine, quinine, caffeine, talc, plaster of Paris, amphetamines, salicylamide, borax, Epsom salts, and heroin. (1,6,8,9) When snorted, lactose can cause postnasal drip, inositol can cause rhinitis, lidocaine and procaine can cause facial numbness, and quinine can cause tinnitus, nausea, headache, and visual impairment. (1,9)

* Infection. Local trauma with superinfection can occur when a cocaine abuser picks at his or her nose with an instrument such as a pen or pencil to relieve discomfort by "scratching the itch."

In addition, we also believe that cocaine's effect on both the immune system and nasal mucociliary transport rates plays an important role in the pathophysiology of CIMDL, although this is not generally stated in the literature. Researchers have shown that cocaine suppresses immune function in mice by decreasing phagocytosis, antibody function, and natural killer-cell and cytotoxic T-cell activity. (11) These decreases result in an increase in tumor growth and a lowering of resistance to various viruses. Wu et al determined how cocaine injected into mice alters the activity of the thymus and influences T-lymphocyte maturation. (12) Their analysis revealed that apoptosis of different subpopulations of T lymphocytes increased as the dosage of cocaine increased.

Cocaine also impairs the body's defense mechanisms against infection by slowing nasal mucociliary transport rates. The nasal mucociliary system moves foreign particles trapped in the mucus of the nose to the nasopharynx, where they can be either swallowed or expectorated. (13) Mucociliary clearance can be monitored by analyzing ciliary beat frequency with in vitro testing. (14) Human nasal mucociliary transport rates can be determined by tracing radioactive particles or the clearance times of colored dyes or saccharin from the anterior end of the inferior turbinate to the oropharynx. (13) It has been demonstrated that normal mucociliary clearance takes less than 20 minutes. (13) Ingels et al found that cocaine decreased ciliary beat frequency at concentrations of 1.75% and greater. (15) Mason et al (13) found that 2.5% cocaine decreased the mucociliary transport rate by 58%, and Ingles et al (15) also found that partially reversible cilia stasis occurs with 7.0% cocaine. This latter finding was in agreement with those of earlier studies that showed that 10.0% cocaine is likely to produce irreversible cilia stasis.(13,15) It is therefore logical to assume that the snorting of street cocaine significantly impairs the mucociliary system.

The histopathology seen in most patients with sinonasal tract necrosis following cocaine abuse is consistent with infection and positive cultures. The most common infectious agent is S aureus, which was isolated in our patient 2. (16)

Symptoms. The history and symptoms of the two patients described in this report were similar to those reported by several other authors who examined the detrimental effects of cocaine abuse on the sinonasal tract. (1,6,16,17) Common symptoms of intranasal snorting are excessive sniffing, epistaxis, nasal crusting, a diminished sense of smell and/or taste, sinus infection, frontal headache, nasal obstruction, perforation of the nasal septum and hard palate, and saddle-nose deformity. (1,9,17) Other medical complications of cocaine use include myocardial infarction, cardiac arrhythmias, hypertension, aortic rupture, stroke, clonic-tonic convulsions, septicemia with endocarditis and hepatitis, pulmonary insufficiency, and psychiatric disturbances. (1,2,17) The incidence of osteocartilaginous necrosis is not clear.

Diagnosis. A CIMDL should be considered in the differential diagnosis of any patient who has any of these symptoms that cannot be explained. The primary differential diagnoses include Wegener's granulomatosis, sinonasal lymphomas and other primary neoplasms of the paranasal sinuses, collagen vascular disease, and sinusitis with osteitis secondary to cocaine use. Identification of the specific etiology of midline nasal destruction and perforation is critical to successful management because therapy is quite specific. At one time, the infectious causes of nasal perforation included tuberculosis, typhoid, diphtheria, tertiary syphilis, and leprosy. However, as noted by Kuriloff and Kimmelman, these causes have become exceedingly rare since the introduction of antibiotics. (6)

The work-up should include a chest x-ray to rule out pulmonary infiltrates and laboratory testing to measure the ESR and levels of ANA, c-ANCA, and rheumatoid factor. Various authors have also recommended measuring angiotensin-converting enzyme, (6) fluorescent treponemal antibody-absorbed, (7,8) purified protein derivative, (8) and creatinine (17) levels. Multiple biopsies should be performed to rule out other causes of midline destruction and to ensure that life-threatening but treatable diseases such as Wegener' s granulomatosis are not missed. (6-8) Many authors once listed polymorphic reticulosis, malignant midline reticulosis, lymphomatoid granulomatosis, malignant destructive granuloma, and idiopathic midline destructive disease as separate entities in the etiology of midfacial necrotizing lesions. (8,16-18) However, advances in immunocytochemistry phenotyping and molecular genetics have shown that most of these lesions are lymphomas of the sinonasal tract. (19)

Wegener's granulomatosis. Before treating a patient with a necrotizing lesion of the sinonasal tract, it is crucial to rule out Wegener's granulomatosis because the 2-year survival rate for untreated patients is only 10%. (7) The presence of c-ANCA is not definitive for the presence of Wegener's granulomatosis, but it is strongly suggestive. The presence of c-ANCA is 88% sensitive and 95% specific for active Wegener's granulomatosis, (20) but a diagnosis must be based on a histopathologic demonstration of necrotizing granulomatous vasculitis on biopsy and findings of upper and lower respiratory disease with evidence of glomerulonephritis on clinical assessment. (15)

Although our patient 2 was positive for c-ANCA, multiple intranasal biopsies did not detect any necrotizing granulomatous vasculitis. Moreover, patient 2 did not have any evidence of pulmonary or renal disease. Therefore, based on both the histopathologic findings and the clinical picture, we concluded that the c-ANCA finding in this case was a false positive. Armstrong and Shikani reported a very similar case. (7) They treated a known cocaine abuser who had a nasal septal perforation, an oronasal fistula, and a positive c-ANCA test. After obtaining an appropriate biopsy and consulting with the rheumatology service, they likewise concluded that the c-ANCA finding was a false positive and that their patient did not have Wegener's granulomatosis.

Therefore, it does not appear that a positive c-ANCA without appropriate histopathologic findings is diagnostic of Wegener's granulomatosis in this patient population. Again, to make a definitive determination, it is clearly important to perform multiple biopsies until an adequate sample can be obtained to allow for an appropriate tissue analysis.

Studies have shown that treatment of Wegener's granulomatosis with cyclophosphamide and prednisone results in a complete remission in more than 75% of patients and a marked improvement in another 15%. (20)

References

(1.) Schweitzer VG. Osteolytic sinusitis and pneumomediastinum: Deceptive otolaryngologic complications of cocaine abuse. Laryngoscope 1986;96:206-10.

(2.) Kaplan HI, Sadock BJ, Grebb JA. Substance-related disorders. In: Kaplan HI, Sadock BJ, Grebb JA, eds. Kaplan and Sadock's Synopsis of Psychiatry. 7th ed. Baltimore: Williams and Wilkins, 1994:423-9.

(3.) O'Brien CP. Drug addiction and drug abuse. In: Hardman JG, Limbird LE, Molinoff PB, Ruddon RW, eds. Goodman and Gilman's The Pharmacological Basis of Therapeutics. 9th ed. New York: McGraw-Hill, 1996:569-72.

(4.) Kasemsuwan L, Griffiths MV. Lignocaine with adrenaline: Is it as effective as cocaine in rhinological practice? Clin Otolaryngol 1996;21:127-9.

(5.) Tarver CP, Noorily AD, Sakai CS. A comparison of cocaine vs. lidocaine with oxymetazoline for use in nasal procedures. Otolaryngol Head Neck Surg 1993;109:653-9.

(6.) Kuriloff DB, Kimmelman CP. Osteocartilaginous necrosis of the sinonasal tract following cocaine abuse. Laryngoscope 1989;99:918-24.

(7.) Armstrong M, Jr., Shikani AH. Nasal septal necrosis mimicking Wegener's granulomatosis in a cocaine abuser. Ear Nose Throat J 1996:75:623-6

(8.) Helie F, Fournier J. Destructive lesions of the median line secondary to cocaine abuse. J Otolaryngol 1997;26:67-9.

(9.) Sastry RC, Lee D, Har-El G. Palate perforation from cocaine abuse. Otolaryngol Head Neck Surg 1997;l16:565-6.

(10.) Yanagisawa E, Latorre R. Endoscopic view of cocaine rhinitis. Ear Nose Throat J 1996;75:128-30.

(11.) Shen ML, Luo YD, Hagen K, et al. Immunomodulating activities of cocaine--evaluation of lymphocyte transformation related to other immune functions. Intl J Immunopharmacol 1994;16:311-9.

(12.) Wu YB, Shen ML, Gu GG, et al. The effects of cocaine injections on mouse thymocyte population. Proc Soc Exp Biol Med 1997;214:173-9.

(13.) Mason JD, Aspden TJ, Adler J, et al. Measurement of nasal mucociliary transport rates on the isolated human inferior turbinate. Clin Otolaryngol 1995;20:530-5.

(14.) Edelstein DR. Aging of the normal nose in adults. Laryngoscope 1996;106:1-25.

(15.) Ingels KJ, Nijziel MR, Graamans K, Huizing EH. Influence of cocaine and lidocaine on human nasal cilia. Beat frequency and harmony in vitro. Arch Otolaryngol Head Neck Surg 1994;120:197-201.

(16.) Becker GD, Hill S. Midline granuloma due to illicit cocaine use. Arch Otolaryngol Head Neck Surg 1988;114:90-1.

(17.) Sercarz JA, Strasnick B, Newman A, Dodd LG. Midline nasal destruction in cocaine abusers. Otolaryngol Head Neck Surg 1991;105:694-701.

(18.) Wolff SM. Midline granuloma. In: Isselbacher KJ, ed. Harrison's Principles of Internal Medicine. 13th ed. New York: McGraw-Hill, 1994:1686-7.

(19.) Cleary KR, Batsakis JG. Sinonasal lymphomas. Ann Otol Rhinol Laryngol 1994;103:911-4.

(20.) Fauci AS. Wegener's granulomatosis. In: Isselbacher KJ, ed. Harrison's Principles of Internal Medicine. 13th ed. New York: McGraw-Hill, 1994:1674-6.

(21.) Lennox P, Hern J, Birchall M, Lund V. Local anaesthesia in flexible nasendoscopy. A comparison between cocaine and cophenylcaine. J Laryngol Otol 1996;l10:540-2.

(22.) Feehan HF, Mancusi Ungaro A. The use of cocaine as a topical anesthetic in nasal surgery. A survey report. Plast Reconstr Surg 1976;57:162-5.

(23.) Ashchi M, Wiedemann HP, James KB. Cardiac complication from use of cocaine and phenylephrine in nasal septoplasty. Arch Otolaryngol Head Neck Surg 1995;121:681-4.

RELATED ARTICLE: Questions arise over the use of cocaine in rhinologic practice

The use of cocaine in rhinologic practice has come under scrutiny because of its potential for abuse, its cost, and its side effects. (5,21) Side effects are the result of cocaine's ability to potentiate sympathetic nerve stimulation. In severe cases, this stimulation can result in cardiac arrhythmias, myocardial infarction (MI), high-output congestive heart failure, hypertension, aortic dissection, aortic rupture, intracerebral hemorrhage, and clonic-tonic convulsions. (3)

Feehan and Mancusi Ungaro obtained questionnaire responses from 741 surgeons who had administered cocaine during submucous resection and rhinoplasty procedures (total number of cases: 93,004). (22) Among these cases were 233 mild reactions to cocaine (resuscitation not required), 32 severe reactions (resuscitation required), and five deaths.

Vasoconstriction of the coronary arteries is the primary contributor to cocaine-induced MI. (21,23) MI can occur as a result of cocaine use alone or from the combined use of cocaine and either epinephrine or phenylephrine. Ashchi et al reported the case of a 23-year-old woman without coronary artery disease who experienced an acute non-Q-wave Ml after she had received cocaine anesthesia during an elective nasal septoplasty. (23) For hemostasis, cotton pledgets soaked in phenylephrine had been placed in her nose prior to extubation.

There is no other single substance that provides cocaine's unique dual effect of vasoconstriction and anesthesia. (4,5,21) But in light of its potential risks, researchers have looked for alternatives to cocaine anesthesia. Several authors have studied the effects of combining a traditional vasoconstrictor and an anesthetic. (4,5,21) Tarver et al found that a lidocaine and oxymetazoline combination is a stronger vasoconstrictor than cocaine and provides equally good anesthesia. (5) Lennox et al found that co-phenylcaine forte -- a new preparation that combines lidocaine, phenylephrine, and benzylkonium--provided degrees of vasoconstriction and local anesthesia in the nasal mucosa that were similar to those provided by cocaine. (21) The lidocaine provides effective anesthesia, the phenylephrine induces vasoconstriction, and the benzylkonium is a preservative. In another study, Kasemsuwan and Griffiths reported that the combination of lidocaine and adrenaline was as effective as cocaine. (4)

The abuse potential of cocaine, its potential to cause serious adverse reactions, and the availability of anesthetic and vasoconstriction substitutes should clearly influence our decisions regarding the use of cocaine in otolaryngology.

From the Eye and Ear Institute, the University of Pittsburgh School of Medicine (Dr. Smith), and the Department of Otolaryngology-Head and Neck Surgery, New York Presbyterian Hospital, New York City (Dr. Kacker and Dr. Anand).

Reprint requests: Ashutosh Kacker, MD, 500 E. 77th St., #1033, New York, NY 10162. Phone: [phone]; fax: [phone]; e-mail: [e-mail]