is poor penetration of these agents into these thick-walled recesses (unlike the typical acute effusion). Thus, while we concur that the in vitro activity, general pharmacokinetic profiles, and exceUent record of tolerance make the ftuoroquinolones very attractive choices in the retreatment of patients with multiply resistant tuberculosis, we anticipate that the uncommon, particular circumstances we described wiD very likely imperil the success of any chemotherapeutic regimen, even one incorporating a ftuoroquinolone.

Inflammatory pseudotumors originating from major airways have been polypoid or sessile masses in aU of the reported cases. 1..:J In contrast, the tumor we encountered did not form a mass but was seen to be invading beneath the bronchial epithelium. We believe that this type of inflammatory pseudotumor, which invades the subepithelium, has not been previously reported. mmiyuki Iwaml, M.D., Yasunobu Hirotsu, M.D., ]oft WJkimoto, M.D., and Hidetomo mkunaga, M.D., DlvUion of Re8J1iratory DiseMu, National Minaml-Kyu&hu Hospital,

Michael D. Iseman, M.D., and l.or/e Madsen, R.N.-C.,

National Jewish Center for Immunology and Re8J1iratory Medicine, Denver

Kag08hima, japan

Reprint f"equuU: Dr Iwaml, DlvUion of Re8J1iratory DIsea8ea, National Minaml-Kyu&hu Hospital, KJJjiki-cho Aira-gun, Kago-

shima, japan 899-52

Inflammatory Pseudotumor Arising from the Hilum and Invading the Subeplthellum 1b the Editor:

An inflammatory pseudotumor was found at the hilar portion of a lung in a 68-yeaN>ld woman. A chest roentgenogram showed atelectasis of the left upper lobe. Chest computed tomography demonstrated a massive shadow in the left hilum. Fiberoptic bronchoscopy confirmed the presence of a moderate stenosis in almost two thirds of the left main bronchus, complete obstruction of the upper lobe bronchus, and a severe stenosis of the basal bronchus at the bifurcation of the upper and lower lobe bronchi (Fig 1). Histologic examination of bronchial biopsy specimens disclosed chronic inflammatory findings, but a pneumonectomy was performed. Grossly, a firm, white, solid mass was seen in the left hilum. Microscopically, the tumor was seen to be invading from the outermost layer of the bronchus to the mucosa, but the bronchial epithelium was almost normal.

REFERENCES 1 Armstrong P, Elston C, Sanderson M. Endobronchial histiocytoma. Br J Radioll975; 48:221-22 2 Hakimi M, Pai RP, Fine G, Davila JC. Fibrous histiocytoma of the trachea. Chest 1975; 68:367-68 3 Duncan JD, Greenberg SD, Mattox KL, MadeweU JE. Benign fibrous histiocytoma: a rare endobronchial neoplasm. Int Surg 1986; 7l:11()'11

Spontaneous Pneumomediastinum 1b the Editor:

I read with interest the article by Abolnik et al' on spontaneous pneumomediastinum (SPM), which appeared in the July 1991 issue of Chest. The observations and conclusions in this study are very similar to those published by us a few years ago,' although the annual incidence in our study was more than twofold higher. Abolnik et al included 11 patients who were initiaUy admitted to other hospitals. It was not clear whether the original roentgenograms and records were examined. In addition, they also included three patients with concurrent pneumothorax, a subgroup we excluded from our report. There is no way to determine whether spontaneous pneumothorax preceded SPM or vice versa. We excluded pediatric patients with acute asthma, a group in which the rate of SPM may exceed 5 percent.' The benign nature of SPM has been proved by us' and others.' We have shown that esophageal contrast studies and bronchoscopic procedures are not required unless there is reason to suspect a perforation of the aerodigestive system. Yet such studies were performed in 25 percent of the patients of Abolnik et al. The mean hospitalization of 6.3 days seems to be extremely long. We now observe most cases for 24 to 48 h, and about 25 percent are presently not admitted at aU and are just foUowed up in our outpatient clinic. It was initiaUy believed that SPM and spontaneous pneumothorax are two presentations of the same condition -rupture of subpleural blebs. This is supported by the similar age and sex distribution for both entities. The rarity of recurrence of SPM and the lack of a distinguishing body habitus are consistent with different processes. We are presently evaluating 40 patients with SPM and comparing them radiologically and functionaUy with patients with spontaneous pneumothorax. Although not completed, the study reveals surprising similarities between the two conditions.

Alan YeUin, M.D.,

FICURE 1. Fiberoptic bronchoscopic findings. The surface of the

main bronchus is smooth and undulatory in places. Severe stenosis

is present in the peripheral portion.

1742

Department of Thoracic Surgery,

Sheba Medical Center; 'leI Hashomer; Israel

Convnunlcatlon8 to the Editor

REFERENCES 1 Abolnik I, Lossos IS, Brewer R. Spontaneous pneumomediastinum: a report of 25 cases. Chest 1991; 100:93-5 2 Yellin A, Gapany-Gapanavicious M, Liberman Y. Spontaneous pneumomediastinum: is it a rare cause of chest· pain? Thorax 1983; 38:383-85 3 Eggleston PA, Ward SH, Pierson WE, Bierman C. Radiographic abnormalities in acute asthma in children. Pediatrics 1974; 54:442-

49

4 Pierson DJ. Pneumomediastinum. In: Marray JF, Nadel lA, eds. Textbook of respiratory medicine. Philadelphia: WB Saunders, 1988; 1795-1808

7b the Editor:

I read with interest the article by Abolnik et al.' I would like to call your readers' attention to another potential cause of this entity. W! recently cared for a previously healthy 15-yeaN)ld boy who presented to the emergency deparbnent with a 2-h history of nonradiating left-sided pleuritic chest pain, which was localized in the left midclavicular line above the nipple. The pain worsened with deep inspiration and cough. The patient denied weakness, syncope, or dizziness. The respiratory rate was 22 beats per minute. The chest examination revealed symmetrical breath sounds. No rales, wheezes, or rubs were auscultated, and no jUgular venous distension was present. A chest radiograph revealed a pneumomediastinum, and the patient was admitted to the pediatric intensive care unit. Further questioning revealed that the patient had snorted cocaine several hours prior to presentation. He described performing a Valsalva maneuver, which he stated potentiated the effect of the drug. His symptoms resolved, and he was transferred to the pediatric ward for substance abuse counseling. There has been a recent proliferation of articles describing the cardiac and central nervous system effects of cocaine. There have also been several case reports of pneumomediastinum and pneumothorax, which have appeared in the emergency and otolaryngology literature.... A recent study by Rich et al" reviewed the charts of 146 cocaine-abusing patients and found that 23 patients (16 percent) presented to the emergency deparbnent with cardiopulmonary symptoms, 17 of whom had chest pain. There are several other recent reviews concerning cocaine-related chest pain;"? the reader is referred to these for further details. After nasal insufBation of cocaine, a Valsalva maneuver will allow more rapid absorption and therefore potentiation of the effects of the drug. Pneumomediastinum develops as a result of the increase in intrabronchial and intra-alveolar pressure with a vigorous expiration against a closed glottis (Valsalva maneuver). If a large enough pressure gradient is generated, alveolar rupture will occur, and air will escape into the interstitial tissues and dissect along the bronchovascular sheath (toward the mediastinum) and then possibly along the fascial planes into the deep neck spaces or the pleural cavity to cause pneumothorax. 3 Fortunately, this is usually a selflimiting process, because of the short-lived expiratory air80w obstruction. Cocaine abuse should be considered as a potential cause of pneumomediastinum in any patient who presents with pneumomediastinum without a predisposing condition. It should also be noted that the posteroanterior x-ray 81m may be normal in 50 percent of the patients in whom pneumomediastinum has been demonstrated on a lateral view.·

2 Mir lA, Galvete JY, Plaza MY, Ingles MlA, Alonso EO, Navarro FL. Spontaneous pneumomediastinum after cocaine inhalation. Respiration 1986; 50:230-32 3 Aroesty DJ. Stanley RB, Crockett DM. Pneumomediastinum and cervical emphysema from the inhalation of "free based" cdc8ine: 'report of3 cases. Laryngol Head Neck Surg 1986: 94:37274 4 Luque MA, Cavallaro DL, Torres M, Emmanual P, Hillman Jv. Pneumomediastinum. pneumothorax, and subcutaneous emphysema after alternate cocaine inhalation and marijuana smoking. Pediatr Emerg Care 1987; 3:107-09 5 Rich lA. Singer DE. Cocaine-related symptoms in patients presenting to an urban emergency deparbnent. Ann Emerg Med 1991; 20:38-43 6 Zimmerman JL, Dellinger RP, Majid PA. Cocaine-associated chest pain. Ann Emerg Med 1991; 20:33-7 7 Brody SL, Slovis CM, Wrenn KD. Cocaine-related medical problems: consecutive series of 233 patients. Am J Med 1990; 88:325-31 8 Millard CEo Pneumomediastinum. Chest 1969; 56:297-300

Another Case of Pseudocardlomegaly 7b the Editor: I would like to comment on the case report of the Morgagni hernia presenting as cardiomegaly,' which appeared in the July 1991 issue of Chest. I saw a similar case in an asymptomatic 65yeaN)ld woman whose chest radiograph also suggested cardiomegaly. A chest radiograph taken 8ve years previously had a similar appearance. Diagnosis was con6rmed by barium enema examination only (Fig 1). The authors did not need both a computed tomographic scan and a barium study, since they gave similar information. I did not recommend surgery to my patient since I could 8nd no data suggesting that surgery was necessary. In one large series,· 36 of50 patients with a Morgagni hernia were asymptomatic, and none of the 50 patients developed bowel obstruction or needed an emergency operation. The patient whose case was reported by Chu et al was asymptomatic and should have been treated conservatively

Edward E. COfIWtJy.]r., M.D.• Dioiaion of Hldiatric CrltictJl Care. Mcmtefiore Medical Center; New York

REFERENCES 1 Abolnik I, Lossos I, Breuer R. Spontaneous pneumomediastinum: a report of 25 cases. Chest 1991; 100:93-5

FIGURE 1. Radiograph from barium enema examination con8rming diagnosis of Morgagni hernia. CHEST I 101 I 6 I JUNE. 1992

1743