lb theEditor

being tapered off over a period of months. Moreover, so far as we are aware corticosteroids have never been reported as having any beneficial

effect

in the treatment

of LAM.

The basic problem in LAM is a proliferation of smooth muscle which seems to be estrogen- and progesterone-dependent.@ It is almost exclusively found in women of reproductive age, which is exacerbated by pregnancy and estrogen therapy,@ and which often remits after menopause.5 Currently, surgical oophorectomy and progesterone

therapy,

either

alone

or in combination,

appear

to

offer the greatest chance of benefit.2 Our patient, however, has shown marked clinical improvement following rapid chemical menopause induction resulting from treatment with D-Trp6-LHRH. It suggests to us that, in the future, chemical oophorectomy with an LHRH

analogue

should

receive

consideration

as an alternative

Michiel MM.

Eysvogel, M.D., and

Philip S. Page, M.D., Hospital Medisch Spectrum Twente, Ensehede, Netherlands REFERENCES 1 Eliasson AH, Phillips YY, Tenholder MF. Treatment of lymphan gioleiomyomatosis. A meta-analysis. Chest 1989; 196:1352-55 2 Cutler GB, Hoffman AR, Swerdloff RS, Santen RJ, Meldrum DR, Comite F. Therapeutic applications of luteinizing-hormone releasing hormone and its analogs. Ann Intern Med 1985; 102:64357 3 McCarthy KS Jr. Mossler JA, McLelland R, Sieker HO. Pulmo responsive

to progesterone.

New

4 Shen A, Iseman MD, Waldron JA, King TE. Exacerbation of pulmonary lymphangioleiomyomatosis by exogenous estrogens.

Chest 1987;95:782-85

lb theEditor: Dr. Hofford addresses two distinct points ofdiscussion. The main point emphasizes that tamoxifen, while not only lacking convincing evidence for efficacy in treating lymphangioleiomyomatosis (LAM), may in fact be associated with exacerbation of the disease. In our review of seven reported cases treated with tamoxifen as single therapy; four had to be excluded from the meta-analysis, three because of insufficient data'4 and one because treatment was initiated after car pulmonaie had been documented, too late in the course of disease to allow for therapeutic effect.@ Dr. Hofford provides interesting follow-up on one case3 which we had excluded because

the

duration

of treatment

was

not

given.

However, even with this added history we would still exclude this case from our meta-analysis because one month of therapy with tamoxifen is insufficient to allow judgement of the therapy as a success or failure. LAM may be rapidly progressive in its own right and its own tempo of progression could explain subsequent pneu mothoraces

and symptomatic

the two disorders. In 1983, Martin grouped BML, lymphangiomyo matosis, leiomyomatosis peritonealis disseminata and intravenous leiomyomatosis into one category named leiomyomatosis.@ Perhaps resolution ofthis issue depends upon one@spoint ofview on schemes for classifying diseases; that is, whether one is a “¿lumper― or a “¿splitter―. Our decision to include these two cases in the review of LAM therapy was based on the described clinical and histologic information which, in our view, made these cases appear more The case of LAM reported by Drs. Lizotte, Whitlock, and colleagues represents another well-documented failure of proges terone therapy. The dosage and duration oftreatment were adequate by our meta-analysis criteria, and therapy was given early enough

in the course of the disease to have allowed a chance for therapeutic effect. Progression of disease was objectively verified by roentgen ographic examination. The decision to forego oophorectomy “¿owing to the rapid progression of her pulmonary disease― is not endorsed by the results of our meta-analysis, which suggest that the combination of progesterone and oophorectomy provide the greatest likelihood of response. If a patient is felt to he too ill to undergo surgical oophorectomy, chemical oophorectomy with gonadotropin-releas ing hormone analog is a reasonable alternative.' Transplantation must be considered a therapy of last resort for LAM . Early efforts

should focus on optimal hormonal manipulation. The speculation

about recurrence

of LAM in an allograft is

interesting. Progesterone receptors in lung tissue may not be part

5 Silverstein EF, Ellis K, Wolff M, Jaretzki A. Pulmonary lymphan giomyomatosis. Am J Radiol 1974; 120:832-50

specifically

et al@discuss the overlapping clinical presentation and histologic expressions of their cases, remarking that one case clearly bridged

consistent with LAM.

to surgery in the treatment of LAM.

nary lymphangiomyomatosis Engi J Med 1980; 303:1461-65

analysis were originajly labelled benign metastasizing leiomyoma tosis (BML).'7 Dr. Hofford notes that LAM and BML should be clearly distinguishable entities. While the distinctness of the dis eases is usually present, this is not uniformly the case. Banner

exacerbation.

The one case

ofthe pathologic mechanism underlying LAM. Most cases described in the literature lack information on estrogen and progesterone receptors. However, Dishner's report'° and that of our case are two examples of LAM in which estrogen and progesterone receptors were negative, but good response to progesterone therapy was seen. LAM may involve lungs, mediastinum, or abdomen, separately or together, and it may be that this hyperplastic disease process could involve an allograft. Observation of Dr. Cooper's experience will over time yield answers. We appreciate the report by Drs. Eysvogel and Page describing

the successful application ofgonadotropin-releasing hormone analog in the therapy

of LAM . Their

evaluation

of the patient's

functional

abilities and concomitant laboratory data do substantiate a remark able response to treatment. This is the first report ofsuch treatment

and supports our suggestion of “¿chemical oophorectomy―for LAM. While this single case does not constitute proof, it does convincingly suggest that this noninvasive alternative is effective. In fact, the case provides useful information for Dr. Lizotte et al. Therapy with gonadotropin-releasing hormone analog might obviate their plans for lung transplantation. Further investigation of this agent in the therapy of LAM is indicated.

success

fully treated with tamoxifen' was previously treated with tapering doses ofprednisone over one year and concomitant medroxyproges terone for eight months. This prednisone and progesterone trial

Am H. Eliasson, M.D.; Yancy Y Phillips, M.D., F.C.C.?, and Michael F. Tenholder, M.D., F.C.C.?,

was a clear failure. Subsequent therapy with tamoxifen after

Pulmonary Disease Service Walter Reed Army Medical Center Washington, DC

pleurodesis parameters

resulted in documented improvements in spirometric and increased exercise tolerance. A single case such as

this is insufficient evidence upon which to base future therapeutic recommendations,

especially

in view of other

reported

failures.

We

concur that tamoxifen has no established role in the treatment of LAM and consequently we did not advocate its use in our article.

Dr. Hofford's second point is that two cases included in the meta 1046

Downloaded From: http://journal.publications.chestnet.org/ by Arn Eliasson on 02/23/2016

REFERENCES 1 Sobonya RE, Quan SF, Fleishman JS. Pulmonary lymphangio leiomyomatosis: quantitative analysis oflesions producing airflow limitation. Hum Pathol 1985; 16:1122-28

Communications to the Editor

2 Clemm C, Jehn U, WoIf-Hornung B, Siemon G, Walter C. Lymphangiomyomatosis: a report of three cases treated with tamoxifen. Klin Wochenschr 1987; 65:391-93 3 Shen A, Iseman MD, Waldron JA, King TE. Exacerbation of pulmonary lymphangioleiomyomatosis by exogenous estrogens.

Derbyshire Ro@pilInfirmary Derby, England

Reprint requests: Dr. Shaheen, Derby Rot@ilInfirmary, Derby, UK DE12QY

Chest 1987; 91:782-85 4 Tomasian

A, Greenberg

MS, Rtimerman

lymphangioleiomyothatosis.

N Engl J Med

H. Tamoxifen

for

1982; 306:745-46

5 Luna CM, Gene R, Jolly EC, Nahmod N, Defranchi HA, Patino C, Elsner B. Pulmonary lymphangiomyomatosis associated with tuberous sclerosis. Chest 1985; 88:473-75 6 Lipton JH, Fong it, Burgess KR. Miliary pattern as presenta

lion ofleiomyomatosisofthe lung. Chest 1987;91:781-82 7 Banner AS, Carrington

CB, Emory WB, Kittle F, Leonard

C,

Ringus J, et al. Efficacy ofoophorectomy in lymphangioleiomyo matosis and benign metastasizing leiomyoma. N Engl J Med 1981; 305:204-09 8 Martin E . Leiomyomatous lung lesions: a proposed classification. AJR 1983; 141:269-72 9 Cutler GB, Hoffman AR, Swerdloff RS, Santen RJ, Meldrum DR. Comite F. Therapeutic applications ofluteinizing hormone releasing hormone and its analogs. Ann Intern Med 1985; 102:643-57 10 Dishner W, Cordasco EM, Blackburn J, Demeter 5, Levin H,

REFERENCES 1 Shaheen

MZ, Hassan TB, Windebank

WJ. Misleading

isotope

lung scans in the diagnosis of pulmonary embolism. Scot Med J (in press) 2 Stjernholm MR, Landis GA, Marcus F!, Miale A Jr. Moser KM. Walsh B. Perfusion and ventilation radio-isotope lung scans in stenosis of pulmonary arteries and their branches. Am Heart J 1969; 78:37-42

To the Editor: Drs. Shaheen and Windebank have made an assumption with their statertient: “¿among the group ofpatients involved in this study, there is not even a single patient in whom respiratory symptoms

Carey WD. Pulmonary lymphangiomyomatosis. Chest 1984;

highlysuggestiveofpulmonaryembolismpersisted in the presence

85:796-99

of a normal perfusion scan and no other alternative explanation:' It is not possible to make this conclusion from the data in our article:'

the number of pahents with persistent respiratory symptoms is not described. Our objective was to test the safety of withholding anticoagulant treatment in symptomatic patients with suspected pulmonary embolism and normal perfusion scans, regardless of clinical manifestations. Our study did so. We studied a broad spectrum of patients with a wide range of severity of symptoms.

TestingForPulmonaryEmbolism

We agree that the perfusion lung scan does not have perfect

Tothe Editor: We read with great interest Normal

Perftision

the paper, “¿Clinical Validity of a

Lung Scan in Patients

with Suspected

Pulmonary

Embolism―, by Hull et al (Chest 1990; 97:23-26). Their study was designed to evaluate the accuracy of a normal perfusion lung scan in excluding the diagnosis of pulmonary embolism . Our department of respiratory medicine has a special

sensitivity for pulmonary embolism. Two findings of our study support this conclusion: 1) the finding of proximal-vein thrombosis

at presentation in 1 percent ofpatients (five of493 patients), and 2) the finding ofvenous thromboembolism on follow-up in 0.6 percent of patients (three of 515 patients).' Our findings are supported by the observations of Kipper et al.2

in

Which patients with normal perfusion scans should have pulmo nary angiography? This remains clinical judgement. We would consider performing pulmonary angiography in a patient with a normal perfusion scan if the clinical findings suggested massive saddle embolism. Such patients are rare! Indeed, venous throm boembolism occurred in only eight of 515 patients (1.5 percent) in our study. This outcome compares favorably with the 2 percent rate

the presence of a normal perfusion scan and no other alternative

ofpulmonary embolism reported by others on follow-up in patients

explanation. The majority of patients in this study were outpatients and, therefore, likely to have less severe respiratory symptoms. Moreover, in one-third of patients with normal perfusion scan, no alternative diagnosis was made. Duration and severity of symptoms in this group of patients has not clearly been mentioned. We have documented massive pulmonary embolism on pulmo nary angiography in four patients who had normal perfusion lung

with

interest in venous thromboembolism . We believe that a perfusion lung scan is most valuable when it is normal and, therefore, can exclude the possibility of pulmonary embolism—which is in agree

ment with Hull et al. However, among the group ofpatients involved in this study,

symptoms

there

is not even a single

highly suggestive

1 Pulmonary

angiography

patient

of pulmonary

was

performed

in whom

embolism

only

respiratory

persisted

because

clinical

suspicion of pulmonary embolism was strong and persistent. Pre vious studies have shown that this situation can arise in cases of partial occlusion, recanalization ofthe clot, or with saddle embolus.2 Pulmonary angiography is the only diagnostic test in such situations. Isotope lung scanning is our first-line investigation in patients suspected of pulmonary embolism. If perfusion scan is normal, we stop anticoagulation therapy and look for an alternative diagnosis. However, if a strong clinical suspicion of pulmonary embolism persists, then we always perfurm pulmonary angiography—the gold standard for diagnosis ofpulmonary embolism. We recommend that this policy he considered in such situations.

negative

pulmonary

angiograms.3'4

Thus,

pulmonary

angiog

raphy is unlikely to provide added benefit if the perfusion scan is normal, other than in the exceptional patient. Both Moser@and Secker-Walk& have suggested the most impor tant concept underlying the modern approach to venous thrombo embolism is that pulmonary embolism is not a disease; rather, embolism is merely a complication of deep venous thrombosis.5.6 Recent data7 in patients with pulmonary embolism suggest that if proximal-vein thrombosis is absent, clinically-evident, recurrent venous thromboembolism is rare. An alternative to performing pulmonary angiography is to perform objective testing for proximal vein thrombosis. Noninvasive objective tests for proximal-vein thrombosis (such as impedance plethysmography or B-mode venous ultrasound) are increasingly available and can be performed serially. If proximal-vein thrombosis is detected, therapy is indicated.― If serial testing is negative and the patient has adequate cardiorespi ratory reserve, anticoagulant therapy can he safely withheld.7

M. Z. Shaheen, M.R.C.P[U.K.], and Wj Wzndebank, F.R.C.P,

Russell D. Hull,M.B.B.S.,M.Sc.,F.C.C.?,and Gary E. Raskob, M.Sc.,

University of Calgary, Calgary, Canada CHEST I 98 I 4 I OCTOBER, 1990

Downloaded From: http://journal.publications.chestnet.org/ by Arn Eliasson on 02/23/2016

1047