| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Thoracic Imaging |
1 From the Departments of Radiology (Y.W.C., S.C.J., H.S.S., C.K.H., K.B.J.) and Internal Medicine (S.S.P.), Hanyang University Seoul Hospital, 17 Haengdang-dong, Sungdong-ku, Seoul 133-792, South Korea; and Department of Radiology, Hanyang University Kuri Hospital, Kuri, Kyungki-do, South Korea (C.K.P.). Received July 26, 2001; revision requested September 24; revision received December 20; accepted January 22, 2002. Address correspondence to Y.W.C. (e-mail: ywchoi@hanyang.ac.kr).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: In 16 patients, follow-up chest radiographs (n = 16) and initial (n = 2) and follow-up (n = 9) computed tomographic (CT) scans of the chest were retrospectively reviewed by two radiologists. Patient records (n = 16) and results of percutaneous needle aspiration and/or biopsy (n = 6) were reviewed by one radiologist.
RESULTS: Eighteen episodes of new lesion development were identified on radiographs in 16 patients. Each episode showed single (nine of 18 episodes, 50%) or multiple (nine of 18 episodes, 50%) nodules, most of which were in the peripheral lung (16 of 18 episodes, 89%) ipsilateral to the side of previous effusion (17 of 18 episodes, 94%). On CT scans, all lesions were peripheral pulmonary nodules, not round atelectasis. Needle aspiration and/or biopsy of the lesions showed findings consistent with tuberculosis in all six patients. Lesions usually evolved within 3 months after the start of medication (13 of 18 episodes) and finally disappeared (15 episodes) or left residual opacities (three episodes) 3–18 months later, with continuation of medication.
CONCLUSION: New lung lesions that develop during medication for tuberculous pleural effusion should be considered a transient worsening that ultimately improves with continuation of medication.
© RSNA, 2002
Index terms: Lung, diseases, 60.234 • Lung, effects of drugs on Pleura, diseases, 66.234 • Pleura, infection, 66.234 • Tuberculosis, pulmonary, 60.234
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
We have noted patients with a paradoxical response to antituberculous therapy for tuberculous pleural effusion. Although tuberculous pleural effusion had initially responded to medication with resolution of the pleural fluid, new nodular opacities subsequently developed in the lung but finally cleared with continuation of the original drug regimen in these patients. The purpose of our study was to evaluate our experience with patients who had a paradoxical response (development of new opacities) to treatment for tuberculous pleural effusion not related to acquired immunodeficiency syndrome (AIDS).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Patients received a diagnosis of tuberculous pleural effusion if there were tubercle bacilli in the sputum, pleural fluid, or pleural biopsy specimen; if there were granulomas in the pleura; and if there was radiographic pleural effusion without an alternative explanation for the presence of effusion. A diagnosis was also made in the absence of a positive culture result if the patient had undiagnosed lymphocyte-dominant exudative pleural effusion and if all clinical and radiographic abnormalities resolved after antituberculous chemotherapy (8).
In 135 of the 141 patients, antituberculous treatment started less than a week after radiographic detection of pleural effusion. In 133 of the 141 patients, the drug regimen was an orally administered combination of isoniazid (Yuhan-zid; Yuhan, Seoul, Korea), rifampin (Yuhan), ethambutol (Myambutol; Yuhan), and pyrazinamide (Yuhan) for the entire period of medication (41 patients) or the same drug combination for the first 2–5 months, followed by a combination of isoniazid and one or two other drugs for an additional 7 months or longer (92 patients). Conventional radiographic follow-up was usually performed at least once a week during the 1st month and at 1–3-month intervals thereafter.
One chest radiologist (Y.W.C.) retrospectively reviewed initial and follow-up posteroanterior and lateral chest radiographs of the 141 patients to select those who initially responded to chemotherapy with a decrease of pleural fluid but who subsequently developed new lesions in the lungs. Patients were excluded from the study if new lesions were detected just with resolution of pleural effusion on chest radiographs, because the lesions were considered to have originally been hidden behind the effusion. Also excluded were patients who had not been followed up until the lesions disappeared or stabilized without change in size for at least 3 months on chest radiographs. The study criteria were met by a total of 16 patients (14 men and two women) with a mean age of 27 years (range, 22–33 years). The study included three patients who showed focal opacities with resolution of pleural effusion in the area where the new lesions developed later, because the former lesions were far smaller than the latter. The diagnosis of tuberculosis was confirmed in these 16 patients: acid-fast bacilli were found in sputum and pleural fluid in one and two patients, respectively; histopathologic findings in pleural tissue were consistent with tuberculosis in four patients; and lymphocyte-dominant exudative pleural effusion resolved with antituberculous therapy in all but one patient in whom pleural fluid was not aspirated (patient 1).
Chest Radiography
All 16 patients underwent chest radiography for 12–27 months (mean, 19 months) after initial detection of pleural effusion. Two chest radiologists (Y.W.C., S.C.J.) simultaneously reviewed all chest radiographs; findings were recorded by means of consensus. Initial chest radiographs that showed pleural effusion were reviewed to determine whether the effusion was unilateral or bilateral. Presence or absence of a concomitant pulmonary lesion or mediastinal or hilar lymphadenopathy was determined by reviewing initial and follow-up chest radiographs obtained before free pleural fluid had disappeared. When a lesion was present, its lateralization (ipsilateral or contralateral) to the side of effusion was noted. Appearance and location of the lesion were described, along with consistency with known radiographic features of pulmonary tuberculosis (9).
Follow-up chest radiographs were reviewed for detailed features of the new lesions, which developed after disappearance of free pleural fluid. When recognized, number (one or more), shape, and laterality (ipsilateral or contralateral to the side of initial effusion) of the new lesions were determined. When the new lesions were maximally grown, their diameters were measured. The origin of the new lesions was classified as pulmonary, pleural, or extrapleural by using previously established radiologic criteria (10–13). A lesion was considered of pulmonary origin when it showed a circumferentially defined border, a heterogeneous appearance with air bronchograms, or acute pleural angles (13). A lesion was considered pleural or extrapleural when it showed convexity toward the lung and a sharply marginated border with tapering upper and lower margins; other associated pleural changes were thought to suggest pleural origin. A lesion with rib involvement was considered extrapleural in origin. We noted lesions that showed the incomplete border sign. Although the incomplete border sign was originally used to describe extrapleural masses (10), in our study its use was extended to imply pleural lesions and peripheral pulmonary lesions abutting the pleura, as well as extrapleural lesions. Lesions with the incomplete border sign were recorded as pleural or extrapleural when the recognized borders were sharply marginated or as pulmonary when the borders were not sharply marginated (10–13). A lesion that was difficult to classify was considered uncertain in origin. Pulmonary lesions were further classified as peripheral, central, or undetermined according to their relative location within the lung: Peripheral lesions were those within 1–2 cm of the nearby pleura or abutting the pleura; central lesions were those more than 1–2 cm away from the pleura (14); undetermined lesions were those that were difficult to categorize as either. Pulmonary lesions associated with the incomplete border sign were classified as peripheral.
Whether the new lesions disappeared completely or left residual opacities that did not change for 3 or more months was also recorded. The appearance of the residual opacity was described. Time intervals between initial detection of effusion and subsequent events (ie, initial recognition of new lesions, maximum development in size, and final outcome) were recorded. New lesions appearing in different locations from prior lesions after an interval of more than 1 month were considered to be in different episodes of new lesion development. In patients with more than one episode, time intervals between each episode were also recorded. If a pulmonary or rib lesion or lymphadenopathy was present when pleural effusion was initially detected, change in size was assessed during follow-up.
Computed Tomography
Computed tomographic (CT) scans obtained to assess initial pleural effusion and subsequent new lesions were available in two and nine patients, respectively. Patients 2 and 6 underwent CT at both initial pleural effusion and subsequent new lesion development. CT was performed at full inspiration with a CT/T 9800 system (GE Medical Systems, Milwaukee, Wis) in eight patients (10 scans) and with a Somatom Plus 4 scanner (Siemens Medical Systems, Erlangen, Germany) in one patient. Among these patients, one initial and four follow-up CT examinations were performed both before and after the administration of contrast media. The other six CT examinations were performed only after the administration of contrast media. Contiguous 5- (n = 1) or 10-mm (n = 10) sections were obtained through the chest during intravenous drip infusion of 50 mL of iopromide (Ultravist 300; Schering, Berlin, Germany) following bolus injection of 50 mL in 10 examinations or 100 mL in the remaining examination. In five patients, additional 1.5-mm-collimation CT scans were obtained at 5- or 10-mm intervals in selected areas. All scans were printed with both mediastinal (window width, 350–450 HU; window level, 20–35 HU) and lung (window width, 1,500 HU; window level, -700 HU) windows.
Two chest radiologists (Y.W.C., S.C.J.) reviewed all chest CT scans simultaneously; findings were recorded by means of consensus. Initial CT scans obtained to evaluate pleural effusion were reviewed for any associated abnormality in the lung, hilum, mediastinum, or chest wall. With reference to follow-up radiographs, we tried to determine which lesion, when present, was in the area where the new lesion subsequently developed. When the lesion was pulmonary, its appearance, enhancement pattern, and relative location within the lung were recorded. Follow-up CT scans obtained to evaluate new lesions were reviewed to assess the origin of the lesions. The origins were classified as pulmonary, pleural, or extrapleural by using previously established radiologic criteria (15,16). Lesions with the split pleura sign were classified as pleural in origin. Lesions obtusely abutting the chest wall were classified as pleural or extrapleural; among them, those associated with rib destruction were recorded as extrapleural. Lesions were considered of pulmonary origin when the normal lung interposed between them and the pleura or when they acutely abutted the pleura. Pulmonary lesions were further divided into peripheral and central lesions according to the same criteria used in the radiographic classification. Presence or absence of contrast material enhancement was also recorded. When only contrast-enhanced CT scans were available, enhancement of the lesion was considered present when its attenuation exceeded that of the skeletal muscle. When present, the enhancement pattern (homogeneous or inhomogeneous) of the lesion was also recorded. It was then determined whether lung lesions had previously published CT features of round atelectasis (17,18). CT scans were also assessed for lung lesions not recognized on corresponding chest radiographs. In addition, presence and appearance of pleural, hilar, mediastinal, or chest wall lesions were described.
Clinical Examination
Six of the 16 patients underwent percutaneous needle aspiration and/or biopsy of the new lesions with fluoroscopic (n = 5) or ultrasonographic (n = 1) guidance. One author (C.K.P.) reviewed medical records for microbiologic and histopathologic findings of needle aspiration and/or biopsy. Additional information, when available, was also obtained from medical records: history of tuberculosis, diabetes mellitus, AIDS, steroid medication, or compliance with medication before recognition of new lesions; results of sputum examination; changes in symptoms around the day of new lesion development; and change and duration of the treatment regimen after recognition of new lesions.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
|
A total of 18 episodes of new lesion development were recognized in 16 patients. Two episodes occurred in two patients (patients 15 and 16) each, when new lesions that developed in different locations from prior lesions more than 1 month apart were considered to be in different episodes. Each episode showed single (nine of 18 episodes, 50%) (Figs 1, 2) or multiple (nine of 18 episodes, 50%) (Fig 3) lesions. In the two patients who experienced two episodes each, the lesions were single in the first episode and multiple in the second episode in patient 15 (Fig 3) and multiple in both episodes in patient 16 (Table 2).
|
|
|
|
|
|
|
|
|
|
|
|
Most lesions appeared pulmonary in origin (17 of 18 episodes, 94%). Thirteen episodes showed only pulmonary lesions, and four episodes showed both pulmonary lesions and lesions of uncertain origin. The last episode demonstrated a single lesion of uncertain origin (Fig 2). The lesions of uncertain origin were seen in five episodes, and none of them showed associated rib abnormality. Pulmonary lesions, seen in 17 episodes, were mostly peripheral: They were all peripheral in 15 episodes and partly peripheral and partly undetermined in one episode; the last episode showed only an undetermined pulmonary lesion. No pulmonary lesions appeared to be definitely central (Table 2).
Immediately before the new lesions were identified, chest radiographs demonstrated either no abnormality (eight of 18 episodes, 44%) (Fig 2a) or residual pleural thickening or loculated effusion (10 of 18 episodes, 56%) with (three episodes) (Fig 1b) or without (seven episodes) (Fig 3a) adjacent focal pulmonary lesions in the area (Table 2). In the three episodes with associated pulmonary lesions, the lesions were, however, far smaller in size than the subsequent new lesions (Fig 1).
In all but one episode, the new lesions appeared ipsilateral to the side of previous pleural effusion (17 of 18 episodes, 94%) (Figs 1, 2); in the other episode, the patient developed bilateral lesions after disappearance of right pleural effusion (Fig 3). In 16 of 18 episodes, the new lesions were unilateral (Table 2).
The new lesions were usually recognized within 3 months after detection of pleural effusion, as seen in 13 of 18 episodes (72%), with a range of 1–9 months (mean, 3 months). The lesions were maximal in size at the time of first recognition in 13 of 18 episodes, and lesions in the other five episodes grew for 1–3 months after that. After maximally grown, the new lesions gradually decreased in size over 3–18 months (mean, 9 months) and finally disappeared (15 episodes) or left some residual lesions that did not change for more than 3 months (three episodes). Residual lesions appeared nodular in one episode and bandlike in the other two. In patients 15 and 16 (who experienced two episodes), second-episode lesions were recognized 5 and 6 months after the first episodes, respectively (Table 2). At that time, first-episode lesions were decreasing in size (patient 15) (Fig 3) or had already disappeared (patient 16).
At diagnosis of pleural effusion, concomitant pulmonary lesions were seen in eight of 16 patients (50%) (Fig 3a). All lesions (100%) were in the lung apex and appeared nodular in seven patients and infiltrative in one, and all were consistent with radiographic features of postprimary tuberculosis (9). In seven of these eight patients, the pulmonary lesions were ipsilateral to the side of effusion. The remaining patient had bilateral pulmonary lesions associated with right pleural effusion. In five of these eight patients, the original pulmonary lesions displayed radiographic improvement after antituberculous treatment and were thought to be active pulmonary tuberculosis (Fig 3). In the other three patients, the lesions showed no radiographic change and were thus considered to be stable or inactive (Table 2). Recurrence of pleural effusion or lymph node enlargement was not noted in any patient throughout the follow-up period.
Computed Tomography
Initial chest CT scans obtained to evaluate pleural effusion in two patients showed multiple areas of pulmonary consolidation, which was also noted in the area where the new lesion subsequently developed. All of the pulmonary lesions were homogeneously enhanced and adjacent to pleural effusion and were thus considered pulmonary atelectasis caused by pleural effusion (Fig 1). There was no abnormality in the hilum, mediastinum, or chest wall in any patient (Table 2).
Follow-up CT scans obtained to assess the new lesions that developed during follow-up in nine patients demonstrated that all lesions were peripheral pulmonary nodules or masses. Even uncertain (patients 1, 12, and 14) (Fig 2) and undetermined pulmonary (second episode of patient 15) (Fig 3) lesions on chest radiographs appeared to be of peripheral pulmonary origin on CT scans. In seven patients, all of the new lesions were heterogeneously enhanced (Figs 1–3); in the other two patients, both homogeneously and heterogeneously enhancing lesions were seen (patients 14 and 16) (Table 2). No lesions showed internal air bronchogram or adjacent tree-in-bud appearance. Lesions did not appear to be rounded atelectasis because they had areas of internal low attenuation and did not show characteristic CT features of rounded atelectasis, such as vessels and bronchi curving into and sweeping around the lesions.
Most lesions abutted the normal (n = 1) or thickened (n = 8) pleura. Pulmonary lesions separate from the pleura were seen in three patients. In four patients with multiple lesions, pulmonary lesions abutting the thickened pleura were associated with those abutting the normal pleura (n = 1) or separate from the pleura (n = 3) (Table 2). In two patients (patients 15 and 16), CT depicted additional small peripheral pulmonary lesions that were not recognizable on chest radiographs, even at retrospective review. The thickened pleura that the new lesions abutted enhanced well and had areas of internal low attenuation, suggesting pleural empyema in all patients. No images showed associated hilar, mediastinal, or chest wall lesions.
Clinical Findings
No patients had a history of medication for tuberculosis. No patients had diabetes mellitus, AIDS, or a history of steroid medication or alcoholism.
All patients started to take antituberculous medication less than a week after detection of pleural effusion. All but one patient were considered to have been compliantly taking the medication until the new lesions were detected on chest radiographs. Four patients were in the hospital during the entire period, from admission until appearance of the paradoxical response. During this time, medication was given under the supervision of a nurse to be certain that the antituberculous drugs were taken. In addition, each of these 15 patients and their family members who lived with them insisted on the patients’ compliance in taking the medication. The remaining patient showed a newly developing lesion 3 months after completion of 6-month medication, and then medication started again (patient 1).
Applied therapy protocol was usually a combination of orally administered isoniazid, rifampin, ethambutol, and pyrazinamide for the entire period of medication or the same drug combination for the first 2–5 months, followed by a combination of isoniazid and one or two other drugs for the rest of the time. No other medication was given that might have influenced the response of the pulmonary lesions.
There was moderate to marked clinical improvement, with amelioration of symptoms and a decline of fever in all patients when the new lesions developed. Results of sputum smears were negative for acid-fast bacilli throughout follow-up in 15 patients; in the remaining patient (patient 1), the result was positive at admission due to pleurisy, became negative a month later, and was consistently negative thereafter. When the new lesions were recognized on chest radiographs, sputum examinations did not reveal any bacteriologic findings that might have had an etiologic relationship with the lesions in any patient. Only two patients (patients 1 and 3) developed symptoms (cough and sputum) that might have been attributable to those new lesions. There was no sign of failure of the drug regimen in any patient.
Even after detection of the new lesions, all patients continued the antituberculous medication without changing the regimen until resolution or stabilization of the lesions. The duration of medication ranged from 12 to 27 months (mean, 17 months). The medication was discontinued 6 months before (n = 1), at the time of (n = 3), or 1–10 months (mean, 5 months) after (n = 12) disappearance or stabilization of the new lesions.
Of the 16 patients, two experienced side effects of the medication: Patient 6 had toxic hepatitis caused by pyrazinamide 2 months after the start of medication (ie, 3 months before recognition of the new lesions), and the patient took levofloxacin (Levofexin; Il Dong, Seoul, Korea) instead thereafter; patient 8 experienced drug fever caused by rifampin 2 months after the start of medication (ie, 7 months before recognition of the new lesions), and the drug was discontinued for the following 3 months and then reinstituted afterward.
Six patients underwent fine-needle aspiration and/or biopsy of the new lesions. Pathologic examination of the specimens revealed acid-fast bacilli in one, findings consistent with tuberculosis in four, and both findings in one, demonstrating that all lesions were tuberculous (Table 1).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
In our patients, paradoxical response to antituberculous chemotherapy for non–AIDS-related tuberculous pleural effusion was manifested as single or multiple nodules developing after resolution of pleural effusion. Some patients experienced two episodes of paradoxical response, appearing at an interval of up to 6 months. The paradoxical response was observed in 16 of 141 patients (11%) with tuberculous effusion. CT revealed that all of these new lesions were peripheral pulmonary nodules or masses that mostly abutted the normal or thickened pleura. They usually evolved within 3 months after an initiation of antituberculous medication. When the same medication was continued, they subsequently decreased in size, finally disappearing or leaving some residual lesions 3–18 months later.
The theory that the lung lesions developed as a paradoxical response is strongly supported by the biopsy results, which revealed acid-fast bacilli and/or pathologic findings compatible with tuberculosis, and by the ultimate resolution of the lesions with antituberculous medication. The paradoxical response observed in the present study is unlikely to result from lack of patient compliance in the regular taking of medication, although this possibility still remains in 11 patients, whose statements were the only evidence of their compliance. The radiographic improvement of pleural effusion and original pulmonary tuberculosis before the paradoxical response discounts this possibility.
The new lesions were not those previously hidden behind pleural effusion, because we included only the patients in whom new lesions developed in the area where, before recognition of the lesions, pleural lesions disappeared or decreased in size so that subsequent changes could be easily identified. Although CT scans obtained at the time of pleural effusion in two patients showed peripheral pulmonary consolidations adjacent to pleural effusion, they were considered pulmonary atelectasis associated with pleural effusion, because they were homogeneously enhanced and mostly disappeared later, along with pleural effusion (Fig 1). We cannot say for sure, however, that even incipient lesions of the new lesions were absent behind pleural effusion. Actually, after disappearance of most pleural lesions, three patients (including the two with initial CT scans) showed focal pulmonary opacities in the area where paradoxical response developed later. Although these focal opacities might be the incipient lesions, they were, however, far smaller than the subsequent new lesions.
Paradoxical response in the present study was somewhat similar to that in Bobrowitz’s study (27), in which seven patients developed transient aggravation of lung lesions during initial treatment of pulmonary tuberculosis. However, there are distinct differences between the two studies. First, progression patterns in the two studies are different: There were coalescent mottled areas of infiltration compatible with bronchogenic dissemination of active tuberculosis in Bobrowitz’s study (27), versus peripheral pulmonary nodules in the present study. Second, as compared with pulmonary tuberculosis in Bobrowitz’s study, the initial disease for which antituberculous medication was administered was tuberculous pleural effusion in the present study. Although five patients in our study also had active pulmonary tuberculosis at the time of pleural effusion, they showed the same worsening pattern—that is, peripheral lung lesions—as those without pulmonary tuberculosis. This suggests that the paradoxical response in the present study had nothing to do with the original pulmonary tuberculosis.
To the best of our knowledge, there has been only one report in the English language that is similar to ours (30). Cooke et al (30) reported three young patients with tuberculous pleural effusion who developed peripheral opacities that ultimately disappeared or decreased in size. A Japanese journal published more such cases (31,32). Suzuki et al (31) reported the largest series, in which 26 of 226 patients (11.5%) with tuberculous pleural effusion developed similar transient worsening. In the current study, 16 of 141 patients (11.3%) showed transient worsening. Single nodular lesions were more common, but multiple nodular lesions were not rare in their study, while both incidences were the same in the present study. In all of these studies (30–32), the opacities occurred within 4 months after starting antituberculous chemotherapy and disappeared within 13 months with continuation of the same chemotherapy. Most lesions appeared as subpleural nodules ipsilateral to the side of pleural effusion. These features mostly agree with our results. However, these studies included only few cases in which CT was performed (30–32).
We speculated on the nature of this paradoxical response. Tuberculous pleural effusion is thought to result from rupture of a subpleural pulmonary tuberculous focus into the pleural space, which allows tuberculoproteins to enter the pleural space and generate the hypersensitive reaction responsible for most of the clinical manifestations (3). Supporting evidence comes from the surgical findings of Stead et al (33), who reported that 12 of 15 patients with tuberculous pleuritis had concomitant parenchymal disease that was not obvious radiographically. Hulnick et al (34) also reported that CT showed parenchymal cavities in eight and infiltrates in another six of 14 patients with no obvious parenchymal abnormality on standard chest radiographs. Given this fact, concomitant subpleural pulmonary tuberculous lesions, if not radiographically evident, are not unexpected in patients with tuberculous pleural effusion. If one goes a step further, one can reasonably expect that these radiographically invisible lesions subsequently undergo transient worsening and become radiographically evident, despite antituberculous medication, as in other reported cases with paradoxical response (20–29). This speculation is supported by the facts that all new lesions in the present study were in the subpleural lung and that most of the new lesions arose ipsilateral to the side of pleural effusion.
The mechanism of this paradoxical response remains unclear. It has been suggested that active tuberculosis can result in immunosuppression through an altered cell-mediated response. Once active tuberculosis is under control after appropriate therapy, enhanced focal immune response (immunologic rebound) will recruit lymphocytes and macrophages at the site of these lesions, which then enlarge and become evident (35). Hypersensitivity to tuberculoproteins released from the dying mycobacteria may be a factor, as well (36).
The peripheral pulmonary nodules developing as a paradoxical response in the present study should be differentiated from nontuberculous infection, rounded atelectasis, and malignancy. In the current study, the possibility of nontuberculous infection was excluded by means of intensive microbiologic tests that demonstrated no pathologic organisms other than tuberculosis. Rounded atelectasis may develop after resolution of pleural effusion, and characteristic CT findings, such as vessels and bronchi curving into and sweeping around a mass, are helpful in making a correct diagnosis. In the present study, however, no case showed such imaging features, and most lesions had internal low-attenuating areas suggestive of necrosis; both of these indications discount the diagnosis of rounded atelectasis. The possibility of malignancy should also be considered in patients with newly developing lesions. It is worth mentioning, however, that all of our patients were in their twenties or thirties, when malignancies are not usually expected.
In the present study, percutaneous needle aspiration and/or biopsy of the newly developed lesions was performed in six of 16 patients because we were not aware of the paradoxical response at that time. With the results of the present study in mind, however, it appears to be better for the procedure to be reserved for patients who are 40 years of age or older or who show findings different from those in our patients, such as nonperipheral lung lesions.
Our study has several limitations. First, only two and nine of 16 patients of this study underwent CT at the time of pleural effusion and paradoxical response, respectively. Second, percutaneous needle aspiration and/or biopsy of the new lesions was performed in only six of 16 patients. Third, compliance with medication before the recognition of paradoxical response was not ascertained in all patients. These limitations resulted from the retrospective nature of this study. We believe, however, that these limitations are not substantial enough to disprove the conclusion of this study.
In summary, results of this study demonstrate that there are patients who develop new pulmonary lesions during medication for tuberculous pleural effusion, which implies not failure of the current therapeutic regimen but rather transient worsening of tuberculosis that subsequently disappears with continuation of the same medication. It is important to recognize this clinically impressive but benign paradoxical pulmonary response and to avoid unnecessary invasive procedures or changes of current appropriate therapy.
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, Y.W.C.; study concepts and design, Y.W.C.; literature research, H.S.S., Y.W.C.; clinical studies, S.S.P.; data acquisition, S.S.P.; data analysis/interpretation, Y.W.C., C.K.H.; statistical analysis, Y.W.C.; manuscript preparation, Y.W.C.; manuscript definition of intellectual content, Y.W.C., H.S.S.; manuscript editing, Y.W.C.; manuscript revision/review, S.S.P., C.K.P., K.B.J.; manuscript final version approval, Y.W.C., S.C.J.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y. J. Jeong and K. S. Lee Pulmonary Tuberculosis: Up-to-Date Imaging and Management Am. J. Roentgenol., September 1, 2008; 191(3): 834 - 844. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Pilate and Y. W. Choi Tuberculous Pleural Effusion [letter] * Dr Choi responds: Radiology, April 1, 2004; 231(1): 283 - 284. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
