Utility of Repeated Fiberoptic Bronchoscopy for Suspected Malignancy

Utility of Repeated Fiberoptic Bronchoscopy for Suspected Malignancy* Kenneth G. Torrington, LTC, MC, F.C.C.R; and Ronald K. Poropatich, MAl, MC, F.C.C.R We retrospectively evaluated records of 1598 fiberoptic bronchoscopies (FBs) performed on 1,391 patients (PTs) between Jan 1, 1986 and Dec 31, 1990. We found a progressive increase from 11 percent to 20 percent in the use of repeated fiberoptic bronchoscopy (RFB). Of the 254 RFBs, 151 were done in PTs with known or suspected intrathoracic malignant neoplasms. The 78 (of 151) RFBs performed in PTs with previously diagnosed malignant neoplasms were used to guide additional therapy. The other major indication for RFB (67 of 151) was to evaluate new suspicious lesions that had not been diagnosed on the initial FB. RFB specimens were positive in 36, false-negative in 24, and true-negative in 7 PTs. For some PTs, RFB could probably have been avoided if at initial FB physicians had

fiberoptic bronchoscopy (FB) was introduced Since as a diagnostic procedure in 1968,1 a variety of

diagnostic and therapeutic indications have evolved. 2 Most review articles have not addressed the indications for repeated fiberoptic bronchoscopy (RFB) and report only patients (PTs) who have undergone one procedure. 3 ,4 Two groups of investigators have studied the utility of repeated procedures for PTs infected with the human immunodeficiency virus. 5 ,6 In 1982, Ackart et aF described a series of 1,225 outpatients who were examined with 1,428 FBs; however, they did not provide data on indications or yield of the 203 repeated procedures. 7 During the last several years, we noted that increasing numbers ofRFBs were being performed at our institution, which is an active teaching hospital. We undertook this five-year retrospective review of FB to evaluate the overall use of both FB and RFB, and to analyze RFB's utility for assessing PTs with suspected carcinoma. METHODS

Since institution of FB at Walter Reed Army Medical Center, Washington, DC, in the early 1970s, carbon copy bronchoscopy reports have been completed following every procedure. For this retrospective revie\\~ all FB reports for the five-year period of 1986 through 1990 were evaluated to determine each year's total number of FBs, number of RFBs, percentage of RFBs, total number of PTs, number of PTs having RFB, and percentage of PTs having RFB (Table 1). Bronchoscopy reports from the 207 PTs examined with 254 RFBs were further studied to determine the indication fi>r each repeat *From the Pulmonarv and Critical Care Medicine Service, Walter Reed Army Medical" Center, \Vashington, DC. This work \\'as supported by \VRAMC protocol 1747. Manuscript received January 17, revision accepted February 24. Reprint requests: Dr. Torrington, Pulnwnary Division, Walter Reed Anny Aledical Center, Washington, DC 20307-5000

1080

(1) used fluoroscopy to direct transbronchiallung biopsies

in PTs expected to have normal airways, (2) performed transbronchial needle aspiration in all PTs with extraluminal disease or mediastinal adenopathy, and (3) obtained bronchial biopsy specimens from all PTs with endobronchial lesions. In PTs whose initial FB specimens were nondiagnostic despite visualization of endobronchial or extraluminal abnormalities, RFB was associated with a significant diagnostic yield and obviated the need for more morbid, surgical (Chest 1992; 102:1080-84) staging procedures.

=

=

6beroptic bronchoscopy; RFB repeated fiberoptic bronchoscopy; TBBx transbronchial lung biopsy; TBNA transbronchial needle aspiration

FB

=

=

procedure (Table 2). The most frequent indication for RFB, evaluation for suspected bronchogenic or metastatic carcinoma, was present in 118 PTs (57 percent) \\rho underwent a total of 151 (59 percent) RFBs. For these 118 PTs, we evaluated hospital records, duplicate files in Pulmonary and Thoracic Sllr~ery Clinics, Tumor Board records, and pathology reports to determine the precise indication f(>r each repeat procedure as well as to learn each PT's subsequent diagnosis and eourse. Patients were evaluated by year group (Table 3) and were then divided into the following categories: A, proven cancer; diagnostic first FB, se<.'ond FB to reassess anatomy or nile out recurrence; B, proven cancer; therapeutic procedure (secretion removal, abscess drainage, tumor debulkinglbrachytherapy); C, proven cancer; nondiagnostic first FB, positive second FB; D, proven cancer; hoth FBs n()ndia~n()stic (false-ne~ative); E,

Table 1- Total Number of Patients Undergoing Bronchoscopy and Initial and Repeated Fiberoptic Bronchoscopies Performed (1986-1990)*

1(>tal No. of FB No.ofRFB RFB,% Total No. of PTs No. of PIs with RFB PIs \\'ith RFB, %

1986

1987

1988

1989

1990

Total

225

293 37 13 262 31 12

331 47 14 291 40 14

348

401 81 20 340 61 18

1,598 254

25

11 201 24 12

64

18 297 51 17

1,391 207

*FB = Fiberoptic bronchoscopy; RFB = repeated FB; PI = patient.

Table 2-lndicationfor All Repeated Fiberoptic Bronchoscopies (RFB)

Suspected carcinoma Infection in immunoconlpromised host Ilemoptysis Sarcoidosis Miscellaneous

No. of RFB

%

151 46 16 7 34 254

59 18 6 3 14 100

Utility of Repeated FOB for Suspected Malignancy (Torrington, Poropatich)

Table 3-Number of Repeated Fiberoptic Bronchoscopies (RFB) per Patient (PT) (1986-1990)

Total No. of PTs No.ofRFB No.ofRFB/PT

1986

1987

1988

1989

1990

Total

12 14 1.2

20 26 1.3

28

27

31 45 1.5

118 151

31 1.1

35

1.3

proven cancer; dia~nostic first FB, second FB for mediastinal staging or to obtain specific tumor cell type; and F, no cancer present; both FBs nondia~nostic (true-negative). These data are presented in Table 4. For 151 RFBs, the interval between repeated procedures was measured and PTs were placed in three W'0ups of <1 month, 1 to 6 months, and >6 months (Table 5). The following characteristics were identified as important and were also studied: (1) the presence/absence at FB of endobronchial tumor, extraluminal tumor, or normal airway anatomy; (2) the diagnosis of small cell carcinoma; (3) the indication and yield of transbronchial needle aspiration (TBNA); (4) the use of fluoroscopy; and (5) the number of PTs with pulmonary metastases. We also noted the frequency of "suspicious, but nondiagnostic of malignancy" patholoJ.n· reports, hut considered such specimens negative for the purpose of the study. RESULTS

Between Jan 1, 1986 and Dec 31, 1990, 1,391 PTs underwent a total of 1,598 FBs at Walter Reed Army Medical Center. A total of 207 PTs had two or more FBs. Table 1 demonstrates progressive increases in all measurements during the five-year interval. While the annual number of FBs performed nearly doubled between 1986 and 1990, the percentage of RFBs increased even more rapidly. In 1990, 18 percent of all PTs were submitted for RFB and 20 percent of all FBs were repeated procedures. Of the 207 PTs with RFB, the majority (57 percent or 118 PTs) were performed in PTs suspected of having bronchogenic or metastatic carcinoma. These 118 individuals \vere evaluated with 151 (59 percent) of the 254 RFBs. Other indications for RFB included suspected infection in immunocompromised hosts, hemoptysis, sarcoidosis, and miscellaneous conditions (Table 2). Evaluations of the 118 PTs with suspected carcinoma are presented in Tables 3 and 4. Table 3 demonstrates annual increases in the number of RFBs in these PTs and in the number of RFBs per PT. Table 4 reveals that the increase in RFB occurred primarily Table 4-lndicationfor Repeated Fiberoptic Bronchoscopy in Suspected Cancer Patients (See Text for Description of Groups A through F)

A

B C D E F Totals

1986

1987

1988

1989

1990

lhtal

5

7 2 11

14 1 7

20 8 7

65

I 0

1 2 31

19 2 7 3 1 3

0

4 4 1

0

14

5

26

6

35

6

2

2

45

13

36 24 6

7 151

Table 5- Time Interval Between Initial and Repeated Fiberoptic Bronchoscopy in AU Suspected Cancer Patients (1986-1990)

<1 mo 1-6 mo

>6mo Totals

1986

1987

1988

1989

19(j()

9 1 4 14

15

19 2 10 31

12

23

78

10

R 14

44

8

3

26

13 35

45

l()tal 29

151

because of two indications: (1) follo\\'-up FB of PTs with known cancer, and (2) increased use of therapeutic interventions. There were 49 group A PTs, all of whom had been previously diagnosed as having one or more malignant neoplasms, who underwent 65 RFBs. Forty-two had a prior diagnosis of bronchogenic carcinoma and eight metastatic malignancies (one PT had both renal cell and bronchogenic carcinoma). Fifty of the procedures occurred >6 months after the initial bronchoscopy. Sixty-one RFBs were performed to reevaluate the airways to plan therapeutic interventions, ",hile four were completed as part of a postchemotherapy protocol for smaIl cell bronchogenic carcinoma. Airway anatomy revealed endobronchial lesions (manifested by endobronchial mass, lobar or segmental occlusion, friable mucosa, or mucosal nodularity) in 20 PTs, extraluminal disease (characterized by submucosal edema ± erythema, carinal widening, or extrinsic compression, distortion, or fixation ofthe tracheobronchial tree) in 19 PTs, both endobronchial and extraluminal disease in two PTs, and nornlal findings in 24 PTs. Pathologic specimens were submitted from 45 procedures and nine were positive for malignancy. All five TBNAs from PTs with extraluminal disease were negative. The one positive TBNA was obtained by aspiration of a right paratracheal mass. Group C totaled 36 PTs, of whom 34 required hvo FBs and two required three FBs to prove the presence of malignancy. Of the 38 repeated procedures, 31 were performed within one month and all were performed within six months. Sixteen PTs had endobronchial tumor, 25 had extraluminal tumor, and nine had evidence of both endobronchial and extrahllninal tumor. In four PTs with totally norlnal airways, rualignancy was diagnosed by fluoroscopically guided transbronchial lung biopsy (TBBx). In the seven group C PTs diagnosed as having small cell carcinolua, all had abnormal bronchoscopies with endobronchial tunlor in four and extraluminal tumor in six. TBNAs were performed in 18 PTs. Ten PTs had negative TBNA on the first FB. Repeated TBNAs in eight of the ten \vere positive for malignancy in four. Three of eight PTs undergoing TBNA only on the second FB had positive diagnoses. Overall, seven of 18 PTs with extralulllinal tumor had positive TBNAs. The one TBNA perfornled in a PT with only endobronchial tUluor but no extraCHEST I 102 I 4 I OCTOBER, 1992

1081

luminal disease was negative. The 24 group 0 PTs each had two false-negative bronchoscopies. The presence of malignancy was subsequently verified in all by one of the following: transthoracic needle aspiration, surgical sampling of mediastinal lymph nodes, thoracotomy, or third FB (three PTs). Twenty PTs underwent RFB in < one month, three in one to six months, and one (who refused further evaluation until hospitalized with intracranial metastases) in > six months. Fifteen PTs had evidence of extraluminal tumor and three had concurrent endobronchial tumor. Of these 15 PTs, five had no TBNA, five had one negative TBNA, and five had two nondiagnostic TBNAs. A total of 13 PTs had TBNA performed at one or both FBs. Nine PTs had normal airway anatomy at FB. Despite using fluoroscopically directed TBBx in five of these PTs, a tissue diagnosis of malignancy was not made. Therefore, in the nine group C and D PTs with normal anatomy who underwent fluoroscopically guided TBBx, malignancy was diagnosed in four (44 percent). All three PTs subsequently proven to have small cell carcinoma had abnormal FBs with endobronchial disease in one and extraluminal disease in three (only one had TBNA on both FBs). Nonmalignant conditions were subsequently diagnosed in the seven group F PTs. Of these, three underwent RFB in < one month, two in one to six months, and two in > six months. Table 6 summarizes measurements of the interval between RFBs according to PT group. Most PTs (54167 or 81 percent) with new suspicious lesions (groups C, D, and F) had RFB within one month, whereas most PTs (47nS or 60 percent) with suspected recurrence or new primary lesions (groups A and B) were bronchoscoped > six months after their initial FB. Several other parameters were evaluated in the 118 suspected cancer PTs undergoing RFB. Small cell bronchogenic carcinoma was diagnosed in ten (8 percent) of 118; seven were diagnosed with FB and three PTs required additional surgical procedures. All had abnormal bronchoscopic examinations (five with endobronchial and nine with extraluminal tumor). Of the 118 PTs, nine were diagnosed as having metastatic Table 6- Time Interval between Initial and Repeated Fiberoptic Bronchoscopy (HFB) by Patient Groups (See Text for Description ofGroups A Through F)

A B C D E F lhtals

1082

<1 fin

1-6mo

8

11 5 5 3 1

7 31 20 5 3

74

2 27

>6

nlO

46 1 0 1 0

2

50

malignant neoplasms. Most PTs \vere in group A and had previous diagnoses of cancer. Findings at FB and positive yield of biopsy results did not differ benveen bronchogenic and nletastatic carcinonla PTs. In the 118 initial FBs and 151 RFBs perf()rnlcd in 118 PTs, only five significant conlplications occurred. Postbiopsy hemorrhages of 50 to 200 Inl developed in three PTs. One PT had a 20 percent pneumothorax that required treahnent \\rith a tube thoracostonly. One PT's nasal mucosa \vas injured by a nasotracheal tube advanced over the hronchoscope. There \vere no deaths. For the total 269 FBs, the calculated incidence of major complications \vas 1.9 percent. For the 118 PTs, six (5 percent) had specimens reported suspicious for malignancy.H In three PTs, the initial specimen \\ras "suspicious" and RFB confirmed Inalignancy. In the renlaining three PTs, initial specinlens \"ere negative. Specinlens frolll RFB \\'ere reported as "suspicious" for malignancy, and cancer \vas subsequently surgically confirlned. H

DISCUSSI()N

Review of Walter Reed Army Medical Center records confirmed our inlpression that RFB \'/as performed \\rith increasing frequency behveen 1986 and 1990. During the five-year interval, the annual number of RFBs gre,v fronl 25 to 81 and the percentage of total FBs rose frolll 12 to 20 percent. Since FB is an expensive and invasive procedure \vith a slllall risk of serious complications, \ve evaluated this trend to determine its cause(s) and to learn whether Illodification of our techniques could reduce repeated procedures. We report our analysis of the largest PI group, those suspected of having hronchogenic or Inetastatic

carcinoma. One hundred eighteen PTs underwent a total of 151 RFBs. Patients were divided into six categories. Group A constituted the largest category \vith 65 PTs with previously diagnosed Inalignant neoplasnls. Fifty RFBs \vere perfornled > six months after the initial diagnosis. This interesting group included PTs with second prinlary bronchogenic carcinolnas, locally recurrent or distantly Inetastatic hronchogenic carcinomas, and other extrapultnonary nlalignant neophlslns Inetastatic to the lung. Only four PTs had RFB as part of a formal protocol. Most PTs \\rere treated ,vith additional surgery, radiation therapy, or chelnotherapy follo\\,ing RFB. Clearly, FB is indicated to guide therapy in this group. Group B contained a small but increasing Ilulnber of PTs subjected to therapeutic FB. Techniques such as brachytherapy, electrocautery, or laser therapy have become accepted therapeutic interventions for appropriately selected PIs. RFBs are alnlost invariably required for such procedures. The large number of such PTs \\Till likely provide increasing numbers of Utility of Repeated FOB for Suspected Malignancy (Torrington, Poropatich)

candidates for bronchoscopic therapy in coming years. In 36 group C patients, RFB produced a tissue diagnosis of malignancy. Patients in this group had false-negative initial FB and true-positive RFB. Thirty-two of the 36 patients had endobronchial or extraluminal tumors. For ten patients, a tissue diagnosis might have been achieved in a single procedure, if endobronchial biops~ TBNA, bronchial brushings, and washings had all been performed during the initial procedure. Furthermore, the four PTs expected to have normal airway anatomy could have had their malignancies diagnosed in one FB, if fluoroscopically guided TBBx had been performed initially. On the other hand, we noted that 18 of 32 PTs with endobronchial or extraluminal cancers did not have their conditions diagnosed on the initial FB despite appropriately aggressive biopsy procedures. Such patients can benefit from RFB, because malignancy will likely be diagnosed on the second procedure. If small cell carcinoma or metastatic carcinoma is found (as it was in eight of 36 patients), definitive care can be based on FB results and more invasive surgical testing will be avoided. Therefore, we recommend RFB be performed in suspected cancer PTs when the initial FB reveals endobronchial or submucosal abnormalities, but histologic and cytologic findings of the initial specimens are not diagnostic. The 24 group D PTs had two false-negative FBs. These PTs were indistinguishable from group C PTs except that more had anatomically normal FB and fewer had endobronchial tumor. We believe that diagnostic yield of both initial FB and RFB would have been improved by performing fluoroscopically guided TBBx in all PTs with normal airway anatomy and TBNA in all with extraluminal tumor infiltration or mediastinal adenopathy, as previously recommended by Shure and Fedullo. 8 However, even if maximally aggressive biopsy techniques had been employed, one third of group D PTs would have required additional surgical procedures for definitive diagnosis. All seven RFBs in group E were preventable. In the majority of this group, routine chest roentgenograms suggested mediastinal lymphadenopathy, and RFB was performed to sample mediastinal lymph nodes with TBNA. We recommend that initial FB be delayed in such patients, until chest computed tomography has defined the location and extent of adenopathy for appropriate mediastinal staging during the initial FB. In the other group E PTs, a usecond look" RFB was performed to stage the extent of the endobronchial tumor. At teaching hospitals, occasional PTs will have to undergo repeated bronchoscopies because of physician inexperience. Finally, there were seven group F PTs who underwent RFB to diagnose suspected malignancy, but whose subsequent clinical

courses ruled out cancer. When we evaluated the time that had elapsed between initial and RFBs (Table 6), we found that 54 (81 percent) of 67 PTs with newly discovered lesions suspicious for cancer were evaluated within one month and 64 (96 percent) of 67 were evaluated within six months. On the other hand, 60 percent of PTs with previously diagnosed malignant neoplasms (47n8) underwent RFB > six months after their initial procedure. In this series of suspected cancer PTs undergoing RFB, ten (8 percent) of 118 had their conditions diagnosed as small cell bronchogenic carcinoma. All had abnormal anatomy visualized during bronchoscopy. Although this group is small, it reminds us that small cell bronchogenic carcinoma usually involves the central airways. Despite this, surgical nodal sampling was required to achieve a tissue diagnosis of malignancy in three of ten PTs. Pathologic specimens from six (2.2 percent) of the 269 FBs were reported as ususpicious" for malignancy. Because of the small size of bronchial biopsies and limited amount of cellular material present in TBNA, bronchial brushings and washings, we believe that our 2.2 percent incidence of ususpicious" specimens is acceptable and is coincidentally identical to a previous report. 9 Only five significant complications occurred in the 269 FBs performed on the 118 PTs. There was no difference in morbidity between initial FB and RFB. Specific complications included three postbiopsy hemorrhages (one on the initial FB and two on the RFB), one pneumothorax (on initial FB), and one nasal mucosal injury (on RFB). The calculated incidence of major complications was 1.9 percent, which compares favorably to the review of Fulkerson. 2 If more aggressive biopsies had been performed as we currently re(,~mmend, a higher rate ofcomplications might have occurred. However, the benefit of avoiding the risks inherent in general anesthesia and surgical procedures would justify possibly increased morbidit~ In summary, we have observed a steady increase in the incidence of RFB in the past five years. Most of the increase has occurred in PTs with previously diagnosed cancers (groups A and B), in whom RFB has been performed to guide additional therapy (eg, surgery, radiation, chemotherapy, or endobronchial procedures). As new and increasingly aggressive therapies for intrathoracic carcinoma continue to be described, we expect such indications for RFB to proliferate. On the other hand, the number of PTs requiring RFB to diagnose ne~ suspected intrathoracic malignant neoplasms (groups C, D, and F) did not change appreciably between 1986 and 1990. Our findings demonstrate that some patients could have heen CHEST I 102 I 4 I OCTOBER, 1992

1083

spared RFB, if optimal biopsy specimens had been obtained initially. We propose three specific guidelines when performing FB to evaluate PTs with suspected malignancy. (1) Fluoroscopy should be available to direct TBBx in patients likely to have normal airways (stage I lesions in PTs with no synlptoms of airways involvenlent). (2) Patients with extraluminal disease or mediastinal adenopathy demonstrated with computed tomographic scan should always undergo TBNA in addition to biopsy, brushings, and washings. (3) Patients \vith endobronchial lesions should be evaluated with aggressive bronchial biopsies together with brushings and washings.

4

5

6

7

8

REFERENCES 1 Ikeda 5, Yawai N, Ishikawa S. Flexible bronchofiherscope. ICeio J Med 1968; 17:1 2 Fulkerson \Vj. Fiberoptic bronchoscopy. N Engl J Med 1984; 311:511-15 3 Dreisen RB, Albert RIC. Talley PA, et aI. Flexible fiberoptic

1084

9

bronchoscopy in the teachin~ hospital: yield and complications. Chest 1978; 74:144-49 Abdulaziz A, Ibrahim EM, Ghassab A, et aI. Use of flexible fiberoptic bronchoscopy in pulmonary diseases. Ind J Chest Dis All Sci 1988; 30: 167-75 Colangelo G, Baughman R~ Dohn MN, Frame PT. Follow-up bronchoalveolar lavage in AIDS patients with PneuToocystis carinii pneumonia. Am Rev Respir Dis 1991; 143:1067-71 Barrio JL, Harcup C, Baier HJ, Pitchenik AE. Value of repeat fiberoptic bronchoscopy and significance of nondiagnostic bronchoscopic results in patients with the acquired immunodeficiency syndrome. Am Rev Respir Dis 1987; 135:422-25 Ackart RS, Foreman DR, Klayton HJ, et aI. Fiberoptic bronchoscopy in outpatient facilities, 1982. Arch Intern Med 1983; 143:301 Shure D, Fedullo PF. Transhronchial needle aspiration in the diagnosis of submucosal and peribronchial bronchogenic carcinoma. Chest 1985; 88:48-51 Johnston \VW Ten years of respiratory cytopatholo~y at Duke University Medical Center, III: the significance of inconclusive cytopathologic diagnoses durin~ the years 1970 to 1974. Acta Cytol 1982; 26:759-66

Utility of Repeated FOB for Suspected Malignancy (Torrington, Poropatich)