Positive ¹⁸Fluorodeoxyglucose-Positron Emission Tomography/Computed Tomography Predicts Preinvasive Endobronchial Lesion Progression to Invasive Cancer

To the Editor:

Central airway invasive squamous carcinoma develops from preinvasive dysplasia and carcinoma in situ (CIS) epithelial lesions (1). The natural history of preinvasive airway disease is variable, with a proportion progressing to invasive cancer and as much as 40% regressing to normal histology (2). This unpredictability creates difficulty in developing recommendations for the diagnosis, follow-up, and treatment of the disease (3).

Autofluorescence bronchoscopic (AFB) identification and regular surveillance biopsy of these lesions allow a strategy to intervene at the earliest stages of invasive cancer formation (4); however, a noninvasive imaging biomarker to detect malignant potential could assist the clinician in their decision to treat precancerous tracheal and bronchial lesions with potentially radical approaches.

Computed tomography (CT) lacks sensitivity and specificity in identifying premalignant peripheral bronchial lesions (5). Although ¹⁸fluorodeoxyglucose (¹⁸F-FDG)-positron emission tomography (PET) is the current gold standard imaging modality for staging patients with lung cancer and monitoring treatment response (6), its role in patients with preinvasive airway lesions has yet to be investigated.

We therefore investigated the novel use of ¹⁸F-FDG-PET/CT as part of a surveillance program for patients with preinvasive lesions (CIS and high-, moderate-, and low-grade dysplasia). Over the course of 11 years, 44 patients (30 men; median age, 68 years) with preinvasive endobronchial lesions identified using AFB, underwent ¹⁸F-FDG-PET/CT examination within 6 weeks from the AFB examination. The protocol also included AFB follow-up and routine imaging (2). Endpoints included progression to histopathological invasion of the initial lesion. All patients gave informed consent before entering the study.

Our patient population contained patients with CIS (n = 29), severe dysplasia (n = 6), moderate dysplasia (n = 5), and mild dysplasia (n = 4). Twenty patients had no history of bronchial carcinoma, and 24 patients had previously undergone curative surgery for bronchial carcinoma.

We observed ¹⁸F-FDG uptake in approximately a third of patients at sites of CIS (8/29 patients) or at remote sites in the lung (6 patients) (Figure 1). No focal uptake was observed in any grade of dysplasia. In the eight patients with CIS who had positive ¹⁸F-FDG uptake, seven (88%) developed invasive cancer during surveillance versus 6 of 21 (29%) with negative ¹⁸F-FDG uptake (P = 0.0097) (Table 1). Invasive cancer developed between 3 and 36 months of follow-up at sites of previous CIS. None of the eight patients with CIS who had focal uptake showed spontaneous regression.

Figure 1.

(A) Surveillance protocol for patients with preinvasive lesions involving the bronchial epithelium (modified from Reference 2). Positron emission tomography (PET)/computed tomography (CT) and follow-up imaging were performed before entry in the surveillance program. Patients with ¹⁸fluorodeoxyglucose (¹⁸F-FDG) uptake at the site of carcinoma in situ were rebiopsied. If found positive, they were excluded from the surveillance and referred for treatment. (B-C) A 69-year-old man with carcinoma in situ seen at bronchoscopy in the right upper bronchus. CT shows a background of emphysema with minor bronchial thickening, particularly along the right upper lobe branches (C, arrows). ¹⁸F-FDG-PET/CT shows peribronchial (B, arrowhead) ¹⁸F-FDG uptake immediately adjacent to the subsegmental branch of the posterior segment for the right upper bronchus. (D-E) Previous left upper lobectomy for squamous cell carcinoma. Under surveillance for preinvasive disease in the right lower bronchus. ¹⁸F-FDG PET/CT shows positive uptake in the right lower bronchus despite minor CT bronchial wall thickening (E, arrowheads).

Table 1.

Details of PET-Positive and PET-Negative Patients with CIS Who Developed Invasive Cancer

Patient	Sex, Age (yr)	Smoking Hi	FEV1/Pred	Time to Develop Ca from Bronchoscopy	PET (SUV Maximum)	Mediastinum (SUV Maximum)	Symptoms
PET-positive
1	Female, 79	N	1.69 (83%)	18 mo	3.2	1.8	Nil
2	Male, 75	Y	1.54 (67%)	6 mo	3.9	1.5	Nil
3	Male, 70	Y	0.64 (20%)	6 mo	3.3	2.5	Hemoptysis
4	Male, 65	N	1.06 (35%)	RT treatment (3 mo after PET)	3.3	2.6	Hemoptysis
5	Female, 80	Y	1.81 (89%	Stable after 3 yr	5.9	2.1	Hemoptysis
6	Male, 75	N	2.17 (65%)	6 mo	4.2	2.3	Nil
7	Male, 67	Y	2.60 (71%)	3 yr	4.7	1.8	Nil
8	Male, 78	Y	1.54 (35%)	18 mo	3.5	2	Hemoptysis
PET-negative
1	Male, 80	Y	2.50 (81%)	2 yr	Negative	2	Hemoptysis
2	Male, 77	Y	2.6 (74%)	3 yr	Negative	2.1	Nil
3	Female, 59	Y	1.84 (84%)	2 yr	Negative	1.9	Nil
4	Male, 75	Y	1.95 (73%)	18 mo	Negative	2.2	Nil
5	Male, 78	Y	2.16 (80%)	2 yr	Negative	2.3	Hemoptysis
6	Female, 79	Y	0.9 (50%)	3 yr	Negative	2.1	Nil

Definition of abbreviations: Ca = cancer; CIS = carcinoma in situ; CT = computed tomography; ¹⁸F-FDG = ¹⁸fluorodeoxyglucose; Hi = history; PET = positron emission tomography; Pred = predicted; RT = radiotherapy; SUV = standardized uptake value.

¹⁸F-FDG-PET/CT scans were analyzed for areas of abnormal ¹⁸F-FDG uptake; location and site of ¹⁸F-FDG uptake were recorded. In 9 of 29 patients there was increased ¹⁸F-FDG uptake at the site of CIS (SUV maximum range, 3.2–5.9; average, 3.5). Seven of the eight patients developed invasive cancer on histological follow-up of the lesion, three after 6 months, two at 18 months, and one at 36 months after ¹⁸F-FDG-PET/CT. One patient remained stable with CIS after 3 years. Lesions showing positive uptake were significantly more likely to progress (P = 0.0097; the two-tailed Fisher test was used to assess the association between PET positive/negative findings vs. stable disease/invasive cancer during surveillance; P < 0.05 was taken as significant) to invasive cancer than those with negative ¹⁸F-FDG uptake. One patient did not have pathological confirmation of invasive cancer; the diagnosis was made on the basis of increasing bronchial wall thickening on follow-up CT imaging. This patient was treated with external beam radiotherapy with resolution of ¹⁸F-FDG avidity on a follow-up scan and resolution of CIS at least histology to normal. In 21 patients, there was no ¹⁸F-FDG avidity at sites of CIS. Of these patients, six developed invasive cancer within 3 years of a negative PET. In four patients with negative ¹⁸F-FDG scans, CIS lesion histology reverted to normal over 1–3 years follow-up. In the remaining patients, CIS remains stable during follow-up.

Of six patients with severe dysplasia, no lesions showed positive uptake of ¹⁸F-FDG. One patient, however, developed invasive cancer 1 year later. This lesion remained unchanged in morphological appearances, but a repeat PET showed uptake. The other patients remained free of invasive cancer during follow-up.

There were four patients with mild and five patients with moderate dysplasia. No mild to moderate dysplasia sites showed ¹⁸F-FDG uptake. During follow-up, seven lesions remained stable and two regressed to normal histology.

Six patients presented with ¹⁸F-FDG uptake at sites remote from the known preinvasive bronchial lesion, with two cases of distant cancer, supporting the concept of “field cancerization” proposed to explain the progression of multiple foci of precursor lesions throughout the respiratory epithelium. This hypothesis is further supported by the recent observation that the airways of heavy smokers harbor widespread related mutant clones that are frequently separated by nonmutant normal epithelium (7). The remaining four patients presented hilar and mediastinal nodal uptake resulting from biopsy-proven inflammation.

Our findings suggest that many preinvasive epithelial lesions progress to invasive cancer in a relatively short follow-up (31% of our cohort developed invasive cancer), in keeping with published data (8, 9). To date, there is no biological or radiological biomarker of preinvasive lesion progression; hence, we propose a role for ¹⁸F-FDG-PET/CT in the assessment of risk for progression of CIS lesions to invasion.

Only one previous study has reported the use of ¹⁸F-FDG-PET (but not PET/CT) to distinguish between preinvasive airways disease and occult lung carcinoma (10), and the authors reported high sensitivity and specificity of PET in detecting squamous cell carcinoma. However, no data from previous studies are available to define the possible role of ¹⁸F-FDG-PET/CT at occult preinvasive sites of lung malignancy.

We evaluated all patients under surveillance for preinvasive conditions with no histological evidence of invasive disease. It is recognized as a limitation that some of our cases may be false-negatives for invasive cancer as a result of operator-dependent biopsy technique. However, in such cases, the PET/CT could be helpful in reducing sampling error. Other promising imaging techniques (i.e., optical imaging tomography and endoscopic ultrasound [with high-resolution probes and Doppler]) have been investigated for preclinical cancer imaging applications, yielding exciting new capabilities to probe and monitor cancer progression and response in vivo (11).

Our results suggest important implications for the management of patients undergoing surveillance of preinvasive lesions. We provide data that supports the use of ¹⁸F-FDG-PET in cases of CIS. In such cases with ¹⁸F-FDG–positive findings, it would be worth considering preemptive local treatment with ablation, radiotherapy, or surgical resection.