External beam radiation therapy for carcinoma in situ is also problematic, because these tumors are roentgenographically invisible. It is, therefore, a challenging task for the radiation oncologist to localize clinical target volume (CTV) in space and in time, since there is no gross tumor volume (GTV) be delineated. This inherent uncertainty of CTV must be balanced with the utilization of technological sophistication.
Due to the high efficacy in the treatment of superficial lesions, high-dose brachytherapy could be a definitive treatment for carcinoma in situ or endobronchial occult lung cancer with limited invasion. Endobronchial ultrasonography has been reported to result in a change in management in 36% of tumors thought to be curable by endobronchial therapy after bronchoscopy .
A recent large retrospective study  of 226 patients, of those 60 patients with carcinoma in situ, treated with high-dose brachytherapy alone showed no difference in local control between patients with in situ carcinomas and those with invasive carcinomas, probably because they were classified as non-invasive on biopsies though infiltrating on bronchoscopy. Furthermore, outcomes were worse for proximally located tumors and in patients treated with one catheter. These data suggest that the application of brachytherapy may be limited by bronchoscopic access, proximal tumor location, multiple bronchoscopy for catheter placement, geographic miss in dosimetric coverage of the target volume when using one catheter, and by post-procedural complications such as fatal hemoptysis or fistula formation.
Compared with conventionally planned radiotherapy, brachytherapy has the advantage of application of high-dose irradiation in tumor with rapid dose reduction outside the treatment volume. This favorable dosimetrical characteristic can be emulated by stereotactic guidance of external beam radiotherapy. The idea of use SBRT to deliver high-dose-rate-like hypofractionated radiotherapy came originally from Molla et al. , who suggested a fractionated stereotactic radiotherapy boost for gynecological tumors as an alternative to high-dose brachytherapy and has received some attention in recent years in the treatment of cervical cancer . We suggest that this idea could be potentially adopted in the radiation therapy of roentgenographically occult lung cancer. There is growing evidence, that it is possible to construct SBRT plans that closely emulate high dose brachytherapy dosimetry and deliver the plans non-invasively.
A major issue for the treatment of these lesions with SBRT is, however, that high-dose delivery precision requires a higher degree of accuracy in target volume delineation and in patient set-up during the radiation treatment course. Radiographic markers are well known in the postoperative setting, where there is actually no GTV to be delineated, and are standard tools for image-guided radiotherapy (IGRT) and robotic radiosurgery, and can be used for making invisible lesions “visible” for guiding high-dose SBRT by target volume delineation and positional verification.
In a prospective pilot series conducted by Malfiat et al.  to assess the feasibility and safety of EZ-Clips HX 610–090 (Olympus, Aartselaar, Belgium) in inoperable patients with roentgenographically occult lung cancer, one patients underwent high-dose external radiation therapy. Two clips were used to delineate the clinical target volume on the treatment planning system and to verify the correct positioning of the patients during the treatment course. No adverse events were observed. Both clips had disappeared, one was already lost at the time of treatment planning, and the other was lost during radiation treatment.
Anderson et al.  reported on the placement of 127 fiducials in 32 patients with mediastinal and centrally located thoracic tumors via flexible bronchoscopy as precursor to robotic radiosurgery. Complications included migration after insertion (one fiducial) and dropping in the airways prior to insertion (24 fiducials). Of those, 18 were removed with biopsy forceps, 2 were suctioned, 3 coughed out, and 2 were not retrieved, but were not seen on post-procedure chest x-ray. Three fiducials in two patients embolized via the pulmonary artery without adverse clinical consequence.
Endoscopic Ultrasound-guided fiducials placement as an alternative method has been described by Charabaty-Pishvian et al. . The Doppler function also verifies that there are no interfering vessels between the tumor and the needle, addressing the problem of post-procedural bleeding and embolization via pulmonary artery. These published data suggest that the implantation of radiographic markers via flexible bronchoscopy or EBUS is safe, has no clinically significant side adverse events, and can be offered for most patients. Furthermore, the movement of radiographic markers in space and time has been even studied, and estimated to be clinically acceptable .
Nevertheless, high-dose radiation treatment for centrally located tumors, particularly since the “bronchial exclusion zone” of RTOG 0236 is the main location of roentgenographically occult lung cancer, remains the main issue concerning the toxicity of SBRT. In addition to the fractionation scheme, central location of target lesions also appears to be a predictor of normal tissue toxicity. However, a recent review of 20 studies suggests that safe treatment of central tumors can be achieved with hypofractionated concepts with reduced single dose .