The adrenal glands are common sites of metastatic involvement in cancers arising in the skin (melanoma), lung, breast, and kidneys and other primary sites. These lesions can be bulky and can cause considerable pain, leading to decreased quality of life and significant narcotic usage [15]. Traditional treatment includes surgical resection, radiofrequency ablation, and chemoembolization [68]. However, these therapy approaches, particularly resection of tumor in the challenging to access retroperitoneal region, may not be an optimal option in patients, who are frequently debilitated from metastatic cancer and have a limited life expectancy [7].

Recently, stereotactic body radiotherapy (SBRT) approaches have provided a non-invasive alternative and broadened treatment options for patients with metastatic lesions in the lung, spine, liver, and other anatomic sites [918]. Excellent long-tem results in early stage primary lung cancer with local control rates equivalent to that of surgical resection have sparked interest in expanding the use of SBRT for metastatic lesions [19]. Similar high rates of local control have been recently reported for the use of SBRT in metastatic liver [17, 20, 21] and spine lesions [11, 15, 22].

However, relatively less data exist on the efficacy of SBRT in the management of adrenal metastases from lung cancer. There is no prior experience on the regression rate and response dynamics of these frequently bulky tumors and the role of histology in their regression pattern, which are important considerations for the highly conformal stereotactic therapy delivery. The purpose of this study was to assess local control, overall survival, tumor volume regression pattern, symptomatic relief, and radiation-induced side effects in patients treated with SBRT for adrenal metastases from lung cancer.

Methods and materials

Patient population

Following IRB approval, a retrospective review of nine patients with adrenal metastases from lung cancer was performed at our institution. Four women and five men were identified. Average age was 59.2 ± 9.3 years. All patients had primary lung cancer (five small cell carcinoma (SCLC), four non-small cell carcinoma (NSCLC)) and had received previous therapy for their primary site of cancer. Seven of nine patients received definitive concurrent chemo/radiotherapy for their primary disease. One patient received palliative radiotherapy to the chest followed by chemotherapy and another had metastatic disease at diagnosis therefore received chemotherapy alone. At the time of treatment, systemic disease was stable in five patients and progressive in four. All of the patients were either poor surgical candidates or refused surgical intervention.

A total of ten lesions were treated. Eight patients had unilateral and one had bilateral adrenal metastases. Based on the study by Tanvetyanon et al. [23] that showed differences in survival dependent on disease-free interval (i.e., time from original diagnosis to diagnosis of adrenal lesions), the time between original diagnosis and diagnosis of the adrenal metastasis was scored as either metachronous or synchronous. Lesions were defined as synchronous if they occurred ≤ 6 months from time of diagnosis and metachronous of they were diagnosed ≥ 6 months from time of diagnosis. Five lesions were metachronous and five were synchronous. Four patients presented with symptomatic, painful lesions requiring management with narcotics. Patient demographics are listed in Table 1.

Table 1 Patient characteristics


Patients were immobilized using one of two methods. Five patients were simulated in the Elekta Body Frame (Elekta AB, Stockholm, Sweden) with abdominal compression to minimize respiratory motion. The remaining four patients were unable to tolerate compression due to body habitus, pain, or decreased respiratory function. These patients were simulated in a BlueBAG™ Vacuum Cushion (Elekta AB, Stockholm Sweden) with a knee sponge. All patients underwent 4D computed tomography (CT) for planning purposes. An internal target volume was generated that represented the maximal superior, inferior, and transverse movement of the lesion. This volume was then expanded to create the planning target volume (PTV). Dosimetric variables are presented in Table 2. Seven patients were treated using IMRT, and two were treated using three-dimensional (3D) conformal technique. Seven to ten, noncoplanar, 6 MV photon beams were used. Prior to each fraction, patient setup was confirmed with conebeam CT. Eight of nine patients were treated every other day. One patient received a protracted course of treatment. Two fractions totaling 10 Gy were delivered prior to a break in treatment for a short course of palliative radiotherapy to a symptomatic neck mass. Two more fractions were delivered 1 month later after re-simulation and planning. On the final fraction, the pre-treatment conebeam showed poor alignment therefore the treatment was re-planned, and the final 5-Gy fraction was delivered several weeks later. One patient received simultaneous treatment of 20 Gy in five fractions to the bilateral adrenal glands.

Table 2 Dosimetric variables

Post-therapy surveillance and follow-up

Patients were followed approximately every 3 months with history and physical exam, pain and symptom assessment, and repeat imaging including CT of the abdomen and pelvis. Local disease response was determined by imaging response using Response Evaluation Criteria in Solid Tumors (RECIST)-based criteria on follow-up scans [24]. In addition, 3D lesion volumes on the pre-therapy and serial post-therapy CT were compared over time.

Data analysis

For overall survival, death of any cause was scored, and loss to follow-up was censored. Overall survival was calculated as the time from completion of treatment to the date of death (as determined by clinician notes and Social Security Death Index). Local control was defined as the interval between completion of SBRT and increase in volume greater than 20% of the lesion on follow-up imaging. The rate of treatment response was determined by calculating the slope between the pre-SBRT and first follow-up values of the volumetric and 2D response curves. Survival and local control were calculated using the Kaplan-Meier analysis. Differences between patient groups were assessed with a two-tailed, Student’s t test.


Treatment dose

The prescription dose ranged from 20.0 to 37.5 Gy in five fractions (mean, 24.9 ± 7.6 Gy) prescribed to the perimeter of the adrenal PTV. This corresponded to a BED10 ranging from 28.0 to 65.6 Gy (mean, 41.6 ± 11.6 Gy), utilizing standard linear quadratic modeling.

Objective response classification

Mean follow-up was 7.3 ± 7.8 months (range, 0–26 months). One patient died prior to first follow-up and was inevaluable. When response was assessed by RECIST criteria at 5 months after therapy completion for the remaining nine patients with ten lesions, there were six partial responses (PR), three patients with stable disease (SD), and none with progressive disease (PD), resulting in an objective response rate (CR + PR) of 67%.

However, with assessment by volumetric measurement, response classification at 5 months was CR (0), PR (7), SD(1), and PD (1) for an objective response rate (CR + PR) of 78%.

Among the three lesions (all NSCLC) classified as SD by RECIST, only one was classified as SD by volumetry. One of the three RECIST-based SDs had a 38.5% reduction in baseline 3D tumor volume and was classified as PR by volumetry. One of the three RECIST-based SDs was scored as a PD by volumetry (26.6% increase from baseline).

Response rate

Volumetric and RECIST-based data of imaging response for each patient are presented in Figs. 1 and 2. Given the variability in size of the individual lesions, each measurement is presented as a percentage of the pre-treatment lesion curves for each patient with SCLC (blue curves) and NSCLC (orange). The group average for the first 4 months following treatment is shown in 1(B) and 2(B). At 4 months post-therapy, 3D volume had decreased by 94.4% (Fig. 1a). In contrast, the RECIST-based 2D lesion size had decreased by only by 60% during the same time period.

Fig. 1
figure 1

Volumetric response to SBRT. a Percent change of 3D volume of the adrenal lesion from pre-therapy baseline over time for all patients. Curves for different patient populations are shown as follows: small cell lung cancer (SCLC) blue, non-small cell lung cancer (NSCLC) orange. Response curves for several NSCLC lesions are directly overlapping. Note the rapid and fairly uniform decrease in 3D volume in patients with SCLC. b Average of all volume regression curves in (a) for the first 4 months following treatment completion. c Average volumetric response for patients with SCLC and NSCLC. Patients with SCLC had a more rapid response onset. By 3 months, both groups had dramatic reductions in lesion size

Fig. 2
figure 2

Two-dimensional response to SBRT base on RECIST criteria. a Percent change of the RECIST-based 2D tumor size of the adrenal lesion from pre-therapy baseline over time for all patients. Curves for different patient populations are shown as follows: small cell lung cancer (SCLC) blue, non-small cell lung cancer (NSCLC) orange. Response curves for several NSCLC lesions are directly overlapping. The magnitude of response is much smaller than that obtained by 3D volumetric measurement, shown in Fig. 1. b Average of all curves in (a) for the first 4 months following treatment completion. c Average 2D response for patients with SCLC and NSCLC. As with the individual curves, note the discrepancy in response magnitude with the 2D technique when comparing b and c with Fig. 1b and c

The average tumor 3D volume regression for the SCLC and NSCLC groups are shown in Figs. 1c and 2c. A rapid volumetric response was observed in all patients with SCLC. This rapid response is illustrated by the serial diagnostic scans in a patient with a bulky (591 cm3) left adrenal metastasis that showed a 38% reduction of the 3D tumor volume prior to the third fraction and 88.5% reduction prior to the fifth and final fraction, as shown in Fig. 3. A similar decrease in size was noted in the 2D measurements (Fig. 2c); however, the rate of change was much less than for the volumetric measurements.

Fig. 3
figure 3

Volumetric response in SCLC patient. This patient with a left adrenal metastasis had a protracted course of therapy. She initially received 10 Gy (5 Gy × 2 fractions) prior to a break to complete a course of palliative radiotherapy for a symptomatic neck mass. Approximately 1 month later, she received an additional 10 Gy (5 Gy × 2 fractions) after re-planning. Prior to her fifth and final fraction, pre-treatment conebeam CT showed poor alignment therefore planning was repeated once again to account for the significant changes in tumor volume. The dose at each time point is specified in red next to response curve. A graphical representation of her tumor with corresponding time points is shown at the bottom of the figure

The rate of initial response to SBRT for both volumetric and 2D assessment methods is presented in Table 3. The mean slope of volumetric tumor regression in the SCLC group was −31.0%/month (range, −40.4 to −17.7), compared with −5.9%/month (range, −28.1 to 26.6) for NSCLC patients. This difference approached but did not reach statistical significance (p = 0.06). The mean slope of the 2D tumor response in the SCLC group was −15.8%/month (range, −17.9 to −10.8), compared with −4.9 (range, −14.0 to 9.1) in the NSCLC group (p = 0.05). When the response was compared between assessment methods, the differences within the SCLC group were statistically significant (p = 0.009) while those in the NSCLC group were not (p = 0.94).

Table 3 Slope of tumor response curves for 3D volumetric vs. RECIST-based method of assessment


The 1-year and 2-year overall post-therapy survival for all patients was 52% and 13%, respectively, with a mean survival of 10.2 ± 1.7 months. Local control at 1 year and 2 years for all patients was 44%. When overall survival was analyzed by histological subtype, 1-year and 2-year overall survival was 50% and 15% in the NSCLC group and 50% and 0%, respectively, in the SCLC group.

Patients with metachronous adrenal metastases had a trend for longer 1-year and 2-year overall survival rates of 60% and 20% (mean survival, 10.4 ± 2.1 months) compared with 37.5% and 0% (mean survival, 9.8 ± 3.1 months) for those with synchronous metastases. This difference was not statistically significant (p = 0.77).

Symptomatic response

Four of the nine patients presented with symptoms of pain from their adrenal metastases. Pain relief was achieved in all symptomatic patients either during therapy or by first follow-up, and response was associated with decreased narcotic usage. Each of these patients had large lesions (mean, 332 cm3; range, 133 to 600 cm3) with three of the four patients demonstrating a rapid response to radiotherapy (mean percent reduction in baseline volume 70%; range, 38% to 88%). The fourth patient had bilateral lesions. On first follow-up, the larger of the two lesions (600 cm3) had decreased by 12% while the smaller lesion (139 cm3) had increased by 27%. Despite this, the patient reported pain reduction while still on therapy.

Treatment toxicity

No patient experienced RTOG grade ≥ 3 acute toxicity. Two patients experienced initial nausea and vomiting while on treatment that was well controlled with oral anti-emetics. With a mean post-therapy follow-up of 7.3 ± 7.8 months, no grade ≥3 late toxicities were observed.


SBRT has become established as a viable non-invasive treatment option for patients with medically inoperable stage I non-small cell lung cancer [19, 2527]. More recently, the use of SBRT has expanded to metastatic disease [28]. Initial reports on the use of SBRT in pulmonary [10, 13, 16], liver [9, 12, 14], and spinal [11, 22] metastases attests to the high local control rates and low toxicity profile for SBRT in the treatment of metastatic lesions. Such non-invasive approach with SBRT may provide a viable alternative to surgical resection of adrenal metastases, which has been generally reserved for patients with adrenal oligometastasis, controlled systemic disease, and sufficient performance status and life expectancy to tolerate surgery [6, 7]. In our patient population of nine lung cancer patients, we have applied these principles of SBRT to treat metastatic adrenal lesions.


We found a favorable objective response rate of 67% (0 CR, 6 PR via RECIST criteria) in our patients at 5 months post-therapy. This is overall consistent with the results reported for SBRT in lung and liver metastases [9, 10, 1214, 16]. Although our patient population had shorter follow-up, was entirely composed of patients with metastatic lung cancer, and had a higher rate of extensive systemic disease than other series of oligometastasis, our outcome results are consistent with the few published series of adrenal metastases [15, 29]. These studies are reviewed in Table 4. A prime goal of SBRT in our population was rapid symptom relief and/or prevention of other local tumor complications. Results of symptomatic response in our patient population show that all patients had pain relief, confirming prior reports of favorable pain control rates in other published series (Table 4).

Table 4 Published studies of SBRT for adrenal metastases

Our patients with metachronous lesions had a trend toward improved 1-year and 2-year overall survival of 60% and 20%. Similar results have been published by Oshiro et al. [29] who reported greater 1-year, 2-year, and 5-year overall survival in patients with metachronous (83%, 56%, 56%) versus synchronous (35%, 0%, 0%) adrenal lesions in a more selected cohort. This, in combination with our data, supports the use of SBRT in patients with metachronous adrenal metastasis from lung cancer occurring after a disease-free interval of more than 6 months.


Our prescription dose (mean, 24.9 Gy; range, 20.0–37.5 Gy) and resulting BED10 (mean, 41.6; range, 28 to 65.6) was low when compared with the range of 10–60 Gy (BED10 range, 20–151 Gy) in previous reports of SBRT for adrenal metastases (Table 4). Despite the lower dose, our subjective and objective response rates were in line with the prior studies and favorable subjective and objective responses could be achieved with our relatively lower dose than previously reported and with minimal toxicity. This result suggest that, in view of extensive critical normal structures in close vicinity, lower doses in the range of 25 to 37.5 Gy delivered in five fractions may provide effective palliation, particularly in SCLC histology.

Treatment-related toxicity

Toxicity was minimal, with no grade ≥ 3 early or late adverse effects. This finding is consistent with prior reports (Table 4) and suggests that SBRT is a promising and well-tolerated palliative modality for symptomatic adrenal metastases.

Tumor response assessment and volumetric regression rate

Imaging-based response was profoundly dependent on tumor histology. The much more rapid volumetric response in SCLC compared with NSCLC histology (regression slope −31.0 vs. −5.9%/month, Table 3) is consistent with the overall radiosensitivity and commonly rapid response pattern of small cell histology. This rapid response was likely accentuated by the high-dose ablative fractionation pattern of the SBRT.

We found discrepancies between 2D and 3D volumetric assessment of response in our patients. Assessment with RECIST underestimated the PR rate and classified a substantial proportion of patients as SD, whereas 3D volumetry showed PR in the majority of patients and identified a patient with PD that had been classified as SD with RECIST. Such discrepancy between 3D volume and RECIST-based measurement has been demonstrated before in other tumors [3032]. The lesser ability of RECIST to classify response or progression is likely related to the inability of two-dimensional measurements to accurately assess subtle changes in irregular tumor volumes [30]. Thus, the RECIST classification may give rise to underestimations of both tumor response rate and treatment efficacy in SBRT for adrenal lesions, as well as inadequately identify tumor progression.

Given the rapid response to SBRT seen in our SCLC patients with adrenal metastases, volumetric assessment of response during the course of SBRT delivery may be of importance. Our case of SCLC with rapid response, presented in Fig. 3, illustrates this fact. In this patient, the fast response necessitated re-planning within the course of SBRT to adapt to the rapidly regressing GTV and PTV. The adrenal gland undergoes profound displacement within the retroperitoneal space, when the tumor volume is reduced. In view of the close proximity of equally mobile critical normal structures, including kidney, stomach, and bowel, adaptive planning is of paramount importance to prevent normal tissue toxicity. Our data suggest that such volume changes and displacement of the adrenal gland can be effectively monitored with on-board CT-based imaging to enable re-planning during the treatment course.


SBRT is a useful, non-invasive palliative treatment option for adrenal metastases from lung cancer, providing excellent local control with minimal toxicity in a highly symptomatic and challenging complication of the disease. Volumetric tumor regression can be rapid, particularly in small cell cancer histologies, and this may influence dosimetry and the need for adaptive therapy in some patients.