Prostate cancer is a common malignancy with a low mortality-to-incidence ratio in the more developed regions such as Europe and the United States. The 5‑year relative survival of patients under 80 years in Western countries is more than 97%. The crude rates of incidence and mortality in Germany are 169.7 and 31.2 per 100,000 [1]. This relatively good outcome is partly due to the regulation of improved multidisciplinary treatments as described in the German S3 guidelines [2]. Relating to the following case of a patient with surgically resected T3 prostate cancer with positive resection margins, the guideline recommendation is immediate adjuvant radiotherapy (RT) or salvage RT in case of prostate-specific antigen (PSA) relapse, both with a level A recommendation. Adjuvant RT significantly reduces the risk of PSA relapse, disease recurrence, and can reduce metastasis-free survival and overall survival [3]. The reported late side effects of adjuvant RT are below 10%, for total urinary incontinence 6.5% vs. 2.8%, and rectal complications 3.3% vs. 0% for surgery plus radiotherapy vs. surgery alone, respectively [4]. Radiosensitivity describes the sensitivity of organisms to ionizing radiation. Increased reactions can be caused by impaired DNA double-strand break (dsbs) repair ability and genetic instability. This is known for example in patients with autosomal recessive disorders as ataxia telangiectasia (ATM) and Nijmegen breakage syndrome (NBS) as well as in BRCA1 mutation carriers [5,6,7]. Generally, in case of need for RT in such patients, a reduction of the single dose to one third of the normal dose could be recommended [8]. The identification of highly radiosensitive patients other than those with the known immunodeficiency syndromes is a challenge that is currently still unresolved. We report on such a patient who developed an abnormal course of acute and late toxicity.

Case presentation, clinical follow-up, and examination findings

A 56-year-old man with an initial diagnosis of high-risk prostate cancer, iPSA 59 ng/ml, Gleason score 4 + 5 = 9 in 8/8 biopsies was treated in 2010 with primary radical prostatectomy. Symptoms were slight dysuria and 5 times nycturia 5 months prior to diagnosis and the prostate had palpable indurations. The postoperative tumor stage was described as “minimum” pT3b pN1(3/15) Pn1 pL1 R1. Initial tumor staging with scintigraphy did not reveal distant metastases. The PSA value was 0.2 ng/ml 37 days after prostatectomy. A complete androgen deprivation therapy (ADT) with bicalutamide and enantone was started after surgery. The PSA value was controlled under ADT. After rehabilitation, the functional results were not completely resolved. The patient needed initially ten pads per day and had pain due to postoperative pelvic lymphoceles. Lymphatic drainage and perineal exercises were continued after rehabilitation. Secondary diagnosis was a papillary thyroid carcinoma, which was treated with surgery and radiation iodine therapy without complications one year before RT started. Due to intense hot flash and gain of weight of 20 kg, the patient terminated ADT, and adjuvant RT to the prostate bed and pelvic lymph nodes was offered 32 months after surgery. The PSA value, due to ADT until RT, was in remission until RT was started. RT was administered as intensity-modulated radiotherapy (IMRT) with daily image guidance (IGRT) and the following dose concept: 35 × 2 Gy to the prostate bed (PB), 35 × 1.8 Gy to PB and seminal vesicle bed (SV) and 30 × 1.7 Gy to the pelvic lymph nodes up to the aortic bifurcation (Fig. 1). This was in accordance to the national guidelines that recommend a minimum dose of 66 Gy for salvage therapy and adjuvant irradiation of the pelvic lymph nodes is possible in case of nodal positive prostate cancer [2]. Already 14 days after the beginning of RT, common toxicity criteria (CTC) grade 3 colitis with painful bowel movements, crampy defecation, and rectal bleeding was reported. Treatment was not stopped due to the initial high-risk tumor. At the end of RT, slight rectal incontinence and increased urinary incontinence continued. The proctoscopy at this time already revealed an ulcer; histopathology showed a radiogenic colitis. Mesalazine foam enemas were prescribed and consequently the symptoms improved. There was no increased hematological toxicity.

Five months after the end of RT, neurological symptoms and edema of the legs developed. The patient experienced a growing unsteadiness of gait and insensibility of his legs caudal to the inguinal regions. The tibialis somatosensory evoked potentials could not be verified, a spinal puncture excluded meningeosis carcinomatosa. Due to edema of the pelvic muscles seen in magnetic resonance imaging (MRI) five months after therapy, prednisone medication and lymphatic drainage were prescribed. Prednisone caused a short alleviation of the deficits. The pathological findings in consecutive MRIs aggravated (Fig. 2). Nine months after RT, a deep abscess with Staphylococcus aureus in the floor of the pelvis and Hunter’s canal was split, prednisone was stopped, and antibiotic treatment with clindamycin and piperacillin/tazobactam was continued. Ten months after RT, the patient was dependent on a walking frame and urinary incontinence was unchanged. Rectal incontinence had ended, but he still had rectal tenesmus, rectal bleeding, and perianal ulcers. Sulfasalazin and mesalazine foam enemas were continued. In the electrophysiological measurements, all neurographies of the lower extremities showed pathologically decreased amplitudes. The electromyographies revealed grouped fasciculation and slow myokymias which are often found in radiation plexopathy [9]. From this time on, due to still elevated C‑reactive protein, antibiotic treatment with cefuroxime was started. Due to recurrent abscess formation and high-grade frailty, hyperbaric oxygen therapy or even cortisone treatment were not possible. The patient had to be cared for in a nursing home. Nineteen months after RT, ureter stenoses caused hydronephrosis and nephrostomies were placed on both sides. The entire mucosa of the urinary bladder showed bullous transformations. Two years after RT, repeated hospitalizations were necessary due to recurrent rectal bleeding caused by severe radiogenic proctitis. The PSA value was below 0.05 ng/ml during the whole follow-up. The patient died 25 months after RT in hemorrhagic shock after massive bleeding from rectal ulcers. Re-evaluation of RT (dose calculation, plan check-ups) did not reveal any protocol violation.

Fig. 1
figure 1

Dose distribution of the tomotherapy planning (a–f). In the affected nerves and muscles, no dose higher than 50 Gy was applied. The dose–volume histogram shows sparing of bladder and rectum (g, h)

Full size image
Fig. 2
figure 2

Magnetic resonance imaging findings. Seven months after radiotherapy (RT): Thickening of bladder wall (a). Nine months after RT: Extensive abscess formation of the left pelvic and adductor muscles (b). Twelve months after RT: slightly recessive abscess after surgical splitting and strongly thickened wall of urinary bladder (c)

Full size image

Index patients cellular radiosensitivity

The testing was performed by the three-color fluorescence in situ hybridization (3-C FiSH). Peripheral blood was obtained 13 months after the patient was treated with RT and 12 months before he died. The blood was divided into a 2 Gy irradiated and a nonirradiated (control) sample and the 3‑C FiSH method was performed as previously described [10, 11]. Metaphases were automatically detected by a microscope (Zeiss, Axioplan 2, Göttingen, Germany) and an image of each metaphase was acquired (Metafer 4 V3.10.1, Altlussheim, Germany). This study was approved by the ethics review committees of the Friedrich-Alexander-Universität Erlangen-Nürnberg (No. 2725). The metaphases were analyzed by the occurrence of the following aberrations: translocation, dicentric, acentric and ring chromosomes, insertions, deletions and complex chromosome rearrangements (CCR) (Fig. 3a, b). The different aberrations were evaluated by the average number of chromosomal breakages (B/M) using the imaging software Biomas (Biomas, Erlangen, Germany). The B/M value of the irradiated sample was corrected by the background value of the nonirradiated sample. For comparison B/M values of 30 aged-matched healthy male individuals and 30 patients suffering from different cancer diseases were used [18, 19]. Eleven radiosensitive patients with severe side effects after RT were also included. Detailed information of them can be found in Table 1. The background of 0.61 B/M is clearly increased compared to the mean value of the healthy cohort (0.03 ± 0.04 B/M), the patients (0.09 ± 0.32 B/M), and the radiosensitive patients (mean: 0.45 ± 1.08 B/M; median 0.25 B/M). The 2 Gy corr. B/M value of the patient (0.98 B/M) is clearly increased in comparison to the values of the healthy cohort (0.45 ± 0.19 B/M) and patients (0.43 ± 0.33 B/M). The index patient is in the range of the outliers of the radiosensitive patients (0.72 ± 0.22 B/M; Fig. 3c).

Fig. 3
figure 3

Testing radiosensitivity by 3‑color fluorescence in situ hybridization and a colony formation assay. Normal metaphase of a nonirradiated lymphocyte (a). Metaphase of a lymphocyte after a dose of 2 Gy, with a translocation of chromosome 1 and 2 (2 breaks) and an acentric part of chromosome 2 (1 break) (b). Control and 2 Gy corrected breaks per metaphase (B/M) values of lymphocytes of 30 age-matched heathy individuals and cancer patients, 11 radiosensitive patients and the index patient (red cross). Error bars are 95% confidence intervals (CI). The asterisks and small circles show the outliers of the groups (c). Control and 2 Gy corr. B/M values of fibroblasts of 10 heathy individuals (blue crosses), one Nijmegen breakage syndrome (NBS) patient (green circle), and the index patient (red cross). The average B/M value of the mock irradiated healthy individuals’ lymphocytes is 0.04 B/M and of the irradiated lymphocytes is 0.27 B/M. (d). Colony formation assay of the index patients’ fibroblasts (filled square), two healthy individuals’ fibroblasts lines (circle, triangle), and a homozygous NBS patient (diamond) (e). Scale: 10 µm

Full size image
Table 1 Characteristics of 11 radiosensitive patients
Full size table

In addition, a skin biopsy was taken from the inside of the forearm 2 months after the blood test was carried out. Fibroblasts were cultured from the biopsy and radiosensitivity was tested by 3‑C FiSH and a colony formation assay. The 3‑C FiSH was performed as describe above with a few adjustments. Patients’ values were compared to data of 10 healthy donors and of a highly sensitive homozygous Nijmegen breakage syndrome (NBS) patient. The nonirradiated value with 0.069 B/M is in the range of the healthy individuals (0.04 ± 0.05 B/M) and the NBS patients value (0.036 B/M). The value resulting after irradiation (0.91 B/M) is more than twice as high as the healthy individuals (0.27 ± 0.11 B/M) and in the range of the NBS patient (0.77 B/M; Fig. 3d).

Fibroblasts were seeded in petri dishes for the colony formation assay. After being irradiated by a 120 kV X‑ray tube (Isovolt Titan, General Electrics, Ahrensburg, Germany), the cells were cultured for 2 weeks. Cells were stained with methylene blue and colonies (>50 cells) were counted. Each experiment was triplicated and the whole experiment was repeated for three times. The index patients’ fibroblasts (SF2 0.398) are more radiosensitive than fibroblasts of healthy donors (SF2 0.504 and 0.479; Fig. 2e). At 1.8 Gy, the sensitivity was 1.3 times and at 2 Gy 1.33 times increased. The NBS-/- fibroblasts (SF2 0.307) were 1.6 and 1.7 times increased at the respective doses. The clearly increased values compared to the healthy controls with similar increased values like a highly sensitive NBS patient leads to the conclusion that the index patient has a distinctly increase sensitivity to ionizing radiation.

Discussion and literature review

Both tumor and normal tissue responses to RT vary between patients. Defects in DNA repair, predisposition to different and multiple forms of malignancy, and increased radiosensitivity characterize the rare chromosome breakage syndromes ATM and NBS [12, 13]. However, as in the presented case, patients who are not apparently suffering from these syndromes may also experience extreme side effects from RT and the identification of these patients is an important task for radiation protection. The presented patient was diagnosed with prostate cancer at the age of 56 years, the papillary thyroid cancer was found 2 years later, and colon cancer was found only by autopsy. Examples of tests used to measure individual radiosensitivity are on a DNA level the measurement of DNA damage response proteins as γH2AX, 53BP1 or RAD51 [6, 14], clonogenic assay and chromosome aberrations [15, 16]. The first studies measuring the radiosensitivity of hypersensitive individuals with unusually severe reactions to radiotherapy all involved fibroblast clonogenic assays. Van Oorschot et al. [17] found in ex vivo irradiated lymphocytes (1 Gy) from prostate cancer patients with severe late radiation toxicity significantly higher numbers of γH2AX foci 24 h after irradiation than in those without severe reactions. Moreover, in the severe reacting patients genes of the homologous recombination pathway were less induced as well as genes of the nonhomologous end-joining pathway [17]. Auer et al. [18] estimated individual radiosensitivity in breast cancer patients and healthy individuals by 3‑color fluorescence in situ hybridization (FISH). The number of chromosomal breaks in lymphocytes per metaphase (B/M) after in vitro irradiation (2 Gy) helped to divide the individuals into relatively radioresistant (up to 0.3 B/M), intermediate (0.3 to 0.55 B/M), and sensitive individuals (more than 0.55 B/M). There are individual differences in chromosomal aberrations after in vitro irradiation of cells from healthy individuals, cancer, and cancer susceptibility syndrome patients [19]. The lymphocytes of the presented patient were screened with 3‑color-FISH about one year after irradiation. Already without irradiation, the value of B/M was elevated by 0.614 and may be an indication for the increased susceptibility for impaired DNA damage processing. Therefore, an increased radiosensitivity of 35–55% was suggested which translates into a suspected increase of total dose to 97.2–111.6 Gy instead of 72 Gy that should have been given to a nonradiosensitive patient. The field of radiogenomics has evolved to identify genetic risk factors that are associated with elevated radiosensitivity [20]. Genetic testing was not performed in this patient and no familiar predisposition was known. Auer et al. discuss that due to the time delay the quantification of side effects in correlation to their measurements may need several years. The presented patient suffered from rapidly increasing side effects.

Conclusion

Severe adverse effects after RT occur in less than 10% of individuals. Highly radiosensitive reactions occur in 0.5–5% of cases [21] and the presented case was the first of about 10,000 treated patients in this institution. His dramatic outcome highlights that the timely identification of patients with increased radiosensitivity before treatment is an important task in radiation protection. Effective predictive assays would help to adjust treatment in such patients. A family history of disease syndromes may help identify such patients. The best assay for measuring radiosensitivity has not yet been identified. The validity of the tests is limited if the patient has already been irradiated.