Radiation therapy for canine appendicular osteosarcoma


J. FareseDepartment of Small Animal Clinical Sciences
College of Veterinary Medicine
University of Florida
2015 SW 16th Avenue
FL 32610, USA
e-mail: faresej@vetmed.ufl.edu


Radiation therapy (RT) for the management of canine appendicular osteosarcoma (OSA) can be described as either palliative- or curative intent. Palliative RT uses coarsely fractionated external beam RT or radiopharmaceuticals to provide relief of pain and lameness associated with OSA while resulting in minimal, if any, radiation-induced acute adverse effects. Limb amputation and chemotherapy are considered (together) the standard of care for curative-intent treatment of canine appendicular OSA. When limb amputation is not possible, RT can be used for limb sparing and is supplemented with chemotherapy for presumed micrometastatic disease. Fractionated tumour irradiation with curative intent appears to be ineffective and local disease control can more likely be achieved when stereotactic radiosurgery or intra-operative extracorporeal irradiation is combined with strict case selection and adjunctive chemotherapy. The availability of limb-sparing RT is limited by experience and availability of specialised equipment. When planned and administered appropriately, radiation-associated adverse effects are often mild and self-limiting.


Osteosarcoma (OSA) accounts for up to 98% of all canine primary bone tumours and 5–6% of all canine malignancies.1–3 Appendicular OSA is also a highly malignant tumour with more than 90% of dogs having micrometastatic disease and 15% of dogs having clinically detectable metastasis at the time of initial diagnosis.4 Affected dogs typically present with progressive lameness, bony proliferation or swelling. Acute nonweight-bearing lameness is typically associated with the onset of pathologic fracture.1,2 Characteristic radiographic signs of appendicular OSA include bone lysis, periosteal proliferation, spiculated and sunburst bone formation, subperiosteal bone formation (Codman’s triangle) and soft tissue swelling, with calcification extending into surrounding soft tissues (amorphous new bone).1–3

The management of OSA encompasses palliative- or curative-intent strategies. Palliation can be achieved with combinations of radiation therapy (RT), radiopharmaceutical therapy, bisphosphonate treatment and/or analgesia.1 When analgesics are used as a sole treatment for appendicular OSA in dogs, median survival time (MST) is expected to be 1–3 months.1 Curative-intent strategies include amputation, limb-sparing surgical techniques or RT, each combined with chemotherapy.1,5 Limb amputation remains the current standard of care for local management of primary bone tumours, and MST for dogs having limb amputation alone is 103–175 days.6–10 Importantly, MST for appendicular OSA treated with amputation doubles to 235–366 days when surgery is combined with chemotherapy.7–17 However, some dogs are not considered suitable candidates for amputation because of concurrent severe orthopaedic or neurologic conditions.4 Limb sparing can be achieved with either surgical or radiation techniques or both.18–29 Limb-sparing techniques in the dog have been described for treating tumours involving the humerus, radius, ulna, femur and tibia.18–30 While allograft limb sparing of the distal radius is largely successful despite postoperative complications (infection rate approaching 50%), surgical techniques have been less successful at other locations because of implant failure and poor limb function.3,18,22,28–30

RT can be an effective method for the palliation of pain associated with appendicular OSA.28,31–35 Although no substantial evidence exists, when palliative radiation therapy (PTRT) is combined with chemotherapy, some degree of local tumour control can be achieved.28,31–35 An earlier review by McEntee of RT for canine OSA summarised much of the clinical recommendations still used today;31 however, recently published studies investigating both palliative- and curative-intent RT have prompted this updated review.3,21,27,28,31,36 The purpose of this report was to describe, compare and review the clinical applications of RT currently available for use in the management of canine OSA.

Types of radiation teletherapy

The RT modality most commonly used in veterinary oncology is teletherapy. Teletherapy is external beam RT, where a photon or electron beam is produced from an orthovoltage or megavoltage (high energy) radiation unit. Megavoltage irradiation involves the use of photons or electrons with energy of greater than 1 MeV.37 The higher energy radiation beam penetrates further into tissue (than orthovoltage), allowing treatment of deeper seated tumours such as OSA.37 With megavoltage units, the maximum radiation dose is deposited some depth below the surface resulting in a skin-sparing effect.37

Palliative RT

The goal of PTRT is to provide relief of specific clinical signs (decrease the pain and lameness associated with OSA) while resulting in minimal, if any, radiation-induced acute adverse effects.1,37 This can generally be achieved by delivering multiple large fractions (e.g. 8–10 Gy per fraction). Historically, PTRT has been used most commonly for pain relief in dogs affected with appendicular OSA that are not surgical candidates for amputation (because of concurrent disease or stage III disease with skeletal metastasis) or if an owner has declined curative-intent therapy.32–44 Treatment of OSA in these patients with PTRT can result in local reduction in inflammation, pain relief, slowed progression of osseous metastatic lesions and improved quality of life.1,31 While the exact molecular and cellular mechanisms through which RT reduces bone pain are still unknown, it has been shown that some effect results from acute disruption of inflammatory cells, decreased progression of tumour-induced osteolysis and reduction in tumour size.38,39 The veterinary literature contains reports of PTRT for canine appendicular OSA delivering between 16 and 32 Gy in two, three or four fractions. The majority of these protocols achieve some level of analgesia within 7–14 days after the first dose of radiation, with clinical improvement lasting approximately 2–3 months.1,32–35 Unfortunately, the available literature is limited by variation in treatment protocols, including total radiation dose and fractionation, retrospective studies and sporadic reports of concurrent chemotherapy use. Recently, Boston et al. reported longer survival (MST 130 days) in dogs with metastatic (stage III) appendicular OSA treated with PTRT and chemotherapy, although no details of treatment course and clinical characteristics of treated patients were available36 (Table 1). The reports of PTRT for canine appendicular OSA are detailed below and are stratified based on fractionation protocol.

Table 1.  Summary of palliative, curative-intent and radiopharmaceutical protocols including response, outcome and adjuvant chemotherapy for dogs with appendicular osteosarcoma treated with radiation therapy
ProtocolDose interval (days)Total doseNumber of dogsMedian onset of response (days)Duration of response (days)Response rate (%)MST (days)ChemotherapyReference
  • *

    OPLA-Pt, open cell polylactic acid containing 8% cisplatin.

Palliative protocols
 Two fraction0, 716 Gy37117374122Variable chemotherapy, prolonged duration of response (P < 0.001)33
0, 116 Gy 1,40
 Three fraction0, 7, 2124–30 Gy6*217083No chemotherapy42
0, 7, 2124 Gy54538333 of 54 dogs received carboplatin (day 0, no significant improvement)41
0, 7, 2130 Gy151513080125No chemotherapy34
0, 7, 2124 Gy5035
0, 7, 2130 Gy58117374122Variable chemotherapy, prolonged duration of response (P < 0.001)33
 Four fraction0, 7, 14, 2132 Gy15149593313Carboplatin 240–300 mg m−2 day 0 (n= 2; P= 0.82)32
0, 7, 14, 2124 Gy54538333 of 54 dogs received carboplatin (day 0, no significant improvement)41
Curative-intent protocols
 IORT070 Gy1327477298Variable alternating chemotherapy5
070 Gy4586Variable chemotherapy, frequent, severe complications; not recommended27
 SRS020 Gy52110540% no progression363No chemotherapy; results for all 11 dogs in study28
030 Gy6100% no progressionCarboplatin day 0, then variable chemotherapy28
 Full-course external beam RTMonday to Friday for 19 total fractions48–57 Gy1020950209Variable radiosensitising: 8 of 10 OPLA-Pt, 1 of 10 cisplatin. Variable adjuvant chemotherapy3
 Samarium037 MBq kg−19 58
 Samarium036–57 MBq kg−115 150 59
 Samarium0 3514 6393 55
 Samarium037 MBq kg−15 100 60

Two-fraction protocols

All reports of two-fraction protocols deliver 16 Gy total doses split equally on days 0 and 7. In 1999, Ramirez et al. compared a traditional 0, 7 and 21-day protocol of 10 Gy per fraction of 60Co photon to a day 0 and 7 protocol of 8 Gy per fraction.33 This latter abbreviated protocol was intended for retreatment following recurrence of clinical signs. Interestingly, no significant difference in the median time to onset of pain relief (11 days), rate of response (74%) or duration of response (73 days) was noted.33 Chemotherapy was administered to most dogs receiving both RT protocols and appeared to increase both the probability and the duration of clinical response to PTRT.33 These findings must be interpreted with caution, however, because of the substantial bias and because of inconsistent chemotherapy schedules among cases. Liptak et al. and Mayer et al. both describe experience with two-fraction protocols (8 Gy × 2 on days 0 and 1); however, no details of response or clinical characteristics of treated patients were available.1,40

Three-fraction protocols

Three-fraction RT protocols are the most widely reported variation of PTRT for canine appendicular OSA. Reported protocols deliver 8–10 Gy fractions, typically on a 0-, 7- and 21-day schedule, for a total dose of 24–30 Gy.33,34,41,42 These protocols have achieved an 83% response rate, with median onset of response occurring 11–21 days after the first fraction, and median duration of response lasting 53–180 days.33,34,41,42 Variations of the 0-, 7- and 21-day protocol have been recently described where two consecutive daily fractions of 8 Gy are administered (days 0 and 1), followed by additional 8 Gy fractions on a monthly basis or as required.1,40 To date, however, no response rates or durations of response have been reported for these protocols.1,40

Four-fraction protocols

Because most three-fraction protocols specify 2 weeks between the second and third treatments, Green et al. proposed that the possibility exists for repopulation of the tumour within this 2-week interval. In an effort to eliminate this 2-week gap, reduce the risk of tumour repopulation, and increase the duration of pain relief, Green et al. described a four-fraction (0, 7, 14 and 21 days) 60Co (gamma photons) radiotherapy protocol, delivering 8 Gy per fraction for a total dose of 32 Gy.32 Compared with three-fraction protocols, this technique resulted in a higher response rate (93%), similar onset of response (14 days) and a similar duration of response (95 days).32 A similar rate of pathologic fracture (13%) is reported, and while survival time with this four-fraction protocol appears considerably longer than other PTRT protocols (313 days), limitations in retrospective analyses limit statistical comparison between studies.32 There was no apparent benefit of platinum derivatives (cisplatin or carboplatin), although the timing of administration of drugs was not consistent between cases.32 Interestingly, and unlike earlier protocols described by Ramirez et al., no significant difference in duration of response was noted for tumours that extended (radiographically) either <42% or >42% of bone length.32,33

More recently, Mueller et al. compared three fractions of 8 Gy (electrons; days 0, 7 and 21) and four fractions of 6 Gy (electrons; days 0, 7, 14 and 21) palliative protocols in 54 dogs, both with a total dose of 24 Gy.41 In this study, 45 dogs (83%) experienced pain relief during or following treatment. The median duration of effect was 53 days, with both protocols proving effective for palliation of clinical signs in dogs with appendicular OSA.41

Curative-intent RT

Despite OSA previously being thought to be a radiation-resistant tumour,43 some reports detail significant tumour necrosis in OSA after RT.44–46 These reports have served as the basis for investigating RT for curative-intent purposes. Until recently, no curative-intent RT strategies existed for dogs with appendicular OSA. Since 2004, investigative curative-intent RT for canine OSA has been described with either curative-intent full-course fractionated external beam protocol (CI-F), single megadose RT as part of an intra-operative extracorporeal irradiation (IORT) limb-sparing procedure, and as part of a stereotactic radiosurgery (SRS) protocol.3,18,28 With any of these curative-intent strategies, RT is delivered for local tumour control, while chemotherapy (either platinum compounds or doxorubicin) must be given for metastatic disease.

Curative intent – fractionated external beam protocol

CI-F RT delivers a lower dose per fraction on a daily basis and a higher total dose of radiation in an attempt to offer long-term local tumour control while minimising the late effects of radiation that occur more frequently with large fraction RT protocols. Fractionated RT is commonly used in veterinary medicine, but has only been reported twice for canine OSA, with limited success.3,47 Walter et al. report a response rate of 100%, with 50% of dogs (appendicular OSA) showing no progression of local disease based on radiographs and clinical signs before death or the end of the study.3 Median local control time was 196 days and MST was 209 days for dogs with appendicular OSA.3 These results were obtained from nine dogs treated on a Monday through Friday schedule, with the majority of dogs receiving a total radiation dose of 57 Gy (range 48–57 Gy), with most dogs receiving 3 Gy per fraction (range 3–5 Gy).3 These results did not show a substantial improvement over reported palliative protocols, and although treatments were well tolerated, no substantial local disease control or survival benefit was evident. However, limitations of the study including a small number of cases, variation in RT protocol and lack of consistent and definitive local outcome measures (necropsy and histopathology) may have affected these results. Interestingly, 66% of appendicular OSA evaluated with histopathology after RT showed no evidence of viable tumour cells.3 There is a similar paucity of published data involving the use of chemotherapeutic agents, such as cisplatin or carboplatin, which are often used in conjunction with definitive RT for proposed control of metastatic disease and chemical radiopotentiation.3,47,48 Walter et al. |report 4 of 10 cases of appendicular OSA being free of macroscopically visible metastasis at the time of death or euthanasia.3 The same limitations for interpretation of local disease control apply to metastatic disease also.

The Comparative Oncology Lab at the University of Florida – College of Veterinary Medicine (UF-CVM) has recently confirmed moderate radioresistance of canine OSA cell lines in vitro with a relatively low alpha to beta ratio and high survival fraction at 2 Gy.49 Such findings may explain why OSAs may not respond well to conventional fractionated radiotherapy protocols and suggest that larger doses per fraction are needed to induce greater tumour cell kill. Thus, radiation treatment options that deliver large doses per fraction, while sparing normal surrounding tissue, may be more effective in achieving local tumour control.

Intra-operative extracorporeal radiation and radiation in situ

Limb sparing with extracorporeal IORT has been investigated as a curative-intent, single-fraction protocol that involves isolation and exteriorisation of the bone tumour segment, so that a single fraction of 70 Gy can be safely delivered to it. Extraneous irradiated soft tissues are excised and the irradiated bone is reduced and stabilised with internal fixation18,27 (Figs 1–3). The biologic effect of a single dose of IORT is equivalent to two to four times the same dose of radiation delivered using fractionated external beam radiation protocols.50,51 Furthermore, 70 Gy IORT is tumouricidal to bone tumours and results in considerable necrosis of OSA lesions.18,27 Experimental studies using single-fraction, high-dose IORT have shown that peripheral nerves, muscle and skin are particularly radiation sensitive.52 In contrast, bone matrix, ligaments and articular cartilage are relatively resistant to high doses (50 Gy single dose) of radiation, which may allow preservation of normal joint and limb function.53 A principal advantage of extracorporeal IORT is that the radiation field can be focused on the target volume while sparing adjacent normal and radiosensitive soft tissue structures.18 The clinical benefits of limb-sparing protocols using extracorporeal IORT include the maintenance of autogenous bone scaffold with good anatomic fit, preservation of limb and joint function and, in some cases, good local tumour control.18 Specific surgical and RT techniques for extracorporeal IORT have been reported for appendicular OSA in the distal radius, proximal humerus, distal tibia and intercalary locations.18,27

Figure 1.

Preparation of an OSA of the distal radius for IORT. An osteotomy of the radius has been made several centimetres proximal to the gross tumour margin. The soft tissues have been dissected off the tumour, and the distal radius is suspended for radiation treatment through a sterile cord.

Figure 2.

A gelatin radiation bolus is tied to the dissected tumour to allow delivery of a uniform radiation dose to the entire tumour. The proximal aspect of the isolated bone segment is spared from irradiation to maintain cell viability for osteotomy healing.

Figure 3.

After tumour irradiation, the osteotomy is stabilised in compression with appropriately sized and numbered orthopaedic plates and screws.

Liptak et al. reported 10 of 13 dogs (77%) were improved clinically after IORT and assessed as having good to excellent limb function.18 Median local disease-free interval was 274 days and MST was 298 days when combined with chemotherapy.18 Adjunctive chemotherapeutic protocols were varied among individuals from both IORT studies, and metastatic disease was present in approximately 50% of cases at the end of the study periods, which is comparable to Walter et al.3,18,27 While IORT appears to have comparable success to other definitive and PTRT strategies, a large number of dogs (69–100%) experienced postoperative complications, including deep infection, fracture of irradiated bone and implant failure, especially for distal lesions that have poor soft tissue coverage.18,27 Both Liptak et al. and Boston et al. stress the importance of strict case selection criteria to minimise complications; good soft tissue coverage allowing revasculatisation of irradiated bone by extra-osseous and periosteal vessels, therefore tumours in the diaphysis and upper extremity are preferred; minimal soft tissue involvement and no involvement of the ulna; selecting cases with <50% of the length of the radius involved to ensure adequate greater than three bicortical screws proximal to the osteotomy.18,27

Stereotactic radiosurgery

Conventional RT relies on the use of fractionated protocols to minimise damage to surrounding healthy tissues.28 Conversely, SRS uses multiple, noncoplanar beams of radiation that are stereotactically focused on the target to deliver the entire radiation dose in a single treatment.28 SRS minimises damage to healthy surrounding tissues by relying on the extreme accuracy of radiation delivery to a tumour and a steep dose gradient between the tumour and the surrounding normal tissues28 (Fig. 4). The obvious benefits of this technique over fractionated protocols include fewer anaesthetic episodes and possibly a greater biologic effect on the tumour tissue.28 SRS has been reported in human and veterinary patients for the treatment of intracranial abnormalities, but only one report exists describing the use of SRS for the treatment of appendicular OSA.28 Farese et al. described a technique for SRS for appendicular OSA in 11 dogs in which treatment plans were initially designed to ensure the peripheral border of the tumour was included within the 50% isodose shell (minimum dose of 20 Gy) and that the central portion of the tumour was included within the 75% isodose shell (30 Gy).28 Pretreatment preparation involved placement of a targeting array and contrast-enhanced computed tomography images.28 This technique is well described by Farese et al. and will not be detailed here.28 In this report, two treatment protocols were documented. The first five dogs received SRS alone with lower doses of radiation (minimum dose of 20 Gy) and much less experience in treatment planning for appendicular OSA compared with subsequent cases. With the first protocol, tumour-associated swelling and lameness were subjectively improved within 21 days and this effect continued for at least 3 months after treatment (median time to evidence of tumour progression was 105 days).28 The second group of six dogs received larger doses of radiation (periphery of the tumour receiving 30–35 Gy, central portion of the tumour receiving approximately 40–50 Gy) and carboplatin (300 mg m−2, intravenously [IV]) was administered 30 min before irradiation with either carboplatin alone (carboplatin 300 mg m−2, IV every 3 weeks for four treatments) or doxorubicin (carboplatin 300 mg m−2, IV and doxorubicin 30 mg m−2, IV alternating every 3 weeks for four treatments each). The addition of chemotherapy to the SRS protocol was in response to progressive disease in 60% of dogs treated in the first group. Interestingly, no local disease progression was noted in the follow-up period in the second group of dogs, although follow was short for some dogs in this group.28 Overall MST for these 11 dogs was 363 days.28 The SRS technique has subsequently been modified by attempting to now cover the entire tumour with a 30- to 35-Gy isodose line (Fig. 4). Recently obtained (unpublished) histopathology from a dog with a proximal humeral OSA treated with this protocol showed 100% necrosis of the lesion (Farese JP, University of Florida, Gainesville, FL, personal communication). Although such a treatment plan is usually possible in the proximal humeral location, the proximity of the skin to the outer cortical bone in the distal radius makes delivery of these doses possible only when the diseased tissue is confined to the medullary cavity and endosteum (Farese JP, University of Florida, Gainesville, FL, personal communication) (Figs 5–7).

Figure 4.

A computer tomography-based treatment plan for SRS of a proximal humerus OSA. The radiation dose gradient is represented by colour-coded isodose lines. (Approximate values: blue = 35 Gy, red = 32.5 Gy, pink = 30 Gy and green = 20 Gy; at the centre of the lesion, doses of 50–60 Gy are reached.) Notice the steep dose gradient allowing a tolerable dose to the skin.

Figure 5.

A lateral radiograph of a good SRS candidate. Note the relatively small lesion size and area of minimal of osteolysis. The abundant soft tissue between the edge of the tumour and the skin in the proximal humeral location (versus the distal radius) allows a high radiation dose to be delivered without causing serious skin injury.

Figure 6.

A Great Dane with a distal radial OSA lesion that was too advanced for SRS. The proximity of the large, soft tissue tumour component to the overlying skin makes effective treatment difficult without causing skin necrosis.

Figure 7.

Craniocaudal radiographic image of the lesion similar to that in figure. Note the large tumour size, soft tissue component and extensive degree of tumour lysis.

While the results of this study indicate that SRS is a promising nonsurgical limb-sparing treatment modality for canine appendicular OSA, further clinical investigation on a larger number of dogs is needed to validate SRS as a curative-intent treatment option. Specifically, whether the higher radiation doses currently delivered (Fig. 4) in combination with chemotherapy consistently result in local tumour control. Furthermore, the need for highly specialised equipment and personnel has currently limited its use to two veterinary academic institutions (UF-CVM and CSU-ACC). In addition, patients with stage III (metastatic) disease, lesions with advanced osteolysis and/or large tumour volumes (e.g. greater than 5 cm diameter in the transverse plane; length of tumour involvement along the bone is not a limiting factor) are not considered good candidates (Farese, personal communication).


Radioisotopes offer a unique method of radiation delivery when combined with ligands. Ligands function as targeting agents that allow selective targeting of cancers for radiation dose delivery by the radioisotope. Collectively, these agents are known as radiopharmaceuticals. The most commonly reported therapeutic radiopharmaceutical is Samarium-153-lexidronam.54 Samarium153 is the radioisotope that delivers a radiation dose to the target by beta decay (therapeutic effect is concentrated and limited to tissues within a 2- to 3-mm radius) and has the added property of gamma decay (i.e. the distribution of the radioisotope can be monitored using a gamma camera). The ligand lexidronam (ethylene diamine tetramethylene phosphonate [EDTMP]) is an amino-bisphosphonate that, like the bone scan agent methylene diphosphonate (labelled with technetium – 99 m), localises in areas of increased bone metabolism making the radiopharmaceutical Samarium-153-lexidronam (153Sm-EDTMP) suitable to consider for treatment of skeletal neoplsia.54,55 Lexidronam is not only selective for the cancer but also will localise in areas with red marrow activity.

The first report of the radioparmaceutical Samarium-153-lexidronam (153Sm-EDTMP) to treat canine bone tumours was by Lattimer et al.56,57 They treated 40 dogs with naturally occurring OSA. Dogs were randomised to receive either a single dose of 37 MBq kg−1 or two doses a week apart. No significant differences were found between the two groups as assessed by limb function and radiographic improvement, although early tumours and metastatic lesions appeared to show some response.

Milner et al. treated nine dogs with OSA (single intravenous dose of 37 MBq kg−1153Sm-EDTMP). A single dog had a dramatic response to therapy, but all the others cases showed progression of the disease.58 In 1999, Aas et al. treated 15 dogs with 153Sm-EDTMP.59 Dogs were given between one and four doses of 153Sm-EDTMP at 36-57 MBq kg−1. Their conclusions were that a favourable high tumour dose was achieved in the tumour compared with surrounding tissue and that pain relief was achieved and, in some cases, tumour growth was delayed. Median survival was 150 days and at necropsy 11 of 15 dogs were found to have metastases. No serious side effects were observed. Most recently, Barnard et al. reported 35 dogs receiving 153Sm-EDTMP.55 Of the 32 dogs with appendicular tumours, 20 (63%) had an improvement in the severity of lameness 2 weeks after administration of the first dose of radioactive samarium, eight (25%) had no change in the severity of lameness and four (12%) had a worsening. MST for the 32 dogs with appendicular tumours was 93 days, with three (9.4%) alive after 1 year. This was not significantly different from the MST of 134 days for a historical cohort of 162 dogs with appendicular OSA that underwent amputation as the only treatment.55

Samarium-153-EDTMP has also been administered locally through isolated limb perfusion (ILP).60 In 2006, Ehrhart presented the results from five dogs treated with ILP (1 h) using 37 MBq 153Sm-EDTMP kg−1 body weight. Limb function after this procedure was excellent in all dogs, and minimal local or systemic toxicity was observed.60

Palliation of pain in dogs with appendicular OSA treated with 153Sm-EDTMP appears to be unpredictable but very safe for the patient. Cost availability of 153Sm-EDTMP and of alternative radiotherapies should be considered when prescribing radiopharmaceutical therapy.

Radiation-induced adverse effects

Radiation is most toxic to proliferating cells in the late G2 and M phases of the cell cycle at the time of treatment. Most cells receiving a single radiation treatment in the G1 or S phases will repair cellular damage before progressing through the G2 and M phases and appear unaffected. Neoplastic and normal cells use the same basic mechanisms of cell division and therefore RT kills both neoplastic and normal cells within rapidly dividing renewing tissues.

Acute radiation-induced adverse effects occur during, or shortly after, a treatment course and are caused by death of rapidly dividing normal cells in the renewing tissues. Acute effects of tumour irradiation in dogs with appendicular OSA may include moist desquamation, oedema, ulceration and skin necrosis. These (potential) acute adverse effects usually limit the radiation dose per unit time but will usually resolve with appropriate supportive care and time.

Nonproliferative or slowly renewing tissues such as liver, kidney, bone, connective tissues, muscle and nerves are also affected by RT. Because cellular turnover is very slow in these tissues, late adverse effects may only manifest months or years after treatment. Late effects of appendicular OSA irradiation include muscle fibrosis, alopecia, leucotrichia, bone necrosis and tumour induction and can be more devastating, irreversible and have a greater impact on quality of life than acute effects.52,53 If repair of sublethal damage in nonproliverative or slowly renewing tissues has not (likely) occurred before subsequent treatment, the total cumulative dose must be reduced.

As a general rule, the treatment intensity (dose per unit time) is limited by acute adverse effects to rapidly renewing tissues, while cumulative dose is limited by chronic adverse effects or irreversible toxicity to nonrenewing tissues.

Time-dose fractionation (TDF) is used to describe the development of a RT protocol, considering the total dose of radiation given over a specific period of time (e.g. 10 Gy in a single fraction versus 5 × 2 Gy fractions per week). Large fraction TDF protocols (such as most palliative protocols for OSA in dogs) result in decreased acute adverse effects but a greater incidence of late adverse effects. The converse is true for conventional TDF protocols (more frequent acute adverse effects with fewer late adverse effects). In human patients, no difference in pain palliation is achieved with both single- and multiple-fraction RT, but medical and societal costs are reduced with single-fraction treatments.38 In dogs, PTRT is not associated with significant acute adverse effects, although alopecia, erythema and depigmentation have been reported, but these do not reduce quality of life.1 These adverse sequelae may be more common with high doses per fraction and high total cumulative radiation doses.1,28,41,42 A single fraction of radiation results in greater damage to late-responding healthy tissues than the same dose divided into smaller fractions.28,35 Many dogs with appendicular OSA treated with palliative-intent die or are killed before late adverse effects become apparent. Subsequently, PTRT protocols delivering 8–10 Gy in each fraction are usually tolerated without clinically significant late adverse effects.1,33,42

Definitive RT is associated with a higher complication rate, including alopecia (63%), bone resorption (31%), desquamation (27%), bone necrosis (23%), hyperpigmentation (18%), late muscle fibrosis (15%), leucotrichia (9%), fibrosis of the irradiated bone (8%) and acute skin necrosis (8%).3,5,28 Pathologic fracture is reported in all definitive RT strategies (54% IORT, 36% SRS and 14% full-course external beam RT) and is likely the result of tumour destruction caused by radiotherapy rather than adverse effects of radiation on normal tissues.3,18,28 The incidence of pathologic fracture may be reduced with judicious case selection, such as dogs with minimal cortical destruction and bony lysis.28

In the event of radiation-induced adverse effects, interventions such as stabilisation of pathologic fractures and management of soft tissue wounds may afford a good quality of life. Therefore, these late effects may be acceptable in the context of definitive treatment. Desquamation and acute skin necrosis can be difficult to manage because of the poor healing potential of irradiated soft tissues. A major advantage of SRS is that surrounding soft tissues remain well vascularised and therefore are available to be used as transposition flaps in the event of desquamation and focal skin necrosis.


Appendicular OSA is a relatively common tumour in large breed dogs. RT has a role in both the palliative- and curative-intent treatment of the local tumour in dogs with appendicular OSA. PTRT achieves analgesia with reliable response rates (50–93%) for approximately 100 days (range 53–180 days) at which time many dogs die or killed because of disease progression or pathologic fracture. Unfortunately, many of the PTRT reports included in this review are retrospective in nature. Because of the inherent biases of retrospective analyses (small sample populations, variable treatment regimens, inconsistent follow-up, lack of randomisation), the evaluation of efficacy between the various protocols must be made with caution. However, the major benefit of all PTRT protocols is that during the palliated period, quality of life is improved and RT-induced side effects are rare. Curative-intent RT can achieve good to excellent local tumour control but is also associated with a moderate to high rate of radiation-induced complications, which may be ameliorated with appropriate case selection (small intramedullary lesions for SRS and proximal extremity lesions with good soft tissue coverage for IORT). CI-F therapy is largely unsuccessful and curative-intent RT should not replace current standard of care for canine appendicular OSA (amputation and chemotherapy). When limb amputation is not possible, single-fraction curative-intent RT may be successful at controlling local disease, although therapeutic success is reliant on the incidence and management of surgical complications (IORT) and appropriate patient selection (SRS).