Radiation Oncology Refresher 2026 — Odcinek 4: reirradiacja (wersja angielska)

Adam: Welcome to Radioonkolog dot P L Podcast. Today's subject is reirradiation, not a review of whether a second radiation course is technically possible, but a discussion of when it is clinically defensible, what dose is worth delivering, which structures must be reconstructed, and where the evidence is strong enough to alter practice.

Marta: And we should begin with a warning. Reirradiation has accumulated a large vocabulary of apparently precise cumulative constraints. Some are supported by human dose toxicity data, some by Delphi consensus, and some by institutional habit. Those are not equivalent levels of evidence. A mathematically acceptable composite plan is not automatically an acceptable treatment.

Adam: The major development by 2026 is methodological. We now have the Estro E O R T C classification of reirradiation, formal Estro guidance on cumulative dose evaluation, ReCOG reporting recommendations, new international head and neck guidance endorsed by ReCOG, Estro and Astro, and joint Estro ISRS recommendations for spine S B R T reirradiation. That does not eliminate uncertainty, but it gives us a common language for describing it.

Marta: We will organize the episode by intent: first radical or durable local salvage, then oligometastatic and oligoprogressive treatment, and finally palliation. But before we discuss sites, we need a common planning framework.

Adam: Let us refresh the Estro E O R T C terminology because it is easy to use reirradiation imprecisely. Type 1 reirradiation means that the new irradiated volume geometrically overlaps the previous irradiated volume. Type 2 reirradiation means that there may be no direct geometric overlap, but the cumulative exposure of an organ at risk creates a toxicity concern. If neither condition applies, this is repeat organ irradiation rather than reirradiation in the strict consensus definition.

Marta: Type 2 is clinically important. Two thoracic targets can be spatially separate while exposing different segments of the esophagus, bronchial tree, lung or spinal cord. Two pelvic courses can miss one another geometrically while still consuming the same bladder or bowel reserve. So the PTVs do not overlap is not a cumulative dose assessment.

Adam: The second classification is by clinical intent. We should state in the chart whether we are pursuing radical salvage, oligometastatic or oligoprogressive ablation, or palliation. The distinction determines the acceptable treatment burden, the minimum meaningful tumor dose and the degree of residual uncertainty we can tolerate.

Marta: It also determines the language we use. If respecting an organ at risk limit forces severe target underdosage, we should not describe the course as ablative simply because it uses five fractions. An O A R limited 25 gray in five fractions may be useful palliation, but it is not biologically equivalent to a properly delivered metastasis directed S A B R course.

Adam: My practical pre simulation checklist has eight questions. First: is this definitely viable malignant disease rather than necrosis, infection, postoperative change or fracture? Second: is the recurrence technically resectable or suitable for another local intervention such as ablation, embolization or brachytherapy? Third: is systemic disease controlled well enough that local control matters? Fourth: is the patient likely to live long enough to experience both the benefit and the late toxicity? Fifth: what toxicity remains from the first course? Sixth: can we reconstruct the previous dose with acceptable uncertainty? Seventh: can we deliver a biologically meaningful second dose? And eighth: does the patient understand that the major risks may include necrosis, fistula, hemorrhage, neuropathy, myelopathy or permanent organ dysfunction?

Marta: Those questions are more important than the interval alone. A long interval is favorable, but it is not an organ recovery meter. A patient ten years after irradiation with a persistent pharyngeal ulcer, exposed mandible or severe aspiration may be a worse candidate than someone eighteen months out with a small recurrence, excellent function and a clearly separable target.

Adam: For radical treatment, biopsy should usually be pursued when feasible and when it can alter management. That is particularly important for intracranial enhancement that may represent radionecrosis, for a new lung lesion that could be a second primary rather than recurrence, and for prostate salvage where imaging only recurrence can lead to substantial overtreatment.

Marta: A short interval, progression during the first radiation course, continuing grade 3 or 4 toxicity, active fistula, necrotic mucosa, exposed major vessel, uncontrolled infection, rapidly progressive systemic disease or inability to bound the previous dose are all strong arguments against radical reirradiation. They are not always contraindications to limited palliation.

Adam: The minimum acceptable previous treatment dataset is the planning C T, R T structure, R T dose and R T plan; the actual treatment dates; delivered rather than merely prescribed fractions; any adaptive or boost plans; and previous brachytherapy D 2cc data where relevant.

Marta: A previous discharge letter saying 70 gray to the head and neck is not a dose reconstruction. If the DICOM dataset is unavailable, we should produce a bounded estimate: a likely prior dose and a conservative upper dose scenario. False precision is worse than explicitly acknowledged uncertainty.

Adam: We should also reconstruct the clinical history. Was there grade 3 pneumonitis? A feeding tube? Osteoradionecrosis? Urethral stricture? A vertebral compression fracture? Radiation injury changes the effective reserve even when the numerical dose appears acceptable.

Marta: For the show notes, the working equation remains: E Q D two for alpha over beta equals n times d, multiplied by d plus alpha over beta, divided by two plus alpha over beta. We generally report tumor and acute tissue dose using an alpha over beta of 10 gray, spinal cord and other central neural structures using approximately 2 gray, and many late gastrointestinal or genitourinary endpoints using approximately 3 gray. The selected alpha over beta must be written in the report.

Adam: And one should never add 70 physical Gy from one course to 30 physical Gy from another and call the result 100 gray. Equally, one should not treat an E Q D two value calculated to one decimal place as if the underlying radiobiology were known to one decimal place.

Marta: Particularly after stereotactic fractionation. The linear quadratic model remains the common reporting framework because we need a shared scale, not because it perfectly models vascular injury, mucosal necrosis, immunological effects or very high doses per fraction. The current Estro and ReCOG guidance emphasizes biological conversion, registration uncertainty, quality assurance and transparent reporting rather than pretending that one algorithm resolves the problem.

Adam: Let us make the hierarchy practical. The preferred approach is a verified three dimensional accumulation using rigid or deformable registration when the anatomy is sufficiently comparable and the deformation is credible. If deformation is unreliable, use rigid local registration around the relevant stable anatomy. If neither is defensible, sum regional maximum doses, dose volume metrics or representative points. And when the prior information is incomplete, use a conservative worst case estimate.

Marta: Do not let a deformable map manufacture certainty. Deformable registration may fail clinically in a re expanded lung, after major neck surgery, after vertebral collapse, after removal of a large pelvic mass, with markedly different bladder or rectal filling, or when a different loop of bowel occupies the high dose region.

Adam: That last example is crucial. A beautifully warped bowel dose is not automatically biologically meaningful. The same physical coordinate may contain duodenum during one course and jejunum during another. Conversely, two different loops may each receive a near maximum dose. In mobile bowel, I want both the mapped result and a conservative same functional subunit receives both maxima scenario.

Marta: For serial organs, I also prefer consistent near maximum dose metrics, D 0.03cc, D 0.1cc, D 0.5cc or D1cc, rather than a single unstable voxel, unless the source publication explicitly defines a point maximum.

Adam: Our default is simple: zero recovery unless a site specific protocol or institutional policy explicitly justifies another assumption.

Marta: That is intentionally conservative. A frequently used spinal cord heuristic assumes partial recovery over six to twelve months, but it should not be transferred automatically to optic nerves, carotids, plexus, bowel, airway or ulcerated mucosa.

Adam: Nor should one apply recovery twice. If a published cumulative limit was already developed using an interval dependent recovery assumption, we should not discount the first course again before applying it.

Marta: The interval belongs in the clinical interpretation even when the numerical calculation assumes zero recovery. It is a covariate associated with selection, biology and tissue repair, not a universal conversion factor.

Adam: For each critical structure, I want the contour definition, previous course metric, new course metric, composite metric, alpha over beta, registration method, recovery assumption, interval, best estimate, conservative estimate, and the source of the selected constraint.

Marta: I would add the expected failure mode. Cord D 0.1cc 66 Gy2 is not enough. We should state whether the relevant concern is myelopathy, fistula, carotid blowout, airway necrosis, bowel perforation, urethral stricture or radionecrosis. The patient and the multidisciplinary team need to understand what the number represents.

Adam: Let us do the numbers clinicians most often look up. Start with the spinal cord.

Marta: For spine S B R T reirradiation, the widely used lower risk framework is a cumulative thecal sac maximum of approximately 70 Gy2, a retreatment contribution of no more than approximately 25 Gy2, a retreatment to cumulative E Q D two ratio no greater than 0.5, and an interval of at least five months. These are lower risk criteria derived from the available human data, not a guarantee against myelopathy.

Adam: For de novo S B R T reference, the approximate true cord point maxima associated with a one to five percent myelopathy risk are 12.4 to 14 gray in one fraction, 17 gray in two, 20.3 gray in three, 23 gray in four and 25.3 gray in five. But in reirradiation both the second course limit and the cumulative calculation must be satisfied.

Marta: We also need contour discipline. A limit derived for the true cord is not automatically a cord P R V constraint, and a limit derived for the thecal sac is not automatically a true cord constraint. Record which structure was used.

Adam: The 2025 American Radium Society criteria provide a useful contemporary thoracic framework. The proposed composite limits include esophageal V 60 below 40 percent and maximum below 100 gray E Q D two; bilateral lung V 20 below 40 percent; heart V 40 below 50 percent; aortic and great vessel maximum below 120 gray E Q D two; tracheal and proximal bronchial tree maximum below 110 gray; spinal cord maximum below 57 gray; and brachial plexus maximum below 85 gray.

Marta: Those should be treated as upper planning bounds, not planning objectives. An esophageal composite maximum of 99 gray is not reassuring when the tumor is ulcerating into the esophageal wall, the patient has severe dysphagia and concurrent systemic therapy is planned.

Adam: Similarly, a composite bilateral lung V 20 can conceal a high regional dose to previously injured functioning lung. We should also review mean lung dose, regional overlap, prior pneumonitis distribution, perfusion if available, and the amount of preserved uninvolved lung.

Marta: For the carotid, a cumulative D1cc around 120 Gy3 is a useful retrospective risk boundary. For mandible or other irradiated bone, a cumulative near maximum around 119 to 120 Gy3 has been associated with osteoradionecrosis. Those are risk associated cut points, not validated safe limits.

Adam: In five fraction head and neck S B R T, one retrospective series observed no carotid blowout below a current course carotid D 0.1cc of 47.6 gray. But the number of events was small, and ulceration, infection, exposed artery, encasement and necrosis may matter more than a decimal point dose difference. Nonconsecutive fractionation remains preferable.

Marta: For the brachial plexus, a composite maximum around 80 to 85 Gy2 is a reasonable conservative target. For pharynx, larynx, oral cavity and previously ulcerated mucosa, there is no universally validated cumulative number. Circumferential dose, baseline swallowing, mucosal integrity and treated length often dominate the risk.

Adam: The international pelvic S A B R Delphi provides useful but unevenly supported values. The two stronger consensus points were a bladder D 0.5cc up to approximately 110 Gy3 when the interval is at least twelve months and recovery is assumed, and an optimal cauda equina or sacral plexus D 0.1cc of approximately 67 Gy2 without recovery.

Marta: Other suggested values did not achieve the same consensus. These included small bowel D 0.5cc around 70 Gy3 without recovery or 90 Gy3 with recovery; colon, sigmoid or rectum D 0.5cc around 80 Gy3 without recovery or 100 Gy3 with recovery; and cauda equina or sacral plexus D 0.1cc around 85 Gy2 with recovery. These are institutional planning guardrails, not established tolerance doses.

Adam: And they should not be transferred automatically to duodenum or stomach. For upper abdominal hollow viscera, kidneys and liver, we still lack a sufficiently validated universal cumulative constraint table. Use a residual dose budget approach, conservative accumulation and more fractions when the target contacts a luminal organ.

Marta: There is no universal composite constraint for normal brain, brainstem, chiasm or optic nerves. For repeat S R S, the decision should incorporate the second course high dose normal brain volume, overlap of the previous and new dose gradients, target size, interval and whether the imaging could represent radionecrosis.

Adam: In practice, the optic pathways should be approached as organs with very limited demonstrable recovery. If the first course was already near conventional tolerance, a substantial second high dose should be described to the patient as carrying a material risk of permanent visual loss, not merely as constraint challenging.

Adam: The first principle is that reirradiation should not displace curative salvage surgery when an R 0 resection with acceptable function is realistic.

Marta: Correct. The sequence is usually salvage surgery first for resectable disease, postoperative reirradiation for selected adverse pathology, and definitive reirradiation for unresectable disease or when surgery would be excessively morbid. The 2026 international consensus endorsed by ReCOG, Estro and Astro formalizes patient selection, imaging, target definition, dose reconstruction and toxicity assessment.

Adam: The MIRI recursive partitioning classification remains clinically useful. Class one comprises patients more than two years from the original radiation whose recurrence has been resected. Class two includes unresected disease more than two years later, or an interval of two years or less without baseline organ dysfunction. Class three combines an interval of two years or less with organ dysfunction such as feeding tube or tracheostomy dependence.

Marta: In the original analysis, two year overall survival was approximately 62 percent in class one, 40 percent in class two and 17 percent in class three. That does not make class three an absolute prohibition, but it is a strong warning against six or seven weeks of aggressive treatment without a compelling individualized rationale.

Adam: Let us answer the question planners actually face: what exactly goes into the target?

Marta: For definitive conventional reirradiation, the G T V includes all contrast enhancing, F D G avid, endoscopically or clinically apparent recurrent primary and involved nodal disease. The high risk C T V is usually the G T V plus zero to five millimetres, anatomically edited. It can be extended selectively toward five to ten millimetres along a convincing submucosal, perineural or infiltrative route, but not expanded isotropically through air, intact cortical bone or uninvolved fascial barriers. Only involved nodal levels or specifically threatened adjacent anatomy are treated. Routine elective bilateral neck reirradiation is not recommended. The P T V is generally two to five millimetres, depending on immobilization, daily volumetric imaging and intrafraction stability. Current consensus and multi institutional practice support deliberately compact targets rather than recreating the original fields.

Adam: For skull base disease, the C T V may need to follow the involved foramen or named nerve pathway. For a sharply defined retropharyngeal node, it does not need to include the entire retropharyngeal space. For a recurrence in the larynx, it should follow the involved cartilage, paraglottic space or mucosal compartment rather than being a simple geometric sphere.

Marta: After salvage surgery, the target must be reconstructed from the preoperative disease, not only the postoperative cavity. Include the preoperative gross extent, residual disease, the surgical bed directly at risk, the positive or close margin region, and any nodal bed with extranodal extension. Add a small anatomically edited margin, usually five to ten millimetres. Do not automatically include the entire surgical incision, all dissected neck levels or the entire previously irradiated neck.

Adam: This is one of the most common contouring errors: looking only at the postoperative C T and forgetting where the gross recurrence was attached to the carotid sheath, skull base, prevertebral fascia or mandible.

Marta: For definitive conventional reirradiation, a practical range is 60 to 66 gray in 30 to 33 fractions. In carefully selected patients with gross disease and acceptable organ at risk geometry, at least 66 gray has been associated with better disease control in the MIRI analysis. Postoperatively, 54 to 60 gray is reasonable for a high risk microscopic bed, with 60 to 66 gray considered for gross residual disease, a positive margin or a very high risk extranodal extension region. For a small, well demarcated recurrence suitable for S B R T, 40 to 44 gray in five nonconsecutive fractions is a common framework. For a larger lesion, a lesion touching mucosa, carotid, mandible or plexus, or a target with substantial necrosis, use 10 to 30 fractions rather than forcing a five fraction regimen.

Adam: The MIRI data did not show a clear benefit from routine elective nodal irradiation or routine hyperfractionation across the heterogeneous head and neck cohort. That should not be interpreted as proof that hyperfractionation is biologically irrelevant; rather, it means that universal BID treatment has not been validated for all recurrent head and neck sites.

Marta: The Janot randomized trial remains foundational. After macroscopically complete salvage surgery, patients received observation or postoperative reirradiation to 60 gray with concurrent 5 fluorouracil and hydroxyurea. The combined treatment improved locoregional control and disease free survival, but not overall survival. At two years, grade 3 or 4 late toxicity was reported in 39 percent of surviving patients in the reirradiation arm versus 10 percent with observation.

Adam: The practical interpretation is not that every postoperative recurrence should be reirradiated. It is that patients with positive margins, gross residual disease, extranodal extension or another major local risk feature may gain disease control benefit, at a substantial late toxicity cost.

Marta: And the systemic regimen is historical. The trial supports the oncological principle of postoperative reirradiation in selected high risk patients; it does not establish 5 FU and hydroxyurea as the preferred 2026 concurrent regimen.

Adam: The multicentre phase two trial of I M R T reirradiation with nivolumab is frequently discussed. The radiation dose was 60 to 66 gray in 30 to 33 fractions. Nivolumab began before radiation, continued every two weeks during treatment and then every four weeks for a total of approximately one year.

Marta: Among 51 evaluable patients, the one year progression free survival rate was 61.7 percent, exceeding the prespecified historical benchmark of 43.8 percent. Two year P F S was 43.2 percent, median overall survival was 21 months, and grade 3 or higher treatment related toxicity occurred in 12 percent, without reported treatment related deaths.

Adam: Those results are encouraging, but the trial was nonrandomized, used a historical control and contained a highly selected population: predominantly good performance status, mostly MIRI classes I and II, and many patients treated after salvage surgery.

Marta: So nivolumab plus reirradiation is a reasonable clinical trial or multidisciplinary option, not proof that checkpoint inhibition should be added routinely. We also need longer follow up for late necrosis, dysphagia and vascular events.

Adam: My strongest head and neck contraindications are an exposed or actively bleeding carotid, active fistula, ulcerated tumor with infection and necrosis, uncontrolled osteoradionecrosis, severe baseline aspiration, and a target requiring extensive circumferential reirradiation of the pharynx or larynx.

Marta: Carotid encasement by itself is not always an absolute contraindication. But encasement combined with ulceration, infection, necrosis, prior bleeding and a high cumulative dose is a particularly hazardous phenotype.

Adam: Nasopharyngeal carcinoma deserves its own section because it has unusually strong randomized evidence.

Marta: For limited, surgically accessible local recurrence, endoscopic nasopharyngectomy should be discussed before reirradiation. In a phase three trial, three year overall survival was 85.8 percent with endoscopic surgery versus 68.0 percent with I M R T, with fewer fatal late complications in the surgical arm.

Adam: That result applies to selected resectable recurrences in experienced surgical centres. It does not imply that extensive skull base, intracranial or parapharyngeal recurrence is suddenly a surgical disease.

Marta: The second phase three trial compared hyperfractionated I M R T, 65 gray in 54 fractions of approximately 1.2 gray twice daily, separated by at least six hours, with standard fractionation of 60 gray in 27 daily fractions. At a median follow up of 45 months, grade 3 or higher late complications occurred in 34 percent with hyperfractionation versus 57 percent with standard fractionation. Three year overall survival was 74.6 percent versus 55.0 percent, with a hazard ratio for death of 0.54. Fatal late complications occurred in 7 percent versus 24 percent, respectively.

Adam: That is one of the few reirradiation studies that deserves the term practice changing. Hyperfractionation reduced severe late toxicity and improved survival. But we should not sanitize the result: a 7 percent fatal late complication rate remains substantial.

Marta: Correct. The regimen is a standard option for suitable locally advanced recurrent NPC, but it requires meticulous BID logistics, a minimum six hour interval, compact targets and aggressive management of mucosal, vascular and temporal lobe risk.

Adam: The G T V is the recurrent disease defined by M R I, nasoendoscopy and, where informative, P E T. The C T V is usually G T V plus three to five millimetres, selectively extended along submucosal spread, involved skull base foramina, parapharyngeal extension or a named cranial nerve pathway. Do not automatically re cover the entire original nasopharynx or bilateral elective neck. The P T V is commonly two to three millimetres with robust immobilization and daily volumetric guidance.

Marta: Explicitly contour both carotids, temporal lobes, brainstem, optic nerves and chiasm, cochleae, relevant cranial nerve pathways and the mucosal surface. A plan can satisfy a whole organ temporal lobe metric while creating a clinically important focal hotspot at the skull base interface.

Adam: For radical thoracic retreatment, the minimum staging package is contrast enhanced thoracic C T, F D G P E T or C T, brain imaging and biopsy when feasible. We need to distinguish an isolated in field recurrence from a new primary, nodal progression and widespread systemic disease.

Marta: Pulmonary function should be individualized rather than reduced to one exclusion threshold. A patient with a low FEV1 but a small peripheral target and little additional functioning lung exposure may be treatable. A patient with apparently adequate spirometry but severe prior pneumonitis and a central target may not be.

Adam: The G T V includes the P E T- and C T defined recurrent primary or second primary and any involved nodes. For conventionally fractionated treatment, the C T V is generally zero to five millimetres, edited to airway, vessels, pleura, bone and other barriers. For S B R T, a separate C T V can often be omitted. Motion should be defined with four D C T, breath hold, gating or tracking. The P T V is commonly three to five millimetres. No elective mediastinal or hilar nodal irradiation. The international expert survey specifically supported involved site retreatment rather than elective nodal coverage.

Marta: For a larger locoregional recurrence, accepted once daily schedules include 60 gray in 30 fractions or 55 gray in 20 fractions. For a small peripheral lesion with limited high dose overlap, 45 to 50 gray in five fractions is reasonable. I would reserve 54 gray in three fractions for a genuinely peripheral target with highly favorable composite geometry. For central disease, use approximately 50 to 60 gray in eight to fifteen fractions. For ultracentral disease, especially when the P T V touches the proximal bronchial tree or esophagus, use ten to thirty fractions rather than aggressive three fraction treatment.

Adam: The prescription should not be chosen before reviewing the previous airway, esophagus, vessel, heart, lung and cord dose. Central and ultracentral are not merely location labels in reirradiation; they are descriptions of the likely fatal toxicity mechanism.

Marta: Proton therapy is attractive because it can reduce low- and intermediate dose exposure, but it does not make a dangerous target safe. Range uncertainty, anatomical change, motion, variable biological effectiveness near the distal edge and direct tumor invasion of the airway or esophagus still matter.

Adam: In a phase two study of proton reirradiation followed by pembrolizumab, 22 patients received 60 to 70 gray, most for central disease and most with concurrent chemotherapy. Median progression free survival was 8.8 months and median overall survival 22.8 months. Only one isolated in field failure was reported, but grade 3 or higher toxicity occurred in 45 percent, and two likely reirradiation related grade 5 events occurred: an aorto esophageal fistula and fatal hemoptysis.

Marta: That is the complete interpretation: impressive local control, small sample, selected patients, and a persistent late fatal toxicity signal. It is not evidence that adding pembrolizumab neutralizes the biological risk of central reirradiation.

Adam: A 2025 retrospective cohort of 130 patients treated with 60 to 72 gray reported median overall survival of 17.4 months and median P F S of 8.1 months. Consolidative immunotherapy was associated with longer survival, 27.8 versus 15.8 months, and better locoregional control, but grade 5 toxicity occurred in 6.9 percent. The immunotherapy association was nonrandomized and vulnerable to selection and immortal time bias.

Adam: Consider a fit patient three years after 60 gray in 30 fractions with a 2.5 centimetre left hilar recurrence. P E T shows no distant disease. The recurrence contacts the left main bronchus and lies four millimetres from the esophagus.

Marta: I would not begin by selecting 50 gray in five fractions. First reconstruct the bronchial tree, esophageal, great vessel and cord dose. Contour the gross recurrence without elective nodes. Use four D C T or breath hold. If a meaningful ablative dose cannot be delivered in eight to fifteen fractions without exceeding airway or esophageal tolerance, consider thirty fractions, surgery if feasible, systemic therapy, or a palliative goal.

Adam: And active hemoptysis, airway necrosis, tracheoesophageal fistula, bronchopleural fistula or active grade 2 or higher pneumonitis should stop the automatic radical pathway.

Adam: A major 2026 development is the Estro recommendation for the use, planning and delivery of reirradiation in locally recurrent rectal cancer, endorsed by Astro. It supports carefully selected preoperative or definitive reirradiation and provides a more consistent framework for patient selection, target definition and dose reporting.

Marta: For conventionally fractionated or hyperfractionated treatment, a practical range is 30 to 40 gray, delivered either in 2 gray daily fractions or approximately 1.2 to 1.5 gray twice daily. The objective is often to improve the probability of an R 0 resection rather than sterilize a large pelvic recurrence with radiation alone.

Adam: The G T V is the recurrent disease defined on pelvic M R I, planning C T, P E T when informative, endoscopy and operative findings. For conventional or preoperative reirradiation, the C T V is usually the G T V plus approximately five to ten millimetres, but it must follow the involved anatomical compartment. For a presacral recurrence, cover the involved presacral soft tissue and threatened sacral interface, not the entire presacral space. For a pelvic sidewall recurrence, follow the involved muscle, vessel or nerve interface. For an anastomotic recurrence, include the directly threatened anastomotic wall and contiguous mesorectal or pelvic tissue, not the entire previous mesorectum. For a sciatic notch recurrence, follow the gross disease and involved nerve or neural foramen pathway.

Marta: For pelvic S A B R, the C T V is generally zero to five millimetres and may be omitted for a sharply defined nodal target. Do not retreat the elective whole pelvis. The P T V is commonly three to five millimetres with daily CBCT. Two to three millimetres may be defensible with validated online adaptive treatment and reliable intrafraction monitoring.

Adam: The international Delphi accepted 30 to 37.5 gray in five or six fractions or 21 to 27 gray in three fractions, generally delivered on alternate days. Daily volumetric guidance or reliable fiducial guidance was considered necessary.

Marta: The Delphi also stated that organ at risk constraints generally take priority over complete P T V coverage. But that principle must not become an excuse for meaningless treatment. If a large fraction of the target receives a nonablative dose, reconsider the intent, fractionation, surgery or another modality.

Adam: The minimum pelvic contour set should include bladder, rectum, sigmoid and colon, small bowel, cauda equina, sacral plexus and femoral heads. Depending on level, add ureters, kidneys, genital structures, vagina, prostate, urethra and relevant major vessels.

Marta: Direct invasion of a mobile luminal organ, active fistula, uncontrolled infection, active pelvic necrosis and persistent grade 3 or higher late toxicity are strong contraindications to highly hypofractionated pelvic retreatment.

Adam: Prior grade 2 toxicity is not necessarily prohibitive, but it should lower the dose, enlarge the fraction number or change the treatment objective.

Marta: For a localized central or vaginal recurrence after previous pelvic radiation, image guided interstitial or intracavitary brachytherapy may be preferable to broad external beam reirradiation. The G T V is the M R I- and examination defined recurrence. The high risk C T V includes gross disease, directly infiltrated vaginal, parametrial or paracervical tissue, and the relevant pretreatment extent. Previous E B R T and brachytherapy D2cc values to bladder, rectum, sigmoid and bowel must be reconstructed.

Adam: Interstitial techniques are preferred when the lesion is thick, asymmetric, parametrial, periurethral or poorly covered by an intracavitary applicator. Prescription schedules are highly individualized, often approximately 24 to 36 gray in three to six HDR fractions, but implant geometry and remaining organ at risk reserve are more important than selecting a standard nominal dose.

Marta: The major unanswered question is whether preoperative reirradiation improves R 0 resection, quality of life and survival enough to justify its morbidity in the era of more effective systemic therapy. Trials such as GRECCAR 15 are designed to address chemotherapy plus pelvic reirradiation versus chemotherapy alone before surgery, while proton and carbon ion series remain difficult to interpret because of selection, centre expertise and biological dose model differences.

Adam: For selected ipsilateral breast tumor recurrence after previous lumpectomy and whole breast irradiation, second lumpectomy plus partial breast reirradiation is an established alternative to mastectomy.

Marta: Suitable patients generally have a unifocal recurrence no larger than approximately two to three centimetres, negative margins, clinically node negative disease, a visible cavity and no inflammatory recurrence, extensive skin involvement or diffuse suspicious calcifications. A long interval is preferable, although R T O G 1014 allowed more than one year.

Adam: The target construction is worth remembering. The cavity or G T V is defined by surgical clips, seroma and architectural change. The C T V is the cavity plus 15 millimetres, cropped to remain at least five millimetres below the skin and excluding posterior chest wall structures. The P T V is the C T V plus 10 millimetres, with an evaluation P T V excluding tissue outside the patient and the superficial build up region. The prospective regimen was 45 gray in 30 fractions of 1.5 gray twice daily, with at least six hours between fractions.

Marta: At five years, ipsilateral in breast re recurrence was approximately 5 percent, breast preservation 90 percent, overall survival 95 percent, and late grade 3 toxicity 7 percent, with no grade 4 events.

Adam: Moderate and ultrahypofractionated partial breast schedules, such as 40.05 gray in 15 fractions or 30 gray in five, are increasingly used, but their reirradiation evidence is less mature than the R T O G 1014 regimen. Convenience is not yet equivalent to long term validation.

Marta: Salvage prostate reirradiation should usually require biopsy confirmation when feasible, multiparametric M R I, and metastatic exclusion with P S M A P E T. The patient should have sufficient life expectancy, acceptable urinary function and no severe residual proctitis, incontinence, urethral stricture, fistula or previous grade 3 or higher GU or GI toxicity.

Adam: The Estro ACROP consensus provides a patient selection and S B R T framework, but the evidence remains largely nonrandomized.

Marta: For focal salvage, the G T V is the concordant M R I-, P S M A P E T- and biopsy defined recurrence. The C T V is generally G T V plus three to five millimetres, edited from the rectal wall and urethra when oncologically reasonable. The P T V may be approximately two to three millimetres posteriorly and three to five millimetres elsewhere, depending on tracking, fiducials, rectal stabilization and online adaptation.

Adam: If M R I, P E T and biopsy disagree, do not solve the disagreement with a larger empiric target. Re review imaging, pathology and sampling.

Marta: For whole gland salvage, the C T V is the prostate. Include proximal seminal vesicle only when involved. Do not add elective pelvic nodes unless a separate nodal recurrence is being treated. Common salvage S B R T schedules are 30 to 36.25 gray in five fractions. HDR examples include 13.5 gray twice or 10 gray in three fractions, although urethral, bladder neck and rectal constraints vary considerably between protocols.

Adam: There is no universally accepted composite E Q D two limit for rectum, bladder, urethra or penile bulb in this setting. The previous distribution must be reconstructed, particularly the urethral and anterior rectal wall dose. Focal treatment is attractive when the recurrence is clearly localized, but its oncological weakness is the possibility of untreated occult intraprostatic disease.

Marta: The research priority is not another small retrospective series showing feasibility. It is comparative prospective evidence defining focal versus whole gland treatment, S B R T versus HDR, the role of androgen deprivation, patient reported urinary outcomes and appropriate molecular imaging selection.

Adam: For recurrent glioblastoma, reirradiation remains a disease modifying palliative intervention rather than reliably curative salvage. Suitable patients generally have a KPS of at least 60 to 70, localized measurable recurrence, an interval of approximately six months or more, reasonable neurologic function, manageable steroid requirements and a lesion generally no larger than five to six centimetres.

Marta: The G T V is the T 1 post gadolinium enhancing recurrence and, when appropriate, the resection cavity. Routine large FLAIR based expansions are not recommended for focal reirradiation. A separate C T V can be omitted for a sharply defined recurrence or limited to zero to five millimetres according to institutional protocol. The P T V is generally no more than three millimetres with reliable stereotactic guidance.

Adam: 35 gray in ten fractions is the recognizable randomized trial benchmark. Other common schedules are 30 gray in five, 25 to 30 gray in five, or 30 to 36 gray in ten, selected according to size, previous dose and proximity to optic structures and brainstem.

Marta: A small nodular recurrence away from critical structures may receive 24 to 30 gray in three to five fractions. A larger infiltrative recurrence should not be compressed into three fractions merely to shorten treatment.

Adam: R T O G 1205 randomized bevacizumab alone against bevacizumab plus 35 gray in ten fractions. Median overall survival was 10.1 months with reirradiation versus 9.7 months without it, hazard ratio 0.98. Median P F S was 7.1 versus 3.8 months, and six month P F S was 54.3 percent versus 29.1 percent. The combination was generally tolerable and did not produce a delayed high grade C N S toxicity signal in the reported follow up.

Marta: The interpretation is precise: reirradiation improved disease control and P F S, not overall survival. That may still be clinically valuable for neurologic stability, steroid reduction, symptom control and postponing further progression. But it should not be presented as proven life prolonging therapy.

Adam: As of July 2026, E O R T C 2227 BTG LEGATO remains recruiting. It compares lomustine against lomustine plus 35 gray in ten fractions at first progression of IDH wild type glioblastoma.

Marta: The planned population is 411 patients, generally at least six months from first radiation, with measurable disease no larger than five centimetres and performance status zero to two. Overall survival is the primary endpoint. This is the trial that may finally determine whether focal reirradiation adds a survival benefit to active systemic therapy rather than only improving P F S.

Adam: Oligoprogression is not merely a few metastases. It is a state in which most disease remains controlled while a limited number of resistant clones progress. Local ablation is rational when it can prolong the benefit of an effective systemic regimen.

Marta: But the selection is vulnerable to lead time and biology bias. Patients chosen for metastasis directed reirradiation often have inherently favorable disease. A local control rate does not prove an overall survival benefit.

Adam: Therefore we should define the objective: delay systemic therapy change, prevent a predicted local complication, eradicate all visible disease, or obtain durable control of one resistant site.

Marta: The 2026 Estro ISRS recommendations reviewed 20 studies including more than 1,500 treated spine metastases. The pooled one year local control rate was approximately 81 percent, two year local control around 70 percent, vertebral compression fracture risk approximately 5 percent, nerve root injury approximately 6 percent and myelopathy below 2 percent. The guideline supports repeat spine S B R T in selected patients, while explicitly acknowledging low certainty evidence.

Adam: Appropriate candidates have controlled or limited systemic disease, an expected survival usually beyond six to twelve months, M R I defined disease, a spine instability assessment and a realistic benefit from durable control.

Marta: A longer interval, preferably over twelve months, is reassuring, although six to twelve months may be considered. High grade epidural compression, retropulsed bone, symptomatic fracture or substantial mechanical instability requires surgical evaluation. S B R T is not a substitute for stabilization or decompression.

Adam: This is the site where an anatomical C T V is essential. The G T V includes all gross osseous, epidural and paraspinal disease on M R I, C T and, when useful, P E T. The C T V includes the entire involved vertebral anatomical compartment or compartments: vertebral body, right or left pedicle, right or left transverse process, right or left lamina, spinous process, and gross epidural or paraspinal extension. Do not automatically include the entire vertebral level when only one or two compartments are involved.

Marta: If disease extends from the body into the right pedicle and right transverse process, those contiguous compartments should be included. The uninvolved left posterior elements do not automatically require treatment. For postoperative S B R T, include the preoperative tumor extent, residual disease, surgical bed, anatomically compromised vertebral compartments and epidural region at risk. Do not automatically cover the entire incision, all hardware or uninvolved adjacent vertebrae.

Adam: The P T V is usually one to two millimetres with six degree of freedom correction and high quality daily imaging. The true cord should be defined on M R I or C T myelography. Below the conus, contour the thecal sac or cauda equina consistently. A cord or thecal sac P R V of 1.5 to two millimetres may be useful, but the published constraint must be applied to the structure for which it was derived.

Marta: Common repeat S B R T regimens include 16 gray once, 24 gray in two fractions, 27 gray in three, 30 gray in four, and 30 gray in five. Use more fractions for epidural disease, a large target, a high previous cord dose, a target close to esophagus or bowel, or uncertain registration.

Adam: A contemporary multi institutional three fraction series, predominantly using 27 gray in three fractions, reported approximately 85 percent local control at two years. That supports three fraction treatment as an important option, but not as a mandatory regimen.

Marta: When epidural disease abuts the cord, the spinal cord limit takes priority. But we should inspect exactly where the P T V is underdosed. A controlled underdose at the epidural interface may be acceptable; widespread undercoverage of the vertebral disease is not an ablative plan.

Adam: In high grade epidural disease, separation surgery followed by S B R T may provide a better therapeutic ratio than attempting to force an ablative dose against the cord.

Adam: Repeat S R S should begin with a diagnosis problem. Is this viable recurrence or radiation necrosis?

Marta: Use serial M R I, perfusion and diffusion imaging, amino acid P E T where available, and multidisciplinary neuroradiology review. Surgery is valuable when the lesion is large or symptomatic, when tissue would alter systemic therapy, or when imaging uncertainty is clinically decisive.

Adam: The G T V is the contrast enhancing recurrent lesion. The C T V is usually zero. The P T V is one to two millimetres according to the platform and image guidance. For a small favorable lesion, 15 to 18 gray once is commonly used. For a moderate lesion or one with substantial previous high dose overlap, use 24 to 27 gray in three fractions. For a larger, postoperative or organ at risk adjacent lesion, use 25 to 30 gray in five fractions.

Marta: Review the composite high dose normal brain volume, second course V 12 equivalent volume, maximum dose, target size and overlap between the two high dose gradients. A single cumulative whole brain mean dose is not informative for focal radionecrosis risk.

Adam: A systematic review and meta analysis involving 462 lesions estimated one year local control at approximately 69 percent, with radionecrosis in roughly one sixth of treated lesions.

Marta: Those figures are acceptable for selected patients, but the studies are heterogeneous, radionecrosis definitions vary, and symptomatic toxicity is often underreported. Repeat S R S is not simply another standard S R S after recalculating the optic nerve dose.

Adam: For discrete extracranial targets, the general contouring principle is compact and involved site based. A soft tissue or nodal metastasis usually receives G T V or I T V plus a zero to three millimetre C T V, anatomically edited. A bone metastasis requires the involved osseous compartment and gross soft tissue extension. The P T V is typically two to five millimetres. No elective nodal basin or whole organ retreatment.

Marta: Practical starting doses include 45 to 50 gray in five fractions for a peripheral lung lesion; 50 to 60 gray in eight to fifteen for a central lung lesion; 40 to 50 gray in five or 45 to 60 gray in eight fractions for selected liver or adrenal targets; 30 to 40 gray in five for a pelvic or para aortic node; and 24 to 30 gray in three or 30 to 40 gray in five for non spine bone disease.

Adam: Those are not automatic prescriptions. They are starting ranges after the previous dose has been reconstructed.

Marta: For a liver, adrenal or nodal target contacting stomach, duodenum or bowel, I would rather use eight to fifteen fractions or online adaptive treatment than deliberately violate a luminal organ limit in five fractions.

Adam: And if the luminal organ is directly invaded or ulcerated, ablative reirradiation may be inappropriate regardless of what the deformable composite DVH reports.

Adam: Palliative retreatment has a different therapeutic equation. The primary outcomes are pain relief, neurological preservation, convenience and treatment burden, not maximal local control duration at any cost.

Marta: The SC.20 randomized trial compared 8 gray in one fraction with 20 gray in five fractions for painful bone metastasis reirradiation. In the intention to treat analysis, overall pain response at two months was 28 percent with 8 gray and 32 percent with 20 gray in five fractions, satisfying the trial's noninferiority criterion. In the per protocol analysis, response was 45 percent versus 51 percent, and noninferiority was not formally established. Acute toxicity was lower with the single fraction.

Adam: So 8 gray once is a strong default for uncomplicated painful retreatment, particularly with limited prognosis or difficult travel. Twenty Gy in five fractions is reasonable for bulky soft tissue extension, a longer expected survival or concern about durability.

Marta: Other conventional schedules include 24 gray in six fractions and, in selected settings, longer courses. The 2024 Astro bone metastasis guideline supports individualized selection based on prognosis, prior dose, spinal cord tolerance, stability and treatment goals.

Adam: Treat the radiographic pain generator and gross soft tissue extension. For a vertebral lesion, cover the involved vertebral level and gross epidural or paraspinal component. Do not automatically repeat the historical field of one uninvolved vertebra above and below.

Marta: Conversely, avoid an excessively small field that misses the painful cortical, periosteal or paraspinal component merely because the P E T hotspot is focal.

Adam: Reirradiation for cord compression begins with urgent M R I, neurologic examination, steroids when clinically indicated, spinal instability scoring and surgical assessment.

Marta: Surgery should be considered for instability, retropulsed bone, high grade epidural compression, radioresistant disease requiring rapid decompression, or need for tissue diagnosis.

Adam: Conventional retreatment options include 8 gray once, 20 gray in five, 24 gray in six, or a more protracted course in selected patients. S B R T is appropriate after separation surgery or in carefully selected patients with low grade epidural disease and a favorable cumulative cord dose.

Marta: A numerically acceptable spinal cord E Q D two does not make S B R T appropriate for an unstable spine.

Adam: Useful schedules include 8 gray once, 20 gray in five, 30 gray in ten, and the Quad Shot.

Marta: A typical Quad Shot is 14 to 14.8 gray in four twice daily fractions over two days, with at least six hours between fractions. It may be repeated every three to four weeks for up to approximately three cycles, provided there is clinical benefit and acceptable toxicity.

Adam: The G T V is the symptomatic gross tumor. The C T V is zero to five millimetres, and the P T V approximately three to five millimetres. No elective neck.

Marta: Reassess and reduce the target between cycles if the tumor regresses. Quad Shot is particularly useful for bleeding, pain, dysphagia or a bulky symptomatic mass when prolonged treatment is inappropriate. It should not be described as radical S B R T.

Adam: For hemoptysis, cough, focal airway obstruction, chest pain, superior vena cava obstruction or dysphagia from mediastinal disease, practical schedules are 8 gray once, 20 gray in five or 30 gray in ten.

Marta: Treat the gross symptomatic disease. Use a C T V of zero to five millimetres, an I T V based on motion and a P T V usually three to five millimetres. Do not add elective mediastinal or hilar nodes.

Adam: Even with palliative doses, review the previous proximal bronchial tree, esophageal, vessel and spinal cord exposure. Catastrophic toxicity can follow a modest nominal second dose when the first regional dose was high.

Marta: For bleeding, pain, tenesmus, sacral neuropathy or obstructive symptoms, options include 8 gray once, 20 gray in five, 25 gray in five in selected patients, or 30 gray in ten for a patient with longer expected survival.

Adam: The target is the symptom generating gross disease with the smallest anatomically defensible margin. Do not retreat the whole pelvis merely because the original treatment used a whole pelvic field.

Marta: For bleeding from a mucosal lesion, a modest dose may be sufficient. For neuropathic pain from a pelvic sidewall mass, the target should include the involved nerve interface rather than only the central tumor bulk.

Adam: A 62 year old patient received 70 gray with cisplatin four years ago. He now has a 2.2 centimetre lateral oropharyngeal recurrence, biopsy confirmed, no nodal or distant disease, normal swallowing and no feeding tube. The tumor lies seven millimetres from the carotid and does not invade the mandible.

Marta: The first question is salvage surgery. If surgery is not feasible or would be functionally unacceptable, this is a reasonable definitive reirradiation candidate. The G T V is the M R I-, P E T- and endoscopy defined recurrence. The C T V is zero to five millimetres, following the involved pharyngeal compartment and edited off uninvolved air and intact bone. No elective neck. P T V two to three millimetres with daily volumetric guidance.

Adam: Would you use 44 gray in five?

Marta: Possibly, if the target is small, there is no mucosal ulceration, the carotid and bone doses are acceptable, and treatment is delivered nonconsecutively. If there is substantial circumferential mucosal involvement or the carotid dose is marginal, I would use ten or more fractions. I would not choose five fractions simply because the G T V is 2.2 centimetres.

Adam: A 70 year old received 60 gray in 30 fractions and one year of durvalumab. Thirty months later, there is isolated recurrent disease around the right main bronchus, touching the esophagus and pulmonary artery.

Marta: This is not a routine five fraction salvage. Obtain biopsy if feasible, P E T or C T and brain imaging. Reconstruct the prior airway, esophageal, great vessel, heart, lung and cord dose. The G T V is the recurrent mass and involved node if present. The C T V is zero to five millimetres, anatomically edited. No elective nodes. Motion management is mandatory.

Adam: Prescription?

Marta: I would begin by testing 55 gray in 20 or 60 gray in 30 fractions, not 50 gray in five. If even a conventionally fractionated meaningful dose cannot meet composite airway, esophageal and vessel constraints, the options are surgery in an exceptional candidate, systemic therapy, bronchoscopic intervention, or palliation.

Adam: And consolidation immunotherapy?

Marta: It can be discussed, but the evidence is phase two and retrospective. We should not imply that immunotherapy compensates for an unsafe composite plan.

Adam: A patient with metastatic renal cell carcinoma has otherwise controlled disease on systemic therapy but progression in the T 8 vertebral body and right pedicle. The level previously received 30 gray in ten fractions eighteen months earlier. There is grade 1c epidural extension, no instability and no neurologic deficit.

Marta: This is a reasonable repeat S B R T candidate. The G T V includes body and pedicle disease plus the epidural component. The C T V includes the involved vertebral body, right pedicle and contiguous epidural region. It does not automatically include the left pedicle, both laminae or spinous process. Use a one to two millimetre P T V, M R I defined cord or C T myelography, six degree of freedom correction and a cumulative thecal sac E Q D two analysis.

Adam: A practical prescription would be 27 gray in three fractions or 30 gray in five, depending on epidural geometry and remaining cord reserve.

Marta: Exactly. If the epidural interface requires substantial undercoverage, discuss separation surgery rather than accepting a nominally ablative prescription with inadequate dose to the site most likely to fail.

Adam: The most important technical trend is not a new delivery platform. It is standardized cumulative dose reconstruction.

Marta: Reirradiation needs machine readable prior plans, consistent structure nomenclature, registration quality metrics, uncertainty bounds and prospective toxicity linkage. For bioinformaticians, this is a high value domain: spatial dose response modelling can be more informative than another collection of prescription doses without voxel level anatomy.

Adam: But any predictive model must distinguish anatomical mapping uncertainty from biological uncertainty. A neural network can predict a warped dose distribution very confidently and still be clinically wrong if the bowel, lung or postoperative anatomy is not physically corresponding.

Marta: The 2025 international survey of cumulative reirradiation constraints showed substantial inter centre variation, confirming that there is still no universal reirradiation equivalent of QUANTEC.

Adam: Online adaptive M R guided and CBCT guided treatment is particularly attractive for abdominal and pelvic reirradiation. It can reduce the margin and adapt to the bowel, stomach, bladder and rectum of the day.

Marta: Its main value is not that it raises every prescription. Sometimes adaptation demonstrates that the planned dose is not safely deliverable that day. That is useful information.

Adam: Adaptive treatment should also record the delivered rather than merely planned cumulative dose. Otherwise, the daily anatomy is used for optimization but discarded for outcome analysis.

Marta: Proton therapy can reduce integral dose and may improve feasibility in selected head and neck, thoracic, breast, C N S and pelvic cases. But its advantage should be demonstrated with patient specific comparative planning.

Adam: In reirradiation, robustness analysis should include setup, range, anatomical change and the distal edge near serial organs. A proton plan with an attractive nominal DVH may be less convincing if the distal fall off terminates in the brainstem, esophagus or bowel.

Marta: Carbon ion series are scientifically interesting, especially for radioresistant histologies, but photon E Q D two cannot simply be added to model dependent carbon ion biological dose. RBE assumptions, patient selection and centre effects complicate cross modality comparisons.

Adam: Reirradiation plus checkpoint inhibition is one of the most active research areas. The biological rationale is plausible, and phase two signals in head and neck and thoracic disease are encouraging.

Marta: But the clinical question is not merely efficacy. We need randomized data on whether the combination improves survival or systemic therapy duration, and prospective data on fistula, hemorrhage, radionecrosis, pneumonitis and wound healing.

Adam: The same caution applies to antiangiogenic agents. A treatment may reduce edema or radiographic enhancement while increasing concern about wound healing, perforation or fistula in a previously irradiated hollow organ.

Marta: Drug specific timing should therefore be agreed prospectively with medical oncology rather than managed with a generic hold systemic therapy for one week rule.

Adam: What do molecular biologists realistically contribute today?

Marta: At present, no genomic signature is sufficiently validated to override anatomy, prior toxicity and disease distribution in routine reirradiation selection. But several research directions are credible: molecular distinction between true recurrence and radiation injury, clonality assessment for second primary versus recurrence, circulating tumor D N A as an early response marker, and spatially resolved predictors of normal tissue injury.

Adam: Imaging biomarkers may become clinically useful sooner. Examples include amino acid P E T for intracranial recurrence, P S M A P E T for prostate salvage, functional lung imaging, diffusion and perfusion M R I, and quantitative changes during adaptive treatment.

Marta: Radiomics and dose painting remain research tools. They should not be used to escalate a previously irradiated subvolume beyond validated organ at risk limits merely because it appears biologically resistant.

Adam: What trials are most needed?

Marta: First, randomized or carefully controlled comparisons of local reirradiation plus systemic therapy versus systemic therapy alone in oligoprogression. Second, prospective comparisons of fractionation rather than additional single arm feasibility studies. Third, patient reported swallowing, urinary, bowel, neurologic and cognitive outcomes. Fourth, long follow up. A study with twelve months of follow up may estimate local control, but it cannot reliably characterize carotid blowout, myelopathy, urethral stricture or late radionecrosis.

Adam: Fifth, mandatory reporting of previous dose, cumulative E Q D two, registration method, recovery assumption and actual target coverage. Without that, reirradiation studies remain difficult to reproduce.

Marta: And sixth, publication of failure geography. We need to know whether failure occurred in the fully covered G T V, at the deliberately underdosed O A R interface, in an omitted microscopic risk region or outside the retreatment volume.

Adam: Let us finish with the practical essentials.

Marta: First, classify the case as Type 1 or Type 2 reirradiation and state whether the goal is radical, oligometastatic or palliative.

Adam: Second, retrieve the previous DICOM plan and calculate cumulative biological dose. Never add physical Gy across different fractionations.

Marta: Third, assume zero recovery unless a documented site specific protocol justifies otherwise. The interval is a clinical covariate, not a universal recovery calculator.

Adam: Fourth, keep targets compact and anatomy driven. Include preoperative disease in postoperative targets, vertebral compartments in spine S B R T, and gross symptomatic disease, not historical elective fields, in palliation.

Marta: Fifth, use more fractions near cord, plexus, optic pathways, bowel, stomach, esophagus, airway, carotid or great vessels.

Adam: Sixth, distinguish evidence levels. Hyperfractionation for recurrent NPC has randomized phase three support. R T O G 1205 improves P F S but not overall survival. Spine re S B R T is guideline supported but based largely on nonrandomized evidence. Most pelvic and abdominal cumulative constraints remain consensus based.

Marta: Seventh, organ at risk limits take priority over nominal P T V coverage, but if the resulting tumor dose is no longer meaningful, change the treatment strategy or explicitly change the intent.

Adam: And finally: reirradiation is not the art of finding a numerical justification for another course.

Marta: It is the disciplined selection of a patient whose expected local benefit exceeds a cumulative toxicity risk that has been reconstructed, bounded, documented and honestly discussed.

Adam: This has been Radioonkolog dot P L Podcast.