High-yield points
- Rare: about 1 in 1.7 million births; 2–6% of conjoined twins .
- Classify (Stone–Goodrich): partial vs total (shared sinuses); angular vs vertical; TV types I–III by rotation .
- Map the veins: a circumferential sinus with small rotation, helical with large rotation .
- Stage the separation: historical mortality was over 50% . Multi-stage surgery, tissue expanders and non-emergent timing predict success .
- Partial forms had lower mortality ; vertical conjoining predicted success , and total vertical had the highest success rate (90%) in the 30-year review .
- After separation: watch for hydrocephalus, CSF leak and wound problems, as the venous system remodels .
Landmark papers
- Stone and Goodrich, Brain 2006: the classification (partial and total; angular and vertical; TV types I–III). Total forms did worse, and staged separation lowered mortality .
- Staffenberg and Goodrich, J Craniofac Surg 2012: the rationale for staged separation, to avoid the venous hypertension, swelling and infarction of single-stage surgery .
- Hwang et al., Plast Reconstr Surg 2026: 30-year review. Stone–Goodrich class, staging, tissue expanders and non-emergent surgery predicted success; total vertical had a 90% success rate .
Craniopagus twins are conjoined at the head. The condition is extraordinarily rare: about 1 in 1.7 million live births, and 2–6% of conjoined twins . Separation is among the most complex operations in pediatric neurosurgery.
The central problem is the shared venous drainage. In the most severe forms, the twins share dural venous sinuses, and dividing them risks massive hemorrhage, venous infarction and brain swelling . Modern success depends on staged separation, detailed venous mapping, tissue expansion for closure, and a very large, rehearsed multidisciplinary team .
Classification
- Stone–Goodrich classification (64 well-described pairs) :
- partial forms have no significant shared dural venous sinuses;
- total forms share sinuses and show more compressive distortion of the brains;
- four types:
- partial angular (PA);
- partial vertical (PV);
- total angular (TA);
- total vertical (TV), formerly O'Connell types I–III, with a continuous cranium. Subtypes by axial facial rotation:
- type I: under 40°;
- type II: 140–180°;
- type III: intermediate.
- "Angular" means the twins' long axes meet at less than 140°, whatever the rotation.
- Distribution :
- just over half were total vertical, almost a third total angular, and about a sixth partial;
- roughly 3 in 10 pairs had fused or shared brain tissue; among total angular pairs, a shared posterior fossa was about as common.
Venous anatomy
- Venous anatomy shapes the surgical strategy , although the nature of the shared venous drainage was not significantly associated with success in one systematic review .
- Venous mapping (16 sets of twins from the Goodrich archive) :
- 2 were O'Connell type I, 4 type II and 10 type III;
- with a small rotation angle, the venous systems organize around a common circumferential sinus;
- with a large rotation angle (type II), the superior sagittal sinuses join in a helical sinus;
- as rotation approaches 90°, the circumferential pattern gradually becomes helical.
- Imaging (see caveats):
- CT and MRI;
- CT and MR venography;
- catheter angiography to show which twin drains through which sinus, and whether there is cross-circulation;
- 3D-printed models and virtual reality, increasingly used in planning .
Who does well
- Stone–Goodrich review :
- compared with total forms, partial forms were heavier at birth, were operated on younger (6 vs 11 months) and survived more often, with better results;
- for total forms, multi-staged separation had significantly lower mortality than single-stage surgery.
- Systematic review 1995–2015 (24 attempts) :
- vertical craniopagus was associated with successful separation (both twins alive at 30 days);
- staged surgery and surgery before 12 months showed trends toward success.
- Systematic review 1995–2025 (33 cases) :
- of the three classification systems (O'Connell, Winston, Stone–Goodrich), only Stone–Goodrich predicted outcome;
- total vertical twins had the highest success rate (90%);
- success was associated with multi-stage surgery, tissue expanders and non-emergency operations.
Why staged separation
- Historically, when separation was mostly done in one stage (often after scalp expansion), reported mortality was over 50% even at top centers .
- What went wrong in single-stage cases :
- massive intraoperative hemorrhage;
- cerebral edema;
- venous infarcts;
- flap swelling;
- wound breakdown with CSF leak, meningitis or exposed brain.
- The reasoning: a one-stage division of shared sinuses raises cerebral venous pressure suddenly. Staging divides the venous connections a sector at a time, giving collateral drainage time to develop between operations .
- The staged approach (see caveats):
- several operations, weeks to months apart;
- each divides part of the dura and shared venous channels, with careful assessment of venous pressure, swelling and each twin's physiology;
- tissue expanders go in before the final stage to provide scalp for closure (see caveats); their use is associated with success ;
- the final stage completes the separation and reconstructs each head.
Perioperative issues
- Anesthesia (see caveats):
- two anesthesia teams and two sets of monitoring;
- cross-circulation means drugs, fluids and blood given to one twin reach the other;
- massive transfusion planning (see blood volume).
- The team: neurosurgery, plastic surgery, anesthesia, critical care, interventional neuroradiology, nursing, ethics, social work, and often external experts, with rehearsals on 3D models and in VR .
- Even with excellent planning, deaths happen: in a recent three-case series, both twins of the third pair died intraoperatively from hemorrhage and cardiogenic shock, despite accurate models .
- Long-term problems (see caveats):
- hydrocephalus, as venous drainage rearranges;
- CSF leak;
- wound breakdown;
- cranial reconstruction;
- developmental and neurological deficits.
- Venous remodeling after separation: in a 5-year follow-up of total vertical twins, both developed well. Both received lumboperitoneal shunts, one for hydrocephalus and the other for a wound CSF leak; follow-up CT venography showed the venous system had rearranged .
Ethics
- Separation is elective in most cases, and both children's interests must be weighed separately. One twin may be sacrificed in some anatomies (see caveats).
- Families need honest outcome data, including non-separation as an option. Expect international referrals, media attention and cost (see bioethics; see caveats).
Related topics
- Bioethics: Challenging Situations in Pediatric Neurosurgery: consent and weighing two children's interests.
- Cerebral Sinovenous Thrombosis in Children: venous hypertension and infarction.
- Estimation of Blood Volume in Children: planning for massive blood loss.
- Assessing Perioperative Risk for Neurosurgical Cases in Children: perioperative planning.
- Coagulopathy in Children: massive transfusion and coagulopathy.
References
- Kobets A, Fullerton J, Lober RM, et al. Characterizing the venous anatomy of craniopagus twins. Neurosurgery 2026;98(2):474–478. doi:10.1227/neu.0000000000003617. PMID 40622169. verified
- Stone JL, Goodrich JT. The craniopagus malformation: classification and implications for surgical separation. Brain 2006;129(Pt 5):1084–1095. doi:10.1093/brain/awl065. PMID 16597654. verified
- Staffenberg DA, Goodrich JT. Separation of craniopagus conjoined twins with a staged approach. J Craniofac Surg 2012;23(7 Suppl 1):2004–2010. doi:10.1097/SCS.0b013e318262d3f7. PMID 23154370. verified
- Chodraui FI, Grillo FW, Rondinoni C, et al. Integrating virtual reality and patient-specific 3D models for staged craniopagus twins separation: evolution of a surgical planning pipeline across 3 consecutive cases. Neurosurgery 2026 (published online ahead of print). doi:10.1227/neu.0000000000004293. PMID 43018268. verified
- Harvey DJ, Totonchi A, Gosain AK. Separation of craniopagus twins over the past 20 years: a systematic review of the variables that lead to successful separation. Plast Reconstr Surg 2016;138(1):190–200. doi:10.1097/PRS.0000000000002269. PMID 27348650. verified
- Hwang JJ, Stefanski AL, Zhang A, et al. Separation of craniopagus twins: a 30-year systematic review of surgical outcomes. Plast Reconstr Surg 2026 (published online ahead of print). doi:10.1097/PRS.0000000000012962. PMID 41729154. verified
- Ashari S, Aulia A, Priyambodo A, et al. 5-year follow up after successful craniopagus separation: review on hydrocephalus and venous system re-arrangement. Neurochirurgie 2025;71(6):101717. doi:10.1016/j.neuchi.2025.101717. PMID 40915610. verified
Verification (2026-11-19):
- Each reference's authors, title, journal, year, volume, pages, DOI and PMID were checked on PubMed. References 4 and 6 are online ahead of print.
- Numbers were checked against the PubMed abstracts. Reference 5's 24 attempts are its 19 published plus 5 unpublished cases.
Caveats:
- The following are standard teaching, not drawn from the cited abstracts:
- the imaging workup;
- the details of the staged approach (number and spacing of stages, assessment between stages, timing of tissue expanders);
- the anesthetic and cross-circulation points and the team composition beyond reference 4's description;
- the list of long-term problems beyond reference 7's twins;
- the ethics points, including the possibility of sacrificing one twin.
- All of the evidence is small case series and reviews of published cases, which favor reporting of successful separations.