Intervenciones para el tratamiento de los quistes óseos simples de huesos largos en niños
Resumen
Antecedentes
Los quistes óseos simples, también conocidos como quistes óseos unicamerales o quistes óseos solitarios, son el tipo más común de lesión ósea benigna en los niños en etapa de crecimiento. Los quistes pueden dar lugar a una fractura patológica repetida (fractura que se presenta en un área de hueso debilitado por un proceso de enfermedad). Ocasionalmente, estas fracturas pueden dar lugar a una consolidación defectuosa sintomática. Los objetivos principales del tratamiento son la reducción del riesgo de fracturas patológicas, la mejoría en la curación del quiste y la resolución del dolor. A pesar de los numerosos métodos de tratamiento que se han utilizado para los quistes óseos simples de huesos largos en niños, no hay consenso sobre el mejor procedimiento. Ésta es una actualización de una revisión Cochrane publicada por primera vez en 2014.
Objetivos
Evaluar los efectos (beneficiosos y perjudiciales) de las intervenciones para el tratamiento de los quistes óseos simples de huesos largos en niños, incluidos los adolescentes.
Se pretendía realizar las siguientes comparaciones principales: intervenciones invasivas (por ejemplo, inyecciones, legrado, fijación quirúrgica) frente a intervenciones no invasivas (por ejemplo, observación, escayola, actividad restringida); diferentes categorías de intervenciones invasivas (por ejemplo, inyecciones, legrado, perforaciones y descompresión, fijación quirúrgica y descompresión continuada); diferentes variaciones de cada categoría de intervención invasiva (por ejemplo, diferentes sustancias inyectables: médula ósea autóloga frente a esteroides).
Métodos de búsqueda
Se realizaron búsquedas en el Registro Especializado del Grupo Cochrane de Lesiones Óseas, Articulares y Musculares (Cochrane Bone, Joint and Muscle Trauma Group), el Registro Cochrane Central de Ensayos Controlados (Cochrane Central Register of Controlled Trials), MEDLINE, Embase, la Plataforma Nacional de Infraestructura del Conocimiento de China, registros de ensayos, actas de congresos y listas de referencias. Fecha de la última búsqueda: Abril de 2016.
Criterios de selección
Ensayos controlados aleatorizados y cuasialeatorizados que evaluaran métodos de tratamiento de los quistes óseos simples de huesos largos en niños.
Obtención y análisis de los datos
Dos autores de la revisión seleccionaron de forma independiente los resultados de búsqueda y realizaron la selección de estudios. Se resolvieron las diferencias de opinión entre los autores de la revisión mediante discusión y consulta con un tercer autor de la revisión. Dos autores de la revisión evaluaron de forma independiente el riesgo de sesgo y extrajeron los datos. Se resumieron los datos mediante los riesgos relativos (RR) o las diferencias de medias (DM), según fuese apropiado, y los intervalos de confianza (IC) del 95%. Se utilizó el sistema Grading of Recommendations Assessment, Development, and Evaluation (GRADE) para evaluar la calidad general de la evidencia.
Resultados principales
En esta actualización de 2017, no se identificó ningún nuevo ensayo controlado aleatorizado (ECA) para su inclusión. Se identificó un ensayo en curso que es probable que se incluya en una futura actualización. Por consiguiente, los resultados no han cambiado. El único ensayo incluido es un ECA multicéntrico realizado en 24 lugares de América del Norte y la India que comparó la inyección de médula ósea con la inyección de esteroides (acetato de metilprednisolona) para el tratamiento de quistes óseos simples. Se planificó la aplicación de hasta tres inyecciones para los participantes de cada grupo. El ensayo incluyó a 90 niños (edad media 9,5 años) y presentó los resultados de 77 niños a los dos años de seguimiento. Aunque el ensayo tuvo una ocultación segura de la asignación, estuvo en riesgo alto de sesgo de realización y de desequilibrios graves en las características iniciales. Para reflejar dichas limitaciones del estudio, se disminuyó la calidad de la evidencia en dos niveles a “baja” para la mayoría de los resultados, lo cual significa que no existe seguridad acerca de las estimaciones del efecto. Para los resultados en los que hubo una imprecisión grave, la calidad de la evidencia se disminuyó en un nivel adicional a “muy baja”.
El ensayo proporcionó evidencia de muy baja calidad de que menos niños del grupo de inyección de médula ósea habían evaluado radiográficamente la curación de los quistes óseos a los dos años que en el grupo de inyección de esteroides (9/39 versus 16/38; RR 0,55 a favor de la inyección de esteroides, IC del 95%: 0,28 a 1,09). Sin embargo, el resultado fue incierto y puede ser compatible con ninguna diferencia o un beneficio pequeño a favor de la inyección de médula ósea. Sobre la base de una tasa de éxito ilustrativa de 421 niños que presentaron la curación de los quistes óseos por 1000 niños tratados con inyecciones de esteroides, estas cifras se equiparan a 189 niños menos (IC del 95%: 303 menos a 38 más) con curación de los quistes óseos por 1000 niños tratados con inyecciones de médula ósea. Hubo evidencia de baja calidad de la falta de diferencia entre las dos intervenciones a los dos años en cuanto al resultado funcional, según la puntuación funcional de la Escala de Actividad para Niños (0 a 100; las puntuaciones más altas equivalen a un mejor resultado: IC del 95%: ‐4,26 a 2,46) o en el dolor evaluado mediante la puntuación del dolor de Oucher. Hubo evidencia de muy baja calidad de la falta de diferencias entre las dos intervenciones para los eventos adversos: fractura patológica posterior (9/39 versus 11/38; RR 0,80, IC del 95%: 0,37 a 1,70) o infección superficial (dos casos en el grupo de médula ósea). No se informó la recurrencia de quistes óseos, la consolidación defectuosa inadmisible, el retorno a las actividades normales ni la satisfacción del participante.
Conclusiones de los autores
La evidencia disponible es insuficiente para determinar los efectos relativos de las inyecciones de médula ósea versus de esteroides, aunque las inyecciones de médula ósea son más invasivas. Es interesante destacar que la tasa de curación del quiste óseo según la evaluación radiográfica a los dos años estuvo muy por debajo del 50% para ambas intervenciones. En términos generales, hay una falta de evidencia para determinar el mejor método de tratamiento de los quistes óseos simples de huesos largos en niños. Se necesitan más ECA de tamaño y calidad suficientes para guiar la práctica clínica.
PICO
Resumen en términos sencillos
Tratamientos para los quistes óseos simples de huesos largos en niños
Los quistes óseos simples (quistes óseos “unicamerales” o “solitarios”) son el tipo más común de anomalía ósea (lesión) no cancerosa (benigna) en niños en etapa de crecimiento. Los quistes hacen que la corteza ósea (capa externa dura del hueso) se vuelva delgada y pueden dar lugar a fracturas patológicas repetidas (la que se presenta sin mucho traumatismo en un área de hueso debilitado). Ocasionalmente, dichas fracturas pueden dar lugar al acortamiento de la extremidad y a deformidad. Los objetivos principales del tratamiento de los quistes óseos simples son reducir el riesgo de fractura patológica, ayudar a la curación del quiste y detener el dolor. Existen varios métodos de tratamiento disponibles, pero no hay acuerdo sobre cuál es el mejor. Por lo tanto, se realizó una revisión minuciosa de la evidencia disponible sobre diferentes métodos de tratamiento de los quistes óseos simples de huesos largos en niños para observar si se podía identificar qué método es mejor. Ésta es una versión actualizada de la revisión original publicada en 2014.
Se realizaron búsquedas en varias bases de datos médicas importantes en abril de 2016, así como en registros de ensayos, actas de conferencias y listas de referencias. Se encontró sólo un estudio médico relevante en el cual los niños participantes fueron asignados al azar a diferentes tratamientos. El estudio comparó tratamientos en los cuales se administraron inyecciones de médula ósea o de esteroides (acetato de metilprednisolona) en los quistes óseos simples. Noventa niños con una edad media de 9,5 años participaron en este estudio.
Los resultados estaban disponibles para 77 niños. Dos años después del tratamiento, el examen de rayos X mostró que la curación exitosa de los quistes óseos era más común en los niños que habían recibido inyecciones de esteroides; sin embargo, existe incerteza de que esto sea un hallazgo verdadero. La evidencia de baja calidad dos años después del tratamiento mostró que los niños en los dos grupos de tratamiento presentaron niveles altos similares en la función (medidos con la puntuación Activity Scale for Kids) y niveles bajos de dolor (medidos con la puntuación Oucher). Hubo evidencia de muy baja calidad de que no hubo diferencias entre las dos intervenciones para los eventos adversos, incluida la fractura patológica después del tratamiento. Debido a que la calidad de la evidencia es baja, o muy baja, no es posible establecer la conclusión definitiva de que no hay diferencias entre los tratamientos, y no se sabe si algún tratamiento proporciona mejores resultados y menos complicaciones.
Esta revisión se basa en un ensayo con un número pequeño de participantes. En consecuencia, actualmente hay evidencia insuficiente para determinar el mejor método de tratamiento de los quistes óseos simples de huesos largos en niños. Se necesitan más estudios con mayor número de participantes y que monitoricen a los niños durante períodos de seguimiento más largos.
Authors' conclusions
Summary of findings
| Bone marrow versus steroid injections for simple bone cysts in children | ||||||
| Patient or population: children with simple bone cysts | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk1 | Corresponding risk | |||||
| Steroid injection | Bone marrow injection | |||||
| Success rate (radiographically‐assessed healing of cyst healing based on the Neer/Cole rating system) | Study population | RR 0.55 | 77 | ⊕⊝⊝⊝ | ||
| 421 per 1000 | 232 per 1000 | |||||
| Function: ASK function score Scale from: 0 to 100 (best outcome) | The mean ASK function score in the control group was | The mean ASK function score in the intervention group was | ‐ | 77 | ⊕⊕⊝⊝ | |
| Recurrence of bone cyst | see comment | see comment | ‐ | ‐ | ‐ | Outcome not reported |
| Adverse event: pathological fracture | Study population | RR 0.8 | 77 | ⊕⊝⊝⊝ | ||
| 289 per 1000 | 232 per 1000 | |||||
| Adverse event: infection | see comment | see comment | RR 4.88 | 77 | ⊕⊝⊝⊝ | There were no infections in the steroid group and 2 superficial infections in the bone marrow group. The risk of infection for steroid injection is unknown but is likely to be very low |
| Unacceptable malunion | see comment | see comment | ‐ | ‐ | ‐ | Outcome not reported |
| Oucher pain score Scale from: 0 to 100 (maximum pain) | The mean Oucher pain score in the control group was | The mean Oucher pain score in the intervention groups was | ‐ | 77 | ⊕⊕⊝⊝ | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1. The assumed risk is based on the data reported by the included trial 2. Serious limitations resulting from high risk of performance bias and high risk of bias relating to major imbalances in baseline characteristics; evidence downgraded two levels 3. Imprecision: these results come from a single study with a small number of participants and need confirmation in other larger studies; evidence downgraded one level | ||||||
Background
Description of the condition
Simple bone cysts, also known as a unicameral bone cysts or solitary bone cysts, were first reported by Virchow (Virchow 1876). They are benign, fluid‐filled lesions, mainly located in the metaphyses of the long bones (parts of the bone where growth takes place that are situated between the middle shaft section and each of the two bone ends) in children and adolescents. The most common site for a simple bone cyst is the proximal humerus (top end of the upper arm bone) followed by the proximal femur (top end of the thigh bone). The peak age of occurrence is around 10 years (Capanna 1982). Simple bone cysts amount to around 3% of bone lesions and occur around twice as often in boys than in girls (Boseker 1968). Attempts have been made to classify simple bone cysts into active and inactive types (Norman 1977). Active cysts have been defined as those that are in direct contact with the adjacent growth plate in skeletally‐immature children, while inactive ones are those that are separate from the growth plate.
The causes of simple bone cysts are unclear. The finding by Chigira 1983 of higher osseous pressures (relating to blood circulation within the bone) within the cysts compared with the normal pressure of the bone marrow lends support to the popular theory that obstruction of venous drainage in the bone is the likely cause of these cysts.
On an X‐ray, a cyst appears as a well‐localised and distinct lesion without signs of formation of new bone (periosteal reaction) in the medullary canal. Although most cysts become static or resolve (disappear) near skeletal maturity (Capanna 1982), cysts weaken the bone and may lead to pathological fracture of the bone through the thin cortex. Most children with simple bone cysts present with pain because of a fracture. In some cases, however, the fracture may be more disabling and serious, such as a displaced fracture of the proximal femur. These pathological fractures may result in symptomatic malunion of the bone, particularly in the proximal femur (Chuo 2003). Furthermore, simple bone cysts may result in growth disturbance (Stanton 1998). Although growth arrest is a relatively rare complication, it may occur through many mechanisms. The causes of growth arrest include the disruptive assault of active cyst fluid on the physis (growth plate), direct extension of the cyst through the physis, pathological fracture through the cyst that damages the physis, or as a result of treatment for cysts adjacent to the growth plate (Haims 1997; Stanton 1998). Ultimately, growth retardation in the involved limb may result in angular deformity or discrepancy in limb length, or both (Haims 1997). Parents or surgeons are fearful of possible fractures in children with bone cysts, and this can lead to restrictions to activity and exercise that may harm a child's physical and mental development further.
Description of the intervention
The main goals of treatment are to decrease the risk of pathological fracture, enhance cyst healing and resolve pain. The main interventions used are listed below.
-
Non‐invasive interventions
-
Observation: waiting only, with no other active intervention (Chuo 2003; Neer 1966);
-
Activity restriction and other non‐invasive interventions; these include cast immobilisation for children with a pathological fracture (Garceau 1954).
-
-
Invasive (surgical) interventions
-
Mechanical disruption of cyst wall or lining with or without grafts. Curettage: a small skin incision is made at the level of the thinnest cyst wall. After a curette (tool designed for scraping or removing tissue) is inserted into the cyst by puncturing the thinned cortical bone, systematic curettage of the inner wall of the cavity is performed repeatedly using straight and angled curettes in order to remove the membrane lining of the cyst. The fluid in the cyst cavity is then suctioned away. No substance is injected or placed back into the cyst. Curettage, as described above, is a less invasive procedure than open curettage with grafts. The latter technique consists of the incision of skin and overlying tissue of the bone, opening up the medullary canal, removing the cyst membrane, filling in the void with bone‐graft materials, and, finally, the closure of the incision. Subtotal‐resection involves a more extensive removal and then replacement of overlying bone after removal of the cyst.
-
Curettage only (Canavese 2011)
-
Curettage (opening up the medullary canal) with grafts (Neer 1966)
-
Subtotal‐resection with or without bone graft or muscle graft (Fahey 1973)
-
-
Injection treatment of steroid or autologous (the patient's own) bone marrow: after localisation of the lesion, a sharp needle is introduced percutaneously into the cyst, through the thinnest area of the cyst wall, to inject one of the following materials.
-
Steroid (Scaglietti 1979)
-
Autologous bone marrow (Lokiec 1996)
-
-
Drilling holes and decompression: the technique consists of drainage of cyst fluid, washing the cyst cavity with saline and drilling holes through the cortical and the medullary bone of the cyst wall. If a cannulated screw is used, it is inserted through a small hole to provide continuous decompression (drainage).
-
Drilling holes and continued decompression with a cannulated screw (Saraph 2004)
-
Drilling holes and decompression without a cannulated screw (Komiya 1993)
-
-
Use of several bone substitutes: after drilling a hole, one of the following materials is injected or filled.
-
Calcium sulphate or calcium phosphate, or both (Altermatt 1992; Fillingham 2012)
-
Demineralised bone matrix (Killian 1998)
-
-
Internal fixation and continued decompression: insertion of one of the following internal devices into the cyst through a percutaneous or mini‐open approach. These devices aim to stabilise the affected region, as well as enabling drainage.
-
Kirschner wire (Chigira 1983)
-
Elastic stable intramedullary nail (Santori 1988)
-
-
Combinations of the methods above.
-
How the intervention might work
A simple bone cyst is considered to be a benign, self‐limiting condition that usually heals spontaneously after skeletal maturity (Wilkins 2000). Thus observation is included as a treatment option on the basis of some evidence of cysts healing without treatment, particularly for children with an asymptomatic and inactive cyst (Neer 1966; Wilkins 2000). Curettage can eradicate the source of bone‐destroying enzymes through removal of the cyst membrane. As mentioned above, percutaneous curettage is a minimally invasive technique. Compared with curettage only, the technique of resection or open curettage is relatively straightforward as the cyst cavity is exposed widely and the cyst membrane can be removed under direct vision. Then either autologous (from the person's own body) or allogenic (from an external source) bone‐graft materials can be used to fill in the area of the bone defect. While the mechanism underlying the effects of steroid injection is still unclear, Scaglietti 1979 conjectured, on the basis of experimental work, that microcrystals caused destruction of the connective tissue coat of the cyst wall, thus allowing secondary bone repair to occur. The anti‐inflammatory action of the steroid may also play an important role (Scaglietti 1982). Lokiec 1996 suggested that injection of bone marrow might work due to both the perforation of the cyst wall and the injection of bone marrow; the latter would, by itself, engender the formation of normal bone. Injected methods may have a high recurrence rate, but injection procedures can be carried out repeatedly. Additionally, calcium sulphate, demineralised bone matrix and calcium phosphate bone cement could stimulate new bone formation (Rougraff 2002). Several in situ implants, such as Kirschner wires, intramedullary nails and cannulated screws, have been used to treat simple bone cysts. These implants can achieve drainage and continuous decompression of the intraosseous pressure to promote cyst healing and reduce the rate of recurrence. In addition, flexible intramedullary nailing provides stability, which allows early mobilisation and an early return to normal activity. The continued stability provided by flexible intramedullary nailing should also decrease the risk of pathological fracture (De Sanctis 2006).
Why it is important to do this review
Simple bone cysts are the most common type of benign bone lesion in growing children. Cysts make the bone cortex thin and may lead to repeated pathological fracture. Furthermore, fractures may result in unequal limb lengths and angular deformity. Perceived risk for fracture prevents many children from participating in physical activities until the cyst is resolved. This can disrupt normal childhood for extended periods and limit activities (Lokiec 1998). Therefore, it is important to strike an optimal balance between potential for healing and invasiveness, and risk of complications, when contemplating treatment methods.
Despite the numerous treatment methods that have been used for simple bone cysts, there is no consensus on the best procedure. It is necessary to systematically review the available evidence for the different methods of treating simple bone cysts in the long bones of children in order to inform treatment choice.
Objectives
To assess the effects (benefits and harms) of interventions for treating simple bone cysts in the long bones of children, including adolescents.
We intended the following main comparisons:
-
invasive (e.g. injections, curettage, surgical fixation) versus non‐invasive interventions (e.g. observation, plaster cast, restricted activity);
-
different categories of invasive interventions (i.e. injections, curettage, drilling holes and decompression, surgical fixation and continued decompression);
-
different variations of each category of invasive intervention (e.g. different injection substances: autologous bone marrow versus steroid).
Methods
Criteria for considering studies for this review
Types of studies
We included randomised and quasi‐randomised (method of allocating participants to a treatment that is not strictly random, e.g. by date of birth, hospital record number, alternation) controlled trials evaluating methods for treating simple bone cysts in children.
Types of participants
Children, including adolescents, with simple bone cysts in long bones. There were no restrictions regarding either the activity of the cyst or a history of previous pathological fracture. If applicable in future, trials including a few participants with cysts in other bones will be included.
Types of interventions
We proposed to include all trials comparing different interventions used for treating simple bone cysts in children and adolescents. We chose the less invasive or non‐invasive intervention as our control group. The detailed comparisons we planned were:
-
invasive (e.g. curettage, injections, surgical fixation) versus non‐invasive interventions (e.g. observation, plaster cast, restricted activity);
-
different methods of invasive interventions (e.g. curettage with bone graft versus steroid injection, drilling holes and decompression versus autologous bone marrow, curettage with bone graft versus internal fixation and continued decompression);
-
different methods of mechanical disruption of cyst wall or lining (e.g. open curettage with bone graft versus curettage only);
-
different injected substances (autologous bone marrow versus steroid);
-
different methods of drilling holes and decompression (drilling holes and decompression with versus without a cannulated screw);
-
use of different materials as a filling treatment (e.g. calcium sulphate or calcium phosphate (or both), demineralised bone matrix);
-
different internal fixation devices (elastic stable intramedullary nail versus Kirschner wire);
-
different combinations of multiple interventions;
-
combinations of multiple interventions versus a single method.
Types of outcome measures
Primary outcomes
-
Success rate. The Neer/Cole rating system was used to evaluate success rate (Hashemi‐Nejad 1997; Neer 1966). In this system, the radiographic results are graded as follows: grade 1, cyst clearly visible; grade 2, cyst visible but multilocular and opaque; grade 3, sclerosis around or within a partially visible cyst; grade 4, complete obliteration of the cyst. Grade 1 and grade 2 results are considered to indicate failure and grades 3 and 4 are considered to indicate success (Hashemi‐Nejad 1997; Wright 2008). Since injection procedures, such as injection of steroid and bone marrow, might be repeated, if injection therapy was planned or applied as a series, we planned to base success rate on the outcomes after the last injection.
-
Functional outcome measures, such as the Activity Scale for Kids (ASK) (Young 1996; Young 2000) and Pediatric Orthopaedics Society of North America (POSNA) instruments (Daltroy 1998).
-
Recurrence of bone cyst (reappearance after it has 'healed').
Secondary outcomes
-
Adverse events (e.g. pathological fracture, infection, growth arrest of growth plate).
-
Unacceptable malunion (angular, rotational and shortening), leg‐length discrepancy, limping.
-
Return to normal activities.
-
Patient and parent satisfaction.
-
Pain scores.
We planned also to collect cost and resource use data where available.
Timing of outcome measurement
For the primary outcomes of success rate and recurrence, we stipulated beforehand that we would give preference to data from follow‐ups of over two years.
Search methods for identification of studies
Electronic searches
We searched the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register (24 April 2016), the Cochrane Central Register of Controlled Trials (CENTRAL) (2016, Issue 4), Ovid MEDLINE (including MEDLINE In‐Process & Other Non‐Indexed Citations, MEDLINE Daily and Epub Ahead of Print) (1946 to 26 April 2016), Ovid Embase (1980 to 2016 Week 17) and the China National Knowledge Infrastructure Platform (26 April 2016). We also searched the World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) and ClinicalTrials.gov for ongoing and recently completed studies (24 April 2016). For this update, the search results were limited from December 2013 onwards. Details of the search strategies used for the previous version of the review are given in Zhao 2014. We did not apply any restrictions based on language or publication status.
In MEDLINE, the subject‐specific strategy was combined with the sensitivity‐maximising version of the Cochrane Highly Sensitive Search Strategy for identifying randomised trials (Lefebvre 2011). The search strategies for all databases are reported in Appendix 1.
Searching other resources
We searched reference lists from all relevant articles obtained. We searched The Bone and Joint Journal Orthopaedic Proceedings (26 April 2016) for relevant trials.
Data collection and analysis
Selection of studies
Two review authors independently scanned the titles, abstracts and keywords of every record retrieved, and made inclusion decisions according to the pre‐stated eligibility criteria. Full texts of trials that fulfilled our inclusion criteria and those that were unclear from perusal of the abstracts were obtained. Titles of journals, names of trial authors or supporting institutions were not masked at any stage. Had there been differences in opinion between review authors, these would have been resolved by discussion and by consultation with a third review author.
Data extraction and management
Two review authors independently extracted trial details and data for the included trial using a data collection form. Disagreement was resolved by consensus or a third review author. We did not find it necessary to contact the trial authors to complete data forms or clarify methodology, but would do so if necessary in future.
Assessment of risk of bias in included studies
Two review authors independently assessed the risk of bias for the included trial using the criteria described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). The following items were assessed:
-
randomisation: assessment of selection bias;
-
allocation concealment: assessment of selection bias;
-
blinding of participants and personnel: assessment of performance bias;
-
blinding of outcome assessment: assessment of detection bias;
-
incomplete outcome data: assessment of attrition bias at short‐term and long‐term follow‐up from drop‐out/loss to follow‐up;
-
selective outcome reporting: assessment of reporting bias;
-
other bias: as identified on 'Risk of bias' assessment.
We assessed the risk of bias associated with patient‐rated outcomes and clinician‐rated outcomes for blinding of outcome assessment and incomplete outcome data separately.
We classified items into low risk of bias, high risk of bias and unclear risk of bias (indicating either lack of information or uncertainty over the potential for bias). Any disagreements were resolved by discussion with the third review author.
Measures of treatment effect
We calculated risk ratios (RR) and 95% confidence intervals (CI) for dichotomous outcomes and mean differences (MD) and 95% CIs for continuous outcomes. Had we pooled continuous data from outcomes measured using different scales, we would have calculated standardised mean differences (SMD) and 95% CI.
Unit of analysis issues
There were no unit of analysis issues since the units of analysis in the included study were individual participants. We remained alert to other potential unit of analysis issues, such as repeated observation at more than one time point and multiple observations for the same outcome (e.g. total adverse events).
Dealing with missing data
If necessary in future updates, we would seek missing data, particularly denominators and standard deviations, from the trial authors of the primary studies. Unless we could calculate missing standard deviations from standard errors, exact P values or 95% CIs, we did not impute these. We performed intention‐to‐treat analysis wherever possible.
Assessment of heterogeneity
Besides the visual inspection of forest plot analyses, we intended to examine heterogeneity using the Chi² statistic with significance set at P < 0.1 (Higgins 2003). In our interpretation of the I² results, we planned to follow that suggested in Higgins 2011: 0% to 40% might not be important; 30% to 60% may represent moderate heterogeneity; 50% to 90% may represent substantial heterogeneity; and 75% to 100% may represent considerable heterogeneity.
Assessment of reporting biases
In future if pooled data for our primary outcome are available from at least 10 trials, we will attempt to assess publication bias by preparing a funnel plot.
Our pursuit of trials listed in clinical trial registers and screening of conference abstracts should help to avoid publication bias.
Data synthesis
If data become available in future updates, and when considered appropriate, we will pool the results of comparable groups of trials using both fixed‐effect and random‐effects models. The choice of the model used to report will be guided by a careful consideration of the extent of heterogeneity and whether it could be explained, in addition to other factors such as the number and size of studies included. We will use 95% CIs throughout. We will consider not pooling data where there is considerable heterogeneity (I² > 75%) that cannot be explained by the diversity of methodological or clinical features among the trials. Where it is not appropriate to pool data, we will still present trial data in the analyses or tables for illustrative purposes and report these in the text.
Subgroup analysis and investigation of heterogeneity
If data become available in future updates, we plan to explore the following potential sources of heterogeneity using subgroup analyses.
-
Upper versus lower limb.
-
Active versus inactive (latent) cysts.
-
Children under the age of 10 years versus children aged 10 years or over.
We would investigate whether the results of subgroups were significantly different by inspecting the overlap of confidence intervals and performing the test for subgroup differences available in Review Manager software (RevMan 2011).
Sensitivity analysis
If data become available in future updates, we plan to perform sensitivity analyses to examine various aspects of trial and review methodology. These will include the effects of missing data; inclusion of trials at high or unclear risk of bias (such as from lack of allocation concealment); trials only reported in abstracts; and the selection of statistical model (fixed‐effect versus random‐effects) for meta‐analysis.
Assessing the quality of the evidence
We used the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach to assess the quality of the body of evidence for each outcome (Schünemann 2011). The quality rating 'high' is reserved for a body of evidence based on randomised controlled trials and the quality rating is downgraded to 'moderate', 'low' or 'very low' depending on the presence and extent of five factors: study limitations, inconsistency of effect, imprecision, indirectness or publication bias. We have presented a summary of the results and quality assessments in a 'Summary of findings' table (Schünemann 2011). We presented the evidence for the three primary outcomes (success rate, function, recurrence) plus three secondary outcomes (adverse events (fracture and infection), unacceptable malunion and pain).
Results
Description of studies
See: Characteristics of included studies; Characteristics of excluded studies; Characteristics of studies awaiting classification; Characteristics of ongoing studies.
Results of the search
For this update (December 2013 to April 2016) we screened a total of 225 records from the following databases: Cochrane Bone, Joint and Muscle Trauma Group Specialised Register (0); CENTRAL (20), MEDLINE (68), Embase (72), China National Knowledge Infrastructure Platform (7), WHO ICTRP (20) and ClinicalTrials.gov (38). We did not find any potentially eligible studies from The Bone and Joint Journal Orthopaedic Proceedings.
Most of the 225 articles screened for this updated were excluded after title and abstract screening because the topic was irrelevant or they were clearly not RCTs. We identified seven potentially eligible references for the review. Upon evaluation, six studies were excluded (Akram 2015; El‐Adl 2013; Hou 2016; Kadhim 2016; Traub 2016; Yang 2015), and one study was ongoing (NCT02193841). Thus no new RCTs were included. Based on findings of similarity between a withdrawn registered trial (NCT00459641) and a previously ongoing trial (EUCTR2006‐001294‐18‐GB), we have now excluded the second of these.
Overall, there is one included study (Wright 2008), 23 excluded studies, one ongoing trial (NCT02193841) and two studies that continue to await classification (Abramovic 2006; Imaz 1986). A flow diagram summarising the study selection process is shown in Figure 1. The results from the previous searches (up to December 2013) are reported in Appendix 2.
Study flow diagram.
Included studies
The only included trial was a multicentre randomised clinical trial conducted at 24 centres across North America and India (Wright 2008). This compared bone marrow injection versus steroid (methylprednisolone acetate) injection in children, aged up to 18 years, with simple bone cysts. Of the 90 participants (mean age 9.5 years) included in the study, 45 were allocated to each group. Of the 77 participants included in the two‐year follow‐up analysis, 39 had been allocated bone marrow injection and 38 had been allocated steroid injection. Participants were scheduled for a maximum of three injections, spaced three months apart; the application of second and third injection being determined by the surgeon's assessment of whether the cyst was healing. The primary outcome, which was assessed by a blinded radiologist, was radiographic evidence of healing based on the Neer/Cole rating system. A detailed description of Wright 2008 is provided in Characteristics of included studies.
Excluded studies
Twenty‐three studies (24 references) were excluded for the following main reasons:
-
Seventeen studies were excluded because they were not randomised studies (Akram 2015; Bovill 1989; Brecelj 2007; Chang 2002; Cho 2007; Farber 1990; He 2010; Hou 2010; Hou 2016; Kadhim 2016; Kokavec 2010; Oppenheim 1984; Sakamoto 2010; Stedry 1990; Traub 2016; Tsuchiya 2002; Ulici 2012).
-
One RCT was excluded because it did not investigate the long bones of children (Nagaveni 2010).
-
One RCT was excluded because it recruited both children and adults (Wang 2013). Furthermore, this study included both long bones and other bones. It was not possible to extract either child specific or long‐bone specific data.
-
Two RCTs included other benign bone lesions as well as simple bone cysts (El‐Adl 2013; Yang 2015). Furthermore, El‐Adl 2013 included both long bones and other bones. It was not possible to extract either child specific or long‐bone specific data for these trials.
-
One ongoing RCT was excluded because it was withdrawn prior to patient enrolment (NCT00459641). Another ongoing RCT (EUCTR2006‐001294‐18‐GB) testing the same product by the same manufacturer was excluded since it is likely that the trial had the same fate and product has been superseded.
Ongoing studies
Details of the ongoing trial (NCT02193841) are given in the Characteristics of ongoing studies. This trial is currently recruiting participants and aims to include 160 participants to compare curettage and puncture with or without Vitoss morsels (a bone substitute intended for use as a filler for voids or gaps in bones) for treating simple bone cyst.
Studies awaiting classification
Details of the two studies awaiting classification are given in Characteristics of studies awaiting classification (Abramovic 2006; Imaz 1986). We were not able to find full texts of the two studies and or obtain details of methodology and outcomes. Hence, they are still considered studies awaiting classification.
Risk of bias in included studies
We assessed the only included study, Wright 2008, using the Cochrane 'Risk of bias' assessment tool. Our judgements are summarised in Figure 2.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages in included study.
Allocation
In Wright 2008, an independent biostatistician used a computer to generate the randomisation schedule. Treatment assignments were placed in sequentially numbered opaque envelopes and were assigned by trial managers. Random sequence generation and allocation concealment were therefore at low risk of bias.
Blinding
Performance bias
We judged Wright 2008 to be at high risk of performance bias from lack of blinding of surgeons, who discontinued the protocol because of cyst progression in 13 cases.
Detection bias
Risk of detection bias for radiographic evidence of healing was low as this was assessed by two musculoskeletal radiologists who were blind to the treatment assignment. We judged detection bias for participant‐reported outcomes to be unclear since these were reported two years after treatment.
Incomplete outcome data
Complete two‐year follow‐up data were available for 84.4% of the steroid injection group and 86.7% of the bone marrow injection group. Wright 2008 reported that "statistical analyses were performed on an intention‐to‐treat basis so that patients were analysed according to the treatment group to which they had been randomized". Thirteen participants were not included in the final analysis. Five participants (two in the bone marrow injection group and three in the steroid injection group) did not have two‐year radiographs and eight participants (four in the bone marrow injection group, and four in the steroid injection group) were lost to follow‐up (see Characteristics of included studies). Given that the numbers missing from follow‐up were similar in the two groups, we judged the risk of bias to be unclear for both participant‐ and clinician‐rated outcomes.
Selective reporting
We judged Wright 2008 to be at unclear risk of selective reporting bias because the lack of clarity about the basis of the adjusted analyses gives rise to some concern.
Other potential sources of bias
We judged Wright 2008 to be at high risk from other bias relating to the major differences in baseline characteristics (some not acknowledged) and the post‐randomisation exclusion from the analyses of an unknown, albeit small, number of children who were subsequently found to have an aneurysmal (blood‐filled) bone cyst.
Effects of interventions
Bone marrow versus steroid (methylprednisolone acetate) injections
Wright 2008 compared bone marrow versus steroid (methylprednisolone acetate) injections in 90 children with simple bone cysts. All 45 participants allocated to bone marrow injections started their treatment, but eight did not complete the course of three injections because their surgeon discontinued the treatment protocol. One of the 45 participants allocated to steroid injections did not start treatment because the cyst had healed. A further five participants in the steroid group did not complete their course of treatment (three injections) because their surgeon discontinued the treatment protocol. In all 13 cases, the change in protocol was because of cyst progression. After the exclusion of 13 children who were lost to follow‐up or did not have two‐year radiographs, 77 participants (39 in the bone marrow injection group and 38 in the steroid injection group) were followed up for two years. Wright 2008 reported intention‐to‐treat analyses, analyses that were adjusted for baseline differences in body weight and cyst location with in the bone, and a treatment‐received analysis of success rate. Apart from success rate, we have confined our report to intention‐to‐treat analyses in the results presented below.
Success rate
Fewer children in the bone marrow injection group had radiographically assessed healing of bone cysts (treatment success) at two years than in the steroid injection group (9/39 versus 16/38; RR 0.55 favouring steroid injection; 95% CI 0.28 to 1.09; P value 0.08; Analysis 1.1). Wright 2008 reported that the regression analysis adjusting for baseline differences in body weight and cyst location of the cyst within the bone "indicated that steroid injection was significantly better than bone marrow injection for bringing about healing of bone cysts (P = 0.01)". Wright 2008 reported that the treatment‐received analysis showed similar results "(P = 0.09)" for the two interventions.
Functional outcome (ASK function score)
There was no statistically or clinically significant difference between the two groups in the ASK function score (0 to 100: worst to best function) at two years (MD ‐0.90, 95% CI ‐4.26 to 2.46; P value 0.60; Analysis 1.2).
Recurrence of bone cyst
Recurrence of bone cyst was not reported in Wright 2008.
Adverse events
Wright 2008 reported finding 20 additional fractures and two infections. There was no significant difference between the two groups for additional pathological fracture (9/39 versus 11/38; RR 0.80, 95% CI 0.37 to 1.70; P value 0.56; Analysis 1.3). Similarly, there was no significant difference between the two groups for infection (2/39 versus 0/38; RR 4.88, 95% CI 0.24 to 98.32; Analysis 1.4). Both infections were superficial (Donaldson 2010).
Unacceptable malunion, return to normal activities, and patient and parent satisfaction
Unacceptable malunion, return to normal activities, and patient and parent satisfaction were not reported in Wright 2008.
Pain score
There was no statistically or clinically significant difference between the two groups in the Oucher pain score (Beyer 1988; 0 to 100: no to maximum pain) at two years (MD ‐1.70, 95% CI ‐6.08 to 2.68; 77 children; P value 0.45; Analysis 1.5).
Other outcomes
Wright 2008 reported on the mean number of injections received by the 77 participants followed up for two years (2.1 bone marrow injections versus 1.7 steroid injections; reported P value 0.12). These show that the majority of participants did not have the full course of three injections. Apart from changes to protocol because of 'cyst progression', the reasons for stopping the allocated treatment were not given.
Discussion
Summary of main results
Since our search update identified no new RCTs or evidence for inclusion, our results remain as before.
We found only one relevant RCT (Wright 2008). This compared bone marrow injections with steroid injections for treating simple bone cysts in 90 children, reporting results at two years in 77 children. The evidence is summarised in summary of findings Table for the main comparison. There was very low quality evidence that fewer children in the bone marrow injection group had radiographically assessed healing of bone cysts at two years (success rate) than in the steroid injection group (RR 0.55 favouring steroid injection, 95% CI 0.28 to 1.09). However, the result was uncertain and may be compatible with no difference or small benefit favouring bone marrow injection. Based on an illustrative success rate of 421 children with healed bone cysts per 1000 children treated with steroid injections, this equates to 189 fewer (95% CI 303 fewer to 38 more) children with healed bone cysts per 1000 when treated with bone marrow injections. There was low quality evidence of a lack of differences between the two interventions at two years in terms of functional outcome, based on the ASK function score, and pain, which was assessed using the Oucher pain score. There was very low quality evidence of a lack of differences between the two interventions for the adverse events of subsequent pathological fracture and superficial infection. The trial report did not report on recurrence of bone cyst, unacceptable malunion, return to normal activities, or patient and parent satisfaction.
Overall completeness and applicability of evidence
There is minimal evidence from randomised trials to inform on treatment options for simple bone cysts in children. The only trial provided outcome data for only 77 children, despite being a multicentre trial. It also provided an incomplete picture of outcome: there was no report on recurrence of bone cyst, unacceptable malunion, return to normal activities or patient and parent satisfaction.
The trial population of Wright 2008 (mean age (10 years), gender (predominantly male) and location of bone cysts (predominantly humeral)) is representative of the typical population for children with simple bone cysts who would be considered for the two treatments under test. Similarly relevant to current practice are the two treatments and their application, such as with the use of general anaesthesia.
Wright 2008 suggested that the low success rates in their trial could in part have reflected the more rigorous and independent assessment of bone cyst healing at a longer time period than usual. There was no report in Wright 2008 on what other treatment was provided or planned for children with unhealed bone cysts, either during the two year follow‐up or subsequently. The low success rate points to the need for longer term follow‐up, with recording of treatment, potentially up to skeletal maturity.
The ongoing trial (NCT02193841), is comparing curettage and puncture with or without bone substitute, which are two common surgical procedures for bone cysts used by surgeons today. The estimated completion date of this trial is September 2018.
Quality of the evidence
Using GRADEprofiler software (GRADEpro), we made an assessment of the quality of the evidence for the single comparison based on risk of bias, inconsistency, indirectness, imprecision and publication bias. The description and explanation for our assessment for the quality of the evidence for individual outcomes is provided in the Footnotes of the summary of findings Table for the main comparison. We downgraded the evidence for success rate and adverse events (pathological fracture) by two levels for serious study limitations, reflecting the high risk of performance bias and high risk of bias from major imbalances in baseline characteristics, and one level for imprecision, reflecting the small sample size. The evidence for the participant‐reported outcomes of function (ASK score) and pain (Oucher score) was downgraded two levels for serious study limitations. For both scores, the mean differences and 95% CIs were likely to fall below the minimal clinically significant difference but also, notably, the ASK scores were high indicating good function and the pain scores were low indicating little pain in both groups. Hence, we did not downgrade this outcome for imprecision. Thus, we concluded that the evidence was either of very low quality, meaning that we are very uncertain about these estimates, or low quality, meaning that "further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate".
Potential biases in the review process
We have attempted to reduce bias in the review process through our comprehensive search for trials and adherence to our protocol. However, we cannot rule out the potential for publication bias, which is a major threat to the validity of systematic reviews, especially those that include few and small trials.
Agreements and disagreements with other studies or reviews
We located one review on this topic for treatment of simple bone cysts (Donaldson 2010). Donaldson 2010 identified 16 comparative studies including one RCT (Wright 2008), one prospective comparison and 14 retrospective comparisons. While Donaldson 2010 included two retrospective comparisons of bone marrow versus steroid injections, the authors based their main conclusion on their interpretation of the findings of Wright 2008, and used the results of the analyses selectively adjusted for baseline differences rather than intention‐to‐treat results as we used in this review. Concluding that steroid injections had been shown to be superior in Wright 2008, Donaldson 2010 proposed future prospective trials should compare "steroid injections with other treatments".
Study flow diagram.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages in included study.
Comparison 1 Bone marrow versus steroid injections, Outcome 1 Success rate (radiographic assessment at 2 years).
Comparison 1 Bone marrow versus steroid injections, Outcome 2 ASK function score (0 to 100: best outcome) at 2 years.
Comparison 1 Bone marrow versus steroid injections, Outcome 3 Additional pathological fracture.
Comparison 1 Bone marrow versus steroid injections, Outcome 4 Infection (superficial).
Comparison 1 Bone marrow versus steroid injections, Outcome 5 Oucher pain score (0: no pain to 100: maximum pain) at 2 years.
| Bone marrow versus steroid injections for simple bone cysts in children | ||||||
| Patient or population: children with simple bone cysts | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk1 | Corresponding risk | |||||
| Steroid injection | Bone marrow injection | |||||
| Success rate (radiographically‐assessed healing of cyst healing based on the Neer/Cole rating system) | Study population | RR 0.55 | 77 | ⊕⊝⊝⊝ | ||
| 421 per 1000 | 232 per 1000 | |||||
| Function: ASK function score Scale from: 0 to 100 (best outcome) | The mean ASK function score in the control group was | The mean ASK function score in the intervention group was | ‐ | 77 | ⊕⊕⊝⊝ | |
| Recurrence of bone cyst | see comment | see comment | ‐ | ‐ | ‐ | Outcome not reported |
| Adverse event: pathological fracture | Study population | RR 0.8 | 77 | ⊕⊝⊝⊝ | ||
| 289 per 1000 | 232 per 1000 | |||||
| Adverse event: infection | see comment | see comment | RR 4.88 | 77 | ⊕⊝⊝⊝ | There were no infections in the steroid group and 2 superficial infections in the bone marrow group. The risk of infection for steroid injection is unknown but is likely to be very low |
| Unacceptable malunion | see comment | see comment | ‐ | ‐ | ‐ | Outcome not reported |
| Oucher pain score Scale from: 0 to 100 (maximum pain) | The mean Oucher pain score in the control group was | The mean Oucher pain score in the intervention groups was | ‐ | 77 | ⊕⊕⊝⊝ | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1. The assumed risk is based on the data reported by the included trial 2. Serious limitations resulting from high risk of performance bias and high risk of bias relating to major imbalances in baseline characteristics; evidence downgraded two levels 3. Imprecision: these results come from a single study with a small number of participants and need confirmation in other larger studies; evidence downgraded one level | ||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Success rate (radiographic assessment at 2 years) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 2 ASK function score (0 to 100: best outcome) at 2 years Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 3 Additional pathological fracture Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 4 Infection (superficial) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 5 Oucher pain score (0: no pain to 100: maximum pain) at 2 years Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
