“Osteogenesis imperfecta (OI)” also known as “brittle bone disease” is a rare congenital disease resulting from a defect in the type I collagen synthesis or processing with an incidence of 1:20000. It has a wide spectrum of presentations ranging from almost asymptomatic to severe forms causing increased bone fragility, skeletal deformity, and a range of extra-skeletal manifestations. ¹ Majority of the OI patients have pathogenic mutations in “COL1A1 or COL1A2” genes which code for alpha 1 and 2 chains of type I collagen which are abundant in bones, ligaments, and tendons. Collagen type 1 is produced less frequently and/or abnormally in dominant pathogenic variations. It is commonly known that the variation type, precise location, and implicated gene all affect the phenotypic presentation of these patients. ²

Numerous mutations linked to OI have been found; however, missense mutations mostly cause structural changes in the collagen protein, which results in a more severe phenotype, whereas stop mutations typically result in decreased collagen quantity and a mild phenotype. ³ OI is well known for its clinical manifestations, which include blue sclera, hearing loss, ligament laxity, increased joint mobility, small stature, easy bruising, and normal enamel with dentin abnormalities. Bony manifestations include bone abnormalities, fractures from minor trauma, and the requirement for repeated orthopedic treatments. ³

OI is also associated with easy bruising and bleeding, often attributed to the increased fragility of capillaries and perivascular connective tissue that cannot constrict adequately. The clotting abnormalities in OI patients can be explained due to reduced collagen-induced platelet aggregation response surrounding the exposed sub endothelium, reduced platelet retention, and reduced levels of factor VIII. ^{4, 5} Previous research has also shown enlarged platelets, diminished retention of platelets, and decreased factor VIII (FVIII) as the possible reasons for the bleeding manifestation. ^{6, 7}

However, in OI patients even with normal coagulation profile, bleeding might still happen, which makes intraoperative bleeding unpredictable. ^{8, 9} The literature review revealed no studies involving the pediatric population in India and a few Western studies analyzing the perioperative blood loss in pediatric OI patients undergoing orthopedic surgeries. In a retrospective analysis of 23 OI patients, aged between 6 and 13 years, who underwent osteosynthesis for femoral shaft fractures or correction of femoral deformities, Persiani et al. ¹⁰ from Italy identified predictive risk factors regarding intraoperative bleeding, revealing that patients affected by type III OI have a high risk of severe blood loss during surgery. Similarly, a study conducted in the United States by Pichard et al. ¹¹ examined a retrospective review of 22 pediatric OI patients who had 42 surgeries involving the insertion of 52 femoral rods. The results indicated that an increase in osteotomies was associated with an increase in adjusted mean blood loss (P = 0.05). Therefore, having access to autologous blood donations or a sufficient supply of blood products will help address the potential issue of increased perioperative blood loss, a crucial component of treatment, while organizing the care of patients with OI. Therefore, the purpose of this study is to determine the variables influencing perioperative blood loss and need for blood transfusion in pediatric OI patients undergoing osteotomies.

This was a retrospective case-record-based study conducted in pediatric OI patients who underwent osteotomies at a tertiary care hospital. The research conducted in this study adhered to the principles outlined in the Declaration of Helsinki and was approved by the Kasturba Medical College and Kasturba Hospital Institutional Ethics Committee (IEC 363/2024) on 18th September 2024. A waiver of consent was granted as per our institutional ethics committee due to the retrospective nature of the study.

Individual patient anesthesia records, daily progress notes, initial history, physical examination, and doctor instructions from each hospital stay were obtained through electronic records and analyzed. Each surgical procedure was analyzed as a discrete event, even when multiple procedures were performed on the same patient at different time points. Variables analyzed were age, type of OI, number of osteotomies performed during the surgical session, number of bones operated on during the same anesthesia session, preoperative postoperative hemoglobin level, and requirement for perioperative blood transfusion.

Single osteotomy was defined as one osteotomy performed during a surgical session whereas multiple osteotomies was defined as more than one osteotomy performed during the same anesthesia session. The anatomical site, deformity severity, and complexity grading were not systematically coded and therefore were not included in the analysis.

During our study period, 53 patients who underwent a total of 141 procedures during their stay in hospital were analyzed and each surgery was analyzed as a discrete event. Their mean age was 11.07±5.29 years; 4/53 (7.8%) were phenotypically type I OI, the majority (n=33; 62.2%) were phenotypically type III OI and the remaining 16/53 (30%) were type IV OI. 61/141 (43.3%) surgeries involved a single bone, and 80/141 (56.7%) surgeries involved multiple bones. Of 53 patients requiring osteotomies, 11/53 underwent once, 9/53 underwent twice, 17/53 (majority) underwent thrice, 7/53 underwent four times, 4/53 underwent five times, 3/53 underwent six times and only 2/53 underwent osteotomies seven times ( Table 1).

A.

Drop in hemoglobin 1.

During this study, it was noted that the fall in hemoglobin was statistically significant with multiple osteotomies (p=0.002) when compared to single osteotomy (p=0.297), and the total fall in hemoglobin between the pre-operative (12.22±1.06 g/dL) and post-operative (10.57±1.48 g/dL) period was also statistically significant (p≤0.001) ( Table 2).

“Osteogenesis imperfecta” is a broad category of hereditary collagen type I diseases. Bony deformities, heart valvular lesions, cognitive abnormalities, and metabolic disturbances are frequently linked to OI. In addition to defective collagen synthesis, patients with OI have increased capillary fragility, decreased platelet retention, decreased levels of factor VIII, and deficient collagen-induced platelet aggregation that causes excessive bruising and widespread oozing from wound sites, thus surgical procedures performed on these patients are more likely to result in bleeding complications despite normal coagulation parameters which makes assessment of perioperative blood loss unpredictable and is of a major concern to the surgeons. ^{2, 3, 5}

In this study, 53 patients who underwent a total of 141 surgeries were analyzed (an average of 3 surgeries per patient) during their stay in hospital and each surgery was analyzed as a separate event. Their mean age was 11.07±5.29 years, the majority (n=33; 62.2%) were phenotypically type III OI, and 80/141 (56.7%) surgeries involved multiple bones. This is similar to the study by Ruck et al., ¹² who conducted a retrospective analysis of 60 OI children undergoing femoral rodding, showed a mean age of 4 years which is lower when compared to our study, however, the majority (n=30/60) had type III OI which is similar to our study. The study by Persiani et al. ¹⁰ was conducted on 23 patients aged between 6 and 13 years (mean - 8.9±1.9 years) affected by type III OI, wherein 42 osteotomies were done, and the majority (n=11/23) underwent an average of 3 osteotomies which is similar to our study.

Our study showed a greater fall in hemoglobin in patients with multiple osteotomies done simultaneously (p=0.002) when compared to a single osteotomy and also the total fall in hemoglobin post-operatively was statistically significant (p≤0.001) which is similar to the study done by Persiani et al. ¹⁰ that showed average effective blood loss increased significantly as the number of osteotomies increased (p=0.046). The association between multiple osteotomies and greater hemoglobin decline may reflect increased procedural invasiveness. However, surgical approach (open vs. percutaneous) and surgical duration were not systematically recorded and therefore were not included in the analysis. Both are established determinants of blood loss. Without adjustment for these variables, it is not possible to determine whether the number of osteotomies independently contributes to bleeding risk or serves as a proxy for more extensive surgical exposure. Thus, the use of a structured bleeding survey is more advantageous than laboratory measurements as there is little correlation between the severity of bleeding with the levels of a particular factor, and the standard tests do not accurately reflect in vivo hemostasis due to the unmeasurable contribution of numerous factors (such as vessel fragility or fibrinolysis). ^{13– 15}

This study revealed a negative correlation between the age of the patients and a fall in hemoglobin, suggesting that older children had better tolerance for blood loss when compared to a younger age group. Similar findings were reported in the study by Pichard et al., ¹¹ which involved a retrospective review of 22 patients. The oldest patient, who underwent surgery, was 21 years and 2 months old, and the youngest, who underwent surgery, was 1 year and 7 months old. Of the 42 surgeries examined, the mean blood loss was 197 cc, with older patients generally having lower adjusted mean blood loss, though this difference was not statistically significant (p=0.07). Although younger age was statistically associated with greater hemoglobin decline, the correlation was weak (R ² = 0.076), indicating limited predictive value. This association should therefore not be overinterpreted. The most likely explanation offered is that while a larger bone’s radius of diameter and, hence, its area of bleeding may cause more bleeding, a larger patient with a larger total blood volume may be able to withstand more bleeding. ¹¹

In this study, type III OI showed a significant hemoglobin drop (1.82±0.86 g/dL; p=0.01) and post hoc analysis also confirmed that the hemoglobin fall was significantly associated with type III OI (p=0.008) when compared to the other types of OI which are similar to the study by Persiani et al. ¹⁰ wherein patients affected by type III OI had a high risk of severe blood loss during surgery. The probable explanation is that type III OI is characterized by increased capillary fragility and an altered platelet function caused by platelet dysfunction due to alteration in collagen when compared to other types of OI.

However, these findings must be interpreted with caution. Hemoglobin decline was used as a surrogate marker for perioperative blood loss because standardized documentation of estimated blood loss was not available. Hemoglobin levels may be influenced by hemodilution, intraoperative fluid administration, timing of measurement, and transfusion practices. Therefore, the observed changes reflect perioperative hematologic alterations rather than directly measured blood loss.

The perioperative transfusion requirement in surgeries for OI patients was found to be 23/141 surgeries (16.3%) in this study. Our study is similar to the study by Pichard et al. ¹¹ wherein six blood transfusions were given with a transfusion rate of 14%. The study under reference revealed that the average blood loss among transfused patients was 279 cc. Additionally, patients who underwent transfusion had an adjusted blood loss of 0.330 as opposed to those who did not get blood transfusion, who had an adjusted blood loss of only 0.003. Our findings are in line with studies by Gooijer et al. ¹⁶ and Oakley et al., ¹⁷ which found that 17% of OI patients needed blood transfusions following surgery. For this reason, it is crucial to be aware of the bleeding risk. Despite a normal pre-operative hematological assessment, several studies ^{18– 21} describe severe bleeding in OI patients as a result of surgery, thus in patients with OI, “American Society of Anesthesiologists transfusion guidelines” state, “platelet transfusion may be indicated despite an adequate platelet count if there are known platelet dysfunction and microvascular bleeding.” “Bleeding time and platelet aggregation tests are not useful in the operating room”, and “there is an urgent need for the development of clinically relevant measures of in vivo platelet function and bleeding risk to guide the rational use of platelet transfusion”.

Our study showed incidence of transfusion was higher in those who underwent multiple osteotomies simultaneously (p=0.001) which is similar to the study by Persiani et al. ¹⁰ wherein the perioperative transfusion requirement was more in type III OI but not statistically significant (p=0.387). These results were comparable to those of research by Hathaway et al. ²² and others, ^{23, 24} which discovered aberrant platelet adhesion, poor platelet factor 3 (PF3) release, aberrant platelet aggregation to ADP, and commonly faulty platelet aggregation in type III OI patients, thus corroborating with our finding of increased transfusion requirement in type III OI patients. Since the relationship between genotype and phenotype is frequently less strict than previously believed due to variability in penetrance and expressivity, coinheritance of hemostatic defects, or superimposed genetic modifiers, a genomic search for the molecular basis of inherited clotting and platelet defects may not be as beneficial. ¹⁵ When regular hemostasis testing revealed no repeatable anomaly in a group of individuals with a significant history of bleeding, Obaji et al. ²⁵ administered tranexamic acid or desmopressin, and they observed no bleeding in 90% of the patients at high risk of bleeding receiving an intervention which can be used in OI patients as well before surgery to reduce the bleeding incidence post operatively. Yet another prospective randomized controlled study by Elbaseet HM et al. ²⁶ wherein patients with osteogenesis imperfecta undergoing femoral telescoping nail the use of tranexamic acid (local or intravenous) intraoperatively revealed a significant decrease in blood loss, thus supporting the use of antifibrinolytic agents in these patients.

The most reliable indicators of perioperative bleeding and the need for transfusion in procedures involving OI patients were the patient’s age, the type of OI, and the number of osteotomies. These results imply that higher perioperative hematologic changes in children with OI may be linked to specific patient and procedural characteristics; however, to determine clear risk factors and direct perioperative management strategies, prospective studies involving standardized blood loss measurement, comprehensive surgical variables, and multivariable analysis are needed.