DICH is one of complications of VP shunt surgery and was first reported in 198513. It was considered to be a rare complication with incidence varies from 0.4%-4%4, 5, 14–16. However, the incidence might be underestimated. Patient with small hematoma might be missed if CT scan was not performed frequently4. An article published in 2017 reported that the incidence of postoperative DICH was 7.8% (17/218)17. In our study, the incidence was 17.5%, which was much higher than previous studies. Another article published in 2018 reported that the incidence was 23.7% (34/143) without excluding patients with anticoagulant and antiplatelet therapy18. Just as our study, they included the patients with hematoma volume less than 1 ml which would be ignored easily and maybe it is the reason. The incidence of symptomatic DICH was 3.6% (6/166) in our study.
The mechanisms underlying the DICH secondary to VP shunt are still controversial. Several hypotheses have been proposed: (1) erosion of cerebral vasculature by the insertion of catheter; (2) fragility of cerebral tissue caused by advanced age, craniotomy, trauma or stroke; (3) disseminated intravascular coagulation (DIC) induced by VP shunt; (4) coagulopathy, anticoagulant or antiplatelet therapy; (5) sudden change of intracranial pressure after manipulation of the valve system7, 19. Savitz and Bobroff5 pointed out that the mechanism of DICH was more likely erosion of surface or deeper small vessel by catheter. This opinion was supported by most reports because most hematomas located along the catheters4. The hypercapnia, hypoxia and venous congestion might encourage developing of hematoma at the sites of injury just as the mechanism of traumatic delayed ICH20. A study published in 2017 found that postoperative cerebral edema around the catheter observed on the first CT scan was an independent risk factor for DICH16, which might be a sign of vascular erosion. Nevertheless, this difference was not found in our study (P = 0.674). Advanced age, history of craniotomy were considered to be the risk factors for DICH secondary to VP shunt in several articles15–18, and these factors might increase the fragility of cerebral tissue. However, one article published in 2017 reported that there was no difference between two groups with respect to age17. In our study, we also found that history of craniotomy was an independent risk factor for DICH, while advanced age was not. DIC induced by catheter insertion was considered to be another potential mechanism of DICH. Two cases of DIC associated with VP shunt were reported21, 22. A Korean study found that prolonged partial thromboplastin time was major risk factor of DICH18. Some studies showed that dual antiplatelet therapy and postoperative anticoagulation therapy would increase the risk of DICH17, 23. As for our study, we excluded the patients with anticoagulant or antiplatelet therapy in order to control the confounding factors, and we found that preoperative PT, APTT, INR and PLT were not risk factors for DICH. Two articles presented that postoperative manipulation of valve system might be a risk factor for DICH secondary to VP shunt4, 15, which was not supported (P = 0.382) in our study either.
As the main purpose of the study, we concluded that elevated NLRR could independently predict DICH secondary to VP shunt, which suggested that inflammatory responses might play an important role in the development of DICH. Catheters of VP shunt are made of silicones and may not be immunologically inert24. A study pointed out that immune response might be elicited by VP shunt in some patients and could lead to shunt malfunctions25. Neutrophils and giant cells were found on the surface of catheters by scanning electron microscopy26. In view of the above, we hypothesize that the inflammatory responses may arise from the stimulation of catheter as a foreign body. The acute inflammation phase of foreign body reaction against biomaterials is characterized by migration, adhesion, activation of neutrophils and mast cells, and lasts for hours to few days27. Just as the process of brain injury28, neutrophils’ number increase greatly in the peripheral blood and they can enter central nervous system through the damaged blood brain barrier early11. Recruitment and infiltration of neutrophils around catheter could induce neurotoxicity by following pathways: production of cytotoxic mediators and proinflammatory cytokines, activation of matrix metalloproteinases and increase of oxidate stress29–31. The ensuing further destruction of blood brain barrier, cellular swelling and increased permeability of capillary might trigger cerebral edema and active bleeding32–36. Lymphocytes play an important part in the cellular and humoral immune. It was found that autoreactive T cells could promote vascular reconstruction and healing after cerebral trauma37. Therefore, increased neutrophils and decreased lymphocytes might induce DICH secondary to VP shunt. We suggest that the patients with NLRR > 2.05 should be more carefully observed after VP shunt, and perform CT scans more frequently. However, there was no correlation between the NLRR and hematoma volume. The volume of hematoma might be affected by many factors, such as blood pressure, coagulation function and so on. Since limited understanding of the mechanisms of DICH, our finding would also contribute to identify potential preventive and curative strategies.
NLRR, as a new inflammatory parameter, has smaller variation range than NLR, and can roughly represent the change of inflammatory status due to surgery (including anesthesia) and perioperative treatment. Maybe it could be used as predictors of other diseases requiring surgery, such as postoperative rebleeding of ICH following minimally invasive surgery.
There are several limitations in our study. The first, it is a retrospective study with small sample size, and a quarter of patients were excluded due to incomplete laboratory or radiological data, which may induce potential selection of bias. The second, there were too many variables included in the multivariate logistic regression analysis, the model might be unstable, even though the result was reconfirmed by PSM analysis. The third, postoperative treatments such as hemostatic therapy were not included in our study, which might be confounding factors. Finally, preoperative neutrophils and lymphocytes were collected within 5 days before surgery. In general, patient’s condition was stable before operation, fever or other unusual situation would lead to cancellation of operation, and the laboratory indexes would not change greatly in these days. However, there were still small deviations in these data and they could not represent the preoperative inflammatory status accurately. NLRR’s predictive value should be verified by further larger prospective studies.