Author(s)' Reply - Vitelline Vessel Remnant - Derived Funisitis: Semantics, Severity, and Significance.

Abstract

The authors thank Dr. Hatano for his letter to the editor in response to our paper, as this creates an opportunity for further dialog on an entity which appeared as a surprising finding in another paper on vitelline vessel remnants (VVRs); in that paper, acute inflammation arising from VVRs was seen in 70.3% of 37 cords with VVRs; cord cross-sections were stained with both H&E and CD15 immunostains and a complex grading/staging system was utilized to characterize the scope and distribution of infiltrates. When writing a follow-up paper which described two primary patterns of fetal circulation within VVRs, 70 new cases of cords with VVRs were studied and, as mentioned in the current paper, these same 70 cases were used in the study currently under discussion. Having found no advantage to CD15 immunostaining, we developed a simpler, H&E-based low vs high grade/stage system in the current paper. Some of the ”basic information” that Hatano describes as “missing” is provided in the longer paper using the same 70 cases. Other requested information is clinical and, since our study was purely histological, we did not have access to this information – nor was it needed to make our histological observations. Hatano et al. recently published a new CD15immunostain-based classification system for neutrophilic responses in umbilical cords. Overall, their method is highly resource-intensive, but we will focus only on the portion of their study related to VVR inflammation; they concluded the presence of CD15+ cells surrounding VVRs has “no clinical or pathological relevance ... as per statistical analysis.” Such a definitive statement cannot be justified after comparing groups with sample sizes of 3 (cases without vitelline vasculitis) and 19 (cases with vitelline vasculitis). According to their online Supplemental Table 2, only 3 of 22 placentas with VVRs also showed histologic evidence of acute chorioamnionitis; interestingly, all 3 cords had vitelline vasculitis and all cases with the presence of both a VVR and chorioamnionitis showed vitelline vasculitis. In our 4-fold larger study, we found VVR-derived funisitis had a sensitivity of 88% and specificity of 100% for histological evidence of amniotic fluid infection (AFI) in third-trimester placentas. Many of their clinical indicators that could logically be affected by AFI showed a supportive trend. Incidence of threatened premature delivery, frequency of threatened abortion, bacterial vaginosis, premature rupture of membranes, and elevated median maternal C-reactive protein were all higher among cases with vitelline vasculitis. Regardless, their “vitelline vasculitis” does not fully overlap with our “VVR-derived funisitis.” Hatano et al. relied upon CD15-staining, which is both too sensitive and not specific (figure). It should be noted that antibodies to CD15 also stain CD68+ macrophages, which are present in Wharton’s jelly, especially in the vicinity of VVRs. Figure 3 of their paper shows a paired VVR surrounded by some randomly scattered CD15+ cells, which could be macrophages. In our opinion, only neutrophils which are easily identified in Wharton’s jelly in H&E-stained sections matter. Furthermore, our data suggest that only high grade/stage VVR-derived funisitis, which requires the identification of neutrophils having migrated to the cord’s amniotic surface, has much likelihood of clinical significance. Adding further inaccuracy, their method of determining severity of vasculitis is based upon whether CD15+ cells are present in the intima (mild), the muscularis (moderate), or Wharton’s jelly (severe); while such a grading scheme may be applicable to umbilical arteries/veins, it is clearly not applicable to thin-walled VVRs, which do not have discrete layers; therefore, all of their “vitelline vasculitis” cases were graded as severe, including their Figure 3. Using H&Estained sections to identify neutrophils, our study found an incidence of 54.3%; of these cases, 61% were high grade/stage. Overdiagnosis (i.e., lack of specificity) and overestimation of severity explains both their astronomical incidence of 86.4% (19/22) and their belief that the finding, even if “severe,” is meaningless. As for the issue of diffusion distance raised by Hatano, VVRs normally reside within 1 mm of the cord’s amniotic surface, and so this is not an important variable.