Upper Lip Tie and Breastfeeding: A Systematic Review

Introduction

Upper lip tie (ULT) is believed to be linked to breastfeeding difficulties, and it has been suggested by Kotlow that lip tie release improves breastfeeding. ¹ The ULT is the anatomic entity that tethers the upper lip to the upper gum. Most infants have some degree of ULT, but it has been postulated by Kotlow and others that when the ULT becomes large and tight enough, it may prevent the upper lip from flaring out or curling up during breastfeeding, and subsequently prevents the creation of an adequate seal with the breast.^1,2 It has also been claimed that if the ULT is tight enough, an infant may have trouble feeding even from a bottle. ²

Many practitioners involved with the surgical treatment of tongue-tie also perform ULT release during the same surgical session, with the claim that breastfeeding difficulties may be ameliorated by this double procedure. ² A classification system of the severity of the condition has even been proposed by Kotlow ³ and used in several articles by this author.

In a recent systematic review (SR), ⁴ we described a dramatic increase in the number of articles published every year in relationship to tongue-tie, frenotomy, and breastfeeding. However, despite this increasingly abundant literature, we found that there were many unanswered questions, and that the amount of medical evidence in favor or against routine frenotomy in infants with breastfeeding difficulties and tongue-tie was very sparse. ⁴ Furthermore, since it is known that in many cases, the issue of tongue-tie improves over time, it has been suggested that the rapid rise in frenotomy rates might be a “fad” driven by practitioners for financial gain. ⁵ In neonates, ULT release is usually performed by practitioners who also perform tongue-tie release.

We thus conducted the following SR to evaluate the existing literature on ULT. We aimed to provide an updated review of the literature. Based on evidence, we attempted to define whether or not ULT may impair breast feeding and whether or not ULT release may improve breastfeeding in infants with breastfeeding difficulties.

Materials and Methods

We used the following Internet address: www.ncbi.nlm.nih.gov/entrez to evaluate all Medline articles registered from 01/01/1948 and until 07/05/18, at the time the study was conducted (05/07/18). Key words of upper lip tie OR lip frenum were used for the search. The full text of all articles were retrieved, and each article was classified as either a case report (or series), review, editorial (or opinion), cohort study, clinical trial (nonrandomized) (CT), randomized controlled trial (RCT), basic (or animal) study, or a SR. We did not include any language limitation. Contrary to Medline's own classification of articles, we did not list RCTs as CTs. Also, articles based on a case report and a review of the literature were only listed as case reports. Thus, each article appears only once in our classification.

We also used the same keywords and repeated the search on Google Scholar and Embase. We used the PRISMA checklist to make sure that our study is indeed an adequate SR, following all PRISMA guidelines. ⁶

Results

Thirty-two articles were retrieved by the search (Fig. 1). Twenty-four additional records were identified through the bibliography of these 32 articles, 2 of which were removed because of duplication. The 54 remaining articles were screened, and 29 records were excluded based upon the fact that they were unrelated to the topic studied. The 25 remaining articles were retrieved in a full-text version, and 10 of them were excluded because they were unrelated to the topic studied. Thus, only 15 articles remained in the final qualitative analysis. These articles consisted of one SR, six reviews of surgical techniques for ULT release, four cohort prospective studies of the incidence of ULT (two of which without consideration of breastfeeding difficulties or not), two case reports, and two retrospective cohort studies. There was no single RCT of isolated ULT release in children with breastfeeding difficulties. Thus, the articles we retrieved did not allow for any quantitative analysis. The main findings of the abovementioned studies are as follows:

OPEN IN VIEWER

FIG. 1.

PRISMA flow diagram of literature retrieval.

The SR of Francis et al. ⁷ addressed specifically the issue of ankyloglossia or ankyloglossia associated with concomitant ULT, and concluded that a small body of evidence suggests that tongue-tie frenotomy with ULT release may be associated with improvements in breastfeeding as reported by mothers, and potentially in nipple pain, but did not address at all the additional effect, if any of ULT release.

The six additional reviews about techniques of ULT release dealt with laser soft tissue treatment for pediatric patients (among them ULT release, without any specific study of the efficacy on breastfeeding adequacy,^8–12 and at most related to the personal experience of the operator. ¹³ Four of these articles are by Kotlow,^9–12 who in two of these articles^9,12 describes a classification of ULT severity.

According to Kotlow, the classification was established by the examination of “over 100 children” between the ages of 8 months and 3 years (none of them being a neonate). ⁹ The classification is based upon the insertion points of the lip tie to the maxillary gingival tissue and utilizes both the clinical appearance of the lip and its attachment to the maxillary arch. Kotlow stated that the soft tissue covering the maxillary bone is divided into three zones. Zone 1 involves the movable tissue just under the nasal area and is also called free gingival area. Zone 2 encompasses the tissue attached to the bone and has little freedom of movement. Zone 3, also known as the interdental papilla, extends into the area between the future teeth. According to Kotlow, class I lip tie is when the lip tie “does not appear to create problems during breastfeeding,” and has a high insertion, with “little or no tissue attachment of the lip to the gingival tissue.” However, Kotlow also states that “in a normal, healthy infant, there are no known contraindications to correcting lip tie” (without allusion to its class). A class II lip tie displays an insertion point at the junction of the free and attached gingival margins. A class III lip tie occurs when the frenum inserts between the areas where the maxillary central incisors will erupt, just short of attaching into the anterior incisor. Class IV lip tie, which Kotlow considers to be “the most severe frenum,” involves a lip tie wrapping into the hard palate and into the anterior papilla (a small bump located just behind where the central incisor will erupt).

Kotlow considers that class 2–4 may impede breastfeeding and should be released. In the latter report, Kotlow describes two infants whose ULT was removed, and in whom breastfeeding adequacy greatly improved after the procedure. He also stated that in his “experience of treating over 1,000 infants,” revision of the lip tie has shown an increased ability of the upper lip to flange upward, allowing the infant to develop a more effective latch, thereby improving the breastfeeding experience for both infant and mother. No further details were provided about the “over 1,000 infants” so treated, or about breastfeeding outcomes. In references 3, 9, 10, 11, and 12, no disclosure statements were provided by the author.

However, in a lecture given by Kotlow and distributed on the Internet, ¹² the author states that he has “assisted in the development of a variety of laser products, is an investor in the development of the Solea CO₂ laser, is on their professional advisory board, has been a beta tester of new products, and receives honoria or supplies for his participation.” ¹⁴ This level of involvement may well create a conflict of interest, which should be acknowledged when evaluating the research findings.

Out of the three case reports, one article describes the successful outcome of ULT release in one single infant. ¹⁵ Another one by Kotlow included only two infants ¹ with a similar successful outcome. Again, Kotlow states that he has treated over 1,000 such infants, but no details are provided as to how they were selected, what were their symptoms, and what were the outcomes studied.

The cohort prospective study by Flinck et al. ¹⁶ is the largest description of the rates of ULT and other oral findings in 1,021 Swedish infants. ¹⁶ An ULT of some degree was found in all patients. In 6.7% of them, the frenulum was buccal of the alveolar ridge; in 76.7%, it was to the crest of alveolar ridge; and in 16.7%, it was palatal of the alveolar ridge. This description, as novel as it may be, did not address the issue of breastfeeding difficulty rates in any of the three groups described.

The recent cohort prospective study by Santa Maria et al. ¹⁷ included 100 consecutive infants. The purpose of the study was to determine what is normal and what is abnormal ULT, and to determine the inter- and intraobserver reliability of the classification by using pictures of the patients' ULT. All six trained observers were asked to grade each picture, using the Kotlow classification, twice at different times, and were blinded to the ratings of the other raters as well as to their own previous rating. Santa Maria found that all newborns have labial frenula, with most attached at the gingival margins (83%). Raters had poor intra- and interrater reliability (64% to 74% and 8%, respectively), using the Kotlow classification system.

The four cohort studies that assessed the effect of tie release on clinical outcomes were that of Pransky et al., ¹⁸ Siegel, ¹⁹ Ghaheri et al., ²⁰ and Benoiton et al. ²¹ Neither of these studies was randomized and had control group, which did not undergo release procedures for comparison.

The two latter studies assessed the effect of tongue release on breastfeeding outcomes in mother–infants dyads. These two studies together enrolled only four patients with isolated ULT and report significant improvement in three of the four patients. In both studies, many patients had combined ULT and ankyloglossia frenotomies. Thus the added benefit of ULT release, if any cannot be evaluated. Pransky's study was a retrospective one based upon a population of 618 patients referred to a dedicated breastfeeding difficulty clinic. All patients had breastfeeding difficulties and those identified to have ankyloglossia and/or ULTs (79%) underwent release procedures of the tongue-tie AND the ULT if present. The presence of an isolated ULT was noted in 2% of infants, and 6% had both ULT and ankyloglossia.

Pransky reported an improvement in breastfeeding in 100% of infants with isolated ULT, and in 85–91% of infants with both ankyloglossia and ULT (depending upon the type of ankyloglossia). The study by Siegel focused on the theory that ULT may cause an inadequate seal around the breast, excessive air swallowing (aerophagia), and subsequently may cause or aggravate gastroesophageal reflux (GER). Siegel's study was a retrospective 1 of 1,000 patients recruited over a 5-year period. All infants had been referred to a private oral and maxillofacial surgery practice “with heavy emphasis on breastfeeding medicine and treatment of ankyloglossia in breastfeeding infants.” All infants underwent release of both ULT and tongue-tie, while there is no description in the article of the anatomic or functional severity of either condition. The outcome with regard to breastfeeding is not described, but GER symptoms (which are not defined clearly in the article) are reported to improve in 52.6% of the infants. The latter study was also not randomized, did not report effect of the release on breastfeeding adequacy, had no control group that did not undergo release procedures for comparison, and could not differentiate between ULT and tongue-tie since all infants underwent both release procedures.

Discussion

In this SR, we aimed to provide an update on what is currently known on the potential relationship between ULT and breastfeeding difficulties, and whether or not ULT release may improve breastfeeding in infants with breastfeeding difficulties. We find it very difficult to answer these two questions because of the major flaws in the available literature.

We found that the definition of ULT is unclear. The Kotlow classification method was based upon children aged 8 months to 3 years (thus many of them, if not all, with teeth), and not one single neonate was used to establish this classification. Moreover, this classification is only an anatomic one, and there has been no one single attempt to correlate the anatomic severity of the condition with the prevalence of clinical breastfeeding difficulties. Furthermore, the classification has never been validated on a functional basis. Santa Maria's prospective cohort study questioned the reproducibility of Kotlow' classification; in that, there was poor inter- and intraobserver agreement. They found that the majority of infants had a significant level of attachment of the labial frenulum, and they concluded that “the Kotlow classification of lip tie failed to be reproducible by relevant experts.”

In terms of treatment, we found that contrary to the literature on tongue-tie, there was no one single RCT that compared the efficacy of an ULT release procedure on breastfeeding difficulties. In the absence of such studies, it is impossible to determine with certainty whether short-term and/or long-term outcomes of breastfeeding are improved by the release. Moreover, such studies would be very difficult to perform in a blinded manner, and one cannot know with certainty, in a nonblinded study, whether a reported improvement is real or a placebo procedural effect-related phenomenon.

It is clear that available studies performed in a nonrandomized manner have originated from practices potentially associated with financial gain from the procedure.^18,19 Many practitioners involved with ULT release advertise themselves on the web, while claiming the major beneficial effects of ULT release, the nearly nonexistent complication rate and the morbidity involved with nontreatment. As an example, the website of Palmer speaks of a possible link between ULT and the Sudden Infant Death Syndrome (“I believe there is a link but I do not have scientific evidence to back that up”). ²²

We conclude that ULT needs to be better defined. It would be important to use Santa Maria's methodology of blinded observers (who are shown pictures of infants' upper lips) and to correlate anatomic findings with functional findings (breastfeeding difficulties or not, use of latch scores) in a prospective manner. It would also be important to verify whether ULT has an effect on breastfeeding that is independent of that of tongue-tie. Once this is established, it might be justifiable to conduct a RCT of ULT release and to assess both short-term and long-term effects of the release on outcomes such as nipple pain, latch efficacy, duration of feeds, well-defined symptoms of aerophagia, duration of breastfeeding, and so on.

We strongly believe from the available evidence that there is no justification for routine ULT release in infants with breastfeeding difficulties. If performed, this procedure should be conducted in the context of a RCT.