Low acceptance of osteoanabolic therapy with parathyroid hormone in patients with fragility fracture of the pelvis in routine clinical practice: a retrospective observational cohort study

Patients presenting with an FFP at the emergency department of a trauma centre covering an overaged population of > 300,000 citizens were retrospectively identified. Patients had been admitted to the hospital for pain in the pelvic area after minimal or no trauma. The diagnosis of FFP was based on reported low-energy mechanism of injury combined with radiologic confirmation of a pelvic fracture.

As in all fragility fractures, two major aspects have to be considered: the type of treatment of the specific fracture (operative or non-operative) and secondary fracture prevention. The facts contributing to the decisions for operative vs. non-operative treatment of FFP and the results of the respective treatment are presented elsewhere [24]. In theory, there are five possibilities regarding therapy using the two categories of bone-active drugs available for secondary fracture prevention (Table 1):

Because no patient received osteoanabolic therapy prior to the FFP, there was no “Inverse Switch” group of patients switching from osteoanabolic to antiresorptive therapy. All patients in the “Osteoanabolic” group and in the “Switch” group were invited for dedicated follow-up visits comprising specific blood tests and assessment for adverse reactions (according to Fig. 1).

All patients had been evaluated according to the criteria below:

In our institution, a Fracture Liaison Service (FLS) assures implementation of measures for secondary fracture prevention in all patients suffering from fragility fractures of any anatomical site. The workup for osteoporosis specifically adapted to patients with FFP patients comprised the following:

All the captured data are listed in Table 2.

Demographic data (age, sex) were retrieved from medical records. Descriptive data included performed diagnostics of osteoporosis and therapy for osteoporosis already implemented prior to the incidence of FFP, as well as the individual risk profile with respect to osteoporosis and non-union as documented during the case finding visit by the FLS team.

Newly performed diagnostics of osteoporosis and therapy for osteoporosis implemented after the occurrence of FFP were documented in those patients who attended the scheduled follow-up visit in 2016 at our outpatient clinic. New anteroposterior radiographs of the pelvis were obtained and evaluated for signs of non-unions, migration of osteosynthesis material, and other complications. The incidence of new fractures, adverse events possibly related to therapy of osteoporosis, and revision surgery related to the FFP were documented. Patients unable to attend the outpatient clinic were offered a visit at home or were contacted by telephone call either directly or indirectly via caregivers.

The primary outcome was acceptance or implementation rate of PTH therapy in FFP patients in a real-world setting calculated as the number of patients who were started on this drug divided by the sample size that was available for follow-up.

Those rates identified as secondary outcomes were calculated accordingly and were contrasted by the status of diagnostics and therapy for osteoporosis prior to the FFP in Table 1. The secondary outcomes were: implementation rate of alternative antiresorptive therapy of osteoporosis in FFP patients; rates of participation in a bone metabolic workup, comprising bone mineral testing and laboratory workup. Additional exploratory outcomes were: adverse events possibly related to the use of bone active drugs; manifestation of problems with fracture healing comprising delayed union or non-union; the need for revision surgery in FFP patients; and incidence of new fractures (Table 3).

We identified 79 eligible patients with a mean age of 80.5 years (69 women, 10 men). According to national guidelines, an indication for therapy of osteoporosis was present already prior to FFP in 25 (32%) patients. 30-day mortality was 9% (7 patients), and 13.9% (11 patients) died during the observation period. With 29 patients lost to follow-up or refusal of follow-up investigation, 50 patients could finally be evaluated for all outcomes (Fig. 2).

PTH-related medications increase the cancellous bone mass and reduce the risk for vertebral and non-vertebral fractures [30] mainly by increasing the longevity of osteoblasts. If PTH could make these cells work harder and longer, more patients might reach functional restoration after a fracture at an earlier time point [31]. Therefore, the off-label use of PTH in patients after FFP may be worth exploring.

A series of animal experiments has confirmed the positive effects of PTH on fracture healing in different species, locations, and under various pathological conditions [31, 32]. However, to date, only four controlled studies of PTH and fracture healing have been carried out in humans [8‐11, 33]. Overall, the results of these studies were inconclusive. To define the future role for PTH as an agent for accelerated fracture healing, further clinical studies in various fracture types and patient samples would be needed [6]. However, a double-blind, placebo-controlled study of teriparatide for accelerating fracture healing in femoral neck fractures in men and post-menopausal women had to be terminated early because of low enrolment [34]. This result shows that studies on fracture healing in an aged population are difficult to perform. While our study was not designed to contribute to the little evidence with respect to the effect of PTH on fracture healing, we focused on answering the question of whether there is a real need to initiate these studies on fracture healing in various fracture types as outlined above given that they are difficult to perform. To answer this question, data on patient’s acceptance and experience with comprehensive PTH therapy in routine clinical practice were needed, which—to the best of our knowledge—was not available.

Peichl et al. [10] reported that PTH was implemented in all patients and that all patients participated in a bone metabolic workup representing 100% rate. We can only explain these extremely high rates by the specific study setting in their RCT and would not expect to achieve similar rates in routine clinical practice despite the professional support, for instance, by a FLS. Interestingly, the PTH implementation rate in our study increased when compared to the share of osteoanabolic therapy as reported from its label application [35]. Clearly, there is a need for further clarification regarding the reported 0% rate of pathological lab findings while on therapy, adverse events or patients with early termination of medication application which were also reported by Peichl et al. [10].

There is extensive literature reporting FLS impact on secondary fracture prevention in general by means of key clinical outcomes including bone mineral densitometry testing, bone health referrals, and pharmacotherapy initiation [7]. Because the goal of our study was not to measure FLS team’s performance, we did not document all eligible outcomes mentioned above but rather focused on the rate of bone mineral testing and rates of pharmacotherapy initiation that can be considered as benchmark parameters for assessing the success of FLS teams in other healthcare systems. In our study, these parameters were only assessed to judge the impact of the FLS team’s support in implementing PTH therapy.

Rates of post-fracture bone mineral densitometry in inpatients and outpatients with a fragility fracture seen in a coordinator-based fracture liaison service were reported to be as high as 84% for inpatients and 85% for outpatients. Of the patients who attended their appointments, 73% of inpatients and 52% of outpatients received a prescription of anti-osteoporosis medication [26] in a very advanced FLS program.

Previous FLS studies had recruitment rates for eligible patients ranging from 28.4 to 61.2% and refusal rates between 0.7 and 61.2% [28]. A more recent meta-analysis also reported that—compared with patients receiving usual care—patients receiving care from an FLS program had higher rates of bone mineral density testing (48.0% vs 23.5%) and treatment initiation (38.0% vs 17.2%) [36]. Considering these outcome parameters, the performance of our FLS team may be rated “intermediate” especially when focusing on the overall therapy implementation rate.

With respect to the primary outcome, we attribute low PTH implementation rate to the patient’s burden related with PTH therapy. This is especially true for elderly patients awaiting transfer to a rehab setting and using a lot of pain medication. These factors are arguments against PTH implementation in the acute fracture setting.

With respect to the secondary outcomes, several factors influenced the key outcome parameters of FLS performance. For instance, the anatomical site of the investigated fragility fracture plays a role. Amongst 1270 patients who entered our FLS program between January 2014 and December 2015 [37], only 6% of patients had a FFP compared with 48% presenting with hip fracture. This observation does not seem to be specific to our geographic region but rather the high number of patient recruitment with hip fracture compared to the low numbers of patients with pelvic fractures that can be found throughout the existing literature on post-fracture care of any type. In a study investigating the effect of a coordinator-based screening program for improving osteoporosis management following fragility fractures comprising a total of 147,071 patients, only 5216 patients had a pelvic fracture (3.5%) compared to 61,102 patients or 41.5% with hip fractures [38]. Similarly, a previous multicentre nationwide investigation [39] on secondary fracture prevention in Switzerland included 3667 fragility fracture patients: 4.1% females and 3.8% males suffered from pelvic and sacral fractures, whereas 22.5% females and 22.3% males presented with proximal femoral fractures. Hence, with respect to our primary outcome—that is PTH implementation rate—FFP may not be the easiest fracture model.

Contrary to our consideration mentioned above, we found that all patients with implemented PTH therapy were started as inpatients by the FLS team confirming findings of a recent study. Aubry-Rozier et al. [40] reported initiation of osteoporosis therapy in general with FLS management to be more frequent than with general practitioner management and concluded that the focus should be put on initiation of therapy at the time when patients are still in the hospital rather than delegating this task to the general practitioner. Moreover, to achieve higher therapy implementation rates, it may be necessary to approach the patients repetitively.

A situation frequently described by the FLS team when patients were interviewed in the context of follow-up was that the patients did not know if they had received osteoporosis-specific medication. It is well known that reduced familiarity with treatment decreases adherence. In addition to such patient-level barriers against inclusion in any type of post-fracture care program, there are also provider-level barriers with a potential impact on therapy implementation rates [41]. Information on the suggested therapy may have been unclear or incorrect. Addressing these barriers constitutes a great opportunity for further program improvement at our site. The solutions to these problems reside in dissemination of knowledge on osteoporosis by FLS managers to stakeholders [28], such as general practitioners and other caregivers.