This section describes the following treatment scenarios:

• Postmenopausal women with multiple vertebral fractures (DXA not essential but other destructive diseases should be excluded)
• Postmenopausal women with osteoporosis determined by axial DXA and a history of at least one vertebral fracture
• Postmenopausal women with osteoporosis (determined by axial DXA) with or without a previous non-vertebral fracture
• Frail, elderly women with a diagnosis of osteoporosis, with or without previous osteoporotic fracture
• Men with a diagnosis of osteoporosis determined by axial DXA with or without previous osteoporotic fractures

HRT is discussed within each scenario, and further information about HRT is also given in section 5.

The quick reference guide accompanying this guideline summarises the decision making process for each of these scenarios (except for men).

6.1 The aims of treatment

6.1.1 reduction in the incidence of fractures

The prime objective of treatment for osteoporosis is to reduce the incidence of fractures. Patients with a history of previous fracture are two to eight times more likely to have a fracture at any skeletal site. These patients are readily identifiable and should be prioritised for treatment. Patients with vertebral fractures, whether painful or painless, share the same or possibly greater future fracture risk but only a minority of these are currently recognised and treated.

6.1.2 alleviation of fracture related morbidity

The priority from the patient's perspective may be the provision of an effective analgesic regimen - to reduce the pain associated with vertebral fracture, for example. Treatment of osteoporosis to achieve the secondary prevention of fractures will not influence pain that is currently being experienced by a patient as a result of a fracture. Advice on pain management is given in Section 6.11.

6.1.3 which subsequent fractures can be prevented?

The treatments reviewed in this section can broadly be divided into those that have been shown to have potential to reduce the incidence of vertebral fractures only and those with wider potential efficacy in reducing fractures at any skeletal site (ie vertebral and non-vertebral sites). Some therapies have been shown to reduce the incidence of hip fractures as part of their efficacy in reducing the risk of non-vertebral fractures. In general, the first choice therapeutic option would be a treatment that is effective in reducing both vertebral and non-vertebral fracture risk.

6.1.4 risk factors for fracture and their use in targeting treatment

The key risk factors for fracture are low BMD, past history of fracture, age, and the risk of falling (see Section 2).^35,78 Together these define the baseline fracture risk and ultimately are useful in defining who should be treated. Women aged over 60 with osteoporosis and a history of osteoporotic fracture are at greatest risk of vertebral and non-vertebral fractures.³² Assessment of BMD by axial DXA is a prerequisite for targeting treatment in the vast majority of cases. The notable exception is patients who have had at least two vertebral fractures and are known not to have underlying metastatic tumours, myeloma or other destructive disease, when clinical trials have shown that the bisphosphonates are capable of achieving the secondary prevention of vertebral fractures. In all other situations, BMD assessment by axial DXA would be essential to define a level of fracture risk at which treatment is likely to be effective in reducing the patient's risk of fracturing. Targeting treatment to reduce fracture risk on the basis of clinical risk factors for falling (without measurement of BMD) has been shown to be ineffective as a means of reducing the incidence of fractures.¹¹⁶

Efficacy of therapies in reducing fracture risk has most frequently been assessed in terms of the impact on vertebral fracture risk. The majority of studies have specifically assessed potential reduction in incidence of “morphometric” (ie non-clinical) vertebral fractures. These are usually defined on the basis of reductions in posterior, mid, or anterior vertebral body dimensions and these may be accompanied by a semiquantitative grading of fracture severity. The clinical trials differ in the thresholds of height loss that define incident morphometric vertebral fracture. Studies of risedronate^117,118 and HRT¹⁰⁶ have based the definition of incident vertebral fractures on a decrease of at least 15% in one of these vertebral dimensions, albeit in association with changes in the semiquantitative grading system. In other clinical trials of etidronate,¹¹⁹ alendronate,^36,37,120 raloxifene,¹¹⁰ and calcitonin¹²¹ incident vertebral fractures have been defined on the basis of loss of at least 20% of vertebral body height. This has a key influence on the placebo vertebral fracture rates. Results of the main placebo controlled RCTs of drug treatments to reduce fracture risk are summarised in Table 6.

Selection of specific drug therapy for an individual patient is ultimately at the discretion of the prescribing clinician and should take into account the risk versus benefit of therapy in the context of the patient's health record and their individual concerns. Similarly, therapeutic regimens should be reviewed periodically as the risk/benefit ratio may change over time.

There are two important considerations that should be taken into account when using oral bisphosphonates.

All bisphosphonates are poorly absorbed. Typically, only between one and five per cent of the ingested dose is actually absorbed. Optimal absorption requires all bisphosphonates to be ingested on an empty stomach, either first thing in the morning after overnight fasting with the subsequent avoidance of food for 30 minutes or in the middle of a four hour fast, and they should be washed down with a large glass of water.

All bisphosphonates can potentially be associated with gastrointestinal side effects. For aminobisphosphonates such as alendronate this can on rare occasions present as oesophageal ulceration. The risk of these symptoms can be lessened by the avoidance of lying flat within 30 minutes of ingestion, or by using the once weekly preparations.

Given the complicated nature of the protocols for ingestion, it is doubtful if bisphosphonate therapy would be appropriate for patients who are unlikely to be able to comply with such a regimen. This may apply, for example, where the patient is confused and does not have a resident carer.

Table 6: Summary of RCTs evaluating the effectiveness of drug therapies in reducing fracture incidence in postmenopausal women

*NNT for the prevention of a single fracture are calculated, rather than quoted from the source references. The NNT is not only a function of the efficacy of the therapy but is significantly determined by the baseline risk as reflected in the incidence of fractures (vertebral and non-vertebral) observed in the placebo groups of these trials. As indicated in the Table, the placebo rate varies significantly between trials and direct comparison of the NNTs as a guide to relative efficacy is not appropriate.

6.2 Postmenopausal women with multiple vertebral fractures

Women with multiple vertebral fractures are at greatest risk of future fracture.³⁵ That future risk rises exponentially in association with increased number of prevalent fractures. Only a minority of vertebral fracture patients are ever diagnosed or treated for osteoporosis. Only 50% of vertebral fractures are associated with pain but the morbidity with regard to future potential fracture risk and mortality are the same whether a vertebral fracture has been associated with symptoms or not. Inconsistency in radiology reporting is a factor that contributes to low rates of identification and subsequent poor treatment of patients with osteoporotic vertebral fracture.¹²³

The efficacy of three years' risedronate 5 mg daily (with calcium ± vitamin D) was associated with significant reduction in morphometric vertebral fracture incidence among women with at least two vertebral fractures in two large clinical trials.^117,118 The larger study¹¹⁷ also demonstrated reduction in non-vertebral fracture risk.

The alternative regimen of risedronate 35 mg once weekly is as effective as the 5 mg daily regimen in improving bone density¹²⁴ and by implication has similar efficacy in reducing fracture risk.

Intermittent cyclical etidronate (400 mg cyclical etidronate for 14 days followed by 76 days of 500 mg of calcium per day), is effective in the secondary prevention of vertebral fractures in women with at least two vertebral fractures and radiological osteopaenia but without targeting treatment using DXA.¹¹⁹

6.2.1 options for therapy

Provided underlying destructive disease such as tumour (including myeloma), or infection, has been excluded as the cause of multiple vertebral fractures, there is evidence^117,118,119 that targeting women with at least two vertebral fractures with bisphosphonates is associated with a significant reduction in vertebral fracture risk. Evidence level 1++

Postmenopausal women who have suffered at least two vertebral fractures should be considered for one of the following options:

To reduce fracture risk at all sites: treatment with oral risedronate (5 mg daily or 35 mg once weekly + calcium ± vitamin D).

Although not tested specifically in this scenario in clinical trials, it is likely that alendronate would have equal efficacy to risedronate.

To reduce vertebral fracture risk: treatment with intermittent cyclical etidronate (400 mg daily for 14 days + 500 mg calcium daily for 76 days, repeating 3 monthly cyclical therapy).

In either of these cases treatment can be initiated without prior assessment by DXA scanning.

The selection of specific drug therapy for an individual patient is at the discretion of the prescribing clinician taking into account the patient's health record, but will also be influenced by formulation, cost, tolerability, and patient choice. Other things being equal, however, risedronate or alendronate should be the preferred option for this category of patient as they reduce fracture risk at all sites.

6.3 Postmenopausal women with osteoporosis determined by axial dxa and with a history of at least one vertebral fracture

6.3.1 bisphosphonates

Alendronate 10 mg daily (with 500 mg calcium + 250 IU vitamin D per day) is effective in the secondary prevention of osteoporotic vertebral fractures when targeted at women with at least one vertebral fracture and with BMD at the femoral neck that is lower than a T-score of –1.6.³⁶ Like other studies cited below, this trial used reference data from the independent National Health and Nutrition Survey (NHANES).¹²⁵ In this trial the incidence of morphometric vertebral fractures was significantly reduced. Although overall non-vertebral fracture risk was not reduced, specific fracture subtypes such as hip fracture were significantly reduced. Evidence level 1++

The alternative regimen of alendronate 70 mg once weekly is as effective as the 10 mg daily regimen in increasing BMD (although there are currently no fracture data).¹²⁶ Evidence level 1++

6.3.2 hormone replacement therapy

Only one small RCT of the efficacy of transdermal oestrogen in the secondary prevention of fractures has been identified.¹⁰⁶ Patients were recruited on the basis of low bone density and the number of incident vertebral fractures was used to define efficacy, rather than the number of women suffering a new vertebral fracture as used in most other trials. Evidence level 1+

6.3.3 raloxifene

60 mg raloxifene in association with 500 mg calcium and between 400 and 600 IU vitamin D per day has been shown to be effective in reducing the incidence of morphometric vertebral fractures in women with low BMD,¹¹⁰ and either one moderate or two mild vertebral fractures. Raloxifene has not been shown to reduce the incidence of non-vertebral fractures. Evidence level 1++

6.3.4 calcitonin

200 IU calcitonin intranasally in association with 1000 mg calcium plus 400 IU vitamin D per day has been shown to reduce the incidence of vertebral fractures.¹²¹ Unusually, a dose response relationship was not seen: neither 100 IU per day nor 400 IU per day were associated with a change in the incidence of morphometric vertebral fractures. Calcitonin has not been shown to have efficacy in reducing the incidence of non-vertebral fractures in well conducted RCTs. Evidence level 1++

6.3.5 options for therapy

Postmenopausal women who have suffered at least one vertebral fracture and who have had osteoporosis confirmed by DXA scanning should be considered for one of the following options:

To reduce fracture risk at all sites: treatment with oral alendronate (10 mg daily or 70 mg once weekly + calcium ± vitamin D).

Although not tested specifically in this scenario in clinical trials, it is likely that risedronate would have equal efficacy to alendronate

To reduce vertebral fracture risk: treatment with oral raloxifene (60 mg daily + calcium ± vitamin D).

To reduce vertebral fracture risk: treatment with intranasal calcitonin (200 IU daily + calcium ± vitamin D).

Use of HRT can be considered as a treatment option for osteoporosis to reduce vertebral fracture risk, but the risks and benefits should be discussed with each individual woman before starting treatment (see section 5).

The selection of specific drug therapy for an individual patient is at the discretion of the prescribing clinician taking into account the patient's health record, but will also be influenced by formulation, cost, tolerability, and patient choice. Other things being equal, however, alendronate or risedronate should be the preferred option for this category of patient as it reduces fracture risk at all sites. Calcitonin has been shown to have some effectiveness as an analgesic for acute pain (section 6.11.1) and may be considered for patients suffering pain from vertebral fractures.

6.4 Postmenopausal women with osteoporosis determined by axial dxa, with or without previous non-vertebral fracture

The importance of a non-vertebral fracture is that it at least doubles the potential fracture risk at that or other skeletal sites. The evidence base for treating this group of patients derives from studies that have targeted therapy on the basis of low BMD and have shown efficacy in reducing the risk of subsequent fracture. Treatment should be given on the basis of low BMD. If a patient has a non-vertebral fracture they are at greater risk of future fracture.⁷⁸ It follows that there should be greater benefits from treating this high risk group (patients with low BMD and non-vertebral fracture) due to the higher number of fractures prevented.

6.4.1 bisphosphonates

Alendronate 10 mg per day (with 500 mg calcium + 250 IU vitamin D) has been studied in a large clinical trial³⁷ comprising over 4,400 patients where treatment was targeted on the basis of low BMD alone without previous fracture. During the trial, international standardisation of the hip BMD reference data occurred and this impacted significantly on the actual severity of reduction of BMD in the patients recruited to this trial. Subsequent analyses based on the NHANES reference database led to re-evaluation of the entire clinical trial group with re-categorisation of their bone densities by T-scores. The incidence of vertebral fractures is reduced in women treated at femoral neck T-score =<1.6. If, however, treatment is targeted at those patients with femoral neck T-score =<2.5, a reduction in incidence of vertebral and non-vertebral (including hip) fracture is seen. Evidence level 1++

Similar efficacy in non-vertebral fracture risk reduction has been reported in another clinical trial of alendronate¹²² in which 1900 women were treated with alendronate 10mg daily with 500mg calcium per day on the basis that their lumbar spine BMD T-score was =< -2. Evidence level 1++

Risedronate 5 mg per day (with 1,000 mg calcium plus up to 500 IU vitamin D per day) has been shown to be effective when targeted at elderly women with femoral neck T-scores of =<-4 (equates with NHANES T-scores of around –2.7 to –2.9) or with slightly higher bone density and other skeletal risk factors such as increased hip axis length.¹¹⁶ This study uniquely addressed the primary end point of hip fracture incidence and demonstrated efficacy in reducing non-vertebral fracture risk, and specifically hip fracture risk. The incidence of hip fractures was low at 3.2% in the placebo group over three years. Evidence level 1++

6.4.2 RALOXIFENE

Raloxifene 60 mg per day (with 500 mg calcium and up to 600 IU of vitamin D per day) has been studied in 4500 women who were treated on the basis of femoral neck T-score =<-2.5.¹¹⁰ The relative risk of morphometric vertebral fractures in association with raloxifene was reduced. There was no significant reduction in the incidence of non-vertebral fractures. Evidence level 1++

6.4.3 options for therapy

Postmenopausal women who have had low BMD confirmed by DXA scanning should be considered for one of the following options:

To reduce fracture risk at all sites: treatment with either oral alendronate (10 mg daily or 70 mg once weekly + calcium ± vitamin D) or oral risedronate (5 mg daily or 35 mg once weekly + calcium ± vitamin D).

To reduce vertebral fracture risk: treatment with oral raloxifene (60 mg per day + calcium ± vitamin D).

The selection of specific drug therapy for an individual patient is at the discretion of the prescribing clinician taking into account the patient's health record, but will also be influenced by formulation, cost, tolerability, and patient choice. Other things being equal, however, alendronate or risedronate should be the preferred options for this category of patient as they reduce fracture risk at all sites.

6.5 Frail, elderly (aged 80+years) women with a diagnosis of osteoporosis, with or without previous osteoporotic fractures

Studies assessing the efficacy of the bisphosphonates etidronate, alendronate and risedronate and of the SERM raloxifene, have generally recruited women up to 80-85 years of age and one study¹¹⁶ included a study arm that recruited women of any age over 80 years. Age per se should not therefore preclude treatment with antiresorptive therapies. The same criteria for targeting treatment apply to the elderly. Axial DXA would be a prerequisite to establish that the BMD is sufficiently low before starting treatment with bisphosphonates, unless the patient has suffered multiple vertebral fractures.

For such patients, and others who are perceived to be “frail” such as those who are housebound, another common treatable risk factor for hip fracture is vitamin D deficiency. This is a consequence of lack of exposure to ultraviolet light. There is evidence that treating these frail patients with calcium and vitamin D can reduce the incidence of hip fractures by 35% and non-vertebral fractures by 26% without the need to either measure vitamin D or target this therapy using DXA scanning.¹²⁷

6.5.1 options for therapy

To reduce fracture risk at all sites elderly women who have suffered multiple vertebral fractures or who have had osteoporosis confirmed by DXA scanning, should be considered for treatment with either oral risedronate (5mg daily or 35 mg once weekly+ calcium ± vitamin D) or oral alendronate (10 mg daily or 70 mg once weekly + calcium ± vitamin D).

It is clear that targeting bisphosphonate therapy (risedronate) to patients whose fracture risk is defined on the basis of risk factors for falling, will not reduce fracture risk.¹¹⁶ Bisphosphonates strengthen bone, they do not prevent falls.

Falls risk reduction strategies should be employed to reduce fracture risk for elderly women who have suffered any form of previous fracture.

To reduce hip fracture risk, frail elderly women who are housebound should receive oral calcium 1,000-1,200 mg daily + 800 IU vitamin D.

It is not necessary to measure vitamin D levels before commencing treatment.

6.6 Men with a diagnosis of osteoporosis determined by axial dxa with or without previous osteoporotic fracture

Although osteoporotic fractures are less common in men than women, men experience greater fracture-associated morbidity and mortality.^{128,129,130,131} Men are also at increased risk of osteoporosis from secondary causes.^42,43

In women there is a clear relationship between BMD and fracture risk. Further studies are required to establish whether this is also true for men. It is therefore not certain whether women's and men's bones will fracture at similar BMD levels. There is some evidence that men and women may fracture at similar gender-specific T-scores of BMD,¹³² supporting the WHO criteria as being applicable to men using the average young adult male BMD at peak bone mass as the reference for comparison. There are few studies in males with osteoporosis and more studies are required to establish the efficacy of antiresorptive therapies in achieving primary and secondary prevention of osteoporotic fractures

There is one well conducted RCT¹³³ in men with low BMD and a history of one or more vertebral fractures or one non-vertebral osteoporotic fracture. Alendronate (10 mg daily + 500 mg calcium ± 400 IU vitamin D) was shown to significantly increase lumbar spine and femoral neck BMD and reduce morphometric vertebral fracture risk and height loss. Evidence level 1++

The evidence relating to calcium + vitamin D supplementation in men is inconsistent. The efficacy of calcium + vitamin D in the absence of concurrent antiresorptive therapy in osteoporotic men is not known.

The efficacy of Calcitriol in reducing vertebral fracture risk or changing hip or spinal BMD in men has not been established.

The evidence base for the use of androgens is small and dominated by poor quality studies. Testosterone in hypogonadal men may increase spinal BMD but there are no trial data relating to fracture outcomes. There are no convincing data of efficacy in changing BMD in eugonadal men. No studies have targeted testosterone on the basis of low BMD.

6.6.1 options for therapy

To reduce fracture risk at all sites, men with low BMD and/or a history of one or more vertebral fractures or one non-vertebral osteoporotic fracture should be treated with oral alendronate (10mg + 500mg calcium ± 400 IU vitamin D daily).

70mg weekly oral alendronate has been shown to result in equivalent BMD changes to oral 10mg once daily in women.¹²⁶ It is not currently licensed for use in men. It is likely that both formulations share the same efficacy with regard to fracture risk in men.

6.7 Corticoidsteroid induced osteoporosis

The guideline development group decided not to proceed with a detailed literature review of corticosteroid induced osteoporosis in view of the publication of an evidence based guideline published by the UK Bone and Tooth Society in association with the National Osteoporosis Society and the Royal College of Physicians of London in December 2002.¹³⁴ The clinical practice algorithm from that guideline is reproduced in Annex 5.

6.8 Combination of treatments

In clinical trials bisphosphonates (alendronate, etidronate, or risedronate), raloxifene and calcitonin have usually been assessed in conjunction with calcium +/- vitamin D. Doses of calcium have varied from 500 to 1,000 mg and vitamin D from 6.25 to 20 mcg (250 to 800 IU) per day. Where calcium intake is suboptimal (see section 4.3.2), daily doses of up to 1000mg calcium carbonate plus 20 mcg (800 IU) vitamin D are appropriate for use in association with these drugs (in the absence of conditions associated with hypercalcaemia).

Several clinical trials have reported that the addition of bisphosphonate to HRT ^135,136,137 or of bisphosphonate to raloxifene confers additional benefit regarding BMD compared with monotherapy. Further studies are required to elucidate whether such combinations achieve greater reductions in fracture incidence. Until data are available, combinations of HRT or raloxifene with bisphosphonates are not recommended.

6.9 Duration of treatment

After initiating therapy on the basis of assessment of fracture risk defined using fracture history, usually together with axial DXA measurement in the context of the patient's age, it is likely that treatment would be required on a lifelong basis. Fracture efficacy data, however, exist only for between 1-4 years, the duration of the doubleblinded randomised placebo-controlled trials. Safety data do, however, exist for several years thereafter for bisphosphonates and suggest that there is unlikely to be any cumulative disadvantage to the skeleton even though they are likely to be retained in the skeleton for years. Few data exist regarding BMD or fracture risk after cessation of bisphosphonates, although one study¹³⁸ reported increases in markers of bone turnover, without changes in BMD two years after stopping alendronate and this may indicate reactivation of processes that may ultimately result in bone loss.

Identification of optimal longer term treatment patterns should be the subject of future research. Until such time as this has been clarified, it should be assumed that long term management is required. Further data on the benefit of intermittent regimens is awaited.

6.10 Monitoring treatment effect with DXA

There is evidence of a relationship between therapy associated increases in BMD and the extent of fracture risk reduction.^139,140

The efficacy of “monitoring” BMD responses by DXA has not been evaluated by clinical trial, although it has been a key end point in most clinical trials relating to osteoporosis management. Application of repeat DXA to individual patients requires consideration of the following:

• The lumbar spine trabecular bony site is the preferred site for follow up DXA.
• Increases of BMD of at least 3-4% are required as “the least significant difference” that is
  likely to exceed the error of the measurement.
• Follow-up should normally be undertaken only after at least two years of therapy. (Note that
  more frequent follow up is required to monitor the effects of bone-toxic drugs such as
  chemotherapy.)

There are circumstances in which follow up DXA can be helpful in managing individual patients. Until such time as this issue has been resolved by clinical trial, local policies relating to DXA follow up should be devised by agreement with the DXA service provider, primarily to ensure that allocation of scans for monitoring is feasible within the existing DXA service arrangements.

6.11 Treatment of pain

From the patient's perspective the pain of osteoporotic fracture in both the acute and chronic phase is often their most immediate concern. Osteoporotic vertebral fractures may be painless but progressive vertebral failure may give rise to worsening dorsal kyphosis with possible subsequent chronic pain and debility. The acute pain of fracture can vary widely and chronic pain is associated with significant physical dysfunction and decreased quality of life.

Treatment must involve an appreciation of both the need to prevent further progression of the osteoporosis and an assessment of analgesic needs. The use of the WHO Analgesic Ladder may be of value in logically achieving adequate analgesia and its use is validated outwith cancer care.¹⁴¹ Outwith conventional analgesic agents, the use of bisphosphonates has not been shown to alleviate pain.

6.11.1 Acute pain

General measures in the acute fracture phase should be undertaken as appropriate including rest, ice, compression and elevation. Conventional analgesics should be used regularly rather than on demand. Other specific measures such as splintage, reduction and plaster immobilisation or fracture fixation need to be utilised as appropriate to the fracture.

As acute pain can be debilitating, admission to hospital may be necessary, both for analgesia and mobilisation. Achieving adequate analgesia may be difficult and the involvement of a Pain Service may be of value. Opiates may be necessary. Care in their use in the elderly should not prevent their use.

Calcitonin, preferably intranasally rather than by injection, has been shown to be of value in difficult cases with unremitting pain due to acute vertebral fracture¹⁴² but it does not hold a product licence as an analgesic agent in the United Kingdom. It is licensed for the treatment of postmenopausal osteoporosis.

Successful analgesia in the acute phase should allow early mobilisation.

6.11.2 chronic pain

Chronic pain should be identified and treated along accepted guidelines and analgesic scales. Recording of pain levels using, for example, a Visual Analogue Scale is of value in assessing achieved analgesia. Adequate control of chronic pain is often difficult and unsatisfactory. Analgesic agents, NSAIDs, and physiotherapy with physical activity programmes are all of value.^90,143 Calcitonin has again been shown to be of value in control and treatment of chronic back pain.¹⁴²

Parathyroid hormone has been used therapeutically in an 18-month RCT which has shown benefit in reducing back pain in postmenopausal women.¹⁴⁴ It is not yet licensed for use in the United Kingdom and the evidence for its efficacy has not therefore been examined in this guideline.

The use of non-pharmacological measures such as acupuncture or Transcutaneous Electrical Nerve Stimulation (TENS) have been shown to be of value and should be considered. Back strengthening exercises are also valuable.

The psychological care of the osteoporotic patient with pain is important as depression and lack of sleep are commonplace. Clinical psychological intervention may be of benefit along with adjunctive therapy of antidepressants.

6.11.3 new developments

At present there is some suggestion that emerging technologies such as vertebroplasty and kyphoplasty¹⁴⁵ may offer a therapeutic role in acute vertebral fracture with significant pain reduction. Their role still needs to be defined through randomised controlled trials and until such evidence becomes available they cannot be recommended.