Hypoventilation is defined as ventilation below that which is needed to maintain adequate gas exchange. It can be a feature in lung diseases as diverse as chronic bronchitis and pulmonary fibrosis but determining whether it is present of not is often complicated by defects in gas exchange. When desaturation occurs during a CPET (i.e. a significant decrease in SaO2 below 95%) this is a strong indication that the primary exercise limitation is pulmonary in nature and from that point the maximum minute ventilation and the Ve-VCO2 slope can show whether the limitation is ventilatory or instead due to a gas exchange defect. But in this circumstance what what does it mean when both the maximum minute ventilation and Ve-VCO2 slope are normal?
Recently a CPET came across my desk for an individual with chronic SOB. The individual recently had a full panel of pulmonary function tests:
| Observed: | %Predicted: | |
| FVC (L): | 1.73 | 62% |
| FEV1 (L): | 1.39 | 66% |
| FEV1/FVC: | 80 | 106% |
| TLC (L): | 2.99 | 62% |
| DLCO (ml/min/mmHg): | 14.66 | 84% |
| DL/VA: | 5.45 | 124% |
| MIP (cm H2O): | 11.5 | 18% |
| MEP(cm H2O): | 21.3 | 24% |
The reduced TLC showed a mild restrictive defect. At the same time the relatively normal DLCO indicates that the restriction is probably not due to interstitial lung disease and more likely either a chest wall or a neuromuscular disorder, both of which can prevent the thorax from expanding completely but where the lung tissue remains normal. The reduced MIP and MEP tends to suggest that a neuromuscular disorder is the more likely of the two.
I take this with a grain of salt however, and that is because this individual never had pulmonary function tests before and for this reason there is no way to know what their baseline DLCO was prior to the restriction. At the same time far too many individuals perform the MIP/MEP test poorly and low results are not definitive, and in this case in particular the results are so low the individual should have been in the ER, not the PFT Lab.
The CPET results were somewhat complicated, in that a close inspection showed both pulmonary and cardiovascular limitations.
| Rest: | %Predicted: | AT: | %Predicted: | Maximum: | %Predicted: | |
| VO2 (LPM): | 0.26 | 24% | 0.49 | 44% | 0.54 | 48% |
| VCO2 (LPM): | 0.21 | 0.43 | 0.65 | |||
| RER: | 0.84 | 0.87 | 1.27 | |||
| SpO2 (%) | 97% | 94% | 92% | |||
| PetCO2 (mm Hg): | 41.2 | 46.5 | 53.4 | |||
| Ve/VCO2: | 49 | 39 | 30 | |||
| Ve (LPM): | 9.7 | 16% | 15.7 | 28% | 18.2 | 31% |
| Vt (L): | 0.37 | 0.46 | 0.48 | |||
| RR (f): | 27 | 36 | 39 | |||
| Heart Rate (BPM: | 72 | 48% | 86 | 58% | 89 | 60% |
| BP (mm Hg) | 156/78 | 162/82 | 170/84 | |||
| O2 Pulse (ml/beat) | 3.6 | 48% | 6.0 | 79% | 6.2 | 80% |
In addition, the chronotropic index was 0.68 (normal 0.80 to 1.30), the Ve-VCO2 slope from rest to AT was 14.5 and the Ve-VCO2 slope from rest to peak exercise was 16.4. Although the maximum minute ventilation and maximum heart rate are well below their ULN of 85% of predicted, the RER of 1.27 at peak exercise indicates that there was an adequate exercise test effort.
That there is a cardiovascular limitation is indicated first by the reduced VO2 at anaerobic threshold. The LLN for this individual based on their age and gender was 54% so the observed value of 44% is well below normal. This shows that there was an oxygen delivery problem and this is usually related to a decrease in cardiac output. In addition, the chronotropic index is also reduced and since the individual is taking metoprolol, a beta blocker, this is not a major surprise. Together these factors could indicate that the individual’s primary exercise limitation was cardiovascular but the reduced SaO2 at maximum exercise indicates that the primary limitation has to be pulmonary.
But even though the individual has a restrictive ventilatory defect the maximum minute ventilation is well below the ULN of 85% so there was a large ventilatory reserve at peak exercise. Moreover, the ULN for the Ve-VCO2 slope from rest to AT is 34 and for the Ve-VCO2 slope from rest to peak exercise it is 40. The respective Ve-VCO2 slope values of 14.5 and 16.4 are well below this ULN and this (as well as the normal DLCO) is a strong indication that gas exchange was normal. So why is did this individual desaturate?
There are two strong clues. First, PetCO2 was elevated, both somewhat at rest and most distinctly at peak exercise. A normal resting PetCO2 is usually between 30 and 35 and at peak exercise between 35 and 45.
Note: A maximum PetCO2 above 40 is usually seen only in subjects with above average fitness and usually only at higher levels of exercise. PetCO2 can become elevated because an increased cardiac output decreases pulmonary capillary transit time and end-expiratory alveolar air therefore begins to reflect the venous PCO2. In these instances however, even though the PetCO2 is elevated, arterial PCO2 is usually normal and less than PetCO2.
In this case a resting PetCO2 of 41 is a touch high but not necessarily abnormal, but the pattern during exercise was. A normal PetCO2 pattern is a peak near AT (usually slightly after) and then a decline thereafter. For this individual however, PetCO2 actually increased throughout exercise. In addition the maximum PetCO2 of 53.4 is well above any normal.
Second, the Ve-VCO2 slopes, although technically within normal limits, are actually much too low. A normal Ve-VCO2 slope from rest to AT should have been around 29 for this individual.
Note: Ve/VCO2 is the relationship between minute ventilation and CO2 production at any one instant whereas the Ve-VCO2 slope is the relationship between the change in minute ventilation and the change in CO2 production over time. Traditionally the Ve/VCO2 at AT has been used to assess ventilatory efficiency and in this case the Ve/VCO2 of 39 at AT is elevated (ULN is 35). However, it has been pointed out that AT is where the nadir of Ve/VO2 occurs, not Ve/VCO2, and that because the AT is in this sense an arbitrary point at which to assess Ve/VCO2 the lowest observed Ve/VCO2 is a better indicator of of ventilatory efficiency. The lowest observed Ve/VCO2 for this individual is 30 and WNL.
In addition, it has been shown that the Ve/VCO2 at AT is a function of the Ve-VCO2 slope from rest to AT and it’s offset (i.e. , what Ve would be if VCO2 was zero). The offset of the Ve-VCO2 is different between individuals and disease states for reasons that remain unclear. Because of all this, and given that the Ve-VCO2 slope is generated from a relatively large number of data points, it’s my opinion that the Ve-VCO2 slope from rest to AT is a more reliable indicator of ventilatory efficiency than Ve/VCO2.
The Ve-VCO2 slopes of 14.5 and 16.4 are abnormally low and indicate little change in minute ventilation when compared to the change in VCO2 and given the large ventilatory reserves at peak exercise there is no apparent reason for this.
Finally, in addition there is one minor additional point and that is that there was only a small increase in tidal volume during exercise. Tidal volume normally doubles or triples during exercise and a low increase in tidal volume along with all the other factors:
- Mild restriction with a relatively preserved DLCO
- Reduced MIP and MEP
- Elevated PetCO2
- Reduced Ve-VCO2 slope
- Elevated ventilatory reserve
is a strong suggestion that the individual in question is hypoventilating due to respiratory muscle weakness. Their exercise limitation may be exacerbated by their chronotropic incompetence, but the primary limitation is still hypoventilation. There are a variety of possible causes for this weakness but the individual was referred to my lab by a physician outside the hospital network and for this reason there is no history available for review (and unfortunately any follow up will probably also occur outside the hospital network so I may never find out what the cause was).
It is moderately unusual to find hypoventilation by itself, without any gas exchange defect, as a primary exercise limitation. Hypoventilation can be a factor in both COPD and interstitial diseases, but in these cases there is usually a mix of both ventilatory and gas exchange limitations. The primary difference between those cases and the present one, and which showed a lack of underlying lung disease, was the Ve-VCO2 slope and PetCO2. Given that there was no ventilatory or gas exchange limitations, the only reasonable cause can be hypoventilation.
PFT Blog by Richard Johnston is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
This is a quite interesting case and I generally agree with your main conclusion that this patient presented with alveolar hypoventilation. There are some points, however, that may merit further interpretative considerations:
1) Cardiovascular limitation is suggested on the grounds of a low LT and beta-blocker induced chronotropic incompetence (CI). A low LT would be expected in a chronic dyspneic patient who is likely sedentary. Nevertheless, it would be interesting to know which plot(s) were used for AT detection. I usually find challenging to reliably identify the LT by the ventilatory methods in hypoventilators (this sounds quite axiomatic….). Thus, I guess you looked primarily at the V-slope plot (VCO2 x VO2). We should not forget, however, that VCO2 is the product of VE and mixed venous CO2 content (CvCO2). If the patient chronic hypoventilate, the CO2 reservoirs are already populated and any relative small increase in PaCO2 induced by exercise would readily impact on CvCO2. This, in turn, would tend to influence substantially on the rate of increase in VCO2 relative to VO2. In this context, pseudo-thresholds might arise (I once measured lactate levels in some few hypoventilators and the lactate breakpoint rarely coincided with the “gas exchange thresholds”) and RER values becomes inordinately high for relatively-short tests. Assuming a normal VO2/WR slope, your patient should have stopped in the unloaded-25W range (unless he or she is a really small person)- which makes more plausible the view that a high VCO2 (and RER) were consequence of severe hypercapnia rather than the buffer of a massive amount of lactate.
2) On the same topic of cardiovascular limitation, we (i.e., those reading a great volume of CPETs, usually with a teaching/academic position) need to start recognizing that we simply do not know the actual physiological meaning of the so-called CI. The founders of the modern technology that eventually led to CPET did not face a large number of subjects with unclear causes of exercise limitation who had impaired chronotropic responses (either pharmacologically- on non pharmacologically-induced). In practice, we now see many patients in whom nothing else apart from “deconditioning” and a blunted HR response to exercise can be pointed out, These are usually obese and detrained guys with diabetes/metabolic syndrome, some with “diastolic dysfunction (whatever that means…). How can we confidently state that CI was mechanistically linked to impaired O2 delivery and early exercise termination? For instance, we simply don’t know if stroke volume was able to compensate for a low HR and/or the fractional O2 extraction increased sufficiently in individual patients. Of course, there are those few extreme cases (e.g., a grossly unresponsive pacemaker in a patient with claudication and light headedness at exercise termination) but they are rather exception than rule.
In summary, items #1 and 2# would make me quite cautious in endorsing a role for CV factors limiting this patient.
3) PETCO2 is a very tricky measurement which is influenced by ventilation, breathing pattern and gas exchange. In most circumstances we are at loose with a low PETCO2 as it might reflect areas of increased V/Q and/or a low PaCO2 “set-point”. For these reasons, dead space (VD)/tidal volume (VT) estimates using PETCO2 are prone to misinterpretation in patients with V/Q abnormalities. In other words, it is safe to state that almost everything conspires against PETCO2 in patients with lung (and, sometimes, heart) disease. On the contrary, I have a great respect for a high PETCO2 which further increases during exercise. However, I always look first to patient’s VT: a too high PETCO2 may merely reflect better sampling of the alveolar air due to a deeper breathing pattern. In your case, however, it was rather the opposite: a superficial and fast breathing pattern would tend to reduce PETCO2 (I always calculate the f/VT ratio – quite reliable with limits of normal established (Neder JA, et al. The pattern and timing of breathing during incremental exercise: a normative study. Eur Respir J. 2003 Mar;21(3):530-8.). Thus, this patient had a too superficial breathing pattern (f/VT ratio near 70; gender-corrected ULN of 25 in males and 35 in females) to overcome his/her VD leading to alveolar hypoventilation.
4) As you stated, VE/VCO2 ratio at its lowest point (nadir) is close to the algebraic summation of VE-VCO2 slope and its intercept (its starting point when VCO2 is extrapolated to zero (“intercept”)). Your patient had a VE/VCO2 nadir of 30 but a VE-VCO2 slope of 16; thus, the VE-VCO2 intercept was amazingly high (around 14 L/min; ULN= 8 L/min). A high intercept might be consequence of a high PaCO2 set point (hypercapnia) and/or a high wasted fraction of the breath (VD/VT ratio). In the present case, high PaCO2 plus a low VT (leading to high VD/VT) may have acted together to increase VE-VCO2 intercept (and its closer physiological surrogate, VE/VCO2 at rest (49)). The effects of the intercept on the nadir, however, were offset by a shallow slope. In other words, severe respiratory muscle weakness likely precluded a normal increase in ventilation as a function of the metabolic demand (thus, a low slope) but its consequences (hypercapnia and a low VT leading to a high VD/VT) led to an upward displacement of the VE-VCO2 relationship. The bottom line is that despite the (potential) absence of an intrinsic lung disease, a high VD/VT indicates that gas exchange efficiency was not normal.
5) Although not mentioned, I believe you also assumed that the most likely reason underling patient’s mild hypoxemia was increased alveolar PCO2.
In summary, items #3, #4 and #5 boils down to the point that contrary to your statement that “there is one minor additional point and that is that there was only a small increase in tidal volume during exercise”, I would say that a low VT due to respiratory muscle weakness was a rather crucial event which linked disrupted mechanics and its sensorial consequences (dyspnea) to impaired alveolar ventilation and poor overall gas exchange efficiency (high VD/VT).
Dr. Neder –
You bring up a lot of points and I will try to answer at least some of them.
AT was determined by the nadir in PETO2 and Ve/VO2. I find the V-slope technique to be difficult to use because it is hard to know which points to include in the upper and lower slopes and although this appears to be something that a computer algorithm should do well, the software that comes with our CPET system does a poor job of finding AT. I agree that the Lactic Acid Threshold is considered the “gold standard” for AT but at the same time have to point out that the blood samples are taken only at intervals and that AT still has to be interpolated. The patient reached 58 watts.
My experience is that the chronotropic index tends to match the fitness of the individual involved, unless obfuscated by beta blockers or other meds, quite well. High chronotropic indexes in the unfit and low in the well conditioned. I will agree however, that when you throw a gas exchange defect into the mix it may be more questionable. In this case despite the desaturation the normal DLCO and low Ve-VCO2 slope indicates that gas exchange is most likely normal. I consider that any cardiovascular limitation this patient had was secondary and although it may contribute to dyspnea, was not a cause of the exercise limitation.
We did not measure PaCO2 so you can only speculate what it and Vd/Vt might be based on the PetCO2. I would also point out that the original Vd/Vt equation is (PACO2 – PECO2)/PACO2 and although PaCO2 is all too routinely used in place of PACO2 it is not PACO2. Vd/Vt calculated with PaCO2 is not Vd/Vt and although it may track similarly in reality it is something else.
Too often I see patients who have an inefficient ventilatory response to exercise (low Vt, high f) for no apparent reason and who also have a normal max VO2 and normal Ve-VCO2 slope. This is why I am cautious about placing to much emphasis on changes in Vt.
Regards, Richard
Fascinating & educational discussion on this case, thanks to by you both. This is a really useful site.
A much more simple comment: One wonders whether in this case, had a direct MVV maneuver been performed, the terminal breathing reserve would have been low. It appears your maximum predicted Ve is approximating FEV1 x ~42. I would bet a direct MVV would yield a lower value in this patient.