Variation in patients’ experiences with pain after similar types of surgical operations is well established (1,2). Possible factors that may influence postoperative pain intensity are preoperative pain intensity (3,4), age (5), sex (4), personality characteristics (4), and education or information about the surgery (6). However, the role of these factors and the neurophysiologic mechanisms by which they alter nociception remain elusive.
Postoperative pain involves not only peripheral mechanisms, most notably the sensitization of nociceptors due to inflammation, but also secondary central mechanisms, including hyperexcitability of nociceptive neurons (i.e., central sensitization) (7,8). These mechanisms play a major role in postoperative pain, including spontaneous pain and allodynia (i.e., pain due to a stimulus, which normally does not provoke pain) and/or hyperalgesia (i.e., an increased response to a stimulus that is normally painful) (7). In particular, peripheral sensitization would explain the hyperalgesia observed at the incision site (primary hyperalgesia), whereas central sensitization would provide a major mechanism of secondary hyperalgesia at distant noninflammatory sites (9–11).
Results from experimental studies in both animals and humans have suggested that heat and punctuate mechanical hyperalgesia are clinical correlates of the peripheral sensitization of nociceptors (mediated by A-delta and C fibers) in the inflamed primary hyperalgesia area (10,12). In contrast, brush-evoked pain (mechanical dynamic allodynia) observed in the noninflamed area of secondary hyperalgesia would best reflect the abnormal excitability of central nociceptive processing (i.e., central sensitization), since it is induced by the activation of large tactile A-beta fibers (10,13). Detailed sensory examinations thus have the potential to identify underlying mechanisms of postoperative pain processing.
Quantitative sensory testing (QST), based on the measurements of detection and pain thresholds, can readily identify and quantify hyperalgesia. These psychophysical tests have been used in the perioperative period, and have proven valuable in assessing primary and secondary hyperalgesia (14–22).
The goal of the present study was to analyze the influence of preoperative pain and hyperalgesia (primary and secondary) on early and chronic postoperative pain, and to identify putative predictive factors of the severity of postoperative pain. We performed a prospective study with QST before and after total knee arthroplasty, combined with clinical evaluation of acute and chronic pain. Total knee arthroplasty was chosen because it is often associated with intense preoperative inflammatory pain that can develop into chronic pain (23).
With approval of the local ethics committee (Comité de Protection des Personnes, Boulogne Billancourt) and written informed consent, we recruited 20 consecutive patients scheduled for total knee arthroplasty. Inclusion criteria were total knee arthroplasty indicated because of knee arthrosis and surgery performed under general anesthesia. The exclusion criteria were previous surgery or trauma of the knee, preoperative use of opioids, or mental disorders preventing an accurate understanding of the tests.
Anesthesia, Surgery, and Postoperative Pain Relief
All patients were given hydoxyzine 100 mg before surgery. Surgery was performed under balanced general anesthesia combining propofol, sufentanil, a muscle relaxant, and sevoflurane. The same surgeon performed each operation. Postoperative pain was controlled by IV morphine patient-controlled analgesia (PCA) in combination with IV acetaminophen (1 g every 6 h; Perfalgan® UPSA-BMS laboratory, Rueil Malmaison, France) and nefopam, a non-opioid analgesic drug (20 mg every 4 h; Acupan® Biocodex Laboratory, Paris, France). IV treatment was discontinued 48 h after surgery. No patient was given a peripheral nerve block or nonsteroidal anti-inflammatory drugs.
Clinical and quantitative sensory evaluations were performed on the operative knee, the contralateral knee, and the right hand 1 day before surgery (D0), then at 1 day (D1), 4 days (D4), 1 mo (M1), and 4 mo (M4) after surgery.
The circumference of both knees, as well as the surface temperature of the hand and both knees, were monitored at each follow-up visit. Skin temperature at each site was measured at the time of pain testing with Thermopoint device (Protechnique, Quebec). Pain was evaluated at rest and on movement (flexion/extension of the knee during physical therapy) by a 100-mm Visual Analog Scale (VAS) graduated from 0 (no pain) to 100 mm (worst imaginable pain). Pain scores were recorded the day before surgery, then every 4 h for 48 h after surgery, at 4 days after surgery, and at the follow-up visits at M1 and M4, and during all physical therapy sessions. The cumulative doses of morphine consumed via PCA were recorded at 24 and 48 h. Physical therapy was started 24 h after surgery with passive and active mobilization of the operative knee. The active angle of flexion was recorded during hospitalization and at the follow-up visits at M1 and M4. Chronic pain was defined as an operative knee VAS pain score >30 mm at the M4 visit.
Psychophysical testing was performed in a quiet room at a constant temperature (22°C) by the same investigator (V.M.). Measurements included determination of thermal (heat and cold) and punctuate mechanical pain thresholds, and the responses to suprathreshold thermal stimuli.
Mechanical pain thresholds were measured on the operated knee in the middle of the patella; 1 cm lateral to the midline. This location was chosen because in our patients it was always located in the area of maximal inflammation determined clinically (i.e., redness, swelling). Measurements were also taken in the adjacent noninflamed area in the proximal direction, which in our patients was always located at least 5 cm above the top of the incision of the operative knee. We also investigated tactile allodynia (dynamic pain) using a paintbrush (three strokes). We considered tactile allodynia to be present if stroking the skin provoked a distinctly painful sensation.
Thermal pain threshold was measured on the operative knee in the area of maximal inflammation (i.e., in the middle of the patella; 1 cm lateral to the midline).
Control measurements for mechanical and thermal pain threshold were performed at two remote sites: the contralateral knee (stimulation on the patella; 1 cm lateral to the midline) and the palmar aspect of the right hand.
Mechanical Pain Threshold
Pain thresholds for punctuate mechanical stimuli were assessed using calibrated von Frey hairs (Bioseb, Chaville, France). Care was taken to avoid stroking the skin with the hair and to apply only a pressure stimulus. The patients were instructed to close their eyes during the procedure. The von Frey filaments were applied (at least twice) in ascending and descending order of stiffness. The pain threshold was defined as the lowest pressure the patient considered painful. The force required to bend the filaments (0.057–140 g) was converted into log units.
Thermal Pain Threshold
Thermal sensations were assessed with a Somedic thermotest (Somedic AB, Stockholm, Sweden), using the Marstock method (24). Briefly, a contact thermode of Peltier elements measuring 25 × 50 mm2 was applied to the skin. The baseline temperature of the thermode was adjusted to the patient’s skin temperature. Thresholds were measured according to the method of limits described by Fruhstorfer et al. (24). Stimuli of increasing or decreasing intensities were applied. For each stimulus, the subject was instructed to press a button that reversed the thermal stimulation as soon as the stimulation became painful, indicating the pain thresholds. The interval between stimuli was 15 to 20 s for hot stimuli and 20 to 30 s for cold stimuli. The maximum and minimum temperatures were set at 50°C and 4°C. A thermal rate of change of 1°C/s was used. All thresholds were calculated as the average of three successive determinations.
Supraliminal Thermal Stimulation
A series of suprathreshold cold and hot thermal stimuli were applied according to a previously described method (25). Each stimulus lasted for 2 s. The intensity was increased above the pain threshold by 2°C and 4°C for hot stimuli and decreased by 5°C below the pain threshold for cold stimuli. After each stimulus, patients were asked to rate the pain intensity on a VAS. The patients could stop the stimulus at any time. If a VAS score of 80 or more was reported with a lower intensity stimulus, greater stimuli were not applied. In these cases, the same VAS score was assigned to the higher stimulus intensity to allow analysis of the cumulative group data.
Data are expressed as mean ± sem. We used paired t-tests with a Bonferroni adjustment for multiple comparisons, for comparison of (circumference, temperature, pain thresholds) and Wilcoxon’s signed ranked test for comparison of VAS scores. Relationships between two variables were tested using the Spearman correlation test. Repeated-measures ANOVA was used to analyze the stimulus-response curves obtained for suprathreshold mechanical or thermal stimuli. P < 0.05 was considered statistically significant.
Twenty patients (one male, 19 female) were included, aged 69 ± 2 yr old; weighing 74 ± 14 kg. The surgical duration was 115 ± 26 min. Before surgery, the average pain induced by movement was severe [mean VAS score: 61 ± 6 mm (10–80)], whereas average pain at rest was mild [mean VAS score: 16 ± 4 mm (0–50)]. Pain had been present for an average of 3.8 ± 3.3 yr (1–15). The circumference of the operative knee was significantly larger than the contralateral knee (43 ± 1.3 cm vs 41.6 ± 1.2 cm, P = 0.0003) (Fig. 1). However, the skin temperature of the operative knee was similar to that of the contralateral knee (32.2 ± 0.3°C vs 31.9 ± 0.5°C) (Fig. 2).
The pain thresholds to heat, cold, and mechanical punctuate stimuli were similar on the operative knee, the contralateral knee, and the hand (Table 1). No brush allodynia was observed in the inflamed or adjacent noninflamed areas.
The responses to suprathreshold heat stimuli were significantly increased on the operative knee compared with the contralateral knee, indicating preoperative heat hyperalgesia (Fig. 3). In contrast, the responses to suprathreshold cold stimuli were similar on the two knees (results not shown).
Postoperative Clinical Data
The intensity of postoperative pain on D1 and D4 was moderate at rest but severe (VAS score >60) during movement (Fig. 4). In contrast, the mean pain intensity, at rest or during movement, was very mild at M1 and M4 (Fig. 4). At M4, four patients reported moderate pain (VAS score <50) during movement. None of the preoperative or postoperative clinical data were predictive of persistent pain at M4.
Morphine use was 60 ± 18 mg (29–83) over the first 24 h after surgery. Active operative knee flexion was 77 ± 5 degrees (40–95) on D4, 96 ± 2 degrees (70–100) at M1 and 106 ± 2 (85–125) degrees at M4.
Compared with the contralateral knee, both the temperature and circumference of the operative knee were significantly greater from D1 to M1 (Figs. 1 and 2). The maximum temperature increase in the operative knee was observed on D1. The maximum increase in circumference occurred on D4. By M4, both the temperature and circumference had returned to baseline values. Temperature was stable for the hand and the contralateral knee throughout the measurement period.
In comparison with preoperative values, mechanical and cold pain thresholds were significantly decreased on the operative knee in the immediate postoperative period, whereas heat pain threshold was not significantly altered (Table 1). These changes, suggesting static punctuate mechanical and cold allodynia, were observed in the inflammatory area, and not in the adjacent area, and were no longer observed at M1 or M4. There was no brush allodynia in the inflammatory operative area or adjacent area.
On D1 and D4, the responses to suprathreshold heat, but not cold, stimuli were increased on the operative side when compared with the contralateral side (Fig. 5A and B). The heat hyperalgesia response was not significantly different from that measured preoperatively and was not observed at M1 (Fig. 5C) or M4.
No significant changes in the thermal, mechanical pain thresholds, or responses to supraliminal thermal stimuli were observed in the postoperative period in the contralateral knee or hand, confirming that hyperalgesia was strictly limited to the inflammatory area on the operated side.
Relationship Between Pre- and Postoperative Clinical and QST Data
No correlation was detected between preoperative clinical and QST data. In particular, there was no relationship between pain intensity (at rest or during movement) and heat hyperalgesia (i.e., VAS scores) or signs of inflammation measured by knee circumference and local temperature.
Preoperative pain intensity during movement directly correlated with postoperative pain intensity during movement at D1 (Rho = 0.6; P = 0.02) (Fig. 6A), but not at D4. No such correlation was observed for pain at rest.
No correlation was observed between preoperative thermal and mechanical pain thresholds and postoperative clinical data. However, preoperative heat hyperalgesia (VAS scores) correlated with PCA morphine use over the first 24 h (Rho = 0.63; P = 0.01) (Fig. 6B).
In this prospective study, we systematically analyzed thermal and mechanical pain thresholds and the responses to suprathreshold thermal stimuli, applied locally and at distant sites, before and for up to 4 mo after total knee arthroplasty. Preoperatively, all patients reported knee pain that was more intense during movement. Preoperative QST demonstrated primary heat hyperalgesia, but not with secondary hyperalgesia. After surgery, intense pain, mostly during movement, was observed in the early, but not late, postoperative period, and coexisted with increased signs of local inflammation. QST demonstrated concurrent primary hyperalgesia (i.e., local mechanical punctate allodynia, and heat hyperalgesia) without secondary hyperalgesia. These data show that heat hyperalgesia, reflecting primary hyperalgesia, is a predominant component of perioperative pain associated with total knee arthroplasty. The pathophysiologic significance of this finding is supported by our finding that the intensity of heat hyperalgesia was predictive of postoperative morphine consumption.
Before surgery, patients had prolonged pain associated with characteristic inflammation of the operative knee, as indicated by swelling and increased temperature. QST revealed that these inflammatory signs were associated with severe local heat hyperalgesia, but no other significant sensory alteration. In particular, there was no mechanical allodynia at distant sites. It has long been demonstrated, both in experimental and pathologic models of inflammation in animals and humans, that heat hyperalgesia is mostly due to the sensitization of peripheral nociceptors by pronociceptive mediators released locally in inflamed tissue (e.g., prostaglandins, cytokines, bradykinin) (10,12). Thus, in our patients, peripheral mechanisms may have been sufficient to explain preoperative chronic knee pain due to osteoarthritis. However, one cannot formally exclude, on the basis of our study, that nonclinically detectable secondary central processes also contributed.
Heat hyperalgesia was still present in the early postoperative period. Thus, peripheral sensitization may still be a predominant mechanism of postoperative pain. In accordance with this hypothesis, we found that preoperative heat hyperalgesia directly correlated with postoperative morphine consumption. Heat hyperalgesia was associated with postoperative punctuate mechanical allodynia in the inflammatory area, which might reflect an increased peripheral sensitization after surgery. In contrast with other studies concerning other types of surgeries (19,20,26,27), we did not detect postoperative segmental secondary hyperalgesia in our patients. This supports the hypothesis that peripheral mechanisms played the predominant pathophysiologic role in early postoperative pain. This may be specific for major knee surgery, in which postoperative pain is mainly induced by peripheral inflammation. In addition, we did not detect any modification of extrasegmental pain threshold. Previous studies testing secondary hyperalgesia in the extrasegmental area also found no modification (28) or inhibition (18,29).
From a clinical perspective, the present data suggest, in accordance with a previous study (30), that measurement of heat hyperalgesia preoperatively may have some predictive value regarding postoperative pain and, therefore, may also predict perioperative analgesic requirement. We also found a correlation between the intensity of pain induced by movement preoperatively and postoperatively, which may also prove clinically useful. These results are in agreement with those of previous studies on the prognostic value of preoperative pain for immediate postoperative pain intensity, with other types of surgery (3,4).
Despite the severity of both preoperative inflammation and pain, the prevalence of chronic pain was relatively low in our patients. This is in accordance with the incidence of complex regional pain syndrome after total knee arthroplasty reported in a previous study (21% at 1 mo and 13% at 3 mo) (23). In contrast, a significantly higher incidence of chronic pain has been reported after other types of surgery with or without preoperative pain and inflammation (31). Moreover, all our patients had severe prolonged pain before surgery, and only 20% described moderate pain on movement 4 mo after surgery. Thus, preoperative pain, inflammation, or the combination of the two do not predict the development of chronic pain. It has been suggested that the presence of postoperative secondary hyperalgesia may be predictive of the development of chronic pain (11,32). In keeping with this hypothesis, our data tend to indicate that the lack of secondary hyperalgesia is associated with a reduced incidence of postoperative chronic pain. This finding should be verified in future studies in larger groups of patients, since it could help guide perioperative analgesic treatment.
1. Lynch EP, Lazor MA, Gellis JE, Orav J, Goldman L, Marcantonio ER. Patient experience of pain after elective noncardiac surgery. Anesth Analg 1997;85:117–23
2. Bisgaard T, Klarskov B, Rosenberg J, Kehlet H. Characteristics and prediction of early pain after laparoscopic cholecystectomy. Pain 2001;90:261–9
3. Slappendel R, Weber EW, Bugter ML, Dirksen R. The intensity of preoperative pain is directly correlated with the amount of morphine needed for postoperative analgesia. Anesth Analg 1999;88:146–8
4. Thomas T, Robinson C, Champion D, McKell M, Pell M. Prediction and assessment of the severity of post-operative pain and of satisfaction with management. Pain 1998;75:177–85
5. Gagliese L, Jackson M, Ritvo P, Wowk A, Katz J. Age is not an impediment to effective use of patient-controlled analgesia by surgical patients. Anesthesiology 2000;93:601–10
6. Owen H, McMillan V, Rogowski D. Postoperative pain therapy: a survey of patients’ expectations and their experiences. Pain 1990;41:303–7
7. Coderre TJ, Katz J, Vaccarino AL, Melzack R. Contribution of central neuroplasticity to pathological pain: review of clinical and experimental evidence. Pain 1993;52:259–85
8. Coderre TJ, Katz J. Peripheral and central hyperexcitability: differential signs and symptoms in persistent pain. Behav Brain Sci 1997;20:404–19; discussion 435–513
9. Woolf CJ, Chong MS. Preemptive analgesia—treating postoperative pain by preventing the establishment of central sensitization. Anesth Analg 1993;77:362–79
10. Koltzenburg M. Neural mechanisms of cutaneous nociceptive pain. Clin J Pain 2000;16:S131–8
11. Wilder-Smith OH, Arendt-Nielsen L. Postoperative hyperalgesia: its clinical importance and relevance. Anesthesiology 2006;104:601–7
12. Ali Z, Meyer RA, Campbell JN. Secondary hyperalgesia to mechanical but not heat stimuli following a capsaicin injection in hairy skin. Pain 1996;68:401–11
13. Koltzenburg M, Lundberg LE, Torebjork HE. Dynamic and static components of mechanical hyperalgesia in human hairy skin. Pain 1992;51:207–19
14. Tverskoy M, Oz Y, Isakson A, Finger J, Bradley EL Jr, Kissin I. Preemptive effect of fentanyl and ketamine on postoperative pain and wound hyperalgesia. Anesth Analg 1994;78:205–9
15. Wilder-Smith OH, Tassonyi E, Senly C, Otten P, Arendt-Nielsen L. Surgical pain is followed not only by spinal sensitization but also by supraspinal antinociception. Br J Anaesth 1996;76:816–21
16. Wilder-Smith OH. Changes in sensory processing after surgical nociception. Curr Rev Pain 2000;4:234–41
17. Wilder-Smith OH, Tassonyi E, Arendt-Nielsen L. Preoperative back pain is associated with diverse manifestations of central neuroplasticity. Pain 2002;97:189–94
18. Wilder-Smith OH, Tassonyi E, Crul BJ, Arendt-Nielsen L. Quantitative sensory testing and human surgery: effects of analgesic management on postoperative neuroplasticity. Anesthesiology 2003;98:1214–22
19. Stubhaug A, Breivik H, Eide PK, Kreunen M, Foss A. Mapping of punctuate hyperalgesia around a surgical incision demonstrates that ketamine is a powerful suppressor of central sensitization to pain following surgery. Acta Anaesthesiol Scand 1997;41:1124–32
20. Ilkjaer S, Bach LF, Nielsen PA, Wernberg M, Dahl JB. Effect of preoperative oral dextromethorphan on immediate and late postoperative pain and hyperalgesia after total abdominal hysterectomy. Pain 2000;86:19–24
21. Hsu YW, Somma J, Hung YC, Tsai PS, Yang CH, Chen CC. Predicting postoperative pain by preoperative pressure pain assessment. Anesthesiology 2005;103:613–18
22. Granot M, Lowenstein L, Yarnitsky D, Tamir A, Zimmer EZ. Postcesarean section pain prediction by preoperative experimental pain assessment. Anesthesiology 2003;98:1422–6
23. Harden RN, Bruehl S, Stanos S, Brander V, Chung OY, Saltz S, Adams A, Stulberg SD. Prospective examination of pain-related and psychological predictors of CRPS-like phenomena following total knee arthroplasty: a preliminary study. Pain 2003;106:393–400
24. Fruhstorfer H, Lindblom U, Schmidt WC. Method for quantitative estimation of thermal thresholds in patients. J Neurol Neurosurg Psychiatr 1976;39:1071–5
25. Bouhassira D, Attal N, Willer JC, Brasseur L. Painful and painless peripheral sensory neuropathies due to HIV infection: a comparison using quantitative sensory evaluation. Pain 1999;80:265–72
26. De Kock M, Lavand’homme P, Waterloos H. ‘Balanced analgesia’ in the perioperative period: is there a place for ketamine? Pain 2001;92:373–80
27. Joly V, Richebe P, Guignard B, Fletcher D, Maurette P, Sessler DI, Chauvin M. Remifentanil-induced postoperative hyperalgesia and its prevention with small-dose ketamine. Anesthesiology 2005;103:147–55
28. Moiniche S, Dahl JB, Erichsen CJ, Jensen LM, Kehlet H. Time course of subjective pain ratings, and wound and leg tenderness after hysterectomy. Acta Anaesthesiol Scand 1997;41:785–9
29. Wilder-Smith CH, Hill L, Dyer RA, Torr G, Coetzee E. Postoperative sensitization and pain after cesarean delivery and the effects of single IM doses of tramadol and diclofenac alone and in combination. Anesth Analg 2003;97:526–33
30. Werner MU, Duun P, Kehlet H. Prediction of postoperative pain by preoperative nociceptive responses to heat stimulation. Anesthesiology 2004;100:115–19; discussion 5A
31. Perkins FM, Kehlet H. Chronic pain as an outcome of surgery. A review of predictive factors. Anesthesiology 2000;93:1123–33
© 2007 International Anesthesia Research Society
32. Eisenach JC. Preventing chronic pain after surgery: who, how, and when? Reg Anesth Pain Med 2006;31:1–3